cannam@127: commit b405994456f9a87f2170ba19536d4c4d8278682f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 30 16:33:22 2016 -0400 cannam@127: cannam@127: update AUTHORS cannam@127: cannam@127: commit 4d0c1894fb37c61b0f0a42b50afd435d226f6b9e cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 30 15:18:06 2016 -0400 cannam@127: cannam@127: Fixes for Windows cross-compilation cannam@127: cannam@127: These days mingw by default produces binaries that depend on cannam@127: libgcc-sjlj-1.dll, which defeats the whole historical point of mingw cannam@127: (produce vanilla win32 binaries with no GNU stuff). cannam@127: cannam@127: Add a hack to link with -static-libgcc, which avoids the problem. cannam@127: cannam@127: commit a17d44eeb3100780ba106a22f497d47a43be7642 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 30 11:39:09 2016 -0400 cannam@127: cannam@127: Misc fixes. cannam@127: cannam@127: * sed s/avx[_- ]128[-_ ]fma/avx-128-fma cannam@127: * avoid some signed/unsigned casts cannam@127: cannam@127: commit f3688be112ed0099b4c57970db74c08373f3604d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 30 10:52:53 2016 -0400 cannam@127: cannam@127: Fix SIMD autodetection cannam@127: cannam@127: * AVX was not testing for OSXSAVE support cannam@127: cannam@127: * AVX2 was broken (issuing XGETBV without checking for its presence---failing cannam@127: on atom) cannam@127: cannam@127: * AVX512 was broken in the same way as AVX2, I have guessed a fix but cannam@127: I have no way to test it. cannam@127: cannam@127: commit 7fce2ae37f8338bd7e021b1a406c75b213c31c77 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 29 07:48:10 2016 -0400 cannam@127: cannam@127: document fftw_make_planner_thread_safe() cannam@127: cannam@127: commit 6167b92e3362f2d116274daa561c0d788fb670d4 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 29 07:28:03 2016 -0400 cannam@127: cannam@127: rm README-bench cannam@127: cannam@127: It appears in tests/README cannam@127: cannam@127: commit cc9640cbbaa70e6645a0ea46be0508268905c2ba cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 29 07:27:25 2016 -0400 cannam@127: cannam@127: Add README-bench cannam@127: cannam@127: commit d82fe4f3e06bdbf92b09324e36f4d477bc5fe376 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 29 07:25:00 2016 -0400 cannam@127: cannam@127: Do not enable avx128-fma unless the user asks for it. cannam@127: cannam@127: Adding SIMD instruction sets automatically is user-hostile behavior. cannam@127: cannam@127: Also, update the manual to reflect the new SIMD support cannam@127: cannam@127: commit dc32329871d304de8d95ad290973844dfbc6101f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 29 07:00:55 2016 -0400 cannam@127: cannam@127: Update NEWS for 3.3.5 cannam@127: cannam@127: commit 2ed010c62b1bc8ca6b23bfda2e09b8c28e1e8bcc cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 5 07:07:15 2016 -0400 cannam@127: cannam@127: Clean up some int<->size_t confusion cannam@127: cannam@127: commit ea86c49ac7470a646d1e6a4fa007ecbda6ab56c4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 4 20:33:15 2016 -0400 cannam@127: cannam@127: Unused Variable Police cannam@127: cannam@127: commit d9a3f48343bda0a88c8a87cab329d95426ddfcb9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 4 20:30:12 2016 -0400 cannam@127: cannam@127: Integral Type Police cannam@127: cannam@127: clear some int/unsigned/size_t confusions cannam@127: cannam@127: commit 29cee6cc95d434321292d013d6a7be4c55379a49 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 4 19:50:10 2016 -0400 cannam@127: cannam@127: Cast Police cannam@127: cannam@127: Eliminate some useless (but harmless) int<->size_t conversions. cannam@127: cannam@127: commit d7c566eb98523c7c0bafae734c7894a5a3595771 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 13 17:50:45 2016 -0400 cannam@127: cannam@127: Clarify ambiguous/wrong documentation of halfcomplex output format. cannam@127: cannam@127: commit 6543818e3091ea788a1aac41d06ca343e672f103 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 20 18:18:14 2016 -0500 cannam@127: cannam@127: Cleanup cannam@127: cannam@127: Rewrite Unique.make in more idiomatic caml style, cannam@127: strongly typed. cannam@127: cannam@127: commit 4965e33c6c98484b66787f1891cfe4f689becee5 cannam@127: Merge: 119aa4c f8a7359 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 20 15:14:32 2016 -0800 cannam@127: cannam@127: Merge pull request #53 from artemkin/master cannam@127: cannam@127: Fixed unique token generation in genfft cannam@127: cannam@127: commit f8a73593a499efc751103460ff2f07d8b1e2ff0c cannam@127: Author: Stanislav Artemkin cannam@127: Date: Thu Jan 21 01:17:30 2016 +0400 cannam@127: cannam@127: Fixed unique token generation in genfft cannam@127: cannam@127: Unique token generation was based on the assumption that OCaml compiler cannam@127: won't inline a given piece of code. Starting from 4.02.0 it does more cannam@127: aggressive inlining and breaks this functionality. cannam@127: cannam@127: commit 119aa4c4a893f32dfd837a84fac9453b6dae6680 cannam@127: Merge: 8c7a7af e41df2c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 30 15:38:00 2015 -0400 cannam@127: cannam@127: Merge pull request #48 from rleonid/master cannam@127: cannam@127: Replace depracted usage of Pervasives or. cannam@127: cannam@127: commit e41df2c3cac7c3e69586c07f80f1bb0a24dccd5a cannam@127: Author: Leonid Rozenberg cannam@127: Date: Wed Sep 30 15:22:16 2015 -0400 cannam@127: cannam@127: Replace depracted usage of Pervasives or. cannam@127: cannam@127: commit 8c7a7af184a63064325fa542a8d1d7f4e3b4b8aa cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 8 10:35:18 2015 -0400 cannam@127: cannam@127: Clarify how to bootstrap fftw from the git repository cannam@127: cannam@127: commit a0cbff67eae9ab66f6f2b4cf2ea79de6c95d7d61 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 8 10:28:49 2015 -0400 cannam@127: cannam@127: Clarify that ocamlbuild is necessary for --enable-maintainer-mode cannam@127: cannam@127: commit f6339eadef8a62432ea2f2017ce0b4a1954ea738 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Aug 5 11:11:26 2015 -0400 cannam@127: cannam@127: fix LaTeX typo, thanks to Gael Lorieul cannam@127: cannam@127: commit 8cd9bfa347289143a00fa0d5eea30f4766192d46 cannam@127: Author: Erik Lindahl cannam@127: Date: Wed May 27 00:15:57 2015 +0200 cannam@127: cannam@127: Update VSX SIMD to avoid inline assembly cannam@127: cannam@127: Thanks to some help from Michael Gschwind of cannam@127: IBM, this removes the remaining inline assembly cannam@127: calls and replace the with vector functions. This cannam@127: avoid interfering with the optimizer both on GCC cannam@127: and XLC, and gets us another 3-10% of performance cannam@127: when using VSX SIMD. Tested with GCC-4.9, XLC-13.1 cannam@127: in single and double on little-endian power 8. cannam@127: cannam@127: commit 579cec9a64cc177e673f006eb112d488be21b230 cannam@127: Author: Erik Lindahl cannam@127: Date: Tue May 26 19:27:58 2015 +0200 cannam@127: cannam@127: Enable SSE2 automatically with AVX,AVX2, or AVX512. cannam@127: cannam@127: 256-bit AVX can be significantly slower than cannam@127: 128-bit SIMD. Despite recommendations many cannam@127: distributions appear to only enable AVX, but not cannam@127: SSE. This fixes the problem by also enabling cannam@127: SSE when we use the wider SIMD instructions. cannam@127: cannam@127: commit dd80210ec433938876575e2435e12d7e630872e7 cannam@127: Author: Erik Lindahl cannam@127: Date: Tue May 26 19:09:40 2015 +0200 cannam@127: cannam@127: Turn AVX-128 into AMD-specific AVX-128-FMA cannam@127: cannam@127: The only platform where AVX-128 really matters cannam@127: is AMD (since the compute units can execute a cannam@127: single 256-bit or two 128-bit SIMD instructions), cannam@127: so now we only use it there which means we can cannam@127: also enable FMA instructions. cannam@127: cannam@127: commit b3105ed9529846ca8dd9267e46d7bcd2ebb12ff6 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 17:33:15 2015 -0400 cannam@127: cannam@127: Fix broken avx/32-bit compilation cannam@127: cannam@127: commit d3442a8395e1fc6e77490c3f34c868b1998e4e96 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 17:27:31 2015 -0400 cannam@127: cannam@127: rm hooks api's, add fftw_make_planner_thread_safe() api cannam@127: cannam@127: fftw_make_planner_thread_safe() installs a lock around the planner. cannam@127: It is guaranteed to be atomic and idempotent. cannam@127: cannam@127: I wrote an emulation of pthread mutex initializers on Windows, but I cannam@127: haven't even compiled the Windows code yet. cannam@127: cannam@127: commit 842596fe6bfb277effc8f8f8db7e1c4008bb59e1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 11:11:58 2015 -0400 cannam@127: cannam@127: Add TODO's cannam@127: cannam@127: commit eff7dfcd526e90539f169cfff2374ceb2fd2dd0e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 10:50:21 2015 -0400 cannam@127: cannam@127: add TODOs cannam@127: cannam@127: commit 0e53e3e9d2640a895a69c5aff4a676c156271141 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 10:48:03 2015 -0400 cannam@127: cannam@127: Add argument to planner hooks cannam@127: cannam@127: commit 94ef591d61c8e78e87c65f9779eb003fc7f1ba6b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 10:37:24 2015 -0400 cannam@127: cannam@127: Update shared-version-info cannam@127: cannam@127: We added an API (planner hooks), so the shared version info cannam@127: needs to be bumped. cannam@127: cannam@127: commit 9ef9ec85588c026e6a80b1475df24ba20d098e8e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 25 10:27:25 2015 -0400 cannam@127: cannam@127: Bump version to 3.3.5 cannam@127: cannam@127: commit 593d55932959366918e209fbbd2f4719d39d448c cannam@127: Author: Romain Dolbeau cannam@127: Date: Wed May 13 16:42:23 2015 +0200 cannam@127: cannam@127: Typo ; Fixes #41 cannam@127: cannam@127: commit cd2b27d1600d80ba719f1b70094886e39cf145a3 cannam@127: Author: Erik Lindahl cannam@127: Date: Thu May 7 17:45:43 2015 +0200 cannam@127: cannam@127: Separate routines to query 128-bit AVX support cannam@127: cannam@127: This also disables 256-bit AVX for current AMD processors cannam@127: that work better with 128-bit AVX. Note that this is not cannam@127: detected by the timing routines since the effect is only cannam@127: apparent when using multiple cores. cannam@127: cannam@127: commit a1cf4158dd829853bd9f6b8c4c4951d7495c9e64 cannam@127: Merge: b613508 0331b39 cannam@127: Author: Romain Dolbeau cannam@127: Date: Mon Apr 20 22:01:15 2015 +0200 cannam@127: cannam@127: Merge branch 'master' of github.com:FFTW/fftw3 cannam@127: cannam@127: commit 0331b39cd3641a8ac89be27dbde3e41204fd1888 cannam@127: Merge: 38b93cc d2ea399 cannam@127: Author: Erik Lindahl cannam@127: Date: Mon Apr 20 21:09:44 2015 +0200 cannam@127: cannam@127: Merge branch 'experimental-simd' cannam@127: cannam@127: Merged in new SIMD architectures from separate branch. cannam@127: cannam@127: commit d2ea399c46174db45838ca6a3b917cf880970921 cannam@127: Author: Romain Dolbeau cannam@127: Date: Thu Apr 16 08:54:58 2015 +0200 cannam@127: cannam@127: Fix stack alignment (alloca) for generic256. cannam@127: cannam@127: commit b6135085bd1ee2e2c6c82b06e78d492e4f242cca cannam@127: Author: Romain Dolbeau cannam@127: Date: Tue Apr 14 10:17:38 2015 +0200 cannam@127: cannam@127: Add a configure option to disable building the documentation in doc/. This is useful if some documentation tools are missing on the host (i.e. fig2dev in maintainer mode). cannam@127: cannam@127: commit 38b93ccfc3786d1c23726dc939de558f4dd2a2d3 cannam@127: Author: Romain Dolbeau cannam@127: Date: Mon Apr 13 14:16:28 2015 +0200 cannam@127: cannam@127: In maintainer mode, detect whether 'indent' is available and is GNU indent. cannam@127: cannam@127: BSD 'indent' in e.g. MacOSX doesn't support -kr, which is the default style in FFTW3. cannam@127: Fallback to 'indent' with no option for non-GNU 'indent', or 'cat' if 'indent' is not available. cannam@127: This should fix GitHub issue #13. cannam@127: cannam@127: commit 96eb0ad31c8d0c226a6aeb95e68bda90dd7e6f6a cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Apr 12 13:49:39 2015 +0200 cannam@127: cannam@127: missing AC_ARG_ENABLE for --enable-fma cannam@127: cannam@127: commit 24ff943f4e99458d41db543305a2c945ba2ba429 cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Apr 12 13:47:24 2015 +0200 cannam@127: cannam@127: Revert "reinstate --enable-fma ; not enabled by default for AVX2 & AVX-512 (will ad a warning later)" cannam@127: cannam@127: This reverts commit 40691a49eea40d305405fe527e174e8067606dae. A different fix is needed. cannam@127: cannam@127: commit 37a0dbc10f0199fd431f0b5d8b42143cc2a2a88d cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Apr 12 13:39:28 2015 +0200 cannam@127: cannam@127: Add sanity check & warning in AVX2 & AVX-512 cannam@127: cannam@127: commit 40691a49eea40d305405fe527e174e8067606dae cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Apr 12 13:38:29 2015 +0200 cannam@127: cannam@127: reinstate --enable-fma ; not enabled by default for AVX2 & AVX-512 (will ad a warning later) cannam@127: cannam@127: commit 7960d08a3fe74b38d8dfdd20917efb52d141d53a cannam@127: Author: Erik Lindahl cannam@127: Date: Wed Apr 8 22:55:28 2015 +0200 cannam@127: cannam@127: Improved compiler flags for OS X cannam@127: cannam@127: Separate detection for AVX/AVX2 on gcc and clang. cannam@127: Clang works for AVX, but AVX2 leads to a compiler cannam@127: crash. Issue 20471870 has been filed with Apple. cannam@127: When using gcc, we now request to use the external cannam@127: system assembler, or the AVX/AVX2 instructions will cannam@127: cause errors. cannam@127: cannam@127: commit 91928338b767b84742e8ec86da6b4864381ed889 cannam@127: Author: Erik Lindahl cannam@127: Date: Wed Apr 8 22:54:41 2015 +0200 cannam@127: cannam@127: Fix alignments for generic simd. cannam@127: cannam@127: commit eaaec9b6ea9dc0f0656d953639c325855cb3bbee cannam@127: Author: Erik Lindahl cannam@127: Date: Wed Apr 8 21:16:13 2015 +0200 cannam@127: cannam@127: Made api versions more verbose for 128-bit AVX. cannam@127: cannam@127: commit 4b3dbf7009b020bffe7c9c96a5b24c87496fd058 cannam@127: Author: Erik Lindahl cannam@127: Date: Wed Apr 8 21:09:50 2015 +0200 cannam@127: cannam@127: Make 128/256 bit generic simd separate options cannam@127: cannam@127: These will only be used on esoteric and/or new cannam@127: architectures, which likely also miss cycle counters. cannam@127: In this case the widest simd would be picked automatically cannam@127: based on flops estimates, so to give the user more cannam@127: control it is better to provide separate options cannam@127: to enable/disable these two choices. cannam@127: cannam@127: commit cbe2a4a64064d12b9b817235906d61a996c00be1 cannam@127: Author: Romain Dolbeau cannam@127: Date: Mon Mar 30 13:35:47 2015 +0200 cannam@127: cannam@127: AVX-512: minor fix(sp)&improvement(dp) to VDUPL/VDUPH cannam@127: cannam@127: commit 5379243044ea4113b9cbde25fd097195817b3653 cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Mar 29 14:17:13 2015 +0200 cannam@127: cannam@127: AVX-512: fix typo; fix shuffle parameter in SP; _mm512_set1 exists now. cannam@127: cannam@127: commit 90f9610ee6708efc11c848b5e078dd92997ffa25 cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Mar 29 14:14:02 2015 +0200 cannam@127: cannam@127: Fix typo in KCvi cannam@127: cannam@127: commit 2b44c9213a11816506b1bd3d6b7316ed1ed65a15 cannam@127: Merge: 5c5bed2 1f28d2d cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Mar 29 10:49:42 2015 +0200 cannam@127: cannam@127: Merge branch 'experimental-simd' of github.com:FFTW/fftw3 into experimental-simd cannam@127: cannam@127: commit 5c5bed2365693b5f57503f6aad35264a3b3d86b9 cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Mar 29 10:48:13 2015 +0200 cannam@127: cannam@127: typo for AVX-512 cannam@127: cannam@127: commit 1f28d2d3693bd919c674e08dc76726cf56d3648f cannam@127: Author: Romain Dolbeau cannam@127: Date: Sun Mar 29 10:48:13 2015 +0200 cannam@127: cannam@127: typo cannam@127: cannam@127: commit a8845007ecb07fd0cc91994c574b1008c64708e0 cannam@127: Author: Romain Dolbeau cannam@127: Date: Thu Mar 19 19:22:44 2015 +0100 cannam@127: cannam@127: Generic SIMD support for gcc cannam@127: cannam@127: While not as optimized as the specific SIMD ports, this cannam@127: enables the usage of gcc's generic vector representation, cannam@127: which usually gets implemented with SIMD instructions on cannam@127: most hardware. Double precision implementations for 256 cannam@127: and 128 bits by Romain Dolbeau, merged into a single cannam@127: generic SIMD implementation and single precision added cannam@127: by Erik Lindahl. The option --enable-generic-simd will cannam@127: turn on both 128 and 256 bit versions, and the timers will cannam@127: choose the fastest codelets. cannam@127: cannam@127: commit 56bbdbab2f03ebae92fd2c52b9509fb3b8ffe226 cannam@127: Author: Erik Lindahl cannam@127: Date: Tue Mar 24 19:35:31 2015 +0100 cannam@127: cannam@127: Added Power8 VSX SIMD support cannam@127: cannam@127: Power8 is a descendent of Power7, but the switch to little endian cannam@127: means the old altivec SIMD will not work due to shifts on load cannam@127: and store, and the new VSX instructions are much improved. cannam@127: This adds support for both single and double precision VSX SIMD, cannam@127: using either gcc (tested with version 4.9) or IBM xlC (tested with cannam@127: version 13.1.2, slower than gcc). Clang from llvm-3.7 is still too cannam@127: buggy to compile VSX code correctly, but flags and detection has cannam@127: been added so it will work with a correct clang. cannam@127: cannam@127: commit 8aa91763af07767f3ebb71a9836a69e3b3385cab cannam@127: Author: Romain Dolbeau cannam@127: Date: Tue Feb 24 09:27:07 2015 +0100 cannam@127: cannam@127: Double precision Neon SIMD for aarch64 cannam@127: cannam@127: --enable-neon now works in double precision for 64-bit Arm. cannam@127: Support added for the generic timer virtual counter in armv7a cannam@127: (optional, available in A15 & A7) & armv8. They are privileged, but cannam@127: should be made user-readable in recent linux (> 3.19.1 for v7a, cannam@127: most for v8). cannam@127: Architecture Reference Manual ARMv7-A and ARMv7-R edition: cannam@127: E.7.16 CNTVCT, Virtual Count register, system level cannam@127: Architecture Reference Manual ARMv8, for ARMv8-A architecture cannam@127: profile Beta: D7.5.17 CNTVCT_EL0, Counter-timer Virtual Count register cannam@127: cannam@127: commit aa26395250c9c4d6831e8e5017650ea70af56a28 cannam@127: Author: Romain Dolbeau cannam@127: Date: Thu Sep 5 10:53:42 2013 +0200 cannam@127: cannam@127: AVX-512 SIMD support cannam@127: cannam@127: New configure flag --enable-avx512 cannam@127: cannam@127: commit de81bfdb66b9bc867e389bbaf67b56490ca2e2cd cannam@127: Author: Erik Lindahl cannam@127: Date: Wed Mar 25 15:49:33 2015 +0100 cannam@127: cannam@127: 128-bit AVX2 SIMD support cannam@127: cannam@127: Add 128 bit support for AVX2. Similar to AVX-128, this cannam@127: improves slightly on SSE2 due to more efficient instructions, cannam@127: and the shorter SIMD width is beneficial in some cases. Both cannam@127: 128- and 256-bit flavors will be built automatically with cannam@127: --enable-avx2, and the timing routines will chose the best one cannam@127: automatically. cannam@127: cannam@127: commit da988fa4c53fb63fafe2eeff3da4abad93e7d014 cannam@127: Author: Erik Lindahl cannam@127: Date: Sat Mar 28 12:52:52 2015 +0100 cannam@127: cannam@127: AVX2 kernels and CPUID support cannam@127: cannam@127: Initial AVX2 code from Romain Dolbeau. Modifications, cannam@127: cpuid and more compiler flags from Erik Lindahl. cannam@127: New --enable-avx2 configure flag