Filewatcher File Search File Search
Catalog
Content Search
» » » » » libfftw3-dev_3.3.2-3.1_sparc.deb » Content »
pkg://libfftw3-dev_3.3.2-3.1_sparc.deb:1559192/usr/share/doc/libfftw3-dev/  info  control  downloads

libfftw3-dev - Library for computing Fast Fourier Transforms - development…  more info»

changelog.gz

Sat Apr 28 10:55:09 EDT 2012  athena@fftw.org
  * Fix libfftw3/libfftw3_threads chicken-egg problem
  
  On most systems we want to build libfftw3 first, so that
  libfftw3_threads can depend upon libfftw3.  When producing a single
  combined-thread library (e.g. on Windows) we want the opposite,
  so that libfftw3 can include libfftw3_threads.
  

    M ./Makefile.am -1 +18
    M ./threads/Makefile.am +5

Sat Apr 28 10:11:28 EDT 2012  athena@fftw.org
  * updated NEWS for 3.3.2

    M ./NEWS +15

Thu Apr 26 19:36:11 EDT 2012  athena@fftw.org
  * change revision to 3.3.2

    M ./configure.ac -2 +2

Thu Apr 26 19:31:02 EDT 2012  athena@fftw.org
  * Remove old aligned_main() hack.
  
  On i386, in our benchmark program we used to manually aligned the
  stack to 16-byte boundary via asm trickery.  This was a good idea in
  1999 (and it was actually necessary to make things work) but the hack
  is now obsolete and it seems to break gcc-4.7.  So the hack is now
  gone.
  

    M ./libbench2/Makefile.am -6 +6
    R ./libbench2/aligned-main.c
    M ./libbench2/main.c -1 +1

Thu Mar 29 16:26:16 EDT 2012  athena@fftw.org
  * Bugfix: a couple of uninitialized values in the benchmark program

    M ./libbench2/problem.c -1 +1

Tue Mar 20 19:03:47 EDT 2012  athena@fftw.org
  * make libfftw{threads,mpi} depend upon libfftw for libtool purposes
  
  Thanks Julian Taylor for the patch

    M ./Makefile.am -1 +1
    M ./m4/acx_mpi.m4 -3 +3
    M ./mpi/Makefile.am +1
    M ./threads/Makefile.am +1

Tue Mar  6 04:44:00 EST 2012  stevenj@fftw.org
  * formatting tweak

    M ./doc/reference.texi -1 +1

Mon Mar  5 21:05:27 EST 2012  athena@fftw.org
  * destroying => overwriting

    M ./doc/tutorial.texi -1 +1

Fri Mar  2 10:31:20 EST 2012  stevenj@fftw.org
  * note that WISDOM_ONLY is a documented flag

    M ./api/fftw3.h -1 +1

Fri Mar  2 10:27:08 EST 2012  stevenj@fftw.org
  * check for icc pretending to be gcc before including quad-precision decls; thanks to Michael Anselmi for the bug report

    M ./api/fftw3.h +1

Fri Mar  2 10:23:19 EST 2012  stevenj@fftw.org
  * foo_CFLAGS needs to manually include AM_CFLAGS; thanks to Henry Gomersall for the Windows bug report

    M ./threads/Makefile.am -2 +2

Sat Feb 25 15:21:39 EST 2012  athena@fftw.org
  * update for latest mingw

    M ./support/BUILD-MINGW32.sh -3 +3
    M ./support/BUILD-MINGW64.sh -2 +7

Sat Feb 25 15:21:17 EST 2012  athena@fftw.org
  tagged fftw-3.3.1


Mon Feb 20 23:06:13 EST 2012  stevenj@fftw.org
  * added Fortran NEWS

    M ./NEWS +6

Mon Feb 20 23:00:13 EST 2012  stevenj@fftw.org
  * move non-portable extended/quad precision F03 interfaces into separate .f03 files (while keeping double/single in fftw3.f03 for minimal ABI breakage)

    M ./api/Makefile.am -3 +10
    M ./api/f03api.sh -2 +4
    M ./api/genf03.pl -5 +5
    M ./doc/modern-fortran.texi -2 +39
    M ./doc/mpi.texi -2 +2
    M ./mpi/Makefile.am -5 +11
    M ./mpi/f03api.sh -2 +4

Mon Feb 20 11:21:57 EST 2012  athena@fftw.org
  * rm mpi/fftw3-mpi.f03 at make clean time; thanks Tyler Luchko for the bug report.

    M ./mpi/Makefile.am +1

Mon Feb 20 11:18:24 EST 2012  athena@fftw.org
  * Disable a Visual Studion warning that was obnoxious enough for Sebastian Schuberth to send us a patch.

    M ./simd-support/simd-sse2.h +14

Mon Feb 20 11:18:06 EST 2012  athena@fftw.org
  * Change version to 3.3.1

    M ./configure.ac -1 +1

Mon Feb 20 11:03:15 EST 2012  athena@fftw.org
  * Integrated Visual Studio AVX patches by Carsten Steger

    M ./NEWS +3
    M ./simd-support/amd64-cpuid.h -10 +40
    M ./simd-support/simd-avx.h +6
    M ./simd-support/simd-sse2.h +8
    M ./simd-support/x86-cpuid.h +7

Wed Nov  9 10:13:32 EST 2011  stevenj@fftw.org
  * typo

    M ./doc/modern-fortran.texi -2 +2

Tue Nov  8 22:45:09 EST 2011  stevenj@fftw.org
  * add missing F77 set_timelimit function; thanks to Martin Diehl for the bug repory

    M ./api/f77funcs.h +5

Sun Sep 25 10:54:56 EDT 2011  athena@fftw.org
  * note requirement of /machine:x64 in windows x64 README

    M ./support/BUILD-MINGW64.sh +7

Sun Sep 18 09:28:20 EDT 2011  athena@fftw.org
  * AVX detection for MSVC

    M ./simd-support/amd64-cpuid.h -2 +10

Tue Sep 13 14:58:29 EDT 2011  athena@fftw.org
  * compile with C89

    M ./dft/bluestein.c -1 +1
    M ./dft/generic.c -1 +1
    M ./rdft/dht-rader.c -1 +1

Sat Sep  3 16:25:50 EDT 2011  athena@fftw.org
  * use the same search pruning heuristics for threaded plans as for nonthreaded plans

    M ./threads/ct.c -1 +3
    M ./threads/hc2hc.c -1 +3

Sat Sep  3 16:12:11 EDT 2011  athena@fftw.org
  * shorten ESTIMATE planning time for certain weird sizes
  
  FFTW includes a collection of "solvers" that apply to a subset of
  "problems".  Assume for simplicity that a "problem" is a single 1D
  complex transform of size N, even though real "problems" are much more
  general than that.  FFTW includes three "prime" solvers called
  "generic", "bluestein", and "rader", which implement different
  algorithms for prime sizes.
  
  Now, for a "problem" of size 13 (say) FFTW also includes special code
  that handles that size at high speed.  It would be a waste of time to
  measure the execution time of the prime solvers, since we know that
  the special code is way faster.  However, FFTW is modular and one may
  or may not include the special code for size 13, in which case we must
  resort to one of the "prime" solvers.  To address this issue, the
  "prime" solvers (and others) are proclaimed to be SLOW".  When
  planning, FFTW first tries to produce a plan ignoring all the SLOW
  solvers, and if this fails FFTW tries again allowing SLOW solvers.
  
  This heuristic works ok unless the sizes are too large.  For example
  for 1044000=2*2*2*2*2*3*3*5*5*5*29 FFTW explores a huge search tree of
  all zillion factorizations of 1044000/29, failing every time because
  29 is SLOW; then it finally allows SLOW solvers and finds a solution
  immediately.
  
  This patch proclaims solvers to be SLOW only for small values of N.
  For example, the "generic" solver implements an O(n^2) DFT algorithm;
  we say that it is SLOW only for N<=16.
  
  The side effects of this choice are as follows.  If one modifies FFTW to
  include a fast solver of size 17, then planning for N=17*K will be
  slower than today, because FFTW till try both the fast solver and the
  generic solver (which is SLOW today and therefore not tried, but is no
  longer SLOW after the patch).  If one removes a fast solver, of size say
  13, then he may still fall into the current exponential-search behavior
  for "problems" of size 13*HIGHLY_FACTORIZABLE_N.
  
  If somebody had compleined about transforms of size 1044000 ten years
  ago, "don't do that" would have been an acceptable answer.  I guess the
  bar is higher today, so I am going to include this patch in our 3.3.1
  release despite their side-effects for people who want to modify FFTW.
  

    M ./dft/bluestein.c -5 +14
    M ./dft/generic.c -15 +6
    M ./dft/rader.c -7 +7
    M ./kernel/ifftw.h +8
    M ./kernel/primes.c +6
    M ./rdft/dht-rader.c -7 +8
    M ./rdft/generic.c -14 +4

Sat Aug 27 13:55:24 EDT 2011  athena@fftw.org
  * Fix typo fftw_execute_dft_r2r => fftw_execute_r2r
  
  Thanks KIU Shueng Chuan for the bug report.

    M ./doc/reference.texi -1 +1

Fri Aug 26 06:13:55 EDT 2011  athena@fftw.org
  * In Rader's algorithm, compute the generator lazily.  
  
  The planner was spending a lot of time computing generators for
  plans that were immediately discarded.  Now we compute generators
  only when absolutely needed.

    M ./NEWS +5
    M ./dft/rader.c -3 +4
    M ./rdft/dht-rader.c -3 +4

Sun Aug 21 16:27:31 EDT 2011  athena@fftw.org
  tagged fftw-3.3.1-beta1


Sun Aug 21 16:16:38 EDT 2011  athena@fftw.org
  * Release notes for 3.3.1-beta1

    M ./NEWS +4
    M ./configure.ac -1 +1
    M ./doc/install.texi -8 +18
    M ./doc/other.texi -2 +3

Fri Aug 19 19:59:17 EDT 2011  stevenj@alum.mit.edu
  * make fftw_mpi_block routine 10x faster, since it is being called zillions of times (thanks to Tom Vacek for the profiling)

    M ./mpi/block.c -6 +2

Thu Aug 18 14:19:36 EDT 2011  athena@fftw.org
  * Implement autodetection of NEON extensions

    M ./simd-support/neon.c -4 +51

Sun Aug 14 14:12:29 EDT 2011  athena@fftw.org
  * Update the FSF address.
  
  The FSF moved downtown.