supported. cannam@127: cannam@127: commit d7d9b9d2b71bd93bc4d4fa82d46a9c013291b7fe cannam@127: Author: Romain Dolbeau cannam@127: Date: Tue Sep 3 10:02:53 2013 +0200 cannam@127: cannam@127: KCvi [Knight Corner Vector Instructions] SIMD support cannam@127: cannam@127: This adds SIMD support for the first generation of Xeon Phi. cannam@127: cannam@127: commit b606e3191e5b65e2e13f67ef7dad5b1e7c40206c cannam@127: Author: Erik Lindahl cannam@127: Date: Wed Mar 25 01:44:17 2015 +0100 cannam@127: cannam@127: Improved AVX SIMD cannam@127: cannam@127: Previously, some kernels were actually faster with the old SSE2 cannam@127: SIMD, which made it necessary to compile with both sse2 and avx cannam@127: for good performance. This adds 128-bit AVX kernels which are cannam@127: enabled together with the standard AVX kernels. Apart from cannam@127: being encoded with AVX rather than SSE instructions cannam@127: (depending on compiler flags), it also uses a couple of new cannam@127: instructions only available with AVX that use fewer micro-ops. cannam@127: These instructions have also been added to the 256-bit AVX SIMD cannam@127: implementation. No new configure flags needed, it is just faster. cannam@127: cannam@127: commit 131027afcd3ed5d7c0185611036431c1035a734a cannam@127: Merge: 0ea3051 56af330 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 22 16:24:29 2015 -0400 cannam@127: cannam@127: Merge pull request #37 from maxlevesque/patch-1 cannam@127: cannam@127: add indent to requirements cannam@127: cannam@127: commit 56af330fc1600a856241968482ecd443bc2c26aa cannam@127: Author: Maximilien Levesque cannam@127: Date: Sat Mar 21 22:41:26 2015 +0100 cannam@127: cannam@127: add indent to requirements cannam@127: cannam@127: Without indent, `make` reports errors in somewhere in a directory called codelets. cannam@127: It also reports "/bin/bash: indent: command not found" not far away. cannam@127: cannam@127: For my Ubuntu 14.04 flavor, a simple `sudo apt-get install indent` made `make` work as expected. cannam@127: cannam@127: commit 0ea3051f99f2931a46a66aef8862517cd9c5f3c7 cannam@127: Merge: 506c163 69a82a6 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 16 06:36:46 2015 -0500 cannam@127: cannam@127: Merge pull request #32 from psteinb/2d_mem_layout_complying_to_text cannam@127: cannam@127: changed ny to n1 and nx to n0 so that the labels match the text cannam@127: cannam@127: commit 69a82a6c5ff4d2169a7f8a0afda9f3c68f24eb2a cannam@127: Author: Peter Steinbach cannam@127: Date: Mon Feb 16 09:10:23 2015 +0100 cannam@127: cannam@127: changed ny to n1 and nx to n0 so that the labels match the text, it's quite confusing otherwise cannam@127: cannam@127: commit 506c16346f9fc57444b179e542e88225e3c3e923 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 30 15:02:17 2015 -0500 cannam@127: cannam@127: fix #29 cannam@127: cannam@127: commit d94666815b9a0073e4ac8cc48f88e18ac931bd45 cannam@127: Merge: ded0051 9831bbd cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Dec 7 16:39:15 2014 -0500 cannam@127: cannam@127: Merge pull request #27 from mpip/master cannam@127: cannam@127: avoid segfaults due to double free cannam@127: cannam@127: commit 9831bbd14ca5b963ad1dba260c86151c94e000ee cannam@127: Author: Michael Pippig cannam@127: Date: Sat Dec 6 15:03:33 2014 +0100 cannam@127: cannam@127: avoid segfaults due to double free cannam@127: cannam@127: If fftw_mpi_mkplans_posttranspose() fails, the plans cld3, cld2rest, cannam@127: and cld2 are destroyed at nada and must be set to NULL. Otherwise, cannam@127: a second destroy at nada in mkplan() will cause a segfault. cannam@127: cannam@127: commit ded0051238f129fb65846e822191706c9b1f5221 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Nov 16 09:45:16 2014 -0500 cannam@127: cannam@127: Conciseness Police cannam@127: cannam@127: commit 113e1086966fdff4c172672753cc880e6bc74d3d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Nov 16 09:41:05 2014 -0500 cannam@127: cannam@127: add {before,after}_planner_hooks cannam@127: cannam@127: FFTW now calls fftw_before_planner_hook() before creating a plan, and cannam@127: fftw_after_planner_hook() afterwards. This allows users, e.g., to cannam@127: grab a lock. cannam@127: cannam@127: TBD: add arguments. cannam@127: cannam@127: commit 28635e1d5f0a0079af3e7d00cd0678c4745e2c2b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Nov 16 09:22:47 2014 -0500 cannam@127: cannam@127: avoid multiple declarations of fftw_alignment_of() cannam@127: cannam@127: fftw_alignment_of() was declared both in the API header file fftw3.h cannam@127: and in the internal header file ifftw.h. While there is nothing wrong cannam@127: with this, it breaks the property that all exported symbols are cannam@127: defined in the API directory. E.g., I am not sure what happens on cannam@127: windows without the proper DLLEXPORT nonsense. cannam@127: cannam@127: To avoid any issues, rename the internal routine to cannam@127: fftw_ialignment_of(), and define an API wrapper. cannam@127: cannam@127: commit 36597576e4c3c5dc3efd7d8b57a1bbad505715aa cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Aug 12 12:29:50 2014 -0400 cannam@127: cannam@127: fix #21 (don't use float128 on Portland compilers, which pretend to be gcc) cannam@127: cannam@127: commit cde4559ba9b822166cb88a84a0994fdb83a2061c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 26 20:09:38 2014 -0400 cannam@127: cannam@127: Avoid transforming uninitalized data. cannam@127: cannam@127: In r2c/c2r transforms when using 4-way SIMD, sometimes FFTW uses the cannam@127: following hack: to transform an odd number of inputs, it copies the cannam@127: input into a buffer that holds space for one extra input; it cannam@127: transforms the buffer (now comprising an even number of inputs, as cannam@127: required by SIMD); it copies back the odd number of transformed cannam@127: inputs, ignoring the padding element. cannam@127: cannam@127: The extra input was uninitialized until now. This is ok because we cannam@127: ignore the transform of the uninitialized input. Transforming cannam@127: uninitialized data may cause floating-point exceptions, an effect that cannam@127: is observable. This patch initializes the additional elements to cannam@127: zero, thus avoiding the problem. cannam@127: cannam@127: This patch also includes a test, but the test is disabled by default cannam@127: because it is nonportable. To observe the FP exception, one must use cannam@127: feenableexcept(), which appears to be a GNU-ism. cannam@127: cannam@127: commit 2493129c332197c5195ecb6796cfeb5e8d92e09a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 10 10:41:38 2014 -0400 cannam@127: cannam@127: fix #19: missing Fortran interface for fftwq_alloc_real cannam@127: cannam@127: commit 07ef78dc1b273a40fb4f7db1797d12d3423b1f40 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 8 11:14:15 2014 -0400 cannam@127: cannam@127: fix #18 (disable float128 for CUDACC) cannam@127: cannam@127: commit 2fd372f31ab7c6417de0634199bcd5b7765df926 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 28 17:17:19 2014 -0400 cannam@127: cannam@127: git rm *~ cannam@127: cannam@127: commit 203e0d610ec1e413bb426a7d60fd5e2a206a2830 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 3 15:46:19 2014 -0400 cannam@127: cannam@127: Fix wrong boolean precedence in hppa cycle counter. cannam@127: cannam@127: Thanks Jens Keiner for the bug report. For some reason nobody noticed cannam@127: this in years. cannam@127: cannam@127: commit f8048af3e30cb3f65befd0aa2f3d16de3eeb5583 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 15:23:11 2014 -0500 cannam@127: cannam@127: more .gitignore additions cannam@127: cannam@127: commit 5a51b3fe98509cc7e7ba5d3e17a3381777ad4731 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 15:21:31 2014 -0500 cannam@127: cannam@127: some fixes for make distcheck cannam@127: cannam@127: commit 836af27f5d780970c87e436da882c9928e09c0f0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 13:42:02 2014 -0500 cannam@127: cannam@127: copyright year update cannam@127: cannam@127: commit 853f9f7cad1a8d3e92e6767562e4cd2d336164d2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 13:39:18 2014 -0500 cannam@127: cannam@127: updates for 3.3.4 cannam@127: cannam@127: commit bf30f5aeea1c2927b302b13dc7579acf9f2adc12 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 13:30:00 2014 -0500 cannam@127: cannam@127: another file in .gitignore cannam@127: cannam@127: commit c82b4fd61796715b1043982b1d4af49047f90238 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 10:39:09 2014 -0500 cannam@127: cannam@127: add .gitignore cannam@127: cannam@127: commit 146fa8d61fca4a06a85c70d7167ac925575df02b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 10:47:40 2014 -0500 cannam@127: cannam@127: added fftw_sprint_plan to output plan info to a string (so that the caller can be more flexible about how it is displayed) cannam@127: cannam@127: commit 53e1fdbc07133b53ffbbd51c56b57a89880c0b21 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 10:26:48 2014 -0500 cannam@127: cannam@127: document fftw_alignment_of (since I found it useful in Julia, other people may too) cannam@127: cannam@127: commit 7dbc7067e99477312acae30a9001c0dffa9bb428 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Nov 19 19:08:44 2013 -0500 cannam@127: cannam@127: group together AC_CHECK_DECLS for functions in stdlib.h cannam@127: cannam@127: commit fb70e413bddca578b2b72e0cac281d5c9a3c4101 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Nov 19 07:33:59 2013 -0500 cannam@127: cannam@127: Be more careful in detecting sinl(), cos(), memalign(), posix_memalign() cannam@127: cannam@127: commit 40f59a1d0f9bf384826595c499b0e7fe99aa1df5 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Nov 19 06:20:43 2013 -0500 cannam@127: cannam@127: Fix autodetection of cosl(), sinl() cannam@127: cannam@127: Autoconf must have changed since we last looked. Thanks cannam@127: Åke Sandgren for the fix. cannam@127: cannam@127: commit 7e66dc5a495edc855dc9e156767172eaeabee335 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Oct 26 17:31:00 2013 -0400 cannam@127: cannam@127: Fix wrong example in fftw-wisdom-to-conf.1 cannam@127: cannam@127: Thanks Julian Taylor for the bug report. cannam@127: cannam@127: commit dd3283ac7d2e916bdeccb7229a669fcc2ef7ff83 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Oct 1 07:03:51 2013 -0400 cannam@127: cannam@127: Use "bench$(EXEEXT)" instead of "bench" cannam@127: cannam@127: Seems to be necessary on Windows. cannam@127: cannam@127: commit c6acf03f53d217c9041eae3d381ad41b942dc9d9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Sep 20 09:26:08 2013 -0400 cannam@127: cannam@127: fix typo cannam@127: cannam@127: commit f230f8cf903f1e2bb1261ed2f8657a99ac12a9ca cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 11 19:28:50 2013 -0400 cannam@127: cannam@127: Fix fftw-wisdom-to-conf cannam@127: cannam@127: Apparently we broke fftw-wisdom-to-conf many years ago and nobody cannam@127: noticed. Thanks Florian Oppermann for the bug report. cannam@127: cannam@127: commit c74775bff0c164611377b29d95b3f6a6e8192005 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 4 10:31:33 2013 -0400 cannam@127: cannam@127: Somehow the NEWS entries for fftw-3.1.[23] were missing from the trunk. cannam@127: cannam@127: commit c87bdc8bfd7cbf753bbe0635ea82613bc271220d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 3 06:59:35 2013 -0400 cannam@127: cannam@127: note fixes for texinfo-5 cannam@127: cannam@127: commit ed390e3a385832e1faa452032f170510be6ed280 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 2 09:10:01 2013 -0400 cannam@127: cannam@127: version.texi should not be in git cannam@127: cannam@127: commit aff23d05642705f738f788648c060085bdc476d6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 2 09:05:50 2013 -0400 cannam@127: cannam@127: Fix the manual to work with both texinfo-4 and texinfo-5. cannam@127: cannam@127: Texinfo has been stable for the first 15 years of FFTW's history. cannam@127: Then some genius, with too much time in his hands and on a mission to cannam@127: deliver the world from the evil of the C language, decided to rewrite cannam@127: makeinfo in Perl, the old C version of makeinfo being, as I said, cannam@127: evil. The official excuse for the rewrite was that now I can have my cannam@127: manual in XML format, as if XML were a feature. cannam@127: cannam@127: The result of this stroke of genius is that texinfo-5 has different cannam@127: rules for macro expansion than texinfo-4 does, specifically regarding cannam@127: whether or not spaces after a macro are ignored. Texinfo-4 had weird cannam@127: rules, but at least they were constant and internally more or less cannam@127: consistent. Texinfo-5 has different rules, and even worse the rules cannam@127: in texinfo-5 are inconsistent between the TeX and HTML output cannam@127: processors. This situation makes it almost impossible for us to cannam@127: produce a manual that works with both texinfo 4 and 5 in all modes cannam@127: (TeX, info, and html). The @noindent/@refill hack is my best shot at cannam@127: patching this situation. cannam@127: cannam@127: commit b0308275bb63a9cb3edb5847fa130f901deaf47e cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 17 11:39:05 2013 -0400 cannam@127: cannam@127: fftw_wisdom.1.in: document the --threads option. cannam@127: cannam@127: commit 7eb9af0354f7663fa89daa56163d5bc5865bcec1 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 27 13:12:15 2013 -0400 cannam@127: cannam@127: Define the version number as M4 macros, so that it is defined only once. cannam@127: cannam@127: We used to have two version strings: the package number FFTW-X.Y.Z and cannam@127: the libtool number CURRENT:REVISION:AGE with the invariant that cannam@127: REVISION==Z. Unfortunately in the fftw-3.3.3 release we forgot to cannam@127: change REVISION, with the result that fftw-3.3.3 and fftw-3.3.2 have cannam@127: the same REVISION number. cannam@127: cannam@127: Hopefully this patch will prevent similar screwups in the future. cannam@127: cannam@127: commit e67d16ed0b81b0e01124c7cdee2320bfea2fd6e4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 17 19:47:57 2013 -0400 cannam@127: cannam@127: note that these scripts are not meant for normal users cannam@127: cannam@127: commit b892c705c7051bcd27c0939070e515fe85f68bb2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 17 19:45:52 2013 -0400 cannam@127: cannam@127: add README.md for github's convenience cannam@127: cannam@127: commit 029db460f692ea9fefc7d0efdf8e11b5d51215b0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 17 19:44:00 2013 -0400 cannam@127: cannam@127: Update README for people who download the git repository. cannam@127: cannam@127: commit 128e9ddbde76a63c7f255b258ff02bd6766cbaa4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 17 19:37:40 2013 -0400 cannam@127: cannam@127: Remove reference to obsolete mailing list. cannam@127: cannam@127: commit 4b1d3b63d64a01216c023389565e40ca03d6be2f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 16 09:47:59 2013 -0400 cannam@127: cannam@127: mkdist.sh: use git instead of darcs cannam@127: cannam@127: Use git instead of darcs to generate the ChangeLog. Also, refuse to cannam@127: create a distribution if git HEAD does not have a tag, to prevent a cannam@127: common mistake. cannam@127: cannam@127: commit fe84f5d9eeef9f773b0785ba91032ab5b7677461 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 16 09:26:03 2013 -0400 cannam@127: cannam@127: remove some junk '*~' files created by the darcs->git conversion cannam@127: cannam@127: commit 2f9d0a41eec86def6fe9160dac1be172bb241eba cannam@127: Author: stevenj cannam@127: Date: Thu Feb 7 14:22:12 2013 -0500 cannam@127: cannam@127: use Win32 threads, not pthreads, if both are present (it's not clear why Windows users would ever want the latter); see also https://github.com/JuliaLang/julia/issues/2015 cannam@127: cannam@127: commit b8623189967ed7eb1ca50e80f2b5ee2d6f3ca0ad cannam@127: Author: stevenj cannam@127: Date: Thu Feb 7 14:13:40 2013 -0500 cannam@127: cannam@127: I found it useful in the Julia interface to call fftw_alignment_of (in order to check plan applicability), in which case we need to IFFTW_EXTERN it for Windows; might be worth considering documenting this function cannam@127: cannam@127: commit 23b1bf3e19198d2c4575bf6f305cd41ebc17124c cannam@127: Author: athena cannam@127: Date: Sat Jan 12 15:35:46 2013 -0500 cannam@127: cannam@127: Increase timing interval to 5000 cycles on x86_64 cannam@127: cannam@127: commit 6e615417da7e1c7cba49b1a73c9edcc15d938cf9 cannam@127: Author: athena cannam@127: Date: Tue Dec 4 15:58:22 2012 -0500 cannam@127: cannam@127: make -lm a private library in fftw.pc.in cannam@127: cannam@127: Julian Taylor says: cannam@127: cannam@127: Make -lm a private library, libfftw is linked against it so clients cannam@127: don't need it. You can use pkg-configs --static flag for static cannam@127: linking. this works on all systems where indirect linking works, cannam@127: probably on all others pkg-config is not supported anyway (wild cannam@127: guess). cannam@127: cannam@127: commit 553849f32bcc82a17de11c76cfa9b2f672bfb89e cannam@127: Author: athena cannam@127: Date: Tue Dec 4 15:54:58 2012 -0500 cannam@127: cannam@127: Escape minus signs in man pages cannam@127: cannam@127: http://lintian.debian.org/tags/hyphen-used-as-minus-sign.html says: cannam@127: cannam@127: By default, "-" chars are interpreted as hyphens (U+2010) by groff, cannam@127: not as minus signs (U+002D). Since options to programs use minus cannam@127: signs (U+002D), this means for example in UTF-8 locales that you cannam@127: cannot cut and paste options, nor search for them easily. cannam@127: cannam@127: Thanks Julian Taylor for the patch. cannam@127: cannam@127: commit 715c7ea8347a5dad7d97a0c5d81a87801e826ada cannam@127: Author: athena cannam@127: Date: Tue Dec 4 15:53:28 2012 -0500 cannam@127: cannam@127: Change texinfo category to Development cannam@127: cannam@127: commit 01810ba2a427ee086a4a5323e991dd19e2d715be cannam@127: Author: stevenj cannam@127: Date: Sat Nov 24 22:37:54 2012 -0500 cannam@127: cannam@127: fixed deadlock bug caused by bogosity flag getting out of synch between processes; thanks to Michael Pippig for the bug report cannam@127: cannam@127: commit 69aa82642e26a8eb5292a8a7b83250e8df619065 cannam@127: Author: athena cannam@127: Date: Wed Nov 21 18:34:29 2012 -0500 cannam@127: cannam@127: Updated NEWS cannam@127: cannam@127: commit e98f888b9457ce5855491279c6c0ef72e23a374c cannam@127: Author: athena cannam@127: Date: Wed Nov 21 18:33:15 2012 -0500 cannam@127: cannam@127: use 2x2 AVX transposition instead of individual stores. cannam@127: cannam@127: This seems to improve single-precision AVX on Sandy Bridge machines. cannam@127: cannam@127: commit 466f579cb8856a0709da1e6c6b5ca03360bc61a0 cannam@127: Author: stevenj cannam@127: Date: Tue Nov 20 12:18:00 2012 -0500 cannam@127: cannam@127: revert part of Taylor patch to acx_mpi.m4: do not link -lmpi if mpicc works without libraries, as -lmpi may be some completely different MPI implementation cannam@127: cannam@127: commit 610460226f6d5d2d7c4c53896b5aff9b1f108e4b cannam@127: Author: stevenj cannam@127: Date: Tue Nov 20 11:44:57 2012 -0500 cannam@127: cannam@127: fix deadlock bug (thanks to Michael Pippig for the bug report and patch, and to Graham Dennis for the bug report) in which some processes called MPI_Alltoall and some called MPI_Alltoallv cannam@127: cannam@127: commit 512d8d783d6af373fca8376f79255b794df5bd31 cannam@127: Author: athena cannam@127: Date: Mon Oct 29 15:20:01 2012 -0400 cannam@127: cannam@127: fix texinfo quirk cannam@127: cannam@127: commit ff329890540002506c47717ebbc3959de30e5066 cannam@127: Author: athena cannam@127: Date: Mon Oct 29 09:16:43 2012 -0400 cannam@127: cannam@127: clarify that padding only applies to in-place transforms cannam@127: cannam@127: commit 905ded711f93fa3c94faa7623a5093525338fdeb cannam@127: Author: athena cannam@127: Date: Sun Oct 28 18:42:48 2012 -0400 cannam@127: cannam@127: make the index-computation logic less paranoid cannam@127: cannam@127: The problem is that for each K and for each expression of the form P[I cannam@127: + STRIDE * K] in a loop, most compilers will try to lift an induction cannam@127: variable PK := &P[I + STRIDE * K]. In large codelets we have many cannam@127: such values of K. For example, a codelet of size 32 with 4 input cannam@127: pointers will generate O(128) induction variables, which will likely cannam@127: overflow the register set, which is likely worse than doing the index cannam@127: computation in the first place. cannam@127: cannam@127: In the past we (wisely and correctly) assumed that compilers will do cannam@127: the wrong thing, and consequently we disabled the induction-variable cannam@127: "optimization" altogether by setting STRIDE ^= ZERO, where ZERO is a cannam@127: value guaranteed to be 0. Since the compiler does not know that cannam@127: ZERO=0, it cannot perform its "optimization" and it is forced to cannam@127: behave sensibly. cannam@127: cannam@127: With this patch, FFTW is a little bit less paranoid. FFTW now cannam@127: disables the induction-variable optimization" only when we estimate cannam@127: that the codelet uses more than ESTIMATED_AVAILABLE_INDEX_REGISTERS cannam@127: induction variables. cannam@127: cannam@127: Currently we set ESTIMATED_AVAILABLE_INDEX_REGISTERS=16. 