    M ./COPYRIGHT -1 +1
    M ./api/api.h -1 +1
    M ./api/apiplan.c -1 +1
    M ./api/configure.c -1 +1
    M ./api/execute-dft-c2r.c -1 +1
    M ./api/execute-dft-r2c.c -1 +1
    M ./api/execute-dft.c -1 +1
    M ./api/execute-r2r.c -1 +1
    M ./api/execute-split-dft-c2r.c -1 +1
    M ./api/execute-split-dft-r2c.c -1 +1
    M ./api/execute-split-dft.c -1 +1
    M ./api/execute.c -1 +1
    M ./api/export-wisdom-to-file.c -1 +1
    M ./api/export-wisdom-to-string.c -1 +1
    M ./api/export-wisdom.c -1 +1
    M ./api/f77api.c -1 +1
    M ./api/f77funcs.h -1 +1
    M ./api/flops.c -1 +1
    M ./api/forget-wisdom.c -1 +1
    M ./api/import-system-wisdom.c -1 +1
    M ./api/import-wisdom-from-file.c -1 +1
    M ./api/import-wisdom-from-string.c -1 +1
    M ./api/import-wisdom.c -1 +1
    M ./api/malloc.c -1 +1
    M ./api/map-r2r-kind.c -1 +1
    M ./api/mapflags.c -1 +1
    M ./api/mkprinter-file.c -1 +1
    M ./api/mktensor-iodims.h -1 +1
    M ./api/mktensor-rowmajor.c -1 +1
    M ./api/plan-dft-1d.c -1 +1
    M ./api/plan-dft-2d.c -1 +1
    M ./api/plan-dft-3d.c -1 +1
    M ./api/plan-dft-c2r-1d.c -1 +1
    M ./api/plan-dft-c2r-2d.c -1 +1
    M ./api/plan-dft-c2r-3d.c -1 +1
    M ./api/plan-dft-c2r.c -1 +1
    M ./api/plan-dft-r2c-1d.c -1 +1
    M ./api/plan-dft-r2c-2d.c -1 +1
    M ./api/plan-dft-r2c-3d.c -1 +1
    M ./api/plan-dft-r2c.c -1 +1
    M ./api/plan-dft.c -1 +1
    M ./api/plan-guru-dft-c2r.h -1 +1
    M ./api/plan-guru-dft-r2c.h -1 +1
    M ./api/plan-guru-dft.h -1 +1
    M ./api/plan-guru-r2r.h -1 +1
    M ./api/plan-guru-split-dft-c2r.h -1 +1
    M ./api/plan-guru-split-dft-r2c.h -1 +1
    M ./api/plan-guru-split-dft.h -1 +1
    M ./api/plan-many-dft-c2r.c -1 +1
    M ./api/plan-many-dft-r2c.c -1 +1
    M ./api/plan-many-dft.c -1 +1
    M ./api/plan-many-r2r.c -1 +1
    M ./api/plan-r2r-1d.c -1 +1
    M ./api/plan-r2r-2d.c -1 +1
    M ./api/plan-r2r-3d.c -1 +1
    M ./api/plan-r2r.c -1 +1
    M ./api/print-plan.c -1 +1
    M ./api/rdft2-pad.c -1 +1
    M ./api/the-planner.c -1 +1
    M ./api/version.c -1 +1
    M ./api/x77.h -1 +1
    M ./commercialize.sh -2 +2
    M ./dft/bluestein.c -1 +1
    M ./dft/buffered.c -1 +1
    M ./dft/codelet-dft.h -1 +1
    M ./dft/conf.c -1 +1
    M ./dft/ct.c -1 +1
    M ./dft/ct.h -1 +1
    M ./dft/dft.h -1 +1
    M ./dft/dftw-direct.c -1 +1
    M ./dft/dftw-directsq.c -1 +1
    M ./dft/dftw-generic.c -1 +1
    M ./dft/dftw-genericbuf.c -1 +1
    M ./dft/direct.c -1 +1
    M ./dft/generic.c -1 +1
    M ./dft/indirect-transpose.c -1 +1
    M ./dft/indirect.c -1 +1
    M ./dft/kdft-dif.c -1 +1
    M ./dft/kdft-difsq.c -1 +1
    M ./dft/kdft-dit.c -1 +1
    M ./dft/kdft.c -1 +1
    M ./dft/nop.c -1 +1
    M ./dft/plan.c -1 +1
    M ./dft/problem.c -1 +1
    M ./dft/rader.c -1 +1
    M ./dft/rank-geq2.c -1 +1
    M ./dft/scalar/n.c -1 +1
    M ./dft/scalar/n.h -1 +1
    M ./dft/scalar/t.c -1 +1
    M ./dft/scalar/t.h -1 +1
    M ./dft/simd/common/genus.c -1 +1
    M ./dft/simd/n1b.h -1 +1
    M ./dft/simd/n1f.h -1 +1
    M ./dft/simd/n2b.h -1 +1
    M ./dft/simd/n2f.h -1 +1
    M ./dft/simd/n2s.h -1 +1
    M ./dft/simd/q1b.h -1 +1
    M ./dft/simd/q1f.h -1 +1
    M ./dft/simd/t1b.h -1 +1
    M ./dft/simd/t1bu.h -1 +1
    M ./dft/simd/t1f.h -1 +1
    M ./dft/simd/t1fu.h -1 +1
    M ./dft/simd/t2b.h -1 +1
    M ./dft/simd/t2f.h -1 +1
    M ./dft/simd/t3b.h -1 +1
    M ./dft/simd/t3f.h -1 +1
    M ./dft/simd/ts.h -1 +1
    M ./dft/solve.c -1 +1
    M ./dft/vrank-geq1.c -1 +1
    M ./dft/zero.c -1 +1
    M ./doc/f77_wisdom.f -1 +1
    M ./doc/license.texi -3 +3
    M ./genfft/algsimp.ml -1 +1
    M ./genfft/algsimp.mli -1 +1
    M ./genfft/annotate.ml -1 +1
    M ./genfft/annotate.mli -1 +1
    M ./genfft/assoctable.ml -1 +1
    M ./genfft/assoctable.mli -1 +1
    M ./genfft/c.ml -1 +1
    M ./genfft/c.mli -1 +1
    M ./genfft/complex.ml -1 +1
    M ./genfft/complex.mli -1 +1
    M ./genfft/conv.ml -1 +1
    M ./genfft/conv.mli -1 +1
    M ./genfft/dag.ml -1 +1
    M ./genfft/dag.mli -1 +1
    M ./genfft/expr.ml -1 +1
    M ./genfft/expr.mli -1 +1
    M ./genfft/fft.ml -1 +1
    M ./genfft/fft.mli -1 +1
    M ./genfft/gen_hc2c.ml -1 +1
    M ./genfft/gen_hc2cdft.ml -1 +1
    M ./genfft/gen_hc2cdft_c.ml -1 +1
    M ./genfft/gen_hc2hc.ml -1 +1
    M ./genfft/gen_mdct.ml -1 +1
    M ./genfft/gen_notw.ml -1 +1
    M ./genfft/gen_notw_c.ml -1 +1
    M ./genfft/gen_r2cb.ml -1 +1
    M ./genfft/gen_r2cf.ml -1 +1
    M ./genfft/gen_r2r.ml -1 +1
    M ./genfft/gen_twiddle.ml -1 +1
    M ./genfft/gen_twiddle_c.ml -1 +1
    M ./genfft/gen_twidsq.ml -1 +1
    M ./genfft/gen_twidsq_c.ml -1 +1
    M ./genfft/genutil.ml -1 +1
    M ./genfft/littlesimp.ml -1 +1
    M ./genfft/littlesimp.mli -1 +1
    M ./genfft/magic.ml -1 +1
    M ./genfft/monads.ml -1 +1
    M ./genfft/number.ml -1 +1
    M ./genfft/number.mli -1 +1
    M ./genfft/oracle.ml -1 +1
    M ./genfft/oracle.mli -1 +1
    M ./genfft/schedule.ml -1 +1
    M ./genfft/schedule.mli -1 +1
    M ./genfft/simd.ml -1 +1
    M ./genfft/simd.mli -1 +1
    M ./genfft/simdmagic.ml -1 +1
    M ./genfft/to_alist.ml -1 +1
    M ./genfft/to_alist.mli -1 +1
    M ./genfft/trig.ml -1 +1
    M ./genfft/trig.mli -1 +1
    M ./genfft/twiddle.ml -1 +1
    M ./genfft/twiddle.mli -1 +1
    M ./genfft/unique.ml -1 +1
    M ./genfft/unique.mli -1 +1
    M ./genfft/util.ml -1 +1
    M ./genfft/util.mli -1 +1
    M ./genfft/variable.ml -1 +1
    M ./genfft/variable.mli -1 +1
    M ./kernel/align.c -1 +1
    M ./kernel/alloc.c -1 +1
    M ./kernel/assert.c -1 +1
    M ./kernel/awake.c -1 +1
    M ./kernel/buffered.c -1 +1
    M ./kernel/cpy1d.c -1 +1
    M ./kernel/cpy2d-pair.c -1 +1
    M ./kernel/cpy2d.c -1 +1
    M ./kernel/ct.c -1 +1
    M ./kernel/debug.c -1 +1
    M ./kernel/extract-reim.c -1 +1
    M ./kernel/hash.c -1 +1
    M ./kernel/iabs.c -1 +1
    M ./kernel/ifftw.h -1 +1
    M ./kernel/kalloc.c -1 +1
    M ./kernel/md5-1.c -1 +1
    M ./kernel/md5.c -1 +1
    M ./kernel/minmax.c -1 +1
    M ./kernel/ops.c -1 +1
    M ./kernel/pickdim.c -1 +1
    M ./kernel/plan.c -1 +1
    M ./kernel/planner.c -1 +1
    M ./kernel/primes.c -1 +1
    M ./kernel/print.c -1 +1
    M ./kernel/problem.c -1 +1
    M ./kernel/rader.c -1 +1
    M ./kernel/scan.c -1 +1
    M ./kernel/solver.c -1 +1
    M ./kernel/solvtab.c -1 +1
    M ./kernel/stride.c -1 +1
    M ./kernel/tensor.c -1 +1
    M ./kernel/tensor1.c -1 +1
    M ./kernel/tensor2.c -1 +1
    M ./kernel/tensor3.c -1 +1
    M ./kernel/tensor4.c -1 +1
    M ./kernel/tensor5.c -1 +1
    M ./kernel/tensor7.c -1 +1
    M ./kernel/tensor8.c -1 +1
    M ./kernel/tensor9.c -1 +1
    M ./kernel/tile2d.c -1 +1
    M ./kernel/timer.c -1 +1
    M ./kernel/transpose.c -1 +1
    M ./kernel/trig.c -1 +1
    M ./kernel/twiddle.c -1 +1
    M ./libbench2/aligned-main.c -1 +1
    M ./libbench2/bench-main.c -1 +1
    M ./libbench2/bench-user.h -1 +1
    M ./libbench2/bench.h -1 +1
    M ./libbench2/can-do.c -1 +1
    M ./libbench2/dotens2.c -1 +1
    M ./libbench2/info.c -1 +1
    M ./libbench2/main.c -1 +1
    M ./libbench2/my-getopt.c -1 +1
    M ./libbench2/my-getopt.h -1 +1
    M ./libbench2/problem.c -1 +1
    M ./libbench2/report.c -1 +1
    M ./libbench2/speed.c -1 +1
    M ./libbench2/tensor.c -1 +1
    M ./libbench2/timer.c -1 +1
    M ./libbench2/useropt.c -1 +1
    M ./libbench2/util.c -1 +1
    M ./libbench2/verify-dft.c -1 +1
    M ./libbench2/verify-lib.c -1 +1
    M ./libbench2/verify-r2r.c -1 +1
    M ./libbench2/verify-rdft2.c -1 +1
    M ./libbench2/verify.c -1 +1
    M ./libbench2/verify.h -1 +1
    M ./libbench2/zero.c -1 +1
    M ./mpi/any-true.c -1 +1
    M ./mpi/api.c -1 +1
    M ./mpi/block.c -1 +1
    M ./mpi/choose-radix.c -1 +1
    M ./mpi/conf.c -1 +1
    M ./mpi/dft-problem.c -1 +1
    M ./mpi/dft-rank-geq2-transposed.c -1 +1
    M ./mpi/dft-rank-geq2.c -1 +1
    M ./mpi/dft-rank1-bigvec.c -1 +1
    M ./mpi/dft-rank1.c -1 +1
    M ./mpi/dft-serial.c -1 +1
    M ./mpi/dft-solve.c -1 +1
    M ./mpi/dtensor.c -1 +1
    M ./mpi/ifftw-mpi.h -1 +1
    M ./mpi/mpi-dft.h -1 +1
    M ./mpi/mpi-rdft.h -1 +1
    M ./mpi/mpi-rdft2.h -1 +1
    M ./mpi/mpi-transpose.h -1 +1
    M ./mpi/rdft-problem.c -1 +1
    M ./mpi/rdft-rank-geq2-transposed.c -1 +1
    M ./mpi/rdft-rank-geq2.c -1 +1
    M ./mpi/rdft-rank1-bigvec.c -1 +1
    M ./mpi/rdft-serial.c -1 +1
    M ./mpi/rdft-solve.c -1 +1
    M ./mpi/rdft2-problem.c -1 +1
    M ./mpi/rdft2-rank-geq2-transposed.c -1 +1
    M ./mpi/rdft2-rank-geq2.c -1 +1
    M ./mpi/rdft2-serial.c -1 +1
    M ./mpi/rdft2-solve.c -1 +1
    M ./mpi/rearrange.c -1 +1
    M ./mpi/testsched.c -1 +1
    M ./mpi/transpose-alltoall.c -1 +1
    M ./mpi/transpose-pairwise.c -1 +1
    M ./mpi/transpose-problem.c -1 +1
    M ./mpi/transpose-recurse.c -1 +1
    M ./mpi/transpose-solve.c -1 +1
    M ./mpi/wisdom-api.c -1 +1
    M ./rdft/buffered.c -1 +1
    M ./rdft/buffered2.c -1 +1
    M ./rdft/codelet-rdft.h -1 +1
    M ./rdft/conf.c -1 +1
    M ./rdft/ct-hc2c-direct.c -1 +1
    M ./rdft/ct-hc2c.c -1 +1
    M ./rdft/ct-hc2c.h -1 +1
    M ./rdft/dft-r2hc.c -1 +1
    M ./rdft/dht-r2hc.c -1 +1
    M ./rdft/dht-rader.c -1 +1
    M ./rdft/direct-r2c.c -1 +1
    M ./rdft/direct-r2r.c -1 +1
    M ./rdft/direct2.c -1 +1
    M ./rdft/generic.c -1 +1
    M ./rdft/hc2hc-direct.c -1 +1
    M ./rdft/hc2hc-generic.c -1 +1
    M ./rdft/hc2hc.c -1 +1
    M ./rdft/hc2hc.h -1 +1
    M ./rdft/indirect.c -1 +1
    M ./rdft/khc2c.c -1 +1
    M ./rdft/khc2hc.c -1 +1
    M ./rdft/kr2c.c -1 +1
    M ./rdft/kr2r.c -1 +1
    M ./rdft/nop.c -1 +1
    M ./rdft/nop2.c -1 +1
    M ./rdft/plan.c -1 +1
    M ./rdft/plan2.c -1 +1
    M ./rdft/problem.c -1 +1
    M ./rdft/problem2.c -1 +1
    M ./rdft/rank-geq2-rdft2.c -1 +1
    M ./rdft/rank-geq2.c -1 +1
    M ./rdft/rank0-rdft2.c -1 +1
    M ./rdft/rank0.c -1 +1
    M ./rdft/rdft-dht.c -1 +1
    M ./rdft/rdft.h -1 +1
    M ./rdft/rdft2-inplace-strides.c -1 +1
    M ./rdft/rdft2-rdft.c -1 +1
    M ./rdft/rdft2-strides.c -1 +1
    M ./rdft/rdft2-tensor-max-index.c -1 +1
    M ./rdft/scalar/hb.h -1 +1
    M ./rdft/scalar/hc2c.c -1 +1
    M ./rdft/scalar/hc2cb.h -1 +1
    M ./rdft/scalar/hc2cf.h -1 +1
    M ./rdft/scalar/hf.h -1 +1
    M ./rdft/scalar/hfb.c -1 +1
    M ./rdft/scalar/r2c.c -1 +1
    M ./rdft/scalar/r2cb.h -1 +1
    M ./rdft/scalar/r2cbIII.h -1 +1
    M ./rdft/scalar/r2cf.h -1 +1
    M ./rdft/scalar/r2cfII.h -1 +1
    M ./rdft/scalar/r2r.c -1 +1
    M ./rdft/scalar/r2r.h -1 +1
    M ./rdft/simd/common/genus.c -1 +1
    M ./rdft/simd/hc2cbv.h -1 +1
    M ./rdft/simd/hc2cfv.h -1 +1
    M ./rdft/solve.c -1 +1
    M ./rdft/solve2.c -1 +1
    M ./rdft/vrank-geq1-rdft2.c -1 +1
    M ./rdft/vrank-geq1.c -1 +1
    M ./rdft/vrank3-transpose.c -1 +1
    M ./reodft/conf.c -1 +1
    M ./reodft/redft00e-r2hc-pad.c -1 +1
    M ./reodft/redft00e-r2hc.c -1 +1
    M ./reodft/reodft.h -1 +1
    M ./reodft/reodft00e-splitradix.c -1 +1
    M ./reodft/reodft010e-r2hc.c -1 +1
    M ./reodft/reodft11e-r2hc-odd.c -1 +1
    M ./reodft/reodft11e-r2hc.c -1 +1
    M ./reodft/reodft11e-radix2.c -1 +1
    M ./reodft/rodft00e-r2hc-pad.c -1 +1
    M ./reodft/rodft00e-r2hc.c -1 +1
    M ./simd-support/altivec.c -1 +1
    M ./simd-support/amd64-cpuid.h -1 +1
    M ./simd-support/avx.c -1 +1
    M ./simd-support/neon.c -1 +1
    M ./simd-support/simd-altivec.h -1 +1
    M ./simd-support/simd-avx.h -1 +1
    M ./simd-support/simd-common.h -1 +1
    M ./simd-support/simd-neon.h -1 +1
    M ./simd-support/simd-sse2.h -1 +1
    M ./simd-support/sse2-nonportable.c -1 +1
    M ./simd-support/sse2.c -1 +1
    M ./simd-support/taint.c -1 +1
    M ./simd-support/x86-cpuid.h -1 +1
    M ./simd/altivec.c -1 +1
    M ./simd/mips_ps.c -1 +1
    M ./simd/mips_ps.h -1 +1
    M ./simd/nonportable/sse.c -1 +1
    M ./simd/nonportable/sse2.c -1 +1
    M ./simd/simd-altivec.h -1 +1
    M ./simd/simd-mips_ps.h -1 +1
    M ./simd/simd-sse.h -1 +1
    M ./simd/simd-sse2.h -1 +1
    M ./simd/simd.h -1 +1
    M ./simd/sse.c -1 +1
    M ./simd/sse2.c -1 +1
    M ./simd/taint.c -1 +1
    M ./simd/x86-cpuid.h -1 +1
    M ./threads/api.c -1 +1
    M ./threads/conf.c -1 +1
    M ./threads/ct.c -1 +1
    M ./threads/dft-vrank-geq1.c -1 +1
    M ./threads/f77api.c -1 +1
    M ./threads/f77funcs.h -1 +1
    M ./threads/hc2hc.c -1 +1
    M ./threads/openmp.c -1 +1
    M ./threads/rdft-vrank-geq1.c -1 +1
    M ./threads/threads.c -1 +1
    M ./threads/threads.h -1 +1
    M ./threads/vrank-geq1-rdft2.c -1 +1
    M ./tools/fftw-wisdom-to-conf.1 -1 +1
    M ./tools/fftw-wisdom-to-conf.in -1 +1
    M ./tools/fftw-wisdom.c -1 +1
    M ./tools/fftw_wisdom.1.in -1 +1

Thu Aug 11 14:54:38 EDT 2011  stevenj@fftw.org
  * allow specifying TRANSPOSE_{IN/OUT} transpose plans, since libbench does not canonicalize rnk=1 n=1 plans as rnk=0

    M ./mpi/mpi-bench.c -2 +2

Thu Aug 11 14:17:24 EDT 2011  stevenj@fftw.org
  * check.pl should occasionally check DESTROY_INPUT problems too (especially since those enable slightly different algorithms in MPI)

    M ./tests/check.pl +1

Thu Aug 11 12:37:51 EDT 2011  stevenj@fftw.org
  * unify post-MPI transpose handling in pairwise and alltoall solvers; should make the former faster in the destroy-input out-of-place case, and the latter more widely applicable

    M ./mpi/mpi-transpose.h +5
    M ./mpi/transpose-alltoall.c -69 +24
    M ./mpi/transpose-pairwise.c -66 +97

Mon Aug  8 10:06:14 EDT 2011  athena@fftw.org
  * Add support for ARM NEON

    M ./Makefile.am -1 +6
    M ./api/version.c +4
    M ./configure.ac -5 +23
    M ./dft/codelet-dft.h +1
    M ./dft/conf.c +4
    M ./dft/simd/Makefile.am -1 +1
    A ./dft/simd/neon/
    A ./dft/simd/neon/Makefile.am
    M ./kernel/ifftw.h +2
    M ./m4/ax_cc_maxopt.m4 +10
    M ./rdft/codelet-rdft.h +1
    M ./rdft/conf.c +4
    M ./rdft/simd/Makefile.am -1 +1
    A ./rdft/simd/neon/
    A ./rdft/simd/neon/Makefile.am
    M ./simd-support/Makefile.am -1 +2
    A ./simd-support/neon.c
    M ./simd-support/simd-common.h +3
    A ./simd-support/simd-neon.h

Fri Aug  5 17:25:32 EDT 2011  stevenj@fftw.org
  * more C++ paranoia

    M ./tests/fftw-bench.h +10

Fri Aug  5 17:02:00 EDT 2011  stevenj@fftw.org
  * tentative version bump for 3.3.1

    M ./NEWS +5
    M ./configure.ac -2 +2

Fri Aug  5 16:52:28 EDT 2011  stevenj@fftw.org
  * fixes so that MPI code compiles when MPICC is a C++ compiler, even if the serial code is compiled with a C compiler; thanks to Kyle Spyksma for the bug report

    M ./api/api.h +12
    M ./api/plan-guru-r2r.h -2
    M ./api/plan-many-r2r.c -2
    M ./dft/dft.h +9
    M ./kernel/ifftw.h +9
    M ./mpi/api.c -3 +1
    M ./mpi/dtensor.c -1 +1
    M ./mpi/ifftw-mpi.h -1 +4
    M ./mpi/mpi-bench.c -1 +1
    M ./rdft/rdft.h +9

Fri Aug  5 16:04:06 EDT 2011  stevenj@fftw.org
  * use correct precision in f03-wrap.c, avoiding a (harmless) implicit pointer cast that prevented compilation under C++; thanks to Kyle Spyksma for the bug report

    M ./mpi/genf03-wrap.pl +2

Fri Aug  5 14:04:32 EDT 2011  stevenj@fftw.org
  * manual typo

    M ./doc/modern-fortran.texi -1 +1

Wed Jul 27 12:50:25 EDT 2011  athena@fftw.org
  tagged fftw-3.3


Tue Jul 26 20:55:45 EDT 2011  athena@fftw.org
  * Honor WITH_OUR_MALLOC in libbench2

    M ./libbench2/util.c -6 +5

Tue Jul 26 20:27:28 EDT 2011  athena@fftw.org
  * fixed typo: incorrect name of combined threads library on Windows

    M ./Makefile.am -1 +1

Mon Jul 25 14:38:20 EDT 2011  stevenj@fftw.org
  * 3.3 version bump & NEWS

    M ./NEWS -5 +16
    M ./configure.ac -1 +1

Mon Jul 25 14:37:48 EDT 2011  stevenj@fftw.org
  * use int(..., C_SIZE_T) rather than declaring another variable in the Fortran examples

    M ./doc/modern-fortran.texi -12 +9

Wed Jul 13 05:02:32 EDT 2011  stevenj@fftw.org
  * typo, thanks to Rhys Ulerich for the comment

    M ./doc/mpi.texi -1 +1

Mon Jul 11 14:39:52 EDT 2011  athena@fftw.org
  * Fix bug in bubblesort
  
  Bubblesort was not sorting.  This was a bug in the benchmark library
  (not in FFTW per se), and it impacted the benchmark program
  with --report-time and --report-mflops causing it to output
  an incorrect value for the median.  (The minimum, maximum, and
  average value were correct.)  Thanks Dima Baksheev of Intel for
  reporting this bug.
  