16 registers ought cannam@127: to be enough for anybody (or so the amd64 and ARM ISA's seem to imply). cannam@127: cannam@127: commit 1dacef5bde5cb6599f9d98e42495f7897f109787 cannam@127: Author: athena cannam@127: Date: Sun Oct 28 18:33:24 2012 -0400 cannam@127: cannam@127: silence warnings cannam@127: cannam@127: commit fb08724b27a05ca890c1da062c8d0385c22c02eb cannam@127: Author: athena cannam@127: Date: Sat Oct 27 09:58:49 2012 -0400 cannam@127: cannam@127: bump version to 3.3.3 cannam@127: cannam@127: commit c4d6abbc1c80eb612b2abccce728a06189780a69 cannam@127: Author: athena cannam@127: Date: Sat Oct 27 09:55:15 2012 -0400 cannam@127: cannam@127: evaluate plans for >1ms when using gettimeofday() cannam@127: cannam@127: The previous limit 10ms was too paranoid, and it made life difficult cannam@127: on machines without an "official" cycle counter, such as ARM. cannam@127: cannam@127: commit 172dd3def821c0898822a5ca72c3f5391553536c cannam@127: Author: athena cannam@127: Date: Sat Oct 27 09:46:04 2012 -0400 cannam@127: cannam@127: use 4-way NEON SIMD instead of 2-way cannam@127: cannam@127: Kai-Uwe Bloem tried to warn me a year ago that 128-bit NEON was better cannam@127: than 64-bit NEON even on machines with a 64-bit pipe, but I foolishly cannam@127: did not listen. Now that 128-bit NEON pipes are starting to appear on cannam@127: the market it is definitely time to switch. cannam@127: cannam@127: commit 1c9c469f8727ab1780533226283746e7e9098694 cannam@127: Author: athena cannam@127: Date: Wed Sep 26 14:21:12 2012 -0400 cannam@127: cannam@127: Note that fftw-3.3 includes MPI support cannam@127: cannam@127: commit 2de12d67e5f4e34d39119f2e730e9d70e4df0c4e cannam@127: Author: athena cannam@127: Date: Wed Jul 18 11:25:40 2012 -0400 cannam@127: cannam@127: remove obsolete unused function cannam@127: cannam@127: commit 6bc94ae7db56490e35c256bd4840608eea6ec150 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 29 15:57:14 2012 -0400 cannam@127: cannam@127: whoops, call omp_get_max_threads; thanks to Hanno Rein for the bug report cannam@127: cannam@127: commit 747ece1503281aad7beb32448af9a7cad05eba52 cannam@127: Author: athena cannam@127: Date: Sat Apr 28 10:55:09 2012 -0400 cannam@127: cannam@127: Fix libfftw3/libfftw3_threads chicken-egg problem cannam@127: cannam@127: On most systems we want to build libfftw3 first, so that cannam@127: libfftw3_threads can depend upon libfftw3. When producing a single cannam@127: combined-thread library (e.g. on Windows) we want the opposite, cannam@127: so that libfftw3 can include libfftw3_threads. cannam@127: cannam@127: commit 4bcfb67da7211171d5f4b80a90845770bbd2e147 cannam@127: Author: athena cannam@127: Date: Sat Apr 28 10:11:28 2012 -0400 cannam@127: cannam@127: updated NEWS for 3.3.2 cannam@127: cannam@127: commit cb553a8315ae9a700558956a190aac4658064b83 cannam@127: Author: athena cannam@127: Date: Thu Apr 26 19:36:11 2012 -0400 cannam@127: cannam@127: change revision to 3.3.2 cannam@127: cannam@127: commit 98229b0d7673cfa15a8c339d305b09367b97d670 cannam@127: Author: athena cannam@127: Date: Thu Apr 26 19:31:02 2012 -0400 cannam@127: cannam@127: Remove old aligned_main() hack. cannam@127: cannam@127: On i386, in our benchmark program we used to manually aligned the cannam@127: stack to 16-byte boundary via asm trickery. This was a good idea in cannam@127: 1999 (and it was actually necessary to make things work) but the hack cannam@127: is now obsolete and it seems to break gcc-4.7. So the hack is now cannam@127: gone. cannam@127: cannam@127: commit 4e4c680e7497ee8bb87bb31451d10d71b8c205e2 cannam@127: Author: athena cannam@127: Date: Thu Mar 29 16:26:16 2012 -0400 cannam@127: cannam@127: Bugfix: a couple of uninitialized values in the benchmark program cannam@127: cannam@127: commit 229d864bde7f95f45ee23608b756926a3a20ddd0 cannam@127: Author: athena cannam@127: Date: Tue Mar 20 19:03:47 2012 -0400 cannam@127: cannam@127: make libfftw{threads,mpi} depend upon libfftw for libtool purposes cannam@127: cannam@127: Thanks Julian Taylor for the patch cannam@127: cannam@127: commit f9a05701f2027906b68ff913713166310e18c8cc cannam@127: Author: stevenj cannam@127: Date: Tue Mar 6 04:44:00 2012 -0500 cannam@127: cannam@127: formatting tweak cannam@127: cannam@127: commit cec6c01b2647796f5909cfe2d90ce040380da5c5 cannam@127: Author: athena cannam@127: Date: Mon Mar 5 21:05:27 2012 -0500 cannam@127: cannam@127: destroying => overwriting cannam@127: cannam@127: commit ad79a0ae3baf83548bc5c9597b0aae94e5226073 cannam@127: Author: stevenj cannam@127: Date: Fri Mar 2 10:31:20 2012 -0500 cannam@127: cannam@127: note that WISDOM_ONLY is a documented flag cannam@127: cannam@127: commit adf4cdab68e50778689170eb91270bb189aea1ac cannam@127: Author: stevenj cannam@127: Date: Fri Mar 2 10:27:08 2012 -0500 cannam@127: cannam@127: check for icc pretending to be gcc before including quad-precision decls; thanks to Michael Anselmi for the bug report cannam@127: cannam@127: commit 99aeb386978a56c1a33f6fe7c9b7942421cf550e cannam@127: Author: stevenj cannam@127: Date: Fri Mar 2 10:23:19 2012 -0500 cannam@127: cannam@127: foo_CFLAGS needs to manually include AM_CFLAGS; thanks to Henry Gomersall for the Windows bug report cannam@127: cannam@127: commit 6060dbccef23e01e603f44d42ab602ab89a9442e cannam@127: Author: athena cannam@127: Date: Sat Feb 25 15:21:39 2012 -0500 cannam@127: cannam@127: update for latest mingw cannam@127: cannam@127: commit 06fff6523418bd7ee2478c081cec418eb73adee5 cannam@127: Author: stevenj cannam@127: Date: Mon Feb 20 23:06:13 2012 -0500 cannam@127: cannam@127: added Fortran NEWS cannam@127: cannam@127: commit 51e33866a2c0cc189954b8197e5053b3275fc564 cannam@127: Author: stevenj cannam@127: Date: Mon Feb 20 23:00:13 2012 -0500 cannam@127: cannam@127: move non-portable extended/quad precision F03 interfaces into separate .f03 files (while keeping double/single in fftw3.f03 for minimal ABI breakage) cannam@127: cannam@127: commit 545c90a1db20c5cd50f30d0f31d2334ec10cf8ac cannam@127: Author: athena cannam@127: Date: Mon Feb 20 11:21:57 2012 -0500 cannam@127: cannam@127: rm mpi/fftw3-mpi.f03 at make clean time; thanks Tyler Luchko for the bug report. cannam@127: cannam@127: commit b7c0fcdee76b9f2efe194ec13d3b1ca97fadf376 cannam@127: Author: athena cannam@127: Date: Mon Feb 20 11:18:24 2012 -0500 cannam@127: cannam@127: Disable a Visual Studion warning that was obnoxious enough for Sebastian Schuberth to send us a patch. cannam@127: cannam@127: commit 4a0de08f4d1f026454fba2e053bd0573afb984d2 cannam@127: Author: athena cannam@127: Date: Mon Feb 20 11:18:06 2012 -0500 cannam@127: cannam@127: Change version to 3.3.1 cannam@127: cannam@127: commit 49783f83e58d9bc7dc3ea7f1822fefe961361935 cannam@127: Author: athena cannam@127: Date: Mon Feb 20 11:03:15 2012 -0500 cannam@127: cannam@127: Integrated Visual Studio AVX patches by Carsten Steger cannam@127: cannam@127: commit bc9dc18402445ce9b357dea5ff2b20720773ddc5 cannam@127: Author: stevenj cannam@127: Date: Wed Nov 9 10:13:32 2011 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 659e85e6065429fc6a6e4a005bc5f5068f707fb3 cannam@127: Author: stevenj cannam@127: Date: Tue Nov 8 22:45:09 2011 -0500 cannam@127: cannam@127: add missing F77 set_timelimit function; thanks to Martin Diehl for the bug repory cannam@127: cannam@127: commit cf1a5563a691fd0c25c67910926adb2ef936cbb7 cannam@127: Author: athena cannam@127: Date: Sun Sep 25 10:54:56 2011 -0400 cannam@127: cannam@127: note requirement of /machine:x64 in windows x64 README cannam@127: cannam@127: commit cb216e1fd71ab751f2de5e083bc0237cc98535c9 cannam@127: Author: athena cannam@127: Date: Sun Sep 18 09:28:20 2011 -0400 cannam@127: cannam@127: AVX detection for MSVC cannam@127: cannam@127: commit 8acf60e9e0c6417a91d9a21dddb1467feb4fd23e cannam@127: Author: athena cannam@127: Date: Tue Sep 13 14:58:29 2011 -0400 cannam@127: cannam@127: compile with C89 cannam@127: cannam@127: commit f3edf46ebdda4a632ccb0ece6c5fa0559014d24e cannam@127: Author: athena cannam@127: Date: Sat Sep 3 16:25:50 2011 -0400 cannam@127: cannam@127: use the same search pruning heuristics for threaded plans as for nonthreaded plans cannam@127: cannam@127: commit f004d764307d0e1815c2f936a9a398825d367f2c cannam@127: Author: athena cannam@127: Date: Sat Sep 3 16:12:11 2011 -0400 cannam@127: cannam@127: shorten ESTIMATE planning time for certain weird sizes cannam@127: cannam@127: FFTW includes a collection of "solvers" that apply to a subset of cannam@127: "problems". Assume for simplicity that a "problem" is a single 1D cannam@127: complex transform of size N, even though real "problems" are much more cannam@127: general than that. FFTW includes three "prime" solvers called cannam@127: "generic", "bluestein", and "rader", which implement different cannam@127: algorithms for prime sizes. cannam@127: cannam@127: Now, for a "problem" of size 13 (say) FFTW also includes special code cannam@127: that handles that size at high speed. It would be a waste of time to cannam@127: measure the execution time of the prime solvers, since we know that cannam@127: the special code is way faster. However, FFTW is modular and one may cannam@127: or may not include the special code for size 13, in which case we must cannam@127: resort to one of the "prime" solvers. To address this issue, the cannam@127: "prime" solvers (and others) are proclaimed to be SLOW". When cannam@127: planning, FFTW first tries to produce a plan ignoring all the SLOW cannam@127: solvers, and if this fails FFTW tries again allowing SLOW solvers. cannam@127: cannam@127: This heuristic works ok unless the sizes are too large. For example cannam@127: for 1044000=2*2*2*2*2*3*3*5*5*5*29 FFTW explores a huge search tree of cannam@127: all zillion factorizations of 1044000/29, failing every time because cannam@127: 29 is SLOW; then it finally allows SLOW solvers and finds a solution cannam@127: immediately. cannam@127: cannam@127: This patch proclaims solvers to be SLOW only for small values of N. cannam@127: For example, the "generic" solver implements an O(n^2) DFT algorithm; cannam@127: we say that it is SLOW only for N<=16. cannam@127: cannam@127: The side effects of this choice are as follows. If one modifies FFTW to cannam@127: include a fast solver of size 17, then planning for N=17*K will be cannam@127: slower than today, because FFTW till try both the fast solver and the cannam@127: generic solver (which is SLOW today and therefore not tried, but is no cannam@127: longer SLOW after the patch). If one removes a fast solver, of size say cannam@127: 13, then he may still fall into the current exponential-search behavior cannam@127: for "problems" of size 13*HIGHLY_FACTORIZABLE_N. cannam@127: cannam@127: If somebody had compleined about transforms of size 1044000 ten years cannam@127: ago, "don't do that" would have been an acceptable answer. I guess the cannam@127: bar is higher today, so I am going to include this patch in our 3.3.1 cannam@127: release despite their side-effects for people who want to modify FFTW. cannam@127: cannam@127: commit 610f7976d8d31f385e2bce4fd8da0a5c770fb877 cannam@127: Author: athena cannam@127: Date: Sat Aug 27 13:55:24 2011 -0400 cannam@127: cannam@127: Fix typo fftw_execute_dft_r2r => fftw_execute_r2r cannam@127: cannam@127: Thanks KIU Shueng Chuan for the bug report. cannam@127: cannam@127: commit 76f7c5558bb5eedb80ff3e67db798b4e493a5872 cannam@127: Author: athena cannam@127: Date: Fri Aug 26 06:13:55 2011 -0400 cannam@127: cannam@127: In Rader's algorithm, compute the generator lazily. cannam@127: cannam@127: The planner was spending a lot of time computing generators for cannam@127: plans that were immediately discarded. Now we compute generators cannam@127: only when absolutely needed. cannam@127: cannam@127: commit e1b527d72aad02ddea04f266f6831fb13768fbc3 cannam@127: Author: athena cannam@127: Date: Sun Aug 21 16:16:38 2011 -0400 cannam@127: cannam@127: Release notes for 3.3.1-beta1 cannam@127: cannam@127: commit 7079b5216c27e2320215f1eb10f6c6554a6c1ac1 cannam@127: Author: stevenj cannam@127: Date: Fri Aug 19 19:59:17 2011 -0400 cannam@127: cannam@127: make fftw_mpi_block routine 10x faster, since it is being called zillions of times (thanks to Tom Vacek for the profiling) cannam@127: cannam@127: commit e5c7931a01f350aa1f756bfa76307b317e2208e1 cannam@127: Author: athena cannam@127: Date: Thu Aug 18 14:19:36 2011 -0400 cannam@127: cannam@127: Implement autodetection of NEON extensions cannam@127: cannam@127: commit 16600d97d52a81152e4ef9ac140c336a5a1ca126 cannam@127: Author: athena cannam@127: Date: Sun Aug 14 14:12:29 2011 -0400 cannam@127: cannam@127: Update the FSF address. cannam@127: cannam@127: The FSF moved downtown. cannam@127: cannam@127: commit 8609b388c7872e1b39baa6d72349dbbb476ade97 cannam@127: Author: stevenj cannam@127: Date: Thu Aug 11 14:54:38 2011 -0400 cannam@127: cannam@127: allow specifying TRANSPOSE_{IN/OUT} transpose plans, since libbench does not canonicalize rnk=1 n=1 plans as rnk=0 cannam@127: cannam@127: commit 341d3e142ea250096b5c76778f77fba4bf6de622 cannam@127: Author: stevenj cannam@127: Date: Thu Aug 11 14:17:24 2011 -0400 cannam@127: cannam@127: check.pl should occasionally check DESTROY_INPUT problems too (especially since those enable slightly different algorithms in MPI) cannam@127: cannam@127: commit 2845a3c0be7059c93b704b624cc482f9e4778188 cannam@127: Author: stevenj cannam@127: Date: Thu Aug 11 12:37:51 2011 -0400 cannam@127: cannam@127: 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 cannam@127: cannam@127: commit f02c57b267809e16e0e29ae8f7c1301323eda1cb cannam@127: Author: athena cannam@127: Date: Mon Aug 8 10:06:14 2011 -0400 cannam@127: cannam@127: Add support for ARM NEON cannam@127: cannam@127: commit 8ec2b52144bcf72a028cb84c273719b0ecc7730e cannam@127: Author: stevenj cannam@127: Date: Fri Aug 5 17:25:32 2011 -0400 cannam@127: cannam@127: more C++ paranoia cannam@127: cannam@127: commit 6047a7079ad3a55d982e68e03cbe158c5e476ee0 cannam@127: Author: stevenj cannam@127: Date: Fri Aug 5 17:02:00 2011 -0400 cannam@127: cannam@127: tentative version bump for 3.3.1 cannam@127: cannam@127: commit 6fbb0639cf17563d4603c6d545e9335d19e3ac43 cannam@127: Author: stevenj cannam@127: Date: Fri Aug 5 16:52:28 2011 -0400 cannam@127: cannam@127: 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 cannam@127: cannam@127: commit d7feb2daba8f23e5bca6c4e4e3f4177ad9a49e87 cannam@127: Author: stevenj cannam@127: Date: Fri Aug 5 16:04:06 2011 -0400 cannam@127: cannam@127: 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 cannam@127: cannam@127: commit 50d12441bd0b3410c799d11784717e76147b5474 cannam@127: Author: stevenj cannam@127: Date: Fri Aug 5 14:04:32 2011 -0400 cannam@127: cannam@127: manual typo cannam@127: cannam@127: commit 9e45ff08aca4e28ec61c947284188a01aed45fe6 cannam@127: Author: athena cannam@127: Date: Tue Jul 26 20:55:45 2011 -0400 cannam@127: cannam@127: Honor WITH_OUR_MALLOC in libbench2 cannam@127: cannam@127: commit 2cfcd40d46731a41a400ed4a4d2eeeb954422568 cannam@127: Author: athena cannam@127: Date: Tue Jul 26 20:27:28 2011 -0400 cannam@127: cannam@127: fixed typo: incorrect name of combined threads library on Windows cannam@127: cannam@127: commit a81ea0083c5cc32c9acecfd5f57f38d0e4f07bb5 cannam@127: Author: stevenj cannam@127: Date: Mon Jul 25 14:38:20 2011 -0400 cannam@127: cannam@127: 3.3 version bump & NEWS cannam@127: cannam@127: commit df2116046636b6ed6b989fa8b706f3e9db53e17e cannam@127: Author: stevenj cannam@127: Date: Mon Jul 25 14:37:48 2011 -0400 cannam@127: cannam@127: use int(..., C_SIZE_T) rather than declaring another variable in the Fortran examples cannam@127: cannam@127: commit 19ebb38d4180b0a70f4579d37dedfd6e678edecd cannam@127: Author: stevenj cannam@127: Date: Wed Jul 13 05:02:32 2011 -0400 cannam@127: cannam@127: typo, thanks to Rhys Ulerich for the comment cannam@127: cannam@127: commit a221f0e99a9206e1edaa8016e04f891261cc5196 cannam@127: Author: athena cannam@127: Date: Mon Jul 11 14:39:52 2011 -0400 cannam@127: cannam@127: Fix bug in bubblesort cannam@127: cannam@127: Bubblesort was not sorting. This was a bug in the benchmark library cannam@127: (not in FFTW per se), and it impacted the benchmark program cannam@127: with --report-time and --report-mflops causing it to output cannam@127: an incorrect value for the median. (The minimum, maximum, and cannam@127: average value were correct.) Thanks Dima Baksheev of Intel for cannam@127: reporting this bug. cannam@127: cannam@127: commit d33f4f7b648b658d0d232f8561c85fd9b007c105 cannam@127: Author: stevenj cannam@127: Date: Fri Jul 8 13:35:59 2011 -0400 cannam@127: cannam@127: small manual typos cannam@127: cannam@127: commit 047c6636b6cbbdde2d8f4e5a62e26013336fd3e7 cannam@127: Author: athena cannam@127: Date: Wed Jul 6 10:49:40 2011 -0400 cannam@127: cannam@127: Detection of altivec.h requires $ALTIVEC_CFLAGS cannam@127: cannam@127: commit 8cb56c732d86e26edca3ea53e63440756a434031 cannam@127: Author: athena cannam@127: Date: Tue Jul 5 19:58:47 2011 -0400 cannam@127: cannam@127: Introduce fake dependency so that my-getopt.c is recompiled cannam@127: cannam@127: my-getopt.c does not depend on anything, and so it is not rebuilt when cannam@127: reconfiguring for a different ISA (e.g., CC="gcc -m32" vs CC="gcc cannam@127: -m64"). Add a fake dependency on so that the file is cannam@127: recompiled. cannam@127: cannam@127: commit e50fbe175fe6b482d4eabf554a9d923fe1cb727a cannam@127: Author: stevenj cannam@127: Date: Tue Jul 5 18:53:36 2011 -0400 cannam@127: cannam@127: 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 cannam@127: cannam@127: commit 1b13a7673c31c9f98151186ab5ad96952f0c8cc2 cannam@127: Author: stevenj cannam@127: Date: Tue Jul 5 16:04:03 2011 -0400 cannam@127: cannam@127: call omp_set_num_threads in fftw-bench so that the number of OpenMP threads corresponds with the number of FFTW threads cannam@127: cannam@127: commit fcd3d63bce6f23ca8274e739ca83a0fcb8b63a99 cannam@127: Author: stevenj cannam@127: Date: Tue Jul 5 16:03:06 2011 -0400 cannam@127: cannam@127: 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 