    M ./libbench2/report.c -2 +2

Fri Jul  8 13:35:59 EDT 2011  stevenj@fftw.org
  * small manual typos

    M ./doc/modern-fortran.texi -2 +2

Wed Jul  6 10:49:40 EDT 2011  athena@fftw.org
  * Detection of altivec.h requires $ALTIVEC_CFLAGS

    M ./configure.ac -5 +4

Tue Jul  5 19:58:47 EDT 2011  athena@fftw.org
  * Introduce fake dependency so that my-getopt.c is recompiled
  
  my-getopt.c does not depend on anything, and so it is not rebuilt when
  reconfiguring for a different ISA (e.g., CC="gcc -m32" vs CC="gcc
  -m64").  Add a fake dependency on <config.h> so that the file is
  recompiled.
  
  

    M ./libbench2/my-getopt.c +1

Tue Jul  5 18:53:36 EDT 2011  stevenj@fftw.org
  * support compiling/installing --enable-threads --enable-openmp at the same time, although in this case the test program only uses the threads variety.  Update documentation accordingly, and in general expand the documentation of the OpenMP support

    M ./Makefile.am -1 +1
    M ./configure.ac -21 +13
    M ./doc/install.texi -14 +13
    M ./doc/threads.texi -19 +43
    M ./mpi/Makefile.am -2 +6
    M ./tests/Makefile.am -2 +6
    M ./tests/fftw-bench.c -1 +3
    M ./threads/Makefile.am -6 +20
    M ./threads/conf.c -4
    M ./threads/openmp.c -4
    M ./threads/threads.c -3
    M ./tools/Makefile.am -4 +8

Tue Jul  5 16:04:03 EDT 2011  stevenj@fftw.org
  * call omp_set_num_threads in fftw-bench so that the number of OpenMP threads corresponds with the number of FFTW threads

    M ./tests/fftw-bench.c +5

Tue Jul  5 16:03:06 EDT 2011  stevenj@fftw.org
  * when --enable-openmp, install as fftw3_omp rather than fftw3_threads, so that both the POSIX threads and OpenMP variants of FFTW can be installed at once

    M ./Makefile.am -1 +1
    M ./configure.ac +4
    M ./threads/Makefile.am -4 +4

Sat Jul  2 02:21:22 EDT 2011  stevenj@fftw.org
  * don't even declare an fftw_execute interface in Fortran, since it is unsafe and we recommend against it anyway; thanks to Arjen Markus for the suggestion

    M ./api/f03api.sh -1 +1
    M ./doc/modern-fortran.texi -2 +4

Fri Jul  1 14:35:44 EDT 2011  athena@fftw.org
  * consistently use the order single, double, long double

    M ./support/BUILD-MINGW32.sh -1 +1
    M ./support/BUILD-MINGW64.sh -1 +1

Wed Jun 29 17:27:06 EDT 2011  athena@fftw.org
  * MSVC AVX 64-bit detection does not work, punt for now.

    M ./simd-support/amd64-cpuid.h -16 +2

Wed Jun 29 15:52:27 EDT 2011  stevenj@fftw.org
  * fixed typo, added note on transposed flags for r2c/c2r; thanks to Rhys Ulerich for the suggestions

    M ./doc/mpi.texi -1 +14

Wed Jun 29 09:41:39 EDT 2011  athena@fftw.org
  * fixes for compiling with MSVC (untested)

    M ./simd-support/amd64-cpuid.h -2 +16
    M ./simd-support/simd-avx.h -11 +16
    M ./simd-support/x86-cpuid.h -1 +7

Tue Jun 28 16:48:36 EDT 2011  athena@fftw.org
  * comment

    M ./simd-support/sse2-nonportable.c +3

Mon Jun 27 21:01:56 EDT 2011  stevenj@fftw.org
  * rm extraneous line break in HTML output ... I hate texinfo

    M ./doc/reference.texi -6 +3

Mon Jun 27 06:43:57 EDT 2011  athena@fftw.org
  tagged fftw-3.3-beta1


Mon Jun 27 00:47:33 EDT 2011  stevenj@fftw.org
  * maintainer-clean should delete html directory (otherwwise we keep obsolete HTML files in the dist tarball, sigh)

    M ./doc/Makefile.am -2 +2

Mon Jun 27 00:45:28 EDT 2011  stevenj@fftw.org
  tagged fftw-3.3-beta1


Sun Jun 26 23:36:32 EDT 2011  stevenj@fftw.org
  * update copyright year in manual

    M ./doc/license.texi -2 +2

Sun Jun 26 22:52:54 EDT 2011  stevenj@fftw.org
  * whoops, don't dist .f03 headers, since those are built by the user's Makefile

    M ./api/Makefile.am -1 +2
    M ./mpi/Makefile.am -1 +2

Sun Jun 26 22:43:49 EDT 2011  stevenj@fftw.org
  * fix embarrassing deadlock/crashing bug in my previous nowisdom_hook fix -- I forgot to handle the case where one process has wisdom and another one doesn't, requiring a nowisdom_hook in the latter case; this should only affect MPI transforms since otherwise these hook functions are NULL

    M ./kernel/ifftw.h +1
    M ./kernel/planner.c -44 +48
    M ./mpi/api.c +14

Sun Jun 26 21:02:15 EDT 2011  stevenj@fftw.org
  * subsubheadings, MPI transpose reference

    M ./doc/mpi.texi -7 +53

Sun Jun 26 20:48:53 EDT 2011  stevenj@fftw.org
  * add MPI plan reference

    M ./doc/mpi.texi -5 +197

Sun Jun 26 17:07:21 EDT 2011  stevenj@fftw.org
  * portions of MPI reference docs; tweaks to NEWS

    M ./NEWS -2 +3
    M ./doc/mpi.texi -12 +264

Sun Jun 26 12:40:43 EDT 2011  stevenj@fftw.org
  * use $(CHECK_PL_OPTS) more consistently

    M ./tests/Makefile.am -1 +1

Sun Jun 26 10:04:54 EDT 2011  athena@fftw.org
  * accept \r\n as well as \n.  Grrr...

    M ./tests/check.pl -1 +2

Sun Jun 26 09:52:11 EDT 2011  athena@fftw.org
  * new configure option --with-incoming-stack-boundary=N
  
  This option selects CFLAGS to align the stack at all externally-callable
  functions.  This currently comprises api/* and threads/*
  

    M ./api/Makefile.am +1
    M ./configure.ac +14
    M ./support/BUILD-MINGW32.sh -1 +1
    M ./threads/Makefile.am +1

Sun Jun 26 09:51:37 EDT 2011  athena@fftw.org
  * add -fomit-frame-pointer back
  
  Somehow -O3 does not imply -fomit-frame-pointer on ia32
  

    M ./m4/ax_cc_maxopt.m4 -2 +3

Sun Jun 26 07:20:27 EDT 2011  athena@fftw.org
  * Note that removal of mips-ps is temporary.

    M ./NEWS -3 +5

Sat Jun 25 23:15:03 EDT 2011  stevenj@fftw.org
  * update copyright year

    M ./COPYRIGHT -2 +2
    M ./api/api.h -2 +2
    M ./api/apiplan.c -2 +2
    M ./api/configure.c -2 +2
    M ./api/execute-dft-c2r.c -2 +2
    M ./api/execute-dft-r2c.c -2 +2
    M ./api/execute-dft.c -2 +2
    M ./api/execute-r2r.c -2 +2
    M ./api/execute-split-dft-c2r.c -2 +2
    M ./api/execute-split-dft-r2c.c -2 +2
    M ./api/execute-split-dft.c -2 +2
    M ./api/execute.c -2 +2
    M ./api/export-wisdom-to-file.c -2 +2
    M ./api/export-wisdom-to-string.c -2 +2
    M ./api/export-wisdom.c -2 +2
    M ./api/f77api.c -2 +2
    M ./api/f77funcs.h -2 +2
    M ./api/fftw3.h -2 +2
    M ./api/flops.c -2 +2
    M ./api/forget-wisdom.c -2 +2
    M ./api/import-system-wisdom.c -2 +2
    M ./api/import-wisdom-from-file.c -2 +2
    M ./api/import-wisdom-from-string.c -2 +2
    M ./api/import-wisdom.c -2 +2
    M ./api/malloc.c -2 +2
    M ./api/map-r2r-kind.c -2 +2
    M ./api/mapflags.c -2 +2
    M ./api/mkprinter-file.c -2 +2
    M ./api/mktensor-iodims.h -2 +2
    M ./api/mktensor-rowmajor.c -2 +2
    M ./api/plan-dft-1d.c -2 +2
    M ./api/plan-dft-2d.c -2 +2
    M ./api/plan-dft-3d.c -2 +2
    M ./api/plan-dft-c2r-1d.c -2 +2
    M ./api/plan-dft-c2r-2d.c -2 +2
    M ./api/plan-dft-c2r-3d.c -2 +2
    M ./api/plan-dft-c2r.c -2 +2
    M ./api/plan-dft-r2c-1d.c -2 +2
    M ./api/plan-dft-r2c-2d.c -2 +2
    M ./api/plan-dft-r2c-3d.c -2 +2
    M ./api/plan-dft-r2c.c -2 +2
    M ./api/plan-dft.c -2 +2
    M ./api/plan-guru-dft-c2r.h -2 +2
    M ./api/plan-guru-dft-r2c.h -2 +2
    M ./api/plan-guru-dft.h -2 +2
    M ./api/plan-guru-r2r.h -2 +2
    M ./api/plan-guru-split-dft-c2r.h -2 +2
    M ./api/plan-guru-split-dft-r2c.h -2 +2
    M ./api/plan-guru-split-dft.h -2 +2
    M ./api/plan-many-dft-c2r.c -2 +2
    M ./api/plan-many-dft-r2c.c -2 +2
    M ./api/plan-many-dft.c -2 +2
    M ./api/plan-many-r2r.c -2 +2
    M ./api/plan-r2r-1d.c -2 +2
    M ./api/plan-r2r-2d.c -2 +2
    M ./api/plan-r2r-3d.c -2 +2
    M ./api/plan-r2r.c -2 +2
    M ./api/print-plan.c -2 +2
    M ./api/rdft2-pad.c -2 +2
    M ./api/the-planner.c -2 +2
    M ./api/version.c -2 +2
    M ./api/x77.h -2 +2
    M ./dft/bluestein.c -2 +2
    M ./dft/buffered.c -2 +2
    M ./dft/codelet-dft.h -2 +2
    M ./dft/conf.c -2 +2
    M ./dft/ct.c -2 +2
    M ./dft/ct.h -2 +2
    M ./dft/dft.h -2 +2
    M ./dft/dftw-direct.c -2 +2
    M ./dft/dftw-directsq.c -2 +2
    M ./dft/dftw-generic.c -2 +2
    M ./dft/dftw-genericbuf.c -2 +2
    M ./dft/direct.c -2 +2
    M ./dft/generic.c -2 +2
    M ./dft/indirect-transpose.c -2 +2
    M ./dft/indirect.c -2 +2
    M ./dft/kdft-dif.c -2 +2
    M ./dft/kdft-difsq.c -2 +2
    M ./dft/kdft-dit.c -2 +2
    M ./dft/kdft.c -2 +2
    M ./dft/nop.c -2 +2
    M ./dft/plan.c -2 +2
    M ./dft/problem.c -2 +2
    M ./dft/rader.c -2 +2
    M ./dft/rank-geq2.c -2 +2
    M ./dft/scalar/n.c -2 +2
    M ./dft/scalar/n.h -2 +2
    M ./dft/scalar/t.c -2 +2
    M ./dft/scalar/t.h -2 +2
    M ./dft/simd/common/genus.c -2 +2
    M ./dft/simd/n1b.h -2 +2
    M ./dft/simd/n1f.h -2 +2
    M ./dft/simd/n2b.h -2 +2
    M ./dft/simd/n2f.h -2 +2
    M ./dft/simd/n2s.h -2 +2
    M ./dft/simd/q1b.h -2 +2
    M ./dft/simd/q1f.h -2 +2
    M ./dft/simd/t1b.h -2 +2
    M ./dft/simd/t1bu.h -2 +2
    M ./dft/simd/t1f.h -2 +2
    M ./dft/simd/t1fu.h -2 +2
    M ./dft/simd/t2b.h -2 +2
    M ./dft/simd/t2f.h -2 +2
    M ./dft/simd/t3b.h -2 +2
    M ./dft/simd/t3f.h -2 +2
    M ./dft/simd/ts.h -2 +2
    M ./dft/solve.c -2 +2
    M ./dft/vrank-geq1.c -2 +2
    M ./dft/zero.c -2 +2
    M ./doc/f77_wisdom.f -2 +2
    M ./genfft/algsimp.ml -2 +2
    M ./genfft/algsimp.mli -2 +2
    M ./genfft/annotate.ml -2 +2
    M ./genfft/annotate.mli -2 +2
    M ./genfft/assoctable.ml -2 +2
    M ./genfft/assoctable.mli -2 +2
    M ./genfft/c.ml -2 +2
    M ./genfft/c.mli -2 +2
    M ./genfft/complex.ml -2 +2
    M ./genfft/complex.mli -2 +2
    M ./genfft/conv.ml -2 +2
    M ./genfft/conv.mli -2 +2
    M ./genfft/dag.ml -2 +2
    M ./genfft/dag.mli -2 +2
    M ./genfft/expr.ml -2 +2
    M ./genfft/expr.mli -2 +2
    M ./genfft/fft.ml -2 +2
    M ./genfft/fft.mli -2 +2
    M ./genfft/gen_hc2c.ml -2 +2
    M ./genfft/gen_hc2cdft.ml -2 +2
    M ./genfft/gen_hc2cdft_c.ml -2 +2
    M ./genfft/gen_hc2hc.ml -2 +2
    M ./genfft/gen_mdct.ml -2 +2
    M ./genfft/gen_notw.ml -2 +2
    M ./genfft/gen_notw_c.ml -2 +2
    M ./genfft/gen_r2cb.ml -2 +2
    M ./genfft/gen_r2cf.ml -2 +2
    M ./genfft/gen_r2r.ml -2 +2
    M ./genfft/gen_twiddle.ml -2 +2
    M ./genfft/gen_twiddle_c.ml -2 +2
    M ./genfft/gen_twidsq.ml -2 +2
    M ./genfft/gen_twidsq_c.ml -2 +2
    M ./genfft/genutil.ml -2 +2
    M ./genfft/littlesimp.ml -2 +2
    M ./genfft/littlesimp.mli -2 +2
    M ./genfft/magic.ml -2 +2
    M ./genfft/monads.ml -2 +2
    M ./genfft/number.ml -2 +2
    M ./genfft/number.mli -2 +2
    M ./genfft/oracle.ml -2 +2
    M ./genfft/oracle.mli -2 +2
    M ./genfft/schedule.ml -2 +2
    M ./genfft/schedule.mli -2 +2
    M ./genfft/simd.ml -2 +2
    M ./genfft/simd.mli -2 +2
    M ./genfft/simdmagic.ml -2 +2
    M ./genfft/to_alist.ml -2 +2
    M ./genfft/to_alist.mli -2 +2
    M ./genfft/trig.ml -2 +2
    M ./genfft/trig.mli -2 +2
    M ./genfft/twiddle.ml -2 +2
    M ./genfft/twiddle.mli -2 +2
    M ./genfft/unique.ml -2 +2
    M ./genfft/unique.mli -2 +2
    M ./genfft/util.ml -2 +2
    M ./genfft/util.mli -2 +2
    M ./genfft/variable.ml -2 +2
    M ./genfft/variable.mli -2 +2
    M ./kernel/align.c -2 +2
    M ./kernel/alloc.c -2 +2
    M ./kernel/assert.c -2 +2
    M ./kernel/awake.c -2 +2
    M ./kernel/buffered.c -2 +2
    M ./kernel/cpy1d.c -2 +2
    M ./kernel/cpy2d-pair.c -2 +2
    M ./kernel/cpy2d.c -2 +2
    M ./kernel/ct.c -2 +2
    M ./kernel/cycle.h -2 +2
    M ./kernel/debug.c -2 +2
    M ./kernel/extract-reim.c -2 +2
    M ./kernel/hash.c -2 +2
    M ./kernel/iabs.c -2 +2
    M ./kernel/ifftw.h -2 +2
    M ./kernel/kalloc.c -2 +2
    M ./kernel/md5-1.c -2 +2
    M ./kernel/md5.c -2 +2
    M ./kernel/minmax.c -2 +2
    M ./kernel/ops.c -2 +2
    M ./kernel/pickdim.c -2 +2
    M ./kernel/plan.c -2 +2
    M ./kernel/primes.c -2 +2
    M ./kernel/print.c -2 +2
    M ./kernel/problem.c -2 +2
    M ./kernel/rader.c -2 +2
    M ./kernel/scan.c -2 +2
    M ./kernel/solver.c -2 +2
    M ./kernel/solvtab.c -2 +2
    M ./kernel/stride.c -2 +2
    M ./kernel/tensor.c -2 +2
    M ./kernel/tensor1.c -2 +2
    M ./kernel/tensor2.c -2 +2
    M ./kernel/tensor3.c -2 +2
    M ./kernel/tensor4.c -2 +2
    M ./kernel/tensor5.c -2 +2
    M ./kernel/tensor7.c -2 +2
    M ./kernel/tensor8.c -2 +2
    M ./kernel/tensor9.c -2 +2
    M ./kernel/tile2d.c -2 +2
    M ./kernel/timer.c -2 +2
    M ./kernel/transpose.c -2 +2
    M ./kernel/trig.c -2 +2
    M ./kernel/twiddle.c -2 +2
    M ./libbench2/dotens2.c -2 +2
    M ./libbench2/my-getopt.c -2 +2
    M ./libbench2/my-getopt.h -2 +2
    M ./libbench2/verify-dft.c -2 +2
    M ./libbench2/verify-lib.c -2 +2
    M ./libbench2/verify-r2r.c -2 +2
    M ./libbench2/verify-rdft2.c -2 +2
    M ./libbench2/verify.h -2 +2
    M ./mpi/any-true.c -2 +2
    M ./mpi/api.c -2 +2
    M ./mpi/block.c -2 +2
    M ./mpi/choose-radix.c -2 +2
    M ./mpi/conf.c -2 +2
    M ./mpi/dft-problem.c -2 +2
    M ./mpi/dft-rank-geq2-transposed.c -2 +2
    M ./mpi/dft-rank-geq2.c -2 +2
    M ./mpi/dft-rank1-bigvec.c -2 +2
    M ./mpi/dft-rank1.c -2 +2
    M ./mpi/dft-serial.c -2 +2
    M ./mpi/dft-solve.c -2 +2
    M ./mpi/dtensor.c -2 +2
    M ./mpi/fftw3-mpi.h -2 +2
    M ./mpi/ifftw-mpi.h -2 +2
    M ./mpi/mpi-dft.h -2 +2
    M ./mpi/mpi-rdft.h -2 +2
    M ./mpi/mpi-rdft2.h -2 +2
    M ./mpi/mpi-transpose.h -2 +2
    M ./mpi/rdft-problem.c -2 +2
    M ./mpi/rdft-rank-geq2-transposed.c -2 +2
    M ./mpi/rdft-rank-geq2.c -2 +2
    M ./mpi/rdft-rank1-bigvec.c -2 +2
    M ./mpi/rdft-serial.c -2 +2
    M ./mpi/rdft-solve.c -2 +2
    M ./mpi/rdft2-problem.c -2 +2
    M ./mpi/rdft2-rank-geq2-transposed.c -2 +2
    M ./mpi/rdft2-rank-geq2.c -2 +2
    M ./mpi/rdft2-serial.c -2 +2
    M ./mpi/rdft2-solve.c -2 +2
    M ./mpi/rearrange.c -2 +2
    M ./mpi/testsched.c -1 +1
    M ./mpi/transpose-alltoall.c -2 +2
    M ./mpi/transpose-pairwise.c -2 +2
    M ./mpi/transpose-problem.c -2 +2
    M ./mpi/transpose-recurse.c -2 +2
    M ./mpi/transpose-solve.c -2 +2
    M ./mpi/wisdom-api.c -2 +2
    M ./rdft/buffered.c -2 +2
    M ./rdft/buffered2.c -2 +2
    M ./rdft/codelet-rdft.h -2 +2
    M ./rdft/conf.c -2 +2
    M ./rdft/ct-hc2c-direct.c -2 +2
    M ./rdft/ct-hc2c.c -2 +2
    M ./rdft/ct-hc2c.h -2 +2
    M ./rdft/dft-r2hc.c -2 +2
    M ./rdft/dht-r2hc.c -2 +2
    M ./rdft/dht-rader.c -2 +2
    M ./rdft/direct-r2c.c -2 +2
    M ./rdft/direct-r2r.c -2 +2
    M ./rdft/direct2.c -2 +2
    M ./rdft/generic.c -2 +2
    M ./rdft/hc2hc-direct.c -2 +2
    M ./rdft/hc2hc-generic.c -2 +2
    M ./rdft/hc2hc.c -2 +2
    M ./rdft/hc2hc.h -2 +2
    M ./rdft/indirect.c -2 +2
    M ./rdft/khc2c.c -2 +2
    M ./rdft/khc2hc.c -2 +2
    M ./rdft/kr2c.c -2 +2
    M ./rdft/kr2r.c -2 +2
    M ./rdft/nop.c -2 +2
    M ./rdft/nop2.c -2 +2
    M ./rdft/plan.c -2 +2
    M ./rdft/plan2.c -2 +2
    M ./rdft/problem.c -2 +2
    M ./rdft/problem2.c -2 +2
    M ./rdft/rank-geq2-rdft2.c -2 +2
    M ./rdft/rank-geq2.c -2 +2
    M ./rdft/rank0-rdft2.c -2 +2
    M ./rdft/rank0.c -2 +2
    M ./rdft/rdft-dht.c -2 +2
    M ./rdft/rdft.h -2 +2
    M ./rdft/rdft2-inplace-strides.c -2 +2
    M ./rdft/rdft2-rdft.c -2 +2
    M ./rdft/rdft2-strides.c -2 +2
    M ./rdft/rdft2-tensor-max-index.c -2 +2
    M ./rdft/scalar/hb.h -2 +2
    M ./rdft/scalar/hc2c.c -2 +2
    M ./rdft/scalar/hc2cb.h -2 +2
    M ./rdft/scalar/hc2cf.h -2 +2
    M ./rdft/scalar/hf.h -2 +2
    M ./rdft/scalar/hfb.c -2 +2
    M ./rdft/scalar/r2c.c -2 +2
    M ./rdft/scalar/r2cb.h -2 +2
    M ./rdft/scalar/r2cbIII.h -2 +2
    M ./rdft/scalar/r2cf.h -2 +2
    M ./rdft/scalar/r2cfII.h -2 +2
    M ./rdft/scalar/r2r.c -2 +2
    M ./rdft/scalar/r2r.h -2 +2
    M ./rdft/simd/common/genus.c -2 +2
    M ./rdft/simd/hc2cbv.h -2 +2
    M ./rdft/simd/hc2cfv.h -2 +2
    M ./rdft/solve.c -2 +2
    M ./rdft/solve2.c -2 +2
    M ./rdft/vrank-geq1-rdft2.c -2 +2
    M ./rdft/vrank-geq1.c -2 +2
    M ./rdft/vrank3-transpose.c -2 +2
    M ./reodft/conf.c -2 +2
    M ./reodft/redft00e-r2hc-pad.c -2 +2
    M ./reodft/redft00e-r2hc.c -2 +2
    M ./reodft/reodft.h -2 +2
    M ./reodft/reodft010e-r2hc.c -2 +2
    M ./reodft/reodft11e-r2hc-odd.c -2 +2
    M ./reodft/reodft11e-r2hc.c -2 +2
    M ./reodft/reodft11e-radix2.c -2 +2
    M ./reodft/rodft00e-r2hc-pad.c -2 +2
    M ./reodft/rodft00e-r2hc.c -2 +2
    M ./simd/altivec.c -2 +2
    M ./simd/nonportable/sse.c -2 +2
    M ./simd/nonportable/sse2.c -2 +2
    M ./simd/simd-altivec.h -2 +2
    M ./simd/simd-sse.h -2 +2
    M ./simd/simd-sse2.h -2 +2
    M ./simd/simd.h -2 +2
    M ./simd/sse.c -2 +2
    M ./simd/sse2.c -2 +2
    M ./simd/taint.c -2 +2
    M ./simd/x86-cpuid.h -2 +2
    M ./simd-support/altivec.c -2 +2
    M ./simd-support/amd64-cpuid.h -2 +2
    M ./simd-support/avx.c -2 +2
    M ./simd-support/simd-altivec.h -2 +2
    M ./simd-support/simd-avx.h -2 +2
    M ./simd-support/simd-common.h -2 +2
    M ./simd-support/simd-sse2.h -2 +2
    M ./simd-support/sse2-nonportable.c -2 +2
    M ./simd-support/sse2.c -2 +2
    M ./simd-support/taint.c -2 +2
    M ./simd-support/x86-cpuid.h -2 +2
    M ./threads/api.c -2 +2
    M ./threads/conf.c -2 +2
    M ./threads/ct.c -2 +2
    M ./threads/dft-vrank-geq1.c -2 +2
    M ./threads/f77api.c -2 +2
    M ./threads/f77funcs.h -2 +2
    M ./threads/hc2hc.c -2 +2
    M ./threads/openmp.c -2 +2
    M ./threads/rdft-vrank-geq1.c -2 +2
    M ./threads/threads.c -2 +2
    M ./threads/threads.h -2 +2
    M ./threads/vrank-geq1-rdft2.c -2 +2
    M ./tools/fftw-wisdom-to-conf.1 -4 +4
    M ./tools/fftw-wisdom-to-conf.in -2 +2
    M ./tools/fftw-wisdom.c -2 +2
    M ./tools/fftw_wisdom.1.in -4 +4