cannam@127: cannam@127: commit 4c8bae967265bf7b4c9705d6efe87cf7e9151fce cannam@127: Author: stevenj cannam@127: Date: Sat Jul 2 02:21:22 2011 -0400 cannam@127: cannam@127: 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 cannam@127: cannam@127: commit b1741fcc12ccd46d1ce538398c78ca8da98b2448 cannam@127: Author: athena cannam@127: Date: Fri Jul 1 14:35:44 2011 -0400 cannam@127: cannam@127: consistently use the order single, double, long double cannam@127: cannam@127: commit 02d76b0f908814ec69eb9f4edf423e6794d63720 cannam@127: Author: athena cannam@127: Date: Wed Jun 29 17:27:06 2011 -0400 cannam@127: cannam@127: MSVC AVX 64-bit detection does not work, punt for now. cannam@127: cannam@127: commit 5a057b2b67c6eac3cb59c5dd555e1e4093d0ce8e cannam@127: Author: stevenj cannam@127: Date: Wed Jun 29 15:52:27 2011 -0400 cannam@127: cannam@127: fixed typo, added note on transposed flags for r2c/c2r; thanks to Rhys Ulerich for the suggestions cannam@127: [empty commit message] cannam@127: cannam@127: commit 21db43d01a6a55f3bce9bbb3bd01fc968a8fb4ac cannam@127: Author: athena cannam@127: Date: Wed Jun 29 09:41:39 2011 -0400 cannam@127: cannam@127: fixes for compiling with MSVC (untested) cannam@127: cannam@127: commit dcbc5ebfe7dd814f3ef8ee85fb5b2ccb4a3671f0 cannam@127: Author: athena cannam@127: Date: Tue Jun 28 16:48:36 2011 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit f71799bf38f03deaeea50a8b4178757e826854c0 cannam@127: Author: stevenj cannam@127: Date: Mon Jun 27 21:01:56 2011 -0400 cannam@127: cannam@127: rm extraneous line break in HTML output ... I hate texinfo cannam@127: [empty commit message] cannam@127: cannam@127: commit 9ae9c2b534eb9064d8153f235e6d013a4c8b50c1 cannam@127: Author: stevenj cannam@127: Date: Mon Jun 27 00:47:33 2011 -0400 cannam@127: cannam@127: maintainer-clean should delete html directory (otherwwise we keep obsolete HTML files in the dist tarball, sigh) cannam@127: [empty commit message] cannam@127: cannam@127: commit f66d29622c87134e4a790fdab1e25413fac8d33d cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 23:36:32 2011 -0400 cannam@127: cannam@127: update copyright year in manual cannam@127: [empty commit message] cannam@127: cannam@127: commit 3799446cb5d30354dc69a36f07e8bdf87ed5cb34 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 22:52:54 2011 -0400 cannam@127: cannam@127: whoops, don't dist .f03 headers, since those are built by the user's Makefile cannam@127: [empty commit message] cannam@127: cannam@127: commit 8c336f8396e94752233e91433a0e64a72e137599 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 22:43:49 2011 -0400 cannam@127: cannam@127: 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 cannam@127: [empty commit message] cannam@127: cannam@127: commit e32aa9704f9a6e0811638809bbf764dc748116fb cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 21:02:15 2011 -0400 cannam@127: cannam@127: subsubheadings, MPI transpose reference cannam@127: [empty commit message] cannam@127: cannam@127: commit e2759a2102797af24072573371a9d94d4943f1ff cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 20:48:53 2011 -0400 cannam@127: cannam@127: add MPI plan reference cannam@127: [empty commit message] cannam@127: cannam@127: commit 5fd0d86ab1801a09997624a79a56029f76e4c718 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 17:07:21 2011 -0400 cannam@127: cannam@127: portions of MPI reference docs; tweaks to NEWS cannam@127: [empty commit message] cannam@127: cannam@127: commit b280b47c049bc941297a69b5d668a7fdf4a81977 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 26 12:40:43 2011 -0400 cannam@127: cannam@127: use $(CHECK_PL_OPTS) more consistently cannam@127: [empty commit message] cannam@127: cannam@127: commit f78b49c77e0442702bb0bb3b1e52b82795c36358 cannam@127: Author: athena cannam@127: Date: Sun Jun 26 10:04:54 2011 -0400 cannam@127: cannam@127: accept \r\n as well as \n. Grrr... cannam@127: cannam@127: commit 4449361639ee599d5221557c1b7021c85954c2d9 cannam@127: Author: athena cannam@127: Date: Sun Jun 26 09:52:11 2011 -0400 cannam@127: cannam@127: new configure option --with-incoming-stack-boundary=N cannam@127: cannam@127: This option selects CFLAGS to align the stack at all externally-callable cannam@127: functions. This currently comprises api/* and threads/* cannam@127: cannam@127: commit 44191f4b3b5109c9e1befb9a3eefb1f34a1fd63e cannam@127: Author: athena cannam@127: Date: Sun Jun 26 09:51:37 2011 -0400 cannam@127: cannam@127: add -fomit-frame-pointer back cannam@127: cannam@127: Somehow -O3 does not imply -fomit-frame-pointer on ia32 cannam@127: cannam@127: commit 94f1e0517794a91b91b81bc46695d0bcf5d23ca9 cannam@127: Author: athena cannam@127: Date: Sun Jun 26 07:20:27 2011 -0400 cannam@127: cannam@127: Note that removal of mips-ps is temporary. cannam@127: cannam@127: commit 6ec5e833bf16b843f2893e894f786a67721cf647 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 23:15:03 2011 -0400 cannam@127: cannam@127: update copyright year cannam@127: [empty commit message] cannam@127: cannam@127: commit 27117ddc70e191d20cc88be0a2285f454a1409cd cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 21:33:13 2011 -0400 cannam@127: cannam@127: updated NEWS cannam@127: [empty commit message] cannam@127: cannam@127: commit 2e1f81718cf9d9073a65e907c6aecebc1333a4a1 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 20:29:55 2011 -0400 cannam@127: cannam@127: fixes to Fortran interface and docs cannam@127: [empty commit message] cannam@127: cannam@127: commit 32e0027e573cbfcae6c39e535a1a3549f602b97a cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 17:43:31 2011 -0400 cannam@127: cannam@127: initial stab at MPI Fortran docs cannam@127: [empty commit message] cannam@127: cannam@127: commit f1b33feb4669f8b92467a448172c0c0734c5ac48 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 16:43:31 2011 -0400 cannam@127: cannam@127: correct description of what MPI standard says about I/O (I can't believe this crap) cannam@127: [empty commit message] cannam@127: cannam@127: commit 0329701daca6ace6a4167366fda97a86d7d3b6f4 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 15:14:07 2011 -0400 cannam@127: cannam@127: more MPI documentation; mention `fftw_alloc' functions earlier in the manual cannam@127: [empty commit message] cannam@127: cannam@127: commit c4a68ffe9d9332bc0fc9e9db4f41a7c0728ec663 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 25 13:40:19 2011 -0400 cannam@127: cannam@127: clarification about --enable-sse2 cannam@127: [empty commit message] cannam@127: cannam@127: commit cbf6b823ec85dfecab28d071db39d5f92cdcb561 cannam@127: Author: athena cannam@127: Date: Sat Jun 25 13:31:25 2011 -0400 cannam@127: cannam@127: Update mingw build scripts for fftw-3.3 cannam@127: cannam@127: commit 1c97317e9689ac1376ba51c408adde1514475140 cannam@127: Author: athena cannam@127: Date: Sat Jun 25 08:52:13 2011 -0400 cannam@127: cannam@127: Fix typo: EXTRADIST => EXTRA_DIST cannam@127: cannam@127: commit 0ab873d4daf2a047ec04e273f0a8046f8919961c cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 23:52:19 2011 -0400 cannam@127: cannam@127: finished draft "modern fortran" chapter cannam@127: [empty commit message] cannam@127: cannam@127: commit 35f278113d2e2b4532514f40b44fe468b5c7d729 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 20:47:49 2011 -0400 cannam@127: cannam@127: include FFTW_EXTERN prototypes for wrappers, so that they are properly exported to DLLs on Windows (sigh) cannam@127: [empty commit message] cannam@127: cannam@127: commit d2a1f24513b42464d64fa4fea5b97326f4251646 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 16:52:30 2011 -0400 cannam@127: cannam@127: use malloc() instead of alloca() for large buffers cannam@127: cannam@127: The proximate cause for this patch is that OpenBSD/i386 reserves 256KB cannam@127: stack size per thread. We were allocating a buffer of size cannam@127: 128*130*sizeof(fftw_complex) that exceeds the stack. cannam@127: cannam@127: While 128*130*sizeof(fftw_complex) = 260KiB is the worst case for cannam@127: normal configurations, it is a good idea to limit stack allocation cannam@127: just in case. Also, the generic solver might in principle generate cannam@127: unbounded buffers, even though it is normally disabled for n > 137. cannam@127: cannam@127: So, as an added precaution, we now never stack-allocate buffers larger cannam@127: than 64KiB, which ought to be enough for anybody. cannam@127: cannam@127: commit 3b1c71b8e61a7fbfa88589ddf418d494a672ed78 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 16:32:30 2011 -0400 cannam@127: cannam@127: don't imply that AVX is available on Pentium III; note that MIPS Paired Single is currently only in FFTW 3.2.x cannam@127: [empty commit message] cannam@127: cannam@127: commit 02153f462b97b2733b47d298a3fc5cc57b45ba86 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 16:05:27 2011 -0400 cannam@127: cannam@127: silence annoying gfortran warnings cannam@127: [empty commit message] cannam@127: cannam@127: commit dff007f6cc14bdd1ea710466aa3f6fb3a5408c0a cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 14:59:30 2011 -0400 cannam@127: cannam@127: 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 cannam@127: [empty commit message] cannam@127: cannam@127: commit ff330ebfa25ed56b29ea61ea99bf293cec079dde cannam@127: Author: athena cannam@127: Date: Fri Jun 24 15:05:05 2011 -0400 cannam@127: cannam@127: Fix libtool shared version info. cannam@127: cannam@127: FFTW-3.3.x should be a direct drop-in replacement for all FFTW-3.x.y cannam@127: versions. cannam@127: cannam@127: commit 6c3c5cd3040d318a184fc0e6ee6ee4e3429ef8ce cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 14:38:47 2011 -0400 cannam@127: cannam@127: add MPI Fortran API and wrappers cannam@127: [empty commit message] cannam@127: cannam@127: commit 02d3e72585a1254f2685014f5f8de5c7730b8ec5 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 14:51:12 2011 -0400 cannam@127: cannam@127: Do not require fig2dev on the user's machine cannam@127: cannam@127: Distribute the manual's figures in PDF/PS/PNG form instead. cannam@127: cannam@127: commit 5169fc22863b9b6ea7bfbaafc0ca523e82114ee1 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 11:52:44 2011 -0400 cannam@127: cannam@127: Remove --enable-portable-binary, --with-gcc-arch from documentation. cannam@127: cannam@127: commit f6d1274e5c9cc173d07c58df0a3535c9f4767e48 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 11:48:48 2011 -0400 cannam@127: cannam@127: Forget about specifying nonportable CFLAGS. Let the user do it if he wants. cannam@127: cannam@127: commit 74872e79034a2379c537c23a7c29d0b6f43d2437 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 11:48:25 2011 -0400 cannam@127: cannam@127: Add "-avx" to version string when appropriate. cannam@127: cannam@127: commit 2d6800ac3b8070da86d6d825d89fea05fbd44b78 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 10:26:38 2011 -0400 cannam@127: cannam@127: change 3.3-alpha => 3.3-beta1 cannam@127: cannam@127: commit 97a4d17f30831d71b4075eec8b7f4b7c12deedb7 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 09:25:49 2011 -0400 cannam@127: cannam@127: Extend OUR_MALLOC16 to larger alignments cannam@127: cannam@127: Make it work for 32-byte alignment and beyond, as needed by AVX. cannam@127: Rename --with-our-malloc16 to --with-our-malloc. Keep old --with-our-malloc16 cannam@127: flag for compatibility. cannam@127: cannam@127: commit fd31e415cd0731c428daafe791386d79ff34b8ca cannam@127: Author: athena cannam@127: Date: Fri Jun 24 09:19:38 2011 -0400 cannam@127: cannam@127: Fix typo cannam@127: cannam@127: commit 4f8a370b687860b92a93c49fed128218e6fb9f9b cannam@127: Author: athena cannam@127: Date: Fri Jun 24 09:10:26 2011 -0400 cannam@127: cannam@127: One pass over the manual. cannam@127: cannam@127: commit ba838fa07395a4f365eb16aa8ba1bb108f533dd5 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 08:19:03 2011 -0400 cannam@127: cannam@127: eliminate the WITH_ALIGNED_STACK hack cannam@127: cannam@127: This is 2011 and I have no system with incorrect stack alignment. cannam@127: cannam@127: commit 7e32fb649dcd2a78a3b2d216140fc218cb69c334 cannam@127: Author: athena cannam@127: Date: Fri Jun 24 07:49:47 2011 -0400 cannam@127: cannam@127: enable both threaded and unthreaded wisdom in tools/fftw-wisdom cannam@127: cannam@127: commit 7543b3029bce4fc595c5efc7e0d31a71b4ec5cce cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 02:40:04 2011 -0400 cannam@127: cannam@127: clarification cannam@127: [empty commit message] cannam@127: cannam@127: commit 81589ce427090a8aea9f4362a9b7f9ba7e76e111 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 02:24:01 2011 -0400 cannam@127: cannam@127: check for error code in example cannam@127: [empty commit message] cannam@127: cannam@127: commit 3c57716a6525c37f0f485e925c7df8f9819c66fd cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 02:22:18 2011 -0400 cannam@127: cannam@127: cleanup - since NATIVE_MALLOC is always malloc, delete this #define cannam@127: [empty commit message] cannam@127: cannam@127: commit 644b3ee0d54eb80e78a35710d2b8027f3104fbe6 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 02:19:44 2011 -0400 cannam@127: cannam@127: document wisdom string import/export in Fortran cannam@127: [empty commit message] cannam@127: cannam@127: commit bee8d24fa16cd40bba5612938edc915399526c03 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 02:11:40 2011 -0400 cannam@127: cannam@127: 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 cannam@127: [empty commit message] cannam@127: cannam@127: commit 119eb3d276de6b3178c5436fd0cee98a35f0aabf cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 01:35:27 2011 -0400 cannam@127: cannam@127: declaration style cannam@127: [empty commit message] cannam@127: cannam@127: commit 8916c213d4b3121e65af9ee4989ed8ae0fc21db9 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 24 01:25:36 2011 -0400 cannam@127: cannam@127: document wisdom file export/import from Fortran; add export/import_to/from_filename functions for convenience cannam@127: [empty commit message] cannam@127: cannam@127: commit 9fb007e826ee94927e9ff4a9de14c6b80ec06e69 cannam@127: Author: stevenj cannam@127: Date: Thu Jun 23 19:19:43 2011 -0400 cannam@127: cannam@127: more fortran docs cannam@127: [empty commit message] cannam@127: cannam@127: commit 3f0d26b4fd7e8e501fbd2f6ff24337ff0ad3c97c cannam@127: Author: stevenj cannam@127: Date: Thu Jun 23 17:50:30 2011 -0400 cannam@127: cannam@127: enforce 132-character line-length limit that is the default in Fortran cannam@127: [empty commit message] cannam@127: cannam@127: commit 6e69de25872aed3a4eb13523a43bcfdf9a4f731c cannam@127: Author: stevenj cannam@127: Date: Wed Jun 22 23:27:31 2011 -0400 cannam@127: cannam@127: 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 cannam@127: [empty commit message] cannam@127: cannam@127: commit a91a52952fb6d32423f351afdda8de2d04b71e38 cannam@127: Author: athena cannam@127: Date: Thu Jun 23 18:12:10 2011 -0400 cannam@127: cannam@127: Add md5 hash of fftw's configuration to wisdom file cannam@127: cannam@127: People were already confused by threaded vs unthreaded wisdom, and now cannam@127: things will be even worse because we enable/disable AVX codelets at cannam@127: runtime. Accept incoming wisdom only if it was produced by the same cannam@127: configuration (modulo MD5). cannam@127: cannam@127: commit 0de6ca5c7f7720457124d7b7a3ca35153d06f761 cannam@127: Author: athena cannam@127: Date: Thu Jun 23 09:01:27 2011 -0400 cannam@127: cannam@127: distribute fftw3.f03.in cannam@127: cannam@127: commit 6422ed65a6cfb1704c1b428f37a6034baeba212c cannam@127: Author: stevenj cannam@127: Date: Wed Jun 22 22:02:18 2011 -0400 cannam@127: cannam@127: more Fortran documentation cannam@127: [empty commit message] cannam@127: cannam@127: commit 9c1e1d48f0f97b49b62e5ee42e92673964e7d7cf cannam@127: Author: stevenj cannam@127: Date: Wed Jun 22 20:10:39 2011 -0400 cannam@127: cannam@127: correct comment cannam@127: [empty commit message] cannam@127: cannam@127: commit ce8bb23e571efe5595d6d9578a45769e48bad7fe cannam@127: Author: athena cannam@127: Date: Wed Jun 22 20:26:18 2011 -0400 cannam@127: cannam@127: Use "sh FOO.sh" instead of "./FOO.sh" to avoid chmod +x. cannam@127: cannam@127: commit 660905636ebbefe3338d405d4a6fb30b89e07912 cannam@127: Author: stevenj cannam@127: Date: Wed Jun 22 19:19:05 2011 -0400 cannam@127: cannam@127: document fftw_alloc_real/complex ... should we switch to using these in the tutorial examples? cannam@127: [empty commit message] cannam@127: cannam@127: commit ea8c9f7fedc0ca183449b70bee978a4251f552e1 cannam@127: Author: stevenj cannam@127: Date: Wed Jun 22 19:07:49 2011 -0400 cannam@127: cannam@127: whoops, added missing file cannam@127: [empty commit message] cannam@127: cannam@127: commit 6caf08feba78084741e749e022c8c2e523c7ab44 cannam@127: Author: athena cannam@127: Date: Wed Jun 22 18:46:01 2011 -0400 cannam@127: cannam@127: Note addition of AVX. cannam@127: cannam@127: commit a17b6a6a493d19c93fd3a6b8eaacef8174033cd7 cannam@127: Author: athena cannam@127: Date: Wed Jun 22 18:43:43 2011 -0400 cannam@127: cannam@127: In SSE2, AVX: use FMA macros when applicable. cannam@127: cannam@127: Makes it easier to play with fma4 and fma3 when it comes out. cannam@127: cannam@127: commit d4dfffc05a4f9c3f7aa7c2c2ef9fa416f9257a5f cannam@127: Author: stevenj cannam@127: Date: Wed Jun 22 18:16:45 2011 -0400 cannam@127: cannam@127: 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) cannam@127: [empty commit message] cannam@127: cannam@127: commit 10a9a86ddf3d94750c57ca8b20c0fca39cea6541 cannam@127: Author: athena cannam@127: Date: Wed Jun 22 13:10:02 2011 -0400 cannam@127: cannam@127: some cleanup of SSE2 macros cannam@127: cannam@127: commit f301a0adb53b56a2e4fd74ef61cf29d66b745286 cannam@127: Author: athena cannam@127: Date: Wed Jun 22 07:38:18 2011 -0400 cannam@127: cannam@127: don't use -xHost on ICC cannam@127: cannam@127: -xHost with ICC is problematic. On icc-12.0.0, "-mavx -xHost" cannam@127: overrides -mavx with -xHost, generating SSE2 code instead of AVX code. cannam@127: ICC does not seem to support -mtune=host or equivalent non-ABI cannam@127: changing flag. cannam@127: cannam@127: commit 9e7758ff431947863cec44354413c27067f0fda8 cannam@127: Author: athena cannam@127: Date: Tue Jun 21 20:35:36 2011 -0400 cannam@127: cannam@127: Complete AVX implementation for split codelets cannam@127: cannam@127: commit a547e3d5dba38863f6e2c9acb45ffb94351fe3a5 cannam@127: Author: stevenj cannam@127: Date: Tue Jun 21 19:37:14 2011 -0400 cannam@127: cannam@127: whoops, missing altivec conf patches cannam@127: [empty commit message] cannam@127: cannam@127: commit 9dc6263714fa685f384fb61ed08d398b38b26329 cannam@127: Author: stevenj cannam@127: Date: Tue Jun 21 19:12:45 2011 -0400 cannam@127: cannam@127: some BSD ar versions (e.g. on MacOS X) give an error if there are no object files, so we cannot build empty libraries cannam@127: [empty commit message] cannam@127: cannam@127: commit 32dcba1b5d68cf9e33ec3f81e0405f2f909389dd cannam@127: Author: stevenj cannam@127: Date: Tue Jun 21 19:12:12 2011 -0400 cannam@127: cannam@127: re-insertion of Altivec code cannam@127: [empty commit message] cannam@127: cannam@127: commit 931617ee1ba9f9257117ad8e1df38dfe055f9cee cannam@127: Author: athena cannam@127: Date: Tue Jun 21 16:26:09 2011 -0400 cannam@127: cannam@127: Implement faster AVX loads/stores. cannam@127: cannam@127: commit f76d6c1d278b34b2ccac7cff57522cab7ec90864 cannam@127: Author: athena cannam@127: Date: Tue Jun 21 16:03:24 2011 -0400 cannam@127: cannam@127: Initial AVX256/single implementation cannam@127: cannam@127: This should be correct but slow. I need to figure out how to implement cannam@127: noncontiguous loads/stores efficiently. cannam@127: cannam@127: commit 164cc4c8fa9bfdf2b02d9cb9364c8f3f36e420e7 cannam@127: Author: athena cannam@127: Date: Tue Jun 21 14:13:57 2011 -0400 cannam@127: cannam@127: fix AVX alignment cannam@127: cannam@127: commit c277833627164aeb649d187ba0409a3fdca9166d cannam@127: Author: athena cannam@127: Date: Tue Jun 21 14:07:28 2011 -0400 cannam@127: cannam@127: rename avx256d -> avx cannam@127: cannam@127: AVX will work in both double and single precision, like SSE2. cannam@127: cannam@127: commit f97162a135eee43630825fccaa29b735ee284fff cannam@127: Author: athena cannam@127: Date: Tue Jun 21 13:52:20 2011 -0400 cannam@127: cannam@127: remove CODELET_OPTIM cannam@127: cannam@127: In the old 32-bit gcc-3.x days we used to play games with gcc to force cannam@127: it to produce decent code. Now gcc has gotten smarter and it produces cannam@127: indecent code no matter what we do, so it is safe to remove these hacks. cannam@127: cannam@127: commit 57baa1a6761a441c6aa47b0b63503bfd5270d9a9 cannam@127: Author: athena cannam@127: Date: Tue Jun 21 09:57:31 2011 -0400 cannam@127: cannam@127: work around gcc/icc quirks cannam@127: cannam@127: commit deeea8205c4cf4511b41a8a7cd85195b1e4eb7f6 cannam@127: Author: athena cannam@127: Date: Tue Jun 21 09:56:07 2011 -0400 cannam@127: cannam@127: Add remarks in places where we work around gcc quirks cannam@127: cannam@127: commit b357b3a90ba2060915fa03d888f0b2306be1f17a cannam@127: Author: stevenj cannam@127: Date: Mon Jun 20 21:17:59 2011 -0400 cannam@127: cannam@127: remove the libbench directory (which we have kept lingering in the repository for years due to CVS's inability to remove directories) cannam@127: [empty commit message] cannam@127: cannam@127: commit 1ff7bbcc3b0dcfa3bd8f16ec1c672e5671cc0f99 cannam@127: Author: stevenj cannam@127: Date: Mon Jun 20 21:17:14 2011 -0400 cannam@127: cannam@127: update URLs cannam@127: [empty commit message] cannam@127: cannam@127: commit ddcac323fe11f273f8e983836edaec002237a3ca cannam@127: Author: stevenj cannam@127: Date: Mon Jun 20 20:53:31 2011 -0400 cannam@127: cannam@127: whoops, forgot to check in alignment change cannam@127: [empty commit message] cannam@127: cannam@127: commit 5eac5571d61bf5515946dab5750f5db82632cd54 cannam@127: Author: athena cannam@127: Date: Mon Jun 20 20:22:23 2011 -0400 cannam@127: cannam@127: "test X = Y" requires spaces around "=" cannam@127: cannam@127: commit 60d4535475d937207b8fd0c6ebef966d95601154 cannam@127: Author: stevenj cannam@127: Date: Mon Jun 20 19:18:52 2011 -0400 cannam@127: cannam@127: indenting cannam@127: [empty commit message] cannam@127: cannam@127: commit 22cdbb3e3f35d6c289f4c250fbeba696090c6cc9 cannam@127: Author: stevenj cannam@127: Date: Mon Jun 20 18:57:10 2011 -0400 cannam@127: cannam@127: 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) cannam@127: [empty commit message] cannam@127: cannam@127: commit 2d767316e1ba0cf9fd4f5eb3134c6341b2d87a29 cannam@127: Author: athena cannam@127: Date: Mon Jun 20 16:02:07 2011 -0400 cannam@127: cannam@127: Implement AVX autodetection (gcc-only so far) cannam@127: cannam@127: commit 1ed535ea5c0ae847edb64b1696c7c40ea6022fbd cannam@127: Author: athena cannam@127: Date: Mon Jun 20 14:25:54 2011 -0400 cannam@127: cannam@127: Add VZEROUPPER at the end of AVX codelets cannam@127: cannam@127: If the Intel Optimization Manual is to be believed, we need to wave a cannam@127: dead chicken before transitioning from AVX code to SSE code. I am cannam@127: supposed to believe that there is a transition penalty for doing so, cannam@127: unless one uses a magic VZEROUPPER instruction that apparently has cannam@127: zero cost. Whatever. cannam@127: cannam@127: commit 1b26ff69ef0065d12689cd77ae65a7a049a37150 cannam@127: Author: athena cannam@127: Date: Mon Jun 20 10:21:25 2011 -0400 cannam@127: cannam@127: Move RDFT to new simd scheme cannam@127: cannam@127: commit 02b63c9ba5acf94a24d0b948436026df702681a9 cannam@127: Author: athena cannam@127: Date: Mon Jun 20 09:23:38 2011 -0400 cannam@127: cannam@127: New SIMD build system cannam@127: cannam@127: We now support multiple SIMD extensions in the same binary, e.g. cannam@127: --enable-sse2 --enable-avx. This patch adds the necessary cannam@127: infrastructure for SSE2/AVX and complex DFT. Later patches will add cannam@127: RDFT and SSE/ALTIVEC/etc. cannam@127: cannam@127: commit 3409ea120286bc180d314be65f949ecb62f954cb cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 12:29:27 2011 -0400 cannam@127: cannam@127: italicize Latin quote cannam@127: [empty commit message] cannam@127: cannam@127: commit 760f9aec6ca8c45219a357605b8216fef71ff04f cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 12:26:34 2011 -0400 cannam@127: cannam@127: 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 cannam@127: [empty commit message] cannam@127: cannam@127: commit f7a34a1e53ec9e1b7c263d6c8a4cf8020c1de89e cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 12:01:39 2011 -0400 cannam@127: cannam@127: conjugate-pair algorithm turns out not to be due to djb, but it was pointed out to us by djb cannam@127: [empty commit message] cannam@127: cannam@127: commit 34e740878c4a449ef31a6df2f538a67387504b57 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 11:59:21 2011 -0400 cannam@127: cannam@127: new-array execute functions are *not* the same for MPI, since the problems are different cannam@127: [empty commit message] cannam@127: cannam@127: commit a0b07a03fd52e6192dd0023054cc6359dd651554 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 11:57:52 2011 -0400 cannam@127: cannam@127: tweaks to MPI manual cannam@127: [empty commit message] cannam@127: cannam@127: commit 9fa05d4bcd100d5ed0ade53a1c28f00a01cd89fc cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 00:47:23 2011 -0400 cannam@127: cannam@127: only call MPI_Init_thread for MPI version >= 2 cannam@127: [empty commit message] cannam@127: cannam@127: commit e2f3cf7c8965653eb94a03c25ab7ce8f4f09ed8e cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 00:35:44 2011 -0400 cannam@127: cannam@127: document quad precision in Fortran cannam@127: [empty commit message] cannam@127: cannam@127: commit f5dd231509aa0a32aa1d0fa1024adf8ebba30aa9 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 00:16:10 2011 -0400 cannam@127: cannam@127: use -lquadmath for quad-precision library in pkgconfig file cannam@127: [empty commit message] cannam@127: cannam@127: commit d0775d11f358dc46f0f2925be4830b3127564ab4 cannam@127: Author: stevenj cannam@127: Date: Sun Jun 19 00:10:33 2011 -0400 cannam@127: cannam@127: document quad precision cannam@127: [empty commit message] cannam@127: cannam@127: commit a6142f7e745906e6b392369e5ffc5bf83f57d573 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 21:26:24 2011 -0400 cannam@127: cannam@127: there is currently no standard quad-precision type in MPI, so don't allow MPI support to be compiled with --enable-quad-precision cannam@127: [empty commit message] cannam@127: cannam@127: commit 34067999298aa71f65b2ac33386693f03bebd725 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 21:25:51 2011 -0400 cannam@127: cannam@127: make quad-precision library libfftwq cannam@127: [empty commit message] cannam@127: cannam@127: commit 28db5c98edf2ce5508cc8a91118b41ede6476183 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 21:19:50 2011 -0400 cannam@127: cannam@127: use --estimate in check script with --enable-random-estimator cannam@127: [empty commit message] cannam@127: cannam@127: commit e16e119568222dd95a717242b191a4adf72ec2de cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 20:36:51 2011 -0400 cannam@127: cannam@127: bug fix - correct crashing interaction between threads and debug-malloc cannam@127: [empty commit message] cannam@127: cannam@127: commit 25a1b5252eb203919634a13d0c5deb8f8ddad940 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 18:41:48 2011 -0400 cannam@127: cannam@127: whoops, typo cannam@127: [empty commit message] cannam@127: cannam@127: commit 2d8b12a9624e30de0054d035d35474c57d115f49 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 18:35:14 2011 -0400 cannam@127: cannam@127: clarification of when fftw_cost may return 0 cannam@127: [empty commit message] cannam@127: cannam@127: commit 9d7a9545df20eb5b0c4ea6535161108225abce21 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 18:31:26 2011 -0400 cannam@127: cannam@127: corrected manual and test program for proper interaction of MPI and threads cannam@127: [empty commit message] cannam@127: cannam@127: commit a950b94a168ed22d850db5394fd633eee3309ebf cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 17:13:52 2011 -0400 cannam@127: cannam@127: need --mpi restriction in mpi+threads check cannam@127: [empty commit message] cannam@127: cannam@127: commit 7a8106a510a8147ce2f8f986a0ea6f1fc935b419 cannam@127: Author: stevenj cannam@127: Date: Sat Jun 18 12:09:04 2011 -0400 cannam@127: cannam@127: split fftw3.texi into multiple files for ease of editing cannam@127: [empty commit message] cannam@127: cannam@127: commit cb26916dace1d5d264f7561b760a89d5ee972d3b cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 17:26:50 2011 -0400 cannam@127: cannam@127: merge recent Cell deletion with MPI branch cannam@127: [empty commit message] cannam@127: cannam@127: commit 1595e9847b428d6b439d7f908d3d6f06b6746d1f cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 01:52:59 2011 -0400 cannam@127: cannam@127: whoops, incorrect assertion cannam@127: [empty commit message] cannam@127: cannam@127: commit 158a22057fd700263ff39e20dafbf506982bad25 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 01:52:51 2011 -0400 cannam@127: cannam@127: comment fix cannam@127: [empty commit message] cannam@127: cannam@127: commit e0d118305f05ad4f429fda2879547b9285c362ea cannam@127: Author: stevenj cannam@127: Date: Thu Jun 16 23:30:27 2011 -0400 cannam@127: cannam@127: check if pln creation failed (e.g. for split input) bbefore calling setup_gather_scatter, to prevent crashes cannam@127: [empty commit message] cannam@127: cannam@127: commit b2254795d8d3f65d1205053f39d5e837bec6e4d0 cannam@127: Author: stevenj cannam@127: Date: Thu Jun 16 23:26:48 2011 -0400 cannam@127: cannam@127: bug fix -- transpose-recurse is only applicable if subtransposes fit in the same space (unless I change the allocation routine, but this would seem to require looking at all possible recursive invocations of transpose-recurse) cannam@127: [empty commit message] cannam@127: cannam@127: commit 25faa34d32d8b7577ad1107769f3e87e7c87cef0 cannam@127: Author: stevenj cannam@127: Date: Mon Apr 11 17:58:03 2011 -0400 cannam@127: cannam@127: yikes, any_true check on subplan creation should be in comm, not comm2, so that all processes know if failure occurred cannam@127: [empty commit message] cannam@127: cannam@127: commit c0b90d9331fbcc167f07c04a3ce1298cc8d88593 cannam@127: Author: stevenj cannam@127: Date: Mon Apr 11 17:00:46 2011 -0400 cannam@127: cannam@127: add wisdom_ok_hook to enforce wisdom synchronization on MPI problems, apparently fixing a longstanding deadlock/crash bug cannam@127: [empty commit message] cannam@127: cannam@127: commit 651a2f59b8fdeecf79246c3f65b776be567b2ddb cannam@127: Author: stevenj cannam@127: Date: Mon Apr 11 15:10:22 2011 -0400 cannam@127: cannam@127: add a check (in DEBUG mode only) that all processes produce the same hash of MPI problems; don't include alignment in MPI problem hash because it may differ between processes for unaligned malloc cannam@127: [empty commit message] cannam@127: cannam@127: commit 23bb1cb665dc571a773eb5125371b1079e0a4243 cannam@127: Author: stevenj cannam@127: Date: Fri Apr 8 18:46:54 2011 -0400 cannam@127: cannam@127: use cost_hook in random_estimate cannam@127: [empty commit message] cannam@127: cannam@127: commit 21229fc46c88d7ca15c9ba06c1f731d01eecd77a cannam@127: Author: stevenj cannam@127: Date: Sun Mar 6 23:33:53 2011 -0500 cannam@127: cannam@127: added mpi new-array execute functions; thanks to Guo Luo for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit 7335ef11cd5e2e9962dbc41c42e5c14e156e6f82 cannam@127: Author: stevenj cannam@127: Date: Wed Feb 9 21:29:17 2011 -0500 cannam@127: cannam@127: MPI may not support tags > 2^15-1 (e.g. Cray MPI requires tags < 2^24); thanks to Jonathan Bentz for the bug report. cannam@127: [empty commit message] cannam@127: cannam@127: commit 155887d9e173f2a72cb63cf5b7b36ff49bf52356 cannam@127: Author: stevenj cannam@127: Date: Wed Feb 2 12:21:30 2011 -0500 cannam@127: cannam@127: fix merge conflicts cannam@127: [empty commit message] cannam@127: cannam@127: commit 56c3bf01ecfcc23f2d76efed77dac234a59b9234 cannam@127: Author: stevenj cannam@127: Date: Sat Nov 15 20:33:33 2008 -0500 cannam@127: cannam@127: version bump for 3.3alpha1 cannam@127: [empty commit message] cannam@127: cannam@127: commit 136cf63d2cdcb5889a38163c4ccf9b1198e47e04 cannam@127: Author: stevenj cannam@127: Date: Sun Oct 26 22:47:07 2008 -0400 cannam@127: cannam@127: re-added mpi/Makefile cannam@127: [empty commit message] cannam@127: cannam@127: commit 3bfab1285c5390285e734dc910b728a328d9f7fd cannam@127: Author: stevenj cannam@127: Date: Sat Oct 25 17:14:42 2008 -0400 cannam@127: cannam@127: re-add MPI to dist cannam@127: [empty commit message] cannam@127: cannam@127: commit b2470178928b190c6b50d3bef353925061db3d08 cannam@127: Author: athena cannam@127: Date: Sat Jun 18 08:50:13 2011 -0400 cannam@127: cannam@127: remove obsolete Cell code cannam@127: cannam@127: commit 38bfc62ffa5051da294faa46a8ab29fc7225a97b cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 23:31:33 2011 -0400 cannam@127: cannam@127: bug fix in accuracy test, which prevented us from consiistently determining accuracy in > double precision cannam@127: [empty commit message] cannam@127: cannam@127: commit 7356645134bcb66286da0c00ad3d993e90e425af cannam@127: Author: athena cannam@127: Date: Fri Jun 17 20:05:13 2011 -0400 cannam@127: cannam@127: do not check for gcc version before checking for gcc cannam@127: cannam@127: commit d94f234b00d3ced13ccfc3551e0a20cd66645922 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 18:56:37 2011 -0400 cannam@127: cannam@127: require gcc 4.6.0 or later for --enable-quad-precision, to match fftw3.h header file; no need to mark this as EXPERIMENTAL (make check passes, and support in gcc 4.6 seems reasonably complete) cannam@127: [empty commit message] cannam@127: cannam@127: commit 28ebf4da08a1fc607b0ea41ed24ea3707a256548 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 18:51:41 2011 -0400 cannam@127: cannam@127: need ugly __attribute__ to use __float128 with _Complex, ugh cannam@127: [empty commit message] cannam@127: cannam@127: commit b00cf20a5c45b953d2e6d954570543727b72cf9a cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 18:23:05 2011 -0400 cannam@127: cannam@127: --verify tolerance in quad precision changed to 1e-29 cannam@127: [empty commit message] cannam@127: cannam@127: commit 015e365952344e8395a0516c9fe0e3c736550b9e cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 18:22:38 2011 -0400 cannam@127: cannam@127: quad-precision F77 api should use "qfftw" prefix cannam@127: [empty commit message] cannam@127: cannam@127: commit af4a1c37d8eb5e14678449741fd185cb65c4c10b cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 18:22:27 2011 -0400 cannam@127: cannam@127: rm extraneous space from fftw3.h cannam@127: [empty commit message] cannam@127: cannam@127: commit 654e73b0cfe2079498eadaf15b19361f71ef18ab cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 18:05:10 2011 -0400 cannam@127: cannam@127: use cosq etcetera with libquadmath in libbench2, so that --verify correctly gives ~33 decimal places in shift test cannam@127: [empty commit message] cannam@127: cannam@127: commit 00bac1ae1e651e5e85c507294c3e215dde1bb85b cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 17:52:51 2011 -0400 cannam@127: cannam@127: libquadmath ships with gcc 4.6.0, so we should require this library for sinq/cosq with --enable-quad-precision; also, include the __float128 FFTW functions in the header file for gcc >= 4.6 on i86/x86_64/ia64 cannam@127: [empty commit message] cannam@127: cannam@127: commit a17a0720767ca177a799d685a5ac7b63331dd559 cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 16:54:01 2011 -0400 cannam@127: cannam@127: typo in manual for fftw_cost cannam@127: [empty commit message] cannam@127: cannam@127: commit 4aab5d51dbc8b798c25f857cfa07ce7f25f9360d cannam@127: Author: stevenj cannam@127: Date: Fri Jun 17 16:48:24 2011 -0400 cannam@127: cannam@127: fix fftw_cost function: pcost needs to be saved in mkapiplan, since the plan is re-created from wisdom cannam@127: [empty commit message] cannam@127: cannam@127: commit c031d561c14a97b9a04197ba07c19cbf769648f9 cannam@127: Author: athena cannam@127: Date: Fri Jun 17 16:42:25 2011 -0400 cannam@127: cannam@127: removed support for the Cell Broadband Engine cannam@127: cannam@127: commit 34f5ad2eaae86075973d26ee5adcc7ceb04f4924 cannam@127: Author: athena cannam@127: Date: Tue May 24 06:51:07 2011 -0400 cannam@127: cannam@127: Undo previous change; the typo was not a typo after all. cannam@127: cannam@127: commit b9b72d1a42b162f74dfe9dc073ce5f5fc873113b cannam@127: Author: athena cannam@127: Date: Mon May 23 05:08:05 2011 -0400 cannam@127: cannam@127: Fix typo in manual cannam@127: cannam@127: commit f41fb798be412eff5c26c539e07385dbbc6ac56a cannam@127: Author: athena cannam@127: Date: Sat May 21 17:37:50 2011 -0400 cannam@127: cannam@127: clarify intent about canonicalization of tensor in tensor_compress_contiguous() cannam@127: cannam@127: commit 438c3b46f428023ce986422a697c9d33687ae9c1 cannam@127: Author: athena cannam@127: Date: Sat May 21 17:30:31 2011 -0400 cannam@127: cannam@127: avoid useless canonicalization in tensor_compress_contiguous() cannam@127: cannam@127: commit 8fbf77ca71756bd32bb5c71e709ec717e388451e cannam@127: Author: athena cannam@127: Date: Sat May 21 17:24:57 2011 -0400 cannam@127: cannam@127: Fix tensor_compress_contiguous cannam@127: cannam@127: tensor_compress_contiguous() was supposed to sort dimensions by cannam@127: descending istride, and then compress adjacent dimensions. This cannam@127: property was lost once we changed the canonical order of strides to be cannam@127: sorted by descending min{istride,ostride}. cannam@127: cannam@127: Change tensor_compress_contiguous() to sort by descending istride cannam@127: again, which is necessary for its correctness, and then canonicalize cannam@127: at the end. cannam@127: cannam@127: commit 5ad70bf5c1bdeadefcaac86cc3e9a76b31f75724 cannam@127: Author: athena cannam@127: Date: Sun May 8 18:47:26 2011 -0400 cannam@127: cannam@127: Don't distribute obsolete .depend cannam@127: cannam@127: commit c48d28bc5b7fb54ea8e037579cb0af3fae339543 cannam@127: Author: athena cannam@127: Date: Sun May 8 18:05:36 2011 -0400 cannam@127: cannam@127: Use ocamlbuild for building genfft cannam@127: cannam@127: Remove the old Makefile cruft to support ocaml, and use ocamlbuild cannam@127: instead. cannam@127: cannam@127: commit d183b35663d030b1ad789795fa441941961472c0 cannam@127: Author: athena cannam@127: Date: Sun May 8 18:03:07 2011 -0400 cannam@127: cannam@127: Do not use __float128 unless BENCHFFT_QUAD is defined cannam@127: cannam@127: Otherwise, compilation fails on compilers that do not support cannam@127: __float128. cannam@127: cannam@127: commit 833ec738fe3c3234382a3fc160c8fe54060dc860 cannam@127: Author: stevenj cannam@127: Date: Fri Apr 8 13:15:54 2011 -0400 cannam@127: cannam@127: fix configure --help string for --disable-alloca (since default is enabled) cannam@127: [empty commit message] cannam@127: cannam@127: commit 68538e1ffa84d206cae95558c636d9fa490888bd cannam@127: Author: stevenj cannam@127: Date: Fri Apr 8 13:09:56 2011 -0400 