Sat Jun 25 21:33:13 EDT 2011  stevenj@fftw.org
  * updated NEWS

    M ./NEWS -4 +53

Sat Jun 25 20:29:55 EDT 2011  stevenj@fftw.org
  * fixes to Fortran interface and docs

    M ./api/genf03.pl -12 +22
    M ./doc/modern-fortran.texi -5 +14
    M ./doc/mpi.texi -13 +16

Sat Jun 25 17:43:31 EDT 2011  stevenj@fftw.org
  * initial stab at MPI Fortran docs

    M ./doc/mpi.texi +138

Sat Jun 25 16:43:31 EDT 2011  stevenj@fftw.org
  * correct description of what MPI standard says about I/O (I can't believe this crap)

    M ./doc/mpi.texi -11 +21

Sat Jun 25 15:14:07 EDT 2011  stevenj@fftw.org
  * more MPI documentation; mention `fftw_alloc' functions earlier in the manual

    M ./doc/fftw3.texi +12
    M ./doc/mpi.texi -30 +158
    M ./doc/other.texi +9
    M ./doc/tutorial.texi -11 +14

Sat Jun 25 13:40:19 EDT 2011  stevenj@fftw.org
  * clarification about --enable-sse2

    M ./doc/install.texi -6 +8

Sat Jun 25 13:31:25 EDT 2011  athena@fftw.org
  * Update mingw build scripts for fftw-3.3

    M ./kernel/kalloc.c -2 +2
    M ./support/BUILD-MINGW32.sh -3 +3
    M ./support/BUILD-MINGW64.sh -4 +4

Sat Jun 25 08:52:13 EDT 2011  athena@fftw.org
  * Fix typo: EXTRADIST => EXTRA_DIST

    M ./mpi/Makefile.am -1 +1

Fri Jun 24 23:52:19 EDT 2011  stevenj@fftw.org
  * finished draft "modern fortran" chapter

    M ./doc/modern-fortran.texi -9 +91

Fri Jun 24 20:47:49 EDT 2011  stevenj@fftw.org
  * include FFTW_EXTERN prototypes for wrappers, so that they are properly exported to DLLs on Windows (sigh)

    M ./mpi/f03-wrap.sh +4

Fri Jun 24 16:52:30 EDT 2011  athena@fftw.org
  * use malloc() instead of alloca() for large buffers
  
  The proximate cause for this patch is that OpenBSD/i386 reserves 256KB
  stack size per thread.  We were allocating a buffer of size
  128*130*sizeof(fftw_complex) that exceeds the stack.
  
  While 128*130*sizeof(fftw_complex) = 260KiB is the worst case for
  normal configurations, it is a good idea to limit stack allocation
  just in case.  Also, the generic solver might in principle generate
  unbounded buffers, even though it is normally disabled for n > 137.
  
  So, as an added precaution, we now never stack-allocate buffers larger
  than 64KiB, which ought to be enough for anybody.
  

    M ./dft/dftw-direct.c -2 +3
    M ./dft/direct.c -2 +3
    M ./dft/generic.c -2 +3
    M ./kernel/ifftw.h -7 +31
    M ./rdft/ct-hc2c-direct.c -2 +3
    M ./rdft/direct-r2c.c -2 +3
    M ./rdft/generic.c -4 +6
    M ./rdft/hc2hc-direct.c -2 +3

Fri Jun 24 16:32:30 EDT 2011  stevenj@fftw.org
  * don't imply that AVX is available on Pentium III; note that MIPS Paired Single is currently only in FFTW 3.2.x

    M ./doc/install.texi -3 +2
    M ./doc/other.texi -3 +4

Fri Jun 24 16:05:27 EDT 2011  stevenj@fftw.org
  * silence annoying gfortran warnings

    M ./api/genf03.pl -1 +6
    M ./configure.ac +14
    M ./mpi/Makefile.am -3 +6

Fri Jun 24 14:59:30 EDT 2011  stevenj@fftw.org
  * a couple MPI Fortran 2003 fixes; changed MPI flags to not use 1<<31 since Fortran (not having unsigned integers) does not allow us to declare that constant in a portable way

    M ./mpi/api.c -1 +1
    M ./mpi/f03api.sh -2 +2
    M ./mpi/fftw3-mpi.h -4 +4

Fri Jun 24 15:05:05 EDT 2011  athena@fftw.org
  * Fix libtool shared version info.
  
  FFTW-3.3.x should be a direct drop-in replacement for all FFTW-3.x.y
  versions.
  

    M ./configure.ac -1 +4

Fri Jun 24 14:38:47 EDT 2011  stevenj@fftw.org
  * add MPI Fortran API and wrappers

    M ./api/Makefile.am -2 +3
    M ./api/f03api.sh -3 +5
    M ./api/genf03.pl -2 +14
    M ./mpi/Makefile.am -3 +14
    R ./mpi/README
    A ./mpi/f03-wrap.sh
    A ./mpi/f03api.sh
    M ./mpi/fftw3-mpi.h -4 +4
    A ./mpi/genf03-wrap.pl

Fri Jun 24 14:51:12 EDT 2011  athena@fftw.org
  * Do not require fig2dev on the user's machine
  
  Distribute the manual's figures in PDF/PS/PNG form instead.

    M ./doc/Makefile.am -10 +16

Fri Jun 24 11:52:44 EDT 2011  athena@fftw.org
  * Remove --enable-portable-binary, --with-gcc-arch from documentation.

    M ./NEWS +3
    M ./doc/install.texi -23 +2

Fri Jun 24 11:48:48 EDT 2011  athena@fftw.org
  * Forget about specifying nonportable CFLAGS.  Let the user do it if he wants.

    M ./m4/Makefile.am -4 +3
    M ./m4/ax_cc_maxopt.m4 -24 +10
    R ./m4/ax_gcc_archflag.m4
    R ./m4/ax_gcc_x86_cpuid.m4

Fri Jun 24 11:48:25 EDT 2011  athena@fftw.org
  * Add "-avx" to version string when appropriate.

    M ./api/version.c +4

Fri Jun 24 10:26:38 EDT 2011  athena@fftw.org
  * change 3.3-alpha => 3.3-beta1

    M ./configure.ac -1 +1

Fri Jun 24 09:25:49 EDT 2011  athena@fftw.org
  * Extend OUR_MALLOC16 to larger alignments
  
  Make it work for 32-byte alignment and beyond, as needed by AVX.
  Rename --with-our-malloc16 to --with-our-malloc.  Keep old --with-our-malloc16
  flag for compatibility.
  

    M ./configure.ac -3 +4
    M ./kernel/kalloc.c -9 +10

Fri Jun 24 09:19:38 EDT 2011  athena@fftw.org
  * Fix typo

    M ./doc/install.texi -1 +1

Fri Jun 24 09:10:26 EDT 2011  athena@fftw.org
  * One pass over the manual.

    M ./NEWS -1 +6
    M ./doc/Makefile.am -5 +2
    M ./doc/install.texi -16 +10
    M ./doc/intro.texi -1 +1
    M ./doc/other.texi -2 +2
    M ./doc/reference.texi -5 +5
    M ./doc/tutorial.texi -32 +32

Fri Jun 24 08:19:03 EDT 2011  athena@fftw.org
  * eliminate the WITH_ALIGNED_STACK hack
  
  This is 2011 and I have no system with incorrect stack alignment.
  

    M ./TODO -2
    M ./api/apiplan.c -9 +4
    M ./api/execute-dft-c2r.c -2 +2
    M ./api/execute-dft-r2c.c -2 +2
    M ./api/execute-dft.c -2 +2
    M ./api/execute-r2r.c -2 +2
    M ./api/execute-split-dft-c2r.c -2 +2
    M ./api/execute-split-dft-r2c.c -2 +2
    M ./api/execute-split-dft.c -2 +2
    M ./api/execute.c -2 +2
    M ./api/f77funcs.h -16 +16
    M ./doc/Makefile.am -2 +2
    M ./doc/fftw3.texi -1 +1
    M ./doc/intro.texi -3 +4
    M ./doc/legacy-fortran.texi -1 +1
    M ./doc/other.texi -51 +4
    M ./doc/reference.texi -1 +1
    M ./doc/tutorial.texi -9 +5
    M ./kernel/ifftw.h -51
    M ./threads/ct.c -2 +2
    M ./threads/dft-vrank-geq1.c -2 +2
    M ./threads/hc2hc.c -2 +2
    M ./threads/rdft-vrank-geq1.c -2 +2
    M ./threads/vrank-geq1-rdft2.c -2 +2

Fri Jun 24 07:49:47 EDT 2011  athena@fftw.org
  * enable both threaded and unthreaded wisdom in tools/fftw-wisdom

    M ./TODO -7 +2
    M ./tools/fftw-wisdom.c -18 +14

Fri Jun 24 02:40:04 EDT 2011  stevenj@fftw.org
  * clarification

    M ./doc/modern-fortran.texi -2 +2

Fri Jun 24 02:24:01 EDT 2011  stevenj@fftw.org
  * check for error code in example

    M ./doc/modern-fortran.texi +2

Fri Jun 24 02:22:18 EDT 2011  stevenj@fftw.org
  * cleanup - since NATIVE_MALLOC is always malloc, delete this #define

    M ./api/export-wisdom-to-string.c -1 +1
    M ./kernel/ifftw.h -2

Fri Jun 24 02:19:44 EDT 2011  stevenj@fftw.org
  * document wisdom string import/export in Fortran

    M ./doc/modern-fortran.texi +49

Fri Jun 24 02:11:40 EDT 2011  stevenj@fftw.org
  * bug fix - NATIVE_MALLOC should always be plain malloc, even in debug_malloc mode, because it is used in the API to return things that should be deallocated with free(); correspondingly, be sure to use free() ansd not X(free) with this

    M ./kernel/ifftw.h -1 +1
    M ./mpi/wisdom-api.c -1 +1

Fri Jun 24 01:35:27 EDT 2011  stevenj@fftw.org
  * declaration style

    M ./doc/modern-fortran.texi -2 +4

Fri Jun 24 01:25:36 EDT 2011  stevenj@fftw.org
  * document wisdom file export/import from Fortran; add export/import_to/from_filename functions for convenience 

    M ./api/export-wisdom-to-file.c +11
    M ./api/fftw3.h +2
    M ./api/import-wisdom-from-file.c +10
    M ./doc/modern-fortran.texi -9 +79
    M ./doc/other.texi -5 +6
    M ./doc/reference.texi -11 +22

Thu Jun 23 19:19:43 EDT 2011  stevenj@fftw.org
  * more fortran docs

    M ./doc/legacy-fortran.texi -5 +6
    M ./doc/modern-fortran.texi -10 +167
    M ./doc/reference.texi +2

Thu Jun 23 17:50:30 EDT 2011  stevenj@fftw.org
  * enforce 132-character line-length limit that is the default in Fortran

    M ./api/genf03.pl -3 +26

Wed Jun 22 23:27:31 EDT 2011  stevenj@fftw.org
  * the F03 standard is ambiguous about whether types can be assigned to wider types as formal parameters with VALUE attributes, and e.g. gfortran interprets it to disallow this code

    M ./doc/modern-fortran.texi -2 +4

Thu Jun 23 18:12:10 EDT 2011  athena@fftw.org
  * Add md5 hash of fftw's configuration to wisdom file
  
  People were already confused by threaded vs unthreaded wisdom, and now
  things will be even worse because we enable/disable AVX codelets at
  runtime.  Accept incoming wisdom only if it was produced by the same
  configuration (modulo MD5).

    M ./kernel/planner.c -2 +33

Thu Jun 23 09:01:27 EDT 2011  athena@fftw.org
  * distribute fftw3.f03.in

    M ./api/Makefile.am -1 +1

Wed Jun 22 22:02:18 EDT 2011  stevenj@fftw.org
  * more Fortran documentation

    M ./doc/modern-fortran.texi -1 +95

Wed Jun 22 20:10:39 EDT 2011  stevenj@fftw.org
  * correct comment

    M ./api/genf03.pl -1 +2

Wed Jun 22 20:26:18 EDT 2011  athena@fftw.org
  * Use "sh FOO.sh" instead of "./FOO.sh" to avoid chmod +x.

    M ./api/Makefile.am -1 +1

Wed Jun 22 19:19:05 EDT 2011  stevenj@fftw.org
  * document fftw_alloc_real/complex ... should we switch to using these in the tutorial examples?

    M ./doc/reference.texi +19

Wed Jun 22 19:07:49 EDT 2011  stevenj@fftw.org
  * whoops, added missing file

    A ./api/genf03.pl

Wed Jun 22 18:46:01 EDT 2011  athena@fftw.org
  * Note addition of AVX.

    M ./NEWS +2

Wed Jun 22 18:43:43 EDT 2011  athena@fftw.org
  * In SSE2, AVX: use FMA macros when applicable.
  
  Makes it easier to play with fma4 and fma3 when it comes out.

    M ./simd-support/simd-avx.h -21 +21
    M ./simd-support/simd-sse2.h -19 +19

Wed Jun 22 18:16:45 EDT 2011  stevenj@fftw.org
  * all modern Fortran compilers can call FFTW's C interfface directly -- support this, and in particular generate a Fortran 2003 interface file from fftw3.h so that Fortran code calling FFTW can be typechecked ((addressing the source of a lot of Fortran-user problems)

     ./doc/fortran.texi -> ./doc/legacy-fortran.texi
    M ./api/Makefile.am -3 +12
    A ./api/f03api.sh
    M ./api/fftw3.h -3 +11
    M ./api/malloc.c +18
    M ./configure.ac +12
    M ./doc/Makefile.am -1 +1
    M ./doc/fftw3.texi -2 +4
    M ./doc/install.texi -2 +2
    M ./doc/intro.texi -5 +6
    M ./doc/legacy-fortran.texi -23 +38
    A ./doc/modern-fortran.texi
    M ./doc/mpi.texi -1 +1
    M ./doc/other.texi -2 +4
    M ./doc/upgrading.texi -1 +1

Wed Jun 22 13:10:02 EDT 2011  athena@fftw.org
  * some cleanup of SSE2 macros

    M ./simd-support/simd-sse2.h -27 +16

Wed Jun 22 07:38:18 EDT 2011  athena@fftw.org
  * don't use -xHost on ICC
  
  -xHost with ICC is problematic.  On icc-12.0.0, "-mavx -xHost"
  overrides -mavx with -xHost, generating SSE2 code instead of AVX code.
  ICC does not seem to support -mtune=host or equivalent non-ABI
  changing flag.
  

    M ./m4/ax_cc_maxopt.m4 -31 +5

Tue Jun 21 20:35:36 EDT 2011  athena@fftw.org
  * Complete AVX implementation for split codelets

    M ./configure.ac -36 +38
    M ./simd-support/simd-avx.h -41 +26

Tue Jun 21 19:37:14 EDT 2011  stevenj@fftw.org
  * whoops, missing altivec conf patches

    M ./dft/codelet-dft.h +1
    M ./dft/conf.c +4
    M ./rdft/codelet-rdft.h +1
    M ./rdft/conf.c +4

Tue Jun 21 19:12:45 EDT 2011  stevenj@fftw.org
  * some BSD ar versions (e.g. on MacOS X) give an error if there are no object files, so we cannot build empty libraries

    M ./dft/simd/altivec/Makefile.am -4 +3
    M ./dft/simd/avx/Makefile.am -3 +4
    M ./dft/simd/sse2/Makefile.am -4 +3
    M ./rdft/simd/altivec/Makefile.am -4 +3
    M ./rdft/simd/avx/Makefile.am -3 +4
    M ./rdft/simd/sse2/Makefile.am -3 +4

Tue Jun 21 19:12:12 EDT 2011  stevenj@fftw.org
  * re-insertion of Altivec code

    M ./Makefile.am -1 +6
    M ./configure.ac -8 +13
    M ./dft/simd/Makefile.am -1 +1
    A ./dft/simd/altivec/
    A ./dft/simd/altivec/Makefile.am
    M ./kernel/ifftw.h -2 +9
    M ./rdft/simd/Makefile.am -1 +1
    A ./rdft/simd/altivec/
    A ./rdft/simd/altivec/Makefile.am
    M ./simd-support/Makefile.am -1 +1
    A ./simd-support/altivec.c
    A ./simd-support/simd-altivec.h
    M ./simd-support/simd-common.h +3

Tue Jun 21 16:26:09 EDT 2011  athena@fftw.org
  * Implement faster AVX loads/stores.

    M ./simd-support/simd-avx.h -21 +28

Tue Jun 21 16:03:24 EDT 2011  athena@fftw.org
  * Initial AVX256/single implementation
  
  This should be correct but slow.  I need to figure out how to implement
  noncontiguous loads/stores efficiently.

    M ./simd-support/simd-avx.h -53 +133

Tue Jun 21 14:13:57 EDT 2011  athena@fftw.org
  * fix AVX alignment

    M ./simd-support/simd-common.h -6 +8

Tue Jun 21 14:07:28 EDT 2011  athena@fftw.org
  * rename avx256d -> avx
  
  AVX will work in both double and single precision, like SSE2.

     ./dft/simd/avx256d -> ./dft/simd/avx
     ./rdft/simd/avx256d -> ./rdft/simd/avx
     ./simd-support/avx256d.c -> ./simd-support/avx.c
     ./simd-support/simd-avx256d.h -> ./simd-support/simd-avx.h
    M ./Makefile.am -4 +4
    M ./configure.ac -11 +6
    M ./dft/codelet-dft.h -1 +1
    M ./dft/conf.c -3 +3
    M ./dft/simd/Makefile.am -1 +1
    M ./dft/simd/avx/Makefile.am -4 +4
    M ./kernel/ifftw.h -2 +2
    M ./rdft/codelet-rdft.h -1 +1
    M ./rdft/conf.c -3 +3
    M ./rdft/simd/Makefile.am -1 +1
    M ./rdft/simd/avx/Makefile.am -4 +4
    M ./simd-support/Makefile.am -1 +1
    M ./simd-support/avx.c -3 +3
    M ./simd-support/simd-avx.h -2 +2

Tue Jun 21 13:52:20 EDT 2011  athena@fftw.org
  * remove CODELET_OPTIM
  
  In the old 32-bit gcc-3.x days we used to play games with gcc to force
  it to produce decent code.  Now gcc has gotten smarter and it produces
  indecent code no matter what we do, so it is safe to remove these hacks.
  

    M ./api/version.c -5 +5
    M ./configure.ac -40
    M ./support/Makefile.codelets -6

Tue Jun 21 09:57:31 EDT 2011  athena@fftw.org
  * work around gcc/icc quirks

    M ./simd-support/simd-avx256d.h -14 +45

Tue Jun 21 09:56:07 EDT 2011  athena@fftw.org
  * Add remarks in places where we work around gcc quirks

    M ./simd-support/simd-sse2.h -7 +16

Mon Jun 20 21:17:59 EDT 2011  stevenj@fftw.org
  * remove the libbench directory (which we have kept lingering in the repository for years due to CVS's inability to remove directories)

    R ./libbench/Makefile.am
    R ./libbench/accopy-from.c
    R ./libbench/accopy-to.c
    R ./libbench/acopy.c
    R ./libbench/allocate.c
    R ./libbench/ascale.c
    R ./libbench/aset.c
    R ./libbench/bench-main.c
    R ./libbench/bench-user.h
    R ./libbench/bench.h
    R ./libbench/caadd.c
    R ./libbench/cacopy.c
    R ./libbench/can-do.c
    R ./libbench/cascale.c
    R ./libbench/caset.c
    R ./libbench/casub.c
    R ./libbench/ccopy-from.c
    R ./libbench/ccopy-to.c
    R ./libbench/copy-c2c-from.c
    R ./libbench/copy-c2c-to.c
    R ./libbench/copy-c2h-1d-fftpack.c
    R ./libbench/copy-c2h-1d-halfcomplex.c
    R ./libbench/copy-c2h-1d-packed.c
    R ./libbench/copy-c2h-1d-unpacked-ri.c
    R ./libbench/copy-c2h-unpacked.c
    R ./libbench/copy-c2h.c
    R ./libbench/copy-c2r-packed.c
    R ./libbench/copy-c2r-unpacked.c
    R ./libbench/copy-c2r.c
    R ./libbench/copy-c2ri.c
    R ./libbench/copy-h2c-1d-fftpack.c
    R ./libbench/copy-h2c-1d-halfcomplex.c
    R ./libbench/copy-h2c-1d-packed.c
    R ./libbench/copy-h2c-1d-unpacked-ri.c
    R ./libbench/copy-h2c-unpacked.c
    R ./libbench/copy-h2c.c
    R ./libbench/copy-r2c-packed.c
    R ./libbench/copy-r2c-unpacked.c
    R ./libbench/copy-r2c.c
    R ./libbench/copy-ri2c.c
    R ./libbench/deallocate.c
    R ./libbench/getopt-utils.c
    R ./libbench/getopt.c
    R ./libbench/getopt.h
    R ./libbench/getopt1.c
    R ./libbench/info.c
    R ./libbench/log2.c
    R ./libbench/main.c
    R ./libbench/mflops.c
    R ./libbench/mp.c
    R ./libbench/ovtpvt.c
    R ./libbench/pow2.c
    R ./libbench/prime.c
    R ./libbench/problem.c
    R ./libbench/report.c
    R ./libbench/speed.c
    R ./libbench/timer.c
    R ./libbench/unnormalize.c
    R ./libbench/util.c
    R ./libbench/verify.c
    R ./libbench/zero.c
    R ./libbench/

Mon Jun 20 21:17:14 EDT 2011  stevenj@fftw.org
  * update URLs

    M ./doc/install.texi -2 +2
    M ./doc/intro.texi -1 +1
    M ./doc/license.texi -1 +1
    M ./doc/mpi.texi -1 +1
    M ./doc/other.texi -1 +1

Mon Jun 20 20:53:31 EDT 2011  stevenj@fftw.org
  * whoops, forgot to check in alignment change

    M ./simd-support/simd-common.h -3 +6

Mon Jun 20 20:22:23 EDT 2011  athena@fftw.org
  * "test X = Y" requires spaces around "="

    M ./configure.ac -1 +1

Mon Jun 20 19:18:52 EDT 2011  stevenj@fftw.org
  * indenting

    M ./simd-support/simd-sse2.h -3 +3

Mon Jun 20 18:57:10 EDT 2011  stevenj@fftw.org
  * merge back in SSE support, now combined with SSE2; --enable-sse2 now works in both single and double precision, and simd-sse2.h contains both the double- and single-precision code (which overlap a lot); in single precision it is still compiled for SSE-only (SSE2 is only required for double)

    M ./configure.ac -16 +9
    M ./simd-support/simd-sse2.h -54 +181
    M ./simd-support/sse2-nonportable.c -2 +6
    M ./simd-support/sse2.c -4 +11

Mon Jun 20 16:02:07 EDT 2011  athena@fftw.org
  * Implement AVX autodetection (gcc-only so far)

    M ./simd-support/Makefile.am -1 +1
    A ./simd-support/amd64-cpuid.h
    M ./simd-support/avx256d.c -2 +32
    M ./simd-support/x86-cpuid.h +33

Mon Jun 20 14:25:54 EDT 2011  athena@fftw.org
  * Add VZEROUPPER at the end of AVX codelets
  
  If the Intel Optimization Manual is to be believed, we need to wave a
  dead chicken before transitioning from AVX code to SSE code.  I am
  supposed to believe that there is a transition penalty for doing so,
  unless one uses a magic VZEROUPPER instruction that apparently has
  zero cost.  Whatever.
  

    M ./genfft/c.ml +2
    M ./genfft/c.mli +1
    M ./genfft/gen_hc2c.ml -1 +1
    M ./genfft/gen_hc2cdft.ml -1 +1
    M ./genfft/gen_hc2cdft_c.ml -1 +1
    M ./genfft/gen_hc2hc.ml -1 +1
    M ./genfft/gen_mdct.ml -1 +1
    M ./genfft/gen_notw.ml -1 +1
    M ./genfft/gen_notw_c.ml -1 +1
    M ./genfft/gen_r2cb.ml -1 +1
    M ./genfft/gen_r2cf.ml -1 +1
    M ./genfft/gen_r2r.ml -1 +1
    M ./genfft/gen_twiddle.ml -1 +1
    M ./genfft/gen_twiddle_c.ml -1 +1
    M ./genfft/gen_twidsq.ml -1 +1
    M ./genfft/gen_twidsq_c.ml -1 +1
    M ./genfft/genutil.ml -2 +2
    M ./genfft/simd.ml +1
    M ./simd-support/simd-avx256d.h +5
    M ./simd-support/simd-sse2.h +2

Mon Jun 20 10:21:25 EDT 2011  athena@fftw.org
  * Move RDFT to new simd scheme

    A ./rdft/simd/avx256d/
    R ./rdft/simd/codelets/Makefile.am
    R ./rdft/simd/codelets/
    A ./rdft/simd/common/
    A ./rdft/simd/sse2/
    M ./Makefile.am -2 +4
    M ./configure.ac +4
    M ./dft/simd/Makefile.am -1 +1
    M ./genfft/gen_hc2cdft_c.ml -2 +2
    M ./kernel/ifftw.h +1
    M ./rdft/Makefile.am -1 +1
    M ./rdft/codelet-rdft.h -1 +2
    M ./rdft/conf.c -2 +7
    M ./rdft/simd/Makefile.am -6 +2
    A ./rdft/simd/avx256d/Makefile.am
    A ./rdft/simd/codlist.mk
    A ./rdft/simd/common/Makefile.am
    A ./rdft/simd/common/genus.c
    R ./rdft/simd/hc2cbv.c
    M ./rdft/simd/hc2cbv.h -2 +2
    R ./rdft/simd/hc2cfv.c
    M ./rdft/simd/hc2cfv.h -2 +2
    A ./rdft/simd/simd.mk
    A ./rdft/simd/sse2/Makefile.am

Mon Jun 20 09:23:38 EDT 2011  athena@fftw.org
  * New SIMD build system
  
  We now support multiple SIMD extensions in the same binary, e.g.
  --enable-sse2 --enable-avx.  This patch adds the necessary
  infrastructure for SSE2/AVX and complex DFT.  Later patches will add
  RDFT and SSE/ALTIVEC/etc.
  

    A ./dft/simd/avx256d/
    A ./simd-support/
    R ./dft/simd/codelets/Makefile.am
    R ./dft/simd/codelets/
    A ./dft/simd/common/
    A ./dft/simd/sse2/
    M ./Makefile.am -11 +13
    M ./configure.ac -47 +74
    M ./dft/codelet-dft.h -4 +2
    M ./dft/conf.c -2 +7
    M ./dft/scalar/codelets/Makefile.am -1 +1
    M ./dft/simd/Makefile.am -6 +3
    A ./dft/simd/avx256d/Makefile.am
    A ./dft/simd/codlist.mk
    A ./dft/simd/common/Makefile.am
    A ./dft/simd/common/genus.c
    R ./dft/simd/n1b.c
    M ./dft/simd/n1b.h -2 +2
    R ./dft/simd/n1f.c
    M ./dft/simd/n1f.h -2 +2
    R ./dft/simd/n2b.c
    M ./dft/simd/n2b.h -2 +2
    R ./dft/simd/n2f.c
    M ./dft/simd/n2f.h -2 +2
    R ./dft/simd/n2s.c
    M ./dft/simd/n2s.h -2 +2
    R ./dft/simd/q1b.c
    M ./dft/simd/q1b.h -2 +2
    R ./dft/simd/q1f.c
    M ./dft/simd/q1f.h -2 +2
    A ./dft/simd/simd.mk
    A ./dft/simd/sse2/Makefile.am
    R ./dft/simd/t.c
    M ./dft/simd/t1b.h -2 +2
    M ./dft/simd/t1bu.h -2 +2
    M ./dft/simd/t1f.h -2 +2
    M ./dft/simd/t1fu.h -2 +2
    M ./dft/simd/t2b.h -2 +2
    M ./dft/simd/t2f.h -2 +2
    M ./dft/simd/t3b.h -2 +2
    M ./dft/simd/t3f.h -2 +2
    R ./dft/simd/ts.c
    M ./dft/simd/ts.h -2 +2
    M ./genfft/gen_twiddle.ml -2 +2
    M ./genfft/genutil.ml -2 +5
    M ./kernel/ifftw.h -2 +6
    M ./rdft/Makefile.am -1 +1
    M ./rdft/conf.c -1 +1
    M ./rdft/scalar/r2cb/Makefile.am +1
    M ./rdft/scalar/r2cf/Makefile.am +1
    M ./rdft/scalar/r2r/Makefile.am +1
    A ./simd-support/Makefile.am
    A ./simd-support/avx256d.c
    A ./simd-support/simd-avx256d.h
    A ./simd-support/simd-common.h
    A ./simd-support/simd-sse2.h
    A ./simd-support/sse2-nonportable.c
    A ./simd-support/sse2.c
    A ./simd-support/taint.c
    A ./simd-support/x86-cpuid.h
    M ./support/Makefile.codelets -21 +22

Sun Jun 19 12:29:27 EDT 2011  stevenj@fftw.org
  * italicize Latin quote

    M ./doc/reference.texi -2 +2

Sun Jun 19 12:26:34 EDT 2011  stevenj@fftw.org
  * work around incredibly annoying makeinfo bug -- for HTML output, in any paragraph ending with an @index command, two blank lines are needed to create a paragraph break  ... our HTML output has apparently been screwed up for years

    M ./doc/acknowledgements.texi +2
    M ./doc/fortran.texi +1
    M ./doc/install.texi +2
    M ./doc/intro.texi +1
    M ./doc/mpi.texi +20
    M ./doc/other.texi +2
    M ./doc/reference.texi +21
    M ./doc/threads.texi +2
    M ./doc/tutorial.texi +13

Sun Jun 19 12:01:39 EDT 2011  stevenj@fftw.org
  * conjugate-pair algorithm turns out not to be due to djb, but it was pointed out to us by djb

    M ./doc/intro.texi -3 +4

Sun Jun 19 11:59:21 EDT 2011  stevenj@fftw.org
  * new-array execute functions are *not* the same for MPI, since the problems are different

    M ./doc/mpi.texi -2 +3

Sun Jun 19 11:57:52 EDT 2011  stevenj@fftw.org
  * tweaks to MPI manual

    M ./doc/mpi.texi -19 +44
    M ./doc/reference.texi -1 +1

Sun Jun 19 00:47:23 EDT 2011  stevenj@fftw.org
  * only call MPI_Init_thread for MPI version >= 2

    M ./mpi/mpi-bench.c +5

Sun Jun 19 00:35:44 EDT 2011  stevenj@fftw.org
  * document quad precision in Fortran

    M ./doc/fortran.texi -9 +12

Sun Jun 19 00:16:10 EDT 2011  stevenj@fftw.org
  * use -lquadmath for quad-precision library in pkgconfig file

    M ./configure.ac +2
    M ./fftw.pc.in -1 +1

Sun Jun 19 00:10:33 EDT 2011  stevenj@fftw.org
  * document quad precision

    M ./doc/install.texi -1 +11
    M ./doc/reference.texi +8

Sat Jun 18 21:26:24 EDT 2011  stevenj@fftw.org
  * there is currently no standard quad-precision type in MPI, so don't allow MPI support to be compiled with --enable-quad-precision

    M ./configure.ac +3
    M ./mpi/ifftw-mpi.h +2
    M ./mpi/mpi-bench.c +2

Sa
Results 1 - 1 of 1
Help - FTP Sites List - Software Dir.
Search over 15 billion files
© 1997-2017 FileWatcher.com