cannam@127: cannam@127: add "random estimator" for debugging purposes; note that this is best used with ESTIMATE_PATIENT mode cannam@127: [empty commit message] cannam@127: cannam@127: commit 56d274a97c1868b43a2294bab6a4d6d245849608 cannam@127: Author: stevenj cannam@127: Date: Tue Apr 5 14:47:56 2011 -0400 cannam@127: cannam@127: add AC_CHECK_DECLS for srand48; thanks to Ralf Wildenhues for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit 50465ef2118b72f9f868e9ec53ee7e53bb050259 cannam@127: Author: stevenj cannam@127: Date: Sat Feb 5 17:00:40 2011 -0500 cannam@127: cannam@127: experimental support for gcc's __float128 quad-precision type cannam@127: [empty commit message] cannam@127: cannam@127: commit 2471f34097daef93ae593743403cf40820a0673a cannam@127: Author: athena cannam@127: Date: Sun Oct 24 14:33:59 2010 -0400 cannam@127: cannam@127: guarantee that "timelimit < 0" means "no timeout" cannam@127: cannam@127: "timelimit < 0" was always meant to be equivalent to cannam@127: "timelimit = HUGENUM", but this was not true in all cases, cannam@127: causing some obscure wisdom behavior. cannam@127: cannam@127: Thanks William Andrew Burnson for the bug report. cannam@127: cannam@127: commit c16bc87b770bb0757dead223c13dc1966e3c2e3e cannam@127: Author: athena cannam@127: Date: Sun Oct 24 14:32:20 2010 -0400 cannam@127: cannam@127: compile with --enable-fma and SSE, SSE2 cannam@127: cannam@127: Allow compilation with --enable-fma and --enable-sse, --enable-sse2. cannam@127: This is a bad idea performance-wise, but people will try anyway. cannam@127: cannam@127: commit eb34fdf7b5233e8f8a5b44c7f275e0c950ead3e7 cannam@127: Author: athena cannam@127: Date: Sun Jul 11 13:34:06 2010 -0400 cannam@127: cannam@127: Make threads.c compiler with c++ cannam@127: cannam@127: commit a80ce9ee8210615480efcaf03989278540ad900e cannam@127: Author: athena cannam@127: Date: Sun Jul 11 10:05:05 2010 -0400 cannam@127: cannam@127: Attempt at clarifying the advanced interface doc. cannam@127: cannam@127: commit 537372cf3446b014e53ad2b2bfd636748abfe44f cannam@127: Author: athena cannam@127: Date: Sun Jul 11 07:37:27 2010 -0400 cannam@127: cannam@127: rename rfftwnd html picture cannam@127: cannam@127: It turns out that texinfo with pdf output reads .png cannam@127: files in preference to .pdf files (when did this change?). cannam@127: I renamed the .png figure to avoid producing an ugly pdf file. cannam@127: cannam@127: commit 7be5dbd77b2c719b804b53731ffc9e27100b48f6 cannam@127: Author: stevenj cannam@127: Date: Tue Mar 30 19:43:22 2010 -0400 cannam@127: cannam@127: added fftw_cost function; this is the second time people have asked for this, and there is a reasonable use for it in comparing e.g. oout-of-place vs. in-place plans cannam@127: [empty commit message] cannam@127: cannam@127: commit 101fc17a6dbc1feb1e5cc7625a251068fac7c202 cannam@127: Author: stevenj cannam@127: Date: Tue Mar 2 18:55:49 2010 -0500 cannam@127: cannam@127: documented that --enable-debug-malloc causes fftw_execute to be thread-unsafe (thanks to Alexis Rohou for the problem report) cannam@127: [empty commit message] cannam@127: cannam@127: commit e9b894f0a2b7d65d559d8fded2d7298bdfc90694 cannam@127: Author: athena cannam@127: Date: Fri Jan 22 19:42:08 2010 -0500 cannam@127: cannam@127: Added FAQ about how to transpose matrices using FFTW. cannam@127: [empty commit message] cannam@127: cannam@127: commit 18462a4c21b99482fcb9b2dc7338b83f8bbd790b cannam@127: Author: stevenj cannam@127: Date: Thu Jan 7 20:16:57 2010 -0500 cannam@127: cannam@127: catch FMS (instead of generating FMA(_,_,NEG(_)) with h -generic-arith option cannam@127: [empty commit message] cannam@127: cannam@127: commit 90015026798cd3bad02d8c4144f527dfdc5f1ca7 cannam@127: Author: athena cannam@127: Date: Fri Dec 11 07:01:26 2009 -0500 cannam@127: cannam@127: note future wisdom enhancements. cannam@127: [empty commit message] cannam@127: cannam@127: commit 03747c3de44cda09224d08e8d580483cc23b6687 cannam@127: Author: athena cannam@127: Date: Mon Oct 19 20:21:05 2009 -0400 cannam@127: cannam@127: Use SIMD flags when checking for xmmintrin.h cannam@127: cannam@127: This prevents an obnoxious warning from configure. cannam@127: cannam@127: commit 9211b77226bd67a674d5be4b26843b466a24d377 cannam@127: Author: athena cannam@127: Date: Sat Aug 29 20:47:56 2009 -0400 cannam@127: cannam@127: new bug cannam@127: [empty commit message] cannam@127: cannam@127: commit 4d7817c4b3f6476892515b47aca61d2830ba2e5c cannam@127: Author: stevenj cannam@127: Date: Sun Jul 26 00:40:11 2009 -0400 cannam@127: cannam@127: typo (s/man1/many) cannam@127: [empty commit message] cannam@127: cannam@127: commit 4de43b59f0ef8a445810e2b96e746d95a63b39b1 cannam@127: Author: stevenj cannam@127: Date: Tue Jul 14 14:19:08 2009 -0400 cannam@127: cannam@127: BUILD-MINGW32 script, updated Windows README cannam@127: [empty commit message] cannam@127: cannam@127: commit fe7ce32f22a1cfe7049d095ecbea5024915e93ac cannam@127: Author: fftw cannam@127: Date: Mon Jul 13 09:40:38 2009 -0400 cannam@127: cannam@127: cleanup BUILD-MINGW64.sh cannam@127: cannam@127: commit 18ddd3f4d23697d8f6dcbf9c122ca281ed17bdde cannam@127: Author: athena cannam@127: Date: Sun Jul 12 06:34:46 2009 -0400 cannam@127: cannam@127: Update NEWS, version number for 3.2.2 release. cannam@127: cannam@127: commit 54f4ad016522892f54c2955ecaa5dc06dbff260a cannam@127: Author: athena cannam@127: Date: Sat Jul 11 22:28:38 2009 -0400 cannam@127: cannam@127: Reintroduce the pruning heuristic in ESTIMATE mode for r2r problems. cannam@127: cannam@127: Somehow, we lost this feature between fftw-3.1.3 and fftw-3.2. cannam@127: cannam@127: commit 18e2d0a785627f6360e756d1cd93c78a8f7b6ba3 cannam@127: Author: athena cannam@127: Date: Thu Jun 25 07:39:04 2009 -0400 cannam@127: cannam@127: don't use pshared=1 in sem_init cannam@127: cannam@127: pshared is really not necessary, and it is not supported on cannam@127: GNU/kFreeBSD. Thanks Petr Salinger for the bug report. cannam@127: cannam@127: commit e58f95716b84021e2175673ccbbd4fe1bc180ad3 cannam@127: Author: fftw cannam@127: Date: Thu Jun 11 19:35:40 2009 -0400 cannam@127: cannam@127: Add mingw64 build file so that we can track it. cannam@127: cannam@127: commit 03c7f0e0ade24de0e4d08bc6899f78db917ffaca cannam@127: Author: fftw cannam@127: Date: Wed Jun 10 12:10:58 2009 -0400 cannam@127: cannam@127: note 3.2.2 NEWS cannam@127: cannam@127: commit ddd2281898dca1a30c48cff89c42c2d9c631faa3 cannam@127: Author: fftw cannam@127: Date: Wed Jun 10 12:04:54 2009 -0400 cannam@127: cannam@127: add --disable-alloca to configure cannam@127: cannam@127: It looks like alloca() is broken on mingw64, and thus cannam@127: we need to disable it explicitly. cannam@127: cannam@127: commit 902651afe12f9de0efb131bffa42db8189516595 cannam@127: Author: athena cannam@127: Date: Sun Apr 26 16:33:10 2009 -0400 cannam@127: cannam@127: Note in FAQ that --enable-k7 has been discontinued. cannam@127: cannam@127: commit 0d50e05674406773adea157318c85b8f9c94d9c0 cannam@127: Author: athena cannam@127: Date: Tue Mar 24 09:16:18 2009 -0400 cannam@127: cannam@127: clarified small confusion in fftw_cleanup documentation cannam@127: cannam@127: commit bea8d6909782b15db9d6a591c8344e8179444746 cannam@127: Author: stevenj cannam@127: Date: Thu Mar 19 13:18:06 2009 -0400 cannam@127: cannam@127: fix documentation of dfftw_init_threads to indicate thaat it takes an argument (since the C version returns a value); thanks t Hans Johnnston for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit d6eaf145d2cc51db18fd69b49ef24cf628313c01 cannam@127: Author: fftw cannam@127: Date: Thu Mar 12 13:12:13 2009 -0400 cannam@127: cannam@127: if possible, use a 128-bit type for copy cannam@127: cannam@127: commit b1d732a8e58c303b9f3be4feb082f5b5ac27628c cannam@127: Author: fftw cannam@127: Date: Tue Mar 10 12:49:51 2009 -0400 cannam@127: cannam@127: add size-128 simd codelets cannam@127: cannam@127: It's about time cannam@127: cannam@127: commit d1f4ac081fe3fa267db692b6da4da3b2023e2e8a cannam@127: Author: athena cannam@127: Date: Mon Mar 9 20:29:16 2009 -0400 cannam@127: cannam@127: copy two floats as a double when possible cannam@127: cannam@127: Resurrect the old hack of copying two floats as a double, cannam@127: which makes some difference in these days of 64 bit boxes. cannam@127: cannam@127: commit aa6865bac3cfb4050d6f662ddf744c86c8324393 cannam@127: Author: athena cannam@127: Date: Sun Mar 8 18:08:04 2009 -0400 cannam@127: cannam@127: fixed (harmless) confusion of strides cannam@127: cannam@127: RS and VS were swapped in dftw-direct.c. This is a bug, but cannam@127: it is harmless unless one uses fixed-stride codelets, which we cannam@127: do not. cannam@127: cannam@127: commit f34f1f3fee5d0bd112c8e55c9292d47bd643552f cannam@127: Author: athena cannam@127: Date: Sun Mar 8 10:29:49 2009 -0400 cannam@127: cannam@127: oops, I checked in debug code accidentally. cannam@127: cannam@127: commit d5a07965857012694d310ac28800d47741abccfd cannam@127: Author: athena cannam@127: Date: Sat Feb 14 19:01:00 2009 -0500 cannam@127: cannam@127: Change TLO email address since Magdalen is no longer there. cannam@127: [empty commit message] cannam@127: cannam@127: commit 8418ed8d856c8c50e6463828e015c9d80be1285d cannam@127: Author: stevenj cannam@127: Date: Sat Feb 14 18:18:45 2009 -0500 cannam@127: cannam@127: quote arguments to bench in test script on the off-chance that '*' would be expanded by the shell into a valid filename, and also to avoid shell confusion on Cygwin that "//" begins the name of a Windows network mountpoint cannam@127: [empty commit message] cannam@127: cannam@127: commit 55f747d248139ddffad00ede4d649fc6eb612839 cannam@127: Author: athena cannam@127: Date: Sun Mar 8 10:02:59 2009 -0400 cannam@127: cannam@127: stricter conditions for Cooley-Tukey being ugly cannam@127: cannam@127: It turns out that m=2 in the leaf of Cooley-Tukey may be cannam@127: advantageous in certain cases, eg. i512v512 on AMD Shanghai: cannam@127: cannam@127: (dft-buffered-512-x128/512-6 cannam@127: (dft-ct-dit/4 cannam@127: (dftw-direct-4/24-x128 "t2fv_4") cannam@127: (dft-vrank>=1-x4/1 cannam@127: (dft-ct-dit/64 cannam@127: (dftw-direct-64/504-x128 "t2fv_64") cannam@127: (dft-vrank>=1-x64/1 cannam@127: (dft-direct-2-x128 "n2fv_2"))))) cannam@127: (dft-r2hc-1 cannam@127: (rdft-rank0-tiled/2-x128-x512)) cannam@127: (dft-nop)) cannam@127: cannam@127: Presumably this works around the 2 way associativity of the L1 cache. cannam@127: cannam@127: commit b2acc4c668acebf2ded61cc3a939606bbc73a3e0 cannam@127: Author: stevenj cannam@127: Date: Mon Feb 9 19:46:00 2009 -0500 cannam@127: cannam@127: disable Windows QueryPerformanceCounter code, since it requires us to pull in windows.h in ifftw.h and causes namespace conflicts; gettimeofday seems to work well enough and has had few complaints cannam@127: [empty commit message] cannam@127: cannam@127: commit 1b3884da38a34cbc1c8f33f78394eeca588f0786 cannam@127: Author: stevenj cannam@127: Date: Wed Feb 4 22:55:54 2009 -0500 cannam@127: cannam@127: version bump for 3.2.1, updated NEWS cannam@127: [empty commit message] cannam@127: cannam@127: commit e12e5cb04667aa089cf606fb163ce788247d1c2b cannam@127: Author: stevenj cannam@127: Date: Wed Feb 4 22:27:28 2009 -0500 cannam@127: cannam@127: recommend that users avoid fftw_execute in Fortran, instead using dfftw_execute_dft and friends so that the compiler knows that the input/output arrays are used cannam@127: [empty commit message] cannam@127: cannam@127: commit db43aa9ae291dd66fc542c13afc2dd577750ef75 cannam@127: Author: stevenj cannam@127: Date: Wed Jan 21 16:02:08 2009 -0500 cannam@127: cannam@127: prefer windows queryperformancecounter to gettimeofday on Windows, thanks to David Price for the suggestion cannam@127: [empty commit message] cannam@127: cannam@127: commit 338b5272f6924179a0b345e70f44fd1e5edafc23 cannam@127: Author: athena cannam@127: Date: Sun Feb 1 14:34:49 2009 -0500 cannam@127: cannam@127: compilation fixes in case snprintf() is defined as a macro. cannam@127: [empty commit message] cannam@127: cannam@127: commit 321141568010f66f31c36912a230005ab10d54d3 cannam@127: Author: athena cannam@127: Date: Wed Jan 28 20:19:04 2009 -0500 cannam@127: cannam@127: Automake does not like continuation lines beginning with a comment. cannam@127: [empty commit message] cannam@127: cannam@127: commit 618225923a36a2ef96214e2f2a0c6c12b0fb89be cannam@127: Author: athena cannam@127: Date: Wed Jan 28 18:24:39 2009 -0500 cannam@127: cannam@127: Add r2cb_2.c cannam@127: cannam@127: r2cb_2.c is needed for problem rb2, which is not equivalent to cannam@127: rf2 (unlike kb2, which is equivalent to kf2). cannam@127: cannam@127: This change would not matter much except that rb2 is generated cannam@127: when reducing backward rdft2 to dft, and the absence of the codelet was cannam@127: preventing radix 2 from being employed at all in this case. cannam@127: cannam@127: commit dc4c6cb9e1ae4df2be85e77c3fea172f24e1523b cannam@127: Author: athena cannam@127: Date: Sat Jan 10 06:47:22 2009 -0500 cannam@127: cannam@127: handle the case vecsz->rnk == 0 correctly. cannam@127: [empty commit message] cannam@127: cannam@127: commit 3ca4f694d0b27bb0b1e84ea028e3dadcfdf5b236 cannam@127: Author: stevenj cannam@127: Date: Fri Dec 19 15:20:36 2008 -0500 cannam@127: cannam@127: Macs are no longer ppc-based; thanks to Charles Collicutt for the FAQ update cannam@127: [empty commit message] cannam@127: cannam@127: commit bbfa5e2c5102a5f923eed3b31d37ec7b75616edd cannam@127: Author: stevenj cannam@127: Date: Mon Dec 8 18:08:33 2008 -0500 cannam@127: cannam@127: use new multiple-nbuf code in rdft/buffered, like for dft/buffered cannam@127: [empty commit message] cannam@127: cannam@127: commit 679ab9ffd5738b9df115adfb64f72fd015fe7d6b cannam@127: Author: stevenj cannam@127: Date: Sat Dec 6 16:34:36 2008 -0500 cannam@127: cannam@127: make x86_cpuid macro work on x86_64 cannam@127: [empty commit message] cannam@127: cannam@127: commit 9771718d2e4e57450b03c89bb0c06663c59242c8 cannam@127: Author: athena cannam@127: Date: Sat Dec 6 09:20:37 2008 -0500 cannam@127: cannam@127: Allow automatic choice of buffer size in dft/buffered.c cannam@127: cannam@127: Try a couple of different buffer sizes in buffered transforms, cannam@127: since this seems to make a difference on some Core2 models. cannam@127: cannam@127: commit c4888a4f4fb2961e0f067c55489989da8f3223f5 cannam@127: Author: athena cannam@127: Date: Tue Dec 2 19:18:30 2008 -0500 cannam@127: cannam@127: libbench2: do not assume that split-complex arrays are stride-1 cannam@127: [empty commit message] cannam@127: cannam@127: commit e05f9234129ed82f1f5094465788742c271d1f4b cannam@127: Author: stevenj cannam@127: Date: Tue Dec 2 18:39:43 2008 -0500 cannam@127: cannam@127: updated NEWS cannam@127: [empty commit message] cannam@127: cannam@127: commit 3cfe589fdc4393ff549c1bacbeb2b23c27562339 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 2 18:30:00 2008 -0500 cannam@127: cannam@127: date fix cannam@127: [empty commit message] cannam@127: cannam@127: commit 81326cf16c11ddc12f3d3c1fda7861853abba308 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 2 18:29:06 2008 -0500 cannam@127: cannam@127: updated icc flags -- now prefer -xHost (-xN etc. seem t be obsolete), check for new spelling -ansi-alias, and use -malign-double like we do for gcc cannam@127: [empty commit message] cannam@127: cannam@127: commit 2e94f29d37f25690825b11ed436d726d5077dba6 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 2 18:28:03 2008 -0500 cannam@127: cannam@127: use $ax_cv_c_compiler_vendor rather than $GCC, as the former is more reliable (icc incorrectly self-identifies as gcc on MacOS where we don't use -no-gcc) cannam@127: [empty commit message] cannam@127: cannam@127: commit a5edcdb0c705b3b42f8aec48e41457fd3771bed7 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 2 17:55:36 2008 -0500 cannam@127: cannam@127: don't use -no-gcc for icc on MacOS cannam@127: [empty commit message] cannam@127: cannam@127: commit 2ce94a7fa1858ca3bfdf38f47f7f2d62ceae5262 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 2 17:34:04 2008 -0500 cannam@127: cannam@127: document some more bench options cannam@127: [empty commit message] cannam@127: cannam@127: commit 315a8ae3389d84d6c062a1afc5c2c4eddee4bb77 cannam@127: Author: stevenj cannam@127: Date: Wed Nov 19 16:55:13 2008 -0500 cannam@127: cannam@127: make it clearer that --enable-openmp and --enable-threads are mutually exclusive; thanks to Long To for his comments cannam@127: [empty commit message] cannam@127: cannam@127: commit 4bca25954f1e56299cf45e61bec3877007f8cbc4 cannam@127: Author: stevenj cannam@127: Date: Mon Nov 17 20:16:28 2008 -0500 cannam@127: cannam@127: version bump to 3.2.1, use explicit Makefile.am for m4 subdirectory so that tarball does not include random files in there when you do 'make dist' cannam@127: [empty commit message] cannam@127: cannam@127: commit 7728d69ca6e4f30747b182a3f0d30ec7c25bf26d cannam@127: Author: stevenj cannam@127: Date: Sat Nov 15 21:12:58 2008 -0500 cannam@127: cannam@127: document behavior of FFTW guru arrays, and in particular the odd behavior of the plan_guru_r2r routine in Fortran (thanks to Alexander Pozdneev for the bug report) cannam@127: [empty commit message] cannam@127: cannam@127: commit ba5c08b8b8a3c0e69ba228e071d71664f72c76ba cannam@127: Author: stevenj cannam@127: Date: Mon Nov 10 20:21:32 2008 -0500 cannam@127: cannam@127: version bump to 3.2, updated copyright year cannam@127: [empty commit message] cannam@127: cannam@127: commit a25226c3cd7b9451e6690a813cc3266b15acd7c2 cannam@127: Author: athena cannam@127: Date: Wed Nov 5 16:40:31 2008 -0500 cannam@127: cannam@127: Store GPLv2 in darcs because automake installs GPLv3 these days. cannam@127: [empty commit message] cannam@127: cannam@127: commit f80a90668573e199b3509a7737ba2d071f4e3974 cannam@127: Author: athena cannam@127: Date: Thu Oct 30 15:03:41 2008 -0400 cannam@127: cannam@127: stylistic changes, comments cannam@127: [empty commit message] cannam@127: cannam@127: commit b8f22edf8aadff2aea6d4e69b7651506951810d0 cannam@127: Author: athena cannam@127: Date: Thu Oct 30 14:40:14 2008 -0400 cannam@127: cannam@127: oops cannam@127: [empty commit message] cannam@127: cannam@127: commit bfaec2f123eb8d8297ce405590e07d682cf80293 cannam@127: Author: athena cannam@127: Date: Thu Oct 30 14:30:08 2008 -0400 cannam@127: cannam@127: simplification of the threading machinery cannam@127: [empty commit message] cannam@127: cannam@127: commit c471651b5bc46868c1e8231a89ec6d459c868854 cannam@127: Author: athena cannam@127: Date: Thu Oct 30 14:22:40 2008 -0400 cannam@127: cannam@127: typo cannam@127: [empty commit message] cannam@127: cannam@127: commit e1f64989da3c427c36c9df3495ef9d24ab01993d cannam@127: Author: athena cannam@127: Date: Thu Oct 30 13:42:07 2008 -0400 cannam@127: cannam@127: [SECOND ATTEMPT] do not assume that a semaphore can be freed just because nobody is using it cannam@127: cannam@127: Let S be a semaphore, initially 0. Let thread A execute UP(S); cannam@127: let thread B execute DOWN(S); free(&S); It is unclear whether this cannam@127: code is correct with posix semaphores. The problem is whether UP() cannam@127: uses S after allowing DOWN() to continue; this seems to be the cannam@127: case in the glibc-2.7 implementation, and thus the pattern above cannam@127: seems to be incorrect. Avoid using such a pattern, and introduce cannam@127: a global semaphore for the unavoidable case when nothing else cannam@127: can be depended upon. cannam@127: cannam@127: commit cddb0450696d51a99ca1d9663d4e4a606c45725e cannam@127: Author: stevenj cannam@127: Date: Wed Oct 29 20:09:39 2008 -0400 cannam@127: cannam@127: updated cpu codes from x86-1.21 cannam@127: [empty commit message] cannam@127: cannam@127: commit 2b8ab85fb21f1fd637cfe7a04ad0acf1610b5713 cannam@127: Author: athena cannam@127: Date: Wed Oct 29 16:24:16 2008 -0400 cannam@127: cannam@127: Previous change was bogus, need to find another way. cannam@127: [empty commit message] cannam@127: cannam@127: commit 0123295bb8dd2064d858a59a48242611219b020d cannam@127: Author: athena cannam@127: Date: Wed Oct 29 12:22:20 2008 -0400 cannam@127: cannam@127: do not assume that a semaphore can be freed just because nobody is using it cannam@127: cannam@127: Let S be a semaphore, initially 0. Let thread A execute UP(S); cannam@127: let thread B execute DOWN(S); free(&S); It is unclear whether this cannam@127: code is correct with posix semaphores. The problem is whether UP() cannam@127: uses S after allowing DOWN() to continue; this seems to be the cannam@127: case in the glibc-2.7 implementation, and thus the pattern above cannam@127: seems to be incorrect. Avoid using such a pattern, and introduce cannam@127: a global semaphore for the unavoidable case when nothing else cannam@127: can be depended upon. cannam@127: cannam@127: commit 1b1dd4a34bb653d4bd63775d2760b435cb4f0d1d cannam@127: Author: stevenj cannam@127: Date: Mon Oct 27 23:38:02 2008 -0400 cannam@127: cannam@127: don't need PROG_AS any more cannam@127: [empty commit message] cannam@127: cannam@127: commit f7f67160af9b208e74304378ef54b4b5608fb96a cannam@127: Author: stevenj cannam@127: Date: Sun Oct 26 23:41:11 2008 -0400 cannam@127: cannam@127: use AC_CONFIG_MACRO_DIR macro cannam@127: [empty commit message] cannam@127: cannam@127: commit d6334fa841da6e5f06a7323ce1d31346fb79b4f4 cannam@127: Author: athena cannam@127: Date: Sun Oct 26 10:08:44 2008 -0400 cannam@127: cannam@127: Remove mpi/Makefile from configure.ac cannam@127: Otherwise, the tarball breaks because mpi/ is not in cannam@127: the distribution. cannam@127: cannam@127: commit dda1fd64e3ab5fbb80578cfe41c69191cdfcc7f5 cannam@127: Author: stevenj cannam@127: Date: Sat Oct 25 17:13:50 2008 -0400 cannam@127: cannam@127: remove MPI from dist until FFTW 3.3 cannam@127: [empty commit message] cannam@127: cannam@127: commit a3a16288f18899e7fc8351da1c500024709174cd cannam@127: Author: stevenj cannam@127: Date: Sat Oct 25 17:12:35 2008 -0400 cannam@127: cannam@127: use MPIRUN even for -np 1 cannam@127: [empty commit message] cannam@127: cannam@127: commit 1b2a86c0ab92772ce43bf6b0c5e0bbae2dfb7fee cannam@127: Author: stevenj cannam@127: Date: Fri Jul 18 17:17:08 2008 -0400 cannam@127: cannam@127: use new gcc arch=native flag as fallback cannam@127: [empty commit message] cannam@127: cannam@127: commit 73944e9472d933cdafcff9c3e5b85efc9661ea5b cannam@127: Author: athena cannam@127: Date: Sat Oct 25 13:36:40 2008 -0400 cannam@127: cannam@127: Use sem_t to implement mutexes cannam@127: Use sem_t instead of pthread_mutex_t to implement mutexes. cannam@127: It seems like pthread mutexes hang on linux-2.6.22 after several cannam@127: days of tests; the hang does not occur on linux >= 2.6.24 cannam@127: or when we use sem_t instead of pthread_mutex_t. The cannam@127: situation is still quite mysterious but this code seems to cannam@127: work. cannam@127: cannam@127: commit 82a0159088bdacf0cb5d5ecb5547f51ae34f77ab cannam@127: Author: athena cannam@127: Date: Thu Oct 23 13:32:45 2008 -0400 cannam@127: cannam@127: print informative message when pstring is NULL. cannam@127: [empty commit message] cannam@127: cannam@127: commit ddc681c6a5de3aee386cf990428dcfba95d32369 cannam@127: Author: athena cannam@127: Date: Sun Oct 19 16:00:07 2008 -0400 cannam@127: cannam@127: Fix incorrect alignment in dftw-generic. cannam@127: cannam@127: Multithreaded dftw-generic is supposed to process only a slice cannam@127: of the array, but we were planning with the alignment of the cannam@127: original array rather than the slice. This led to unaligned cannam@127: accesses in certain obscure situations. cannam@127: cannam@127: commit d35ad1d719daf7a8b1c7658b88bf962e86dd050a cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 18 17:27:26 2008 -0400 cannam@127: cannam@127: Paranoia: do not create OS threads while holding locks. cannam@127: cannam@127: Glibc at least plays silly games such as keeping a global variable cannam@127: that records whether there is more than one thread in the process, and cannam@127: it does not perform atomic operations if the variable says that there cannam@127: is only one thread. Who knows how this interacts with creating cannam@127: threads while holding a lock. Some day some genius will come up with cannam@127: some ``optimization'' that breaks everything. cannam@127: cannam@127: commit 9ae439e37bf0cb024de699e15f98b5f5074d116c cannam@127: Author: athena cannam@127: Date: Wed Aug 6 07:41:46 2008 -0400 cannam@127: cannam@127: Welcome to the quadcore era cannam@127: [empty commit message] cannam@127: cannam@127: commit be6af68ab86ada70645a79ef9ac5da11ea787eba cannam@127: Author: stevenj cannam@127: Date: Mon Jun 16 16:46:39 2008 -0400 cannam@127: cannam@127: backslash is technically not allowed in "echo" arguments; thanks to Debian Bug#486046 for pointing out problem and solution (and Raphael Geissert and Vincent Zweije, in particular) cannam@127: [empty commit message] cannam@127: cannam@127: commit fd1ef499dca6f079bd9980cbecd1499e08fe99d0 cannam@127: Author: stevenj cannam@127: Date: Sun May 4 12:15:24 2008 -0400 cannam@127: cannam@127: note problem with test program in gcc 4.1.2-4.2; thanks to Raymond Rogers for reporting it cannam@127: [empty commit message] cannam@127: cannam@127: commit 771f298272494232c994bdca79978f00cbd0a0ac cannam@127: Author: stevenj cannam@127: Date: Fri May 2 19:21:30 2008 -0400 cannam@127: cannam@127: output count of constants along with other statistics cannam@127: [empty commit message] cannam@127: cannam@127: commit 8a8a1bb47539bb8be624af291c28c77cc541ba4e cannam@127: Author: athena cannam@127: Date: Sat Apr 19 14:15:03 2008 -0400 cannam@127: cannam@127: Lower priority of unaligned SIMD codelets. cannam@127: List t1[fb]uv_* codelets before the corresponding cannam@127: aligned codelets, since the estimator picks the cannam@127: latter ones in case of a tie and aligned codelets cannam@127: are preferable. cannam@127: cannam@127: In other words, this is a hack. cannam@127: cannam@127: commit d4d0ed3f7cee7e5c06409b3162fbcf4bfd42fea9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 19 08:55:46 2008 -0400 cannam@127: cannam@127: There is no point in using higher radices for unaligned codelets. cannam@127: [empty commit message] cannam@127: cannam@127: commit c09bbbbc4d2c7236b861b7b2f67b77d4821bb8fc cannam@127: Author: stevenj cannam@127: Date: Fri Apr 18 19:01:27 2008 -0400 cannam@127: cannam@127: support generating loopless, strideless r2r codelets cannam@127: [empty commit message] cannam@127: cannam@127: commit 5d2811a46f654c7dbbade5d2d65921e056c6b3a2 cannam@127: Author: stevenj cannam@127: Date: Fri Apr 18 19:00:25 2008 -0400 cannam@127: cannam@127: added Magic.threemult to use 3+3 complex-multiply variant when possible cannam@127: [empty commit message] cannam@127: cannam@127: commit b21cf57a0c4d5711ea4cdb085b068b366f93c916 cannam@127: Author: stevenj cannam@127: Date: Thu Apr 10 19:53:31 2008 -0400 cannam@127: cannam@127: fix documentation bug - export_wisdom_to_string returns a string that should be deallocated with free, not fftw_free (thanks to Stein Vidar Hagfors Haugan for the bug report) cannam@127: [empty commit message] cannam@127: cannam@127: commit 2a8ac0ba37e9087af75bd0edc9563279424c909d cannam@127: Author: stevenj cannam@127: Date: Mon Jan 21 01:11:44 2008 -0500 cannam@127: cannam@127: bsd calls x86_64 "amd64"; thanks to Fernando Herrero Carron for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit 9d150bc32eea5f0404562a6e26c8f5af7571174c cannam@127: Author: stevenj cannam@127: Date: Tue Jan 1 12:29:56 2008 -0500 cannam@127: cannam@127: fix typo in manual, thanks to Yinon Ehrlich cannam@127: [empty commit message] cannam@127: cannam@127: commit a46838157abb2d1d08cf36d882b6e6979c90a059 cannam@127: Author: stevenj cannam@127: Date: Mon Dec 3 13:57:13 2007 -0500 cannam@127: cannam@127: note problem with gcc 3.4.4 on x86_64, thanks to Uwe Hollerbach for the report cannam@127: [empty commit message] cannam@127: cannam@127: commit 14def93d5b1ba54d64d86d0b9635dfbb41795197 cannam@127: Author: stevenj cannam@127: Date: Tue Nov 13 16:19:22 2007 -0500 cannam@127: cannam@127: bump shared-lib revision cannam@127: [empty commit message] cannam@127: cannam@127: commit ac56042c777020dd5edd04a142c522d0ea3d55d9 cannam@127: Author: stevenj cannam@127: Date: Tue Nov 13 16:16:49 2007 -0500 cannam@127: cannam@127: update NEWS for alpha3 cannam@127: [empty commit message] cannam@127: cannam@127: commit d90bca8d53eaa2ecde6c54123b290ea2cccfeda1 cannam@127: Author: stevenj cannam@127: Date: Tue Oct 2 13:53:04 2007 -0400 cannam@127: cannam@127: fixed URL cannam@127: [empty commit message] cannam@127: cannam@127: commit 2e992067b2172b9ef10c068d5b3b1f5f0e336790 cannam@127: Author: stevenj cannam@127: Date: Mon Sep 17 19:38:29 2007 -0400 cannam@127: cannam@127: added missing prototype cannam@127: [empty commit message] cannam@127: cannam@127: commit a9a3696f1daa50cfcea8e9264912fd7010f59edf cannam@127: Author: stevenj cannam@127: Date: Tue Aug 14 22:35:06 2007 -0400 cannam@127: cannam@127: terminology tweak cannam@127: [empty commit message] cannam@127: cannam@127: commit 7e0c1a370fdec22df30379eb952943a8b9f16080 cannam@127: Author: stevenj cannam@127: Date: Wed Aug 1 18:44:21 2007 -0400 cannam@127: cannam@127: check for pathscale compilers (thanks to Julian Cummings) cannam@127: [empty commit message] cannam@127: cannam@127: commit dabff4a5303cc1bde1a1ac88508f3301caeb6e9f cannam@127: Author: athena cannam@127: Date: Sat Sep 15 18:02:32 2007 -0400 cannam@127: cannam@127: Avoid possible conflict with Windows include files. cannam@127: [empty commit message] cannam@127: cannam@127: commit 1090ecb91cd0da452cca31e8ef926494895a83bc cannam@127: Author: athena cannam@127: Date: Tue Aug 7 21:26:05 2007 -0400 cannam@127: cannam@127: Distribute codlist.c for SIMD codelets in the commercial tarball. cannam@127: [empty commit message] cannam@127: cannam@127: commit 9c132ba2c88c1d5f9e35ad89c8d9e7b012f3741f cannam@127: Author: stevenj cannam@127: Date: Wed Aug 1 10:33:41 2007 -0400 cannam@127: cannam@127: some documentation clarifications, and documented FFTW_WISDOM_ONLY, at the suggestion of Mario Emmenlauer and Phil Dumont cannam@127: [empty commit message] cannam@127: cannam@127: commit 79a73f23fc62044a1edd421d031c35d9d17345dc cannam@127: Author: stevenj cannam@127: Date: Tue Jul 31 16:52:56 2007 -0400 cannam@127: cannam@127: bug fix in test program for vrank-3 transpose plans with vl=1 cannam@127: [empty commit message] cannam@127: cannam@127: commit 13dcde33151a281f5dd4084f3a65277223d444eb cannam@127: Author: stevenj cannam@127: Date: Sun Jul 29 17:02:46 2007 -0400 cannam@127: cannam@127: only run mpi checks for --enable-mpi cannam@127: [empty commit message] cannam@127: cannam@127: commit da920b9d9649c89291980b342a38cd31e689d04c cannam@127: Author: stevenj cannam@127: Date: Sun Jul 29 16:45:30 2007 -0400 cannam@127: cannam@127: check for NULL return from spe_context_create in case SPE_MAP_PS not supported cannam@127: [empty commit message] cannam@127: cannam@127: commit 8b5208ab2d26b33f10864d23ae032a575877cdb3 cannam@127: Author: stevenj cannam@127: Date: Sun Jul 29 15:56:57 2007 -0400 cannam@127: cannam@127: use problem-state pointer to write SPE mailbox with lower latency (makes a significant performance difference for N < 32k), thanks to Jan Wagner for suggestion cannam@127: [empty commit message] cannam@127: cannam@127: commit 35435685af71440fc9601b845163491e61845b4b cannam@127: Author: stevenj cannam@127: Date: Sun Jul 29 14:22:08 2007 -0400 cannam@127: cannam@127: port cell code to SDK2.1 (libspe2), since libspe1 API is deprecated and can't be used in code that also uses libspe2 API cannam@127: [empty commit message] cannam@127: cannam@127: commit 21dc1b9f90f96120a92469077cabfd80dd7fcb70 cannam@127: Author: stevenj cannam@127: Date: Sun Jul 29 11:46:24 2007 -0400 cannam@127: cannam@127: bug fix: ego->W allocated with cell_aligned_malloc, so deallocate with free, not X(ifree0) cannam@127: [empty commit message] cannam@127: cannam@127: commit 7c9f576ed6672631a9d36698f5d9824d173e06ea cannam@127: Author: stevenj cannam@127: Date: Mon Jul 2 15:57:12 2007 -0400 cannam@127: cannam@127: removed obsolete reference to CVS id cannam@127: [empty commit message] cannam@127: cannam@127: commit 2efeadcf8d2b6562d8c18707a0b7eb2e3e5f73d3 cannam@127: Author: athena cannam@127: Date: Mon May 21 14:25:39 2007 -0400 cannam@127: cannam@127: cycle counter for sun compiler cannam@127: [empty commit message] cannam@127: cannam@127: commit 0b59ce4b61ae5c052d9c673807e7b3386d7bdaa2 cannam@127: Author: stevenj cannam@127: Date: Wed May 9 19:49:11 2007 -0400 cannam@127: cannam@127: use __inline instead of inline for AIX routines (__inline is supported by gcc and xlc, whereas apparently "inline" is only supported by xlc if you specify -qlanglvl=stdc99 or similar); thanks to Jeff Haferman for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit a3dc6f8631568cca3039a6932aa4a66e46456a79 cannam@127: Author: stevenj cannam@127: Date: Mon Apr 30 15:37:56 2007 -0400 cannam@127: cannam@127: fixed incorrect type prefix (fftw_ vs. X(...)) in mpi/wisdom-api.c; thanks to Eric A. Borisch for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit 122d2b4a77a11b949e61e503681975f2da3d7d4f cannam@127: Author: stevenj cannam@127: Date: Wed Apr 25 21:21:39 2007 -0400 cannam@127: cannam@127: some cleanups in MPI make check cannam@127: [empty commit message] cannam@127: cannam@127: commit 4d26d141fb5e018b49133b1e080acbff744d97c0 cannam@127: Author: stevenj cannam@127: Date: Wed Apr 25 21:19:27 2007 -0400 cannam@127: cannam@127: re-enable heuristic in the common case where we are not compiling for Cell cannam@127: [empty commit message] cannam@127: cannam@127: commit 28a27bde79e55d51dd1fb81d82ca418587106a62 cannam@127: Author: athena cannam@127: Date: Tue Apr 24 17:42:43 2007 -0400 cannam@127: cannam@127: Removed duplicate codelet names, was breaking linker. cannam@127: [empty commit message] cannam@127: cannam@127: commit 1a4a3dcbbe5a046f07654fa3734a5b3568d51b32 cannam@127: Author: stevenj cannam@127: Date: Tue Apr 24 11:38:16 2007 -0400 cannam@127: cannam@127: added more codelets of sizes 5/10/20/25 to improve speed for round decimal sizes (speed improvements of 10-20%, at cost of 10-30% in library size) cannam@127: [empty commit message] cannam@127: cannam@127: commit 35443ed785a0ce0a05e16d1f6419769f7641f415 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 24 18:40:47 2007 -0400 cannam@127: cannam@127: for 1d prime sizes, punt and return serial plan cannam@127: [empty commit message] cannam@127: cannam@127: commit 3cf27d0073f9f911a2b15283b0d1acebeaf7b599 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 24 18:24:55 2007 -0400 cannam@127: cannam@127: output reminders of the problem during bench --verify cannam@127: [empty commit message] cannam@127: cannam@127: commit e68227acb0ebf9ad3cb0b022382c6df9fcf0d8b1 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 24 18:10:24 2007 -0400 cannam@127: cannam@127: bug fix - missing solver->destroy initializer in rdft2-rdft cannam@127: [empty commit message] cannam@127: cannam@127: commit 59a3e77ab1cba10afa53a627849abef6dd93152d cannam@127: Author: stevenj cannam@127: Date: Fri Mar 23 11:12:19 2007 -0400 cannam@127: cannam@127: -static, in --enable-debug, doesn't work on MacOS X (according to Daniel Oberhoff) cannam@127: [empty commit message] cannam@127: cannam@127: commit 1bcacddfa919627af62568a28b0713b368549612 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 22:23:06 2007 -0400 cannam@127: cannam@127: fix MPI r2c/c2r to work with howmany > 1 cannam@127: [empty commit message] cannam@127: cannam@127: commit 86e99768530258b6184733a382920feae222ae55 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 18:44:41 2007 -0400 cannam@127: cannam@127: rm MPI version from TODO cannam@127: [empty commit message] cannam@127: cannam@127: commit b55ed34cb35d64e452aac41b5661536d75c492d4 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 18:34:40 2007 -0400 cannam@127: cannam@127: added 'make bigcheck' for MPI (no paranoid-check, unfortunately), and properly get MPIRUN from configure cannam@127: [empty commit message] cannam@127: cannam@127: commit e11b28e739bf5b888cfdf0ec97337166fbb6c425 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 18:23:18 2007 -0400 cannam@127: cannam@127: bug fix - incorrect local_size returned for 1d bigvec case cannam@127: [empty commit message] cannam@127: cannam@127: commit 3c4171a56630a623798d71e1a6218c1400ea3e46 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 03:13:54 2007 -0400 cannam@127: cannam@127: hack to specify MPI_TRANSPOSED_IN/OUT via "[" and "]" in libbench2 problem cannam@127: [empty commit message] cannam@127: cannam@127: commit b6643c4d6de6ac41e771a65accc67af6d515009f cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 02:58:11 2007 -0400 cannam@127: cannam@127: added MPI 'make check', still needs a bit of work cannam@127: [empty commit message] cannam@127: cannam@127: commit 5c4ca6bf40ab9683f717ef89a5bbb2c7da031680 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 02:47:10 2007 -0400 cannam@127: cannam@127: bug fix in r2r transposed-input case cannam@127: [empty commit message] cannam@127: cannam@127: commit 518bfe4ddbe9a727866374bb0b0fb49a2d0a9f2b cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 02:46:25 2007 -0400 cannam@127: cannam@127: don't output more than 300 erroneous outputs (unless verbose > 2) cannam@127: [empty commit message] cannam@127: cannam@127: commit 5ae1f03689a0f37d5db6becf54c8e54395541407 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 01:48:54 2007 -0400 cannam@127: cannam@127: fixed bug in transposed-in c2r MPI transforms ... seems to be working, finally cannam@127: [empty commit message] cannam@127: cannam@127: commit 6ff00891c3f7fcfe5399e652b9aeb1538bf9c8d6 cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 00:41:32 2007 -0400 cannam@127: cannam@127: some fixes to MPI r2c/c2r transforms with transposed output/input cannam@127: [empty commit message] cannam@127: cannam@127: commit fc68b9bdd4975fec0d3f9b9ef6a2d9e052f8e7ce cannam@127: Author: stevenj cannam@127: Date: Wed Mar 21 00:40:25 2007 -0400 cannam@127: cannam@127: typos cannam@127: [empty commit message] cannam@127: cannam@127: commit 75dce53511a209c32f881a782af96bf68fdf41af cannam@127: Author: stevenj cannam@127: Date: Tue Mar 20 19:53:02 2007 -0400 cannam@127: cannam@127: bug fix for mpi-bench with r2c/c2r: allocate a little bit extra to make sure that padding is allocated cannam@127: [empty commit message] cannam@127: cannam@127: commit 1720fcb4dc2220e66e50a2baa9201f6c58913bf4 cannam@127: Author: stevenj cannam@127: Date: Tue Mar 20 19:19:13 2007 -0400 cannam@127: cannam@127: fix typo, thanks to Ernest Turro for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit 3a9a95e347c10d98ad7d7ef0c3dca97217e2137a cannam@127: Author: stevenj cannam@127: Date: Tue Mar 20 01:39:06 2007 -0400 cannam@127: cannam@127: spacing tweaks cannam@127: [empty commit message] cannam@127: cannam@127: commit 27cc0f277fb7839e10996eb97c07e4ea9e5bb94c cannam@127: Author: stevenj cannam@127: Date: Tue Mar 20 00:53:11 2007 -0400 cannam@127: cannam@127: Ralf Wildenhues is the one who pointed out that the self-communication could fill in the stalls in the pairwise schedule cannam@127: [empty commit message] cannam@127: cannam@127: commit 3879a995d7146eb6be6ab3df826ebdf2660451f5 cannam@127: Author: stevenj cannam@127: Date: Tue Mar 20 00:22:25 2007 -0400 cannam@127: cannam@127: add TRANSPOSED_OUT/IN support for r2c/c2r, respectively cannam@127: [empty commit message] cannam@127: cannam@127: commit 5a4f8df7a48f55926d1a2017e234903e75fbf35b cannam@127: Author: stevenj cannam@127: Date: Mon Mar 19 21:45:34 2007 -0400 cannam@127: cannam@127: yikes! fixed likely deadlock bug in MPI cannam@127: [empty commit message] cannam@127: cannam@127: commit e5514b08ccfb3c99dfa7034276872af1e3a15b84 cannam@127: Author: stevenj cannam@127: Date: Mon Mar 19 21:38:52 2007 -0400 cannam@127: cannam@127: comment cannam@127: [empty commit message] cannam@127: cannam@127: commit 2ccef4a6915eeebe969120c75c5790791905fd37 cannam@127: Author: stevenj cannam@127: Date: Mon Mar 19 21:30:44 2007 -0400 cannam@127: cannam@127: s/alpha1/alpha2/ cannam@127: [empty commit message] cannam@127: cannam@127: commit 4c069ca435d517243da0ea52594b8101723303b1 cannam@127: Author: stevenj cannam@127: Date: Mon Mar 19 00:39:47 2007 -0400 cannam@127: cannam@127: include README in dist tarball cannam@127: [empty commit message] cannam@127: cannam@127: commit ea9cd7ed69c82e7b129bf88b99dc58238d856c0c cannam@127: Author: stevenj cannam@127: Date: Mon Mar 19 00:35:43 2007 -0400 cannam@127: cannam@127: added MPI r2c/c2r transforms, some more documentation cannam@127: [empty commit message] cannam@127: cannam@127: commit 539fd2ff41bbf5d9955ff83880d983d422e4f4bd cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 23:14:29 2007 -0400 cannam@127: cannam@127: set version to 3.2alpha2 cannam@127: [empty commit message] cannam@127: cannam@127: commit 5b7625b7daebd8aba2e0a876083a944e2c7b520e cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 19:12:18 2007 -0400 cannam@127: cannam@127: changed --enable-mips_ps to --enable-mips-ps; added Cell section to manual (from README.Cell); many minor updates to manual cannam@127: [empty commit message] cannam@127: cannam@127: commit c8cd95869bb81b8d85d87a91e0d65402f9de2288 cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 15:27:06 2007 -0400 cannam@127: cannam@127: whoops, need to sync costs in problem_mpi_rdft cannam@127: [empty commit message] cannam@127: cannam@127: commit dfc055b714f7d4b63b6615bb4b00c86d1600b7de cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 12:44:49 2007 -0400 cannam@127: cannam@127: documented guru64 interface cannam@127: [empty commit message] cannam@127: cannam@127: commit e6a8b5ed239bd9e150f62c8f773e0dcdc97df31a cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 02:57:46 2007 -0400 cannam@127: cannam@127: typo cannam@127: [empty commit message] cannam@127: cannam@127: commit 4625ba2558f8f51201b06cc14102507dd3f2731d cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 02:45:09 2007 -0400 cannam@127: cannam@127: bumped copyright year to 2007 cannam@127: [empty commit message] cannam@127: cannam@127: commit 66392e6b7c997772c49e9c38d275fe79cc25ed33 cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 01:41:40 2007 -0400 cannam@127: cannam@127: noted CodeSourcery in AUTHORS cannam@127: [empty commit message] cannam@127: cannam@127: commit 2c18cc0507bb6ea17abd5d54bacf15bd7ccbca13 cannam@127: Author: stevenj cannam@127: Date: Sun Mar 18 01:25:00 2007 -0400 cannam@127: cannam@127: more MPI documentation cannam@127: [empty commit message] cannam@127: cannam@127: commit 498f7ef52ac60aa1467d76bbfbd0d2224b9ccb10 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 23:15:04 2007 -0400 cannam@127: cannam@127: added MPI multi-dimensional rdft solvers & tests cannam@127: [empty commit message] cannam@127: cannam@127: commit 871ff1554eded2f68e184ecff1d3befd19aa2679 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 22:52:00 2007 -0400 cannam@127: cannam@127: whoops cannam@127: [empty commit message] cannam@127: cannam@127: commit 5a46acce32021f8bcdf12188ca3d764ce7f1cc85 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 22:43:54 2007 -0400 cannam@127: cannam@127: clarification - fftw_mpi_init should be called before importing wisdom cannam@127: [empty commit message] cannam@127: cannam@127: commit 3e2d1704698a4609579e332b904502f5b30370fa cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 19:49:37 2007 -0400 cannam@127: cannam@127: kindx/y/z -> kind0/1/2 for consistency cannam@127: [empty commit message] cannam@127: cannam@127: commit 73c018e5b29c759aaf1012ee39853b025024334c cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 19:34:02 2007 -0400 cannam@127: cannam@127: typo cannam@127: [empty commit message] cannam@127: cannam@127: commit 1d9eeb0231c02f554470a9b6150b07df35e85a4a cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 19:14:16 2007 -0400 cannam@127: cannam@127: some refactoring in preparation for mpi-rdft cannam@127: [empty commit message] cannam@127: cannam@127: commit 3a5f38381e0f65c6da82ab93eefe1be2789c9749 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 18:12:45 2007 -0400 cannam@127: cannam@127: documented more stuff for MPI cannam@127: [empty commit message] cannam@127: cannam@127: commit f833ef7087c898b684d1e0945fb28164e7d5fc02 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 15:41:23 2007 -0400 cannam@127: cannam@127: added NEWS for 3.2alpha cannam@127: [empty commit message] cannam@127: cannam@127: commit 75d1f8189e12f1104a11d92da913592e69b37227 cannam@127: Author: stevenj cannam@127: Date: Sat Mar 17 14:50:22 2007 -0400 cannam@127: cannam@127: documented MPI transpose routines cannam@127: [empty commit message] cannam@127: cannam@127: commit 4fdc9e45f4c6f587f12edb5c0bbe0c60a499d0a9 cannam@127: Author: athena cannam@127: Date: Sat Mar 17 08:57:30 2007 -0400 cannam@127: cannam@127: Removed unused variables cannam@127: [empty commit message] cannam@127: cannam@127: commit a1bd09375e2342d3dec8dbaba75321c278b50861 cannam@127: Author: athena cannam@127: Date: Fri Mar 16 14:47:10 2007 -0400 cannam@127: cannam@127: Preparing for interim release of Cell code. cannam@127: [empty commit message] cannam@127: cannam@127: commit bd1f6de1d0c69ee9a7b7be715797ae2e5cb28ed0 cannam@127: Author: athena cannam@127: Date: Thu Feb 8 12:23:43 2007 -0500 cannam@127: cannam@127: Added README.Cell cannam@127: [empty commit message] cannam@127: cannam@127: commit 4125ae1b42049d9828b51cb9f45398601fa25e5c cannam@127: Author: athena cannam@127: Date: Sat Mar 10 19:17:40 2007 -0500 cannam@127: cannam@127: Synchronized with main branch cannam@127: [empty commit message] cannam@127: cannam@127: commit fa8f1748c92e9255b456b995a9c2d439110fb1b5 cannam@127: Author: athena cannam@127: Date: Mon Jan 22 17:43:56 2007 -0500 cannam@127: cannam@127: Adapted vrecur heuristic to Cell. cannam@127: [empty commit message] cannam@127: cannam@127: commit 31c0788d89e5d2db56d1949df2e61171360ad282 cannam@127: Author: athena cannam@127: Date: Thu Jan 18 20:29:22 2007 -0500 cannam@127: cannam@127: Increased MAX_N to 32K/sizeof(R). cannam@127: [empty commit message] cannam@127: cannam@127: commit e0e08abd04fec6f16a5aa4b8dbec2f614b845bde cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 18 13:43:51 2007 -0500 cannam@127: cannam@127: Added pointer to solver->destroy which is used in the Cell branch. cannam@127: [empty commit message] cannam@127: cannam@127: commit dbb33d0b51897749feff0ef26e63af7769cd4fa5 cannam@127: Author: athena cannam@127: Date: Thu Jan 18 12:09:26 2007 -0500 cannam@127: cannam@127: Updated copyright notices cannam@127: [empty commit message] cannam@127: cannam@127: commit bdb23b08725ce86827dc72f39ace915e594e0ddd cannam@127: Author: athena cannam@127: Date: Fri Jan 12 12:54:43 2007 -0500 cannam@127: cannam@127: Use mfc_read_tag_status_all() instead of spu_mfcstat(2), since the former seems to be standardized. cannam@127: [empty commit message] cannam@127: cannam@127: commit d958d4081d9d131a8c331795b51c38392e038f5f cannam@127: Author: athena cannam@127: Date: Thu Jan 11 14:55:08 2007 -0500 cannam@127: cannam@127: Silence some int/INT warnings. cannam@127: [empty commit message] cannam@127: cannam@127: commit a465f3a820c88855cec17a8b62093f151cf4a75b cannam@127: Author: athena cannam@127: Date: Wed Jan 10 18:19:53 2007 -0500 cannam@127: cannam@127: Note incompatibility of --enable-cell with --enable-threads cannam@127: [empty commit message] cannam@127: cannam@127: commit f8d67adca8de472032a5bb176caef2df069ac301 cannam@127: Author: athena cannam@127: Date: Wed Jan 10 17:57:10 2007 -0500 cannam@127: cannam@127: forgot to add file cannam@127: [empty commit message] cannam@127: cannam@127: commit 2b494f41afa950fa4174d4588070c49879c89acb cannam@127: Author: athena cannam@127: Date: Wed Jan 10 17:45:16 2007 -0500 cannam@127: cannam@127: 64-bit cleanup cannam@127: [empty commit message] cannam@127: cannam@127: commit d4d591f6547fba8ab96d982a76747e0248e94031 cannam@127: Author: athena cannam@127: Date: Wed Jan 10 13:47:20 2007 -0500 cannam@127: cannam@127: Use -mcpu=cell where appropriate. cannam@127: [empty commit message] cannam@127: cannam@127: commit bbff6b92337e5462c8e01c3c6f200f0841422229 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 21:35:59 2006 -0500 cannam@127: cannam@127: synchronized with main cannam@127: [empty commit message] cannam@127: cannam@127: commit 011961c240152fa9ff6c791c430f0104ca2dada7 cannam@127: Author: athena cannam@127: Date: Sun Dec 24 20:58:25 2006 -0500 cannam@127: cannam@127: synchronized with main branch cannam@127: [empty commit message] cannam@127: cannam@127: commit 563b10fee948f8cf7089fc91f97ccd07b92169ae cannam@127: Author: athena cannam@127: Date: Sun Dec 24 13:47:37 2006 -0500 cannam@127: cannam@127: synchronized with main branch, updated to new sdk. cannam@127: [empty commit message] cannam@127: cannam@127: commit 8a7076d3c731fa116d61cc072416ab46beddfc03 cannam@127: Author: athena cannam@127: Date: Thu Dec 21 17:17:41 2006 -0500 cannam@127: cannam@127: removed obsolete file cannam@127: [empty commit message] cannam@127: cannam@127: commit b4139c9ef725c1910b97628667ab1fa9ab98e635 cannam@127: Author: athena cannam@127: Date: Tue Dec 19 15:17:20 2006 -0500 cannam@127: cannam@127: synchronized with main branch cannam@127: [empty commit message] cannam@127: cannam@127: commit 7df6b5623e1db3fe7e9dff26fd4d23f484495649 cannam@127: Author: athena cannam@127: Date: Tue Dec 19 11:27:38 2006 -0500 cannam@127: cannam@127: Synchronized with main branch cannam@127: [empty commit message] cannam@127: cannam@127: commit 6ea2fd75cd1aae82fe1d1a510f104646ab7aa7c9 cannam@127: Author: athena cannam@127: Date: Fri Dec 15 16:04:31 2006 -0500 cannam@127: cannam@127: resolved conflict with main branch cannam@127: [empty commit message] cannam@127: cannam@127: commit 65515a62ffff8fb5649200627cb0c93ef813a9c0 cannam@127: Author: athena cannam@127: Date: Fri Dec 8 14:43:50 2006 -0500 cannam@127: cannam@127: Fixes for compilation in subdirectories cannam@127: [empty commit message] cannam@127: cannam@127: commit d39013569f9cb742b50238ca622f3bd52cade5b5 cannam@127: Author: athena cannam@127: Date: Fri Dec 8 12:46:00 2006 -0500 cannam@127: cannam@127: Silence warning cannam@127: [empty commit message] cannam@127: cannam@127: commit 565f828f3b33be86ac3263046fae490fda62d059 cannam@127: Author: athena cannam@127: Date: Fri Dec 8 12:24:19 2006 -0500 cannam@127: cannam@127: silence warning cannam@127: [empty commit message] cannam@127: cannam@127: commit 5ae3fbed814693f0172abe16b6875b17df64616d cannam@127: Author: athena cannam@127: Date: Thu Dec 7 15:18:17 2006 -0500 cannam@127: cannam@127: Commented a particularly obscure piece of code. cannam@127: [empty commit message] cannam@127: cannam@127: commit b4f78f41fb5960b0ec300d23d6653e80f498eee2 cannam@127: Author: athena cannam@127: Date: Thu Dec 7 11:53:29 2006 -0500 cannam@127: cannam@127: Reorganized, clarified conditions for applicability of the DFT solver. cannam@127: [empty commit message] cannam@127: cannam@127: commit 4d31f1609b233b3951f5f660bc7d2f8f98d7bd38 cannam@127: Author: athena cannam@127: Date: Mon Dec 4 21:33:49 2006 -0500 cannam@127: cannam@127: Minor changes cannam@127: [empty commit message] cannam@127: cannam@127: commit b4cd386c5a2ebd85f8c53be407642199914c26ed cannam@127: Author: athena cannam@127: Date: Mon Dec 4 17:43:28 2006 -0500 cannam@127: cannam@127: Clarified comment cannam@127: [empty commit message] cannam@127: cannam@127: commit aafb3252fb01b21a852ed938b9bc6b2e8b852517 cannam@127: Author: athena cannam@127: Date: Mon Dec 4 16:49:06 2006 -0500 cannam@127: cannam@127: Less incorrect conditions for fitting into local store. cannam@127: [empty commit message] cannam@127: cannam@127: commit 985d9f04682d20ed877a04bfc1d5c0cb73af903c cannam@127: Author: athena cannam@127: Date: Mon Dec 4 16:08:24 2006 -0500 cannam@127: cannam@127: Implemented DECDIF+TRANSPOSE on Cell cannam@127: [empty commit message] cannam@127: cannam@127: commit 00969378eff1f94e8858105bd22015e622da58d0 cannam@127: Author: athena cannam@127: Date: Fri Dec 1 17:42:55 2006 -0500 cannam@127: cannam@127: relaxed conditions of applicability of SPE cannam@127: [empty commit message] cannam@127: cannam@127: commit 4df1c8eb6351f9f6dd2e869a33044d8b36f8dd54 cannam@127: Author: athena cannam@127: Date: Fri Dec 1 16:28:10 2006 -0500 cannam@127: cannam@127: tweaks cannam@127: [empty commit message] cannam@127: cannam@127: commit 92a232ae0edf2fe3f92ca5485861d866aa4c96b0 cannam@127: Author: athena cannam@127: Date: Fri Dec 1 14:35:17 2006 -0500 cannam@127: cannam@127: Implemented Cell opcounts cannam@127: [empty commit message] cannam@127: cannam@127: commit b2e38a76dac60881d4e14e3d9a6ad2b236086a68 cannam@127: Author: athena cannam@127: Date: Fri Dec 1 13:38:44 2006 -0500 cannam@127: cannam@127: minor cleanup cannam@127: [empty commit message] cannam@127: cannam@127: commit 0ab21aa8b736fd4002a92db39449d9e140c39606 cannam@127: Author: athena cannam@127: Date: Fri Dec 1 11:16:52 2006 -0500 cannam@127: cannam@127: use [c0 s0 c1 s1] format for Cell twiddle factors, rather than [c0 c1 s0 s1]. This makes life easier and there is no speed penalty on Cell (unlike Altivec). cannam@127: [empty commit message] cannam@127: cannam@127: commit d080990161fe731d5e1af92f9534c4bd86e06d1f cannam@127: Author: athena cannam@127: Date: Wed Nov 29 18:02:54 2006 -0500 cannam@127: cannam@127: Implemented SPE-accelerated copies cannam@127: [empty commit message] cannam@127: cannam@127: commit 23f9c35ec5d4c123a664c907e5aaca2d9704888c cannam@127: Author: athena cannam@127: Date: Wed Nov 29 12:11:08 2006 -0500 cannam@127: cannam@127: allow SPEs to compute vrank-0 problems. cannam@127: [empty commit message] cannam@127: cannam@127: commit 2f71518adcd364d1cc45272d4f254028c779ff83 cannam@127: Author: athena cannam@127: Date: Tue Nov 28 18:03:07 2006 -0500 cannam@127: cannam@127: eliminated DMA lists cannam@127: [empty commit message] cannam@127: cannam@127: commit 8620a7ab00d1bd4d288513998c82d14cce47b98f cannam@127: Author: athena cannam@127: Date: Tue Nov 28 14:22:05 2006 -0500 cannam@127: cannam@127: Conservatively force all dimensions to be 0 (mod VL) in cell, since otherwise it is too hard to get all cases right. cannam@127: [empty commit message] cannam@127: cannam@127: commit 8343bf57ddaff7d35756635dd9a4a6aa8d31e964 cannam@127: Author: athena cannam@127: Date: Tue Nov 28 12:39:01 2006 -0500 cannam@127: cannam@127: Check alignment of strides when transposing on Cell. cannam@127: [empty commit message] cannam@127: cannam@127: commit 7e28410dc1fbaa1bbf53007869937b84d44c37f3 cannam@127: Author: athena cannam@127: Date: Tue Nov 28 12:19:09 2006 -0500 cannam@127: cannam@127: consistent usage of FFT_SIGN cannam@127: [empty commit message] cannam@127: cannam@127: commit 66491d40310d02c982e54ebee2f9f6c7cb0ae7db cannam@127: Author: athena cannam@127: Date: Tue Nov 28 11:35:38 2006 -0500 cannam@127: cannam@127: clever transposition algorithm without buffering cannam@127: [empty commit message] cannam@127: cannam@127: commit 33e68acd59c892e969a722ded4292aef57fab0a9 cannam@127: Author: athena cannam@127: Date: Mon Nov 27 14:08:28 2006 -0500 cannam@127: cannam@127: Fixed tracking of dependencies cannam@127: [empty commit message] cannam@127: cannam@127: commit 1eab08a3242ae935de04debda8abee056fb3d978 cannam@127: Author: athena cannam@127: Date: Mon Nov 27 14:03:53 2006 -0500 cannam@127: cannam@127: implemented 1D transforms, various tweaks cannam@127: [empty commit message] cannam@127: cannam@127: commit 14891530341581ba7a2422754d83d0c621c71daa cannam@127: Author: athena cannam@127: Date: Wed Nov 22 15:43:36 2006 -0500 cannam@127: cannam@127: no need to poll mailbox on spu side cannam@127: [empty commit message] cannam@127: cannam@127: commit bc0822718ba87089828bd9362b112c8a9bef878c cannam@127: Author: athena cannam@127: Date: Wed Nov 22 14:08:24 2006 -0500 cannam@127: cannam@127: increased maximum size handled by spe cannam@127: [empty commit message] cannam@127: cannam@127: commit 0fb0144374b505502ed768b7f13c191d775bf870 cannam@127: Author: athena cannam@127: Date: Tue Nov 21 16:23:17 2006 -0500 cannam@127: cannam@127: allow vrank<=2 problems in SPEs to avoid the vecloop overhead (grrr...) cannam@127: [empty commit message] cannam@127: cannam@127: commit 26017102cbb70e6e0292087249267b6560414f1c cannam@127: Author: athena cannam@127: Date: Mon Nov 20 14:41:45 2006 -0500 cannam@127: cannam@127: added emacs mode cannam@127: [empty commit message] cannam@127: cannam@127: commit ae2ed8341f9860cd659dc62def4f0c3712e856ab cannam@127: Author: athena cannam@127: Date: Mon Nov 20 09:34:12 2006 -0500 cannam@127: cannam@127: revised transpose, cleanup cannam@127: [empty commit message] cannam@127: cannam@127: commit 7de0b7799021747ff1eef31aca59f7b229750e93 cannam@127: Author: athena cannam@127: Date: Sun Nov 19 20:20:23 2006 -0500 cannam@127: cannam@127: added file cannam@127: [empty commit message] cannam@127: cannam@127: commit 7383afd328c1e98b5ec25c32094a28b0312fb7b0 cannam@127: Author: athena cannam@127: Date: Sun Nov 19 20:18:35 2006 -0500 cannam@127: cannam@127: removed file cannam@127: [empty commit message] cannam@127: cannam@127: commit fa6b1a88cae92cf2e4e5ab247d7a9d93e722c405 cannam@127: Author: athena cannam@127: Date: Sun Nov 19 20:15:38 2006 -0500 cannam@127: cannam@127: better automake integration cannam@127: [empty commit message] cannam@127: cannam@127: commit d1af1e31717e8df8126e7f44197d10e652adbaab cannam@127: Author: athena cannam@127: Date: Sat Nov 18 20:14:29 2006 -0500 cannam@127: cannam@127: changed algorithm for computing chunk size cannam@127: [empty commit message] cannam@127: cannam@127: commit 631ad019b179b1e260a5197ef0012e38e5adfb9c cannam@127: Author: athena cannam@127: Date: Sat Nov 18 19:18:11 2006 -0500 cannam@127: cannam@127: implemented transpose, various fixes. cannam@127: [empty commit message] cannam@127: cannam@127: commit 4d0d4332d36eb952e188eb44f039249a78dc2545 cannam@127: Author: athena cannam@127: Date: Thu Nov 16 16:33:50 2006 -0500 cannam@127: cannam@127: Added explicit destructor to all solvers to help with the cell port. cannam@127: [empty commit message] cannam@127: cannam@127: commit c668de4bc5e8677b2c78830b34214aa832631281 cannam@127: Author: athena cannam@127: Date: Thu Nov 16 15:22:15 2006 -0500 cannam@127: cannam@127: consistent use of #if vs. #ifdef cannam@127: [empty commit message] cannam@127: cannam@127: commit 34192d4eaa06426a1168b5ac743332adb1cd6039 cannam@127: Author: athena cannam@127: Date: Thu Nov 16 15:15:34 2006 -0500 cannam@127: cannam@127: Additional Cell double codelets, better automake integration cannam@127: [empty commit message] cannam@127: cannam@127: commit 4b19e0b192a7f797d21adc5b8b39b126bf809e53 cannam@127: Author: athena cannam@127: Date: Thu Nov 16 12:43:34 2006 -0500 cannam@127: cannam@127: Use dma lists. cannam@127: [empty commit message] cannam@127: cannam@127: commit 398bcee9091aa5c56d753877957f367e7041e6a7 cannam@127: Author: athena cannam@127: Date: Thu Nov 16 11:03:46 2006 -0500 cannam@127: cannam@127: converted to automake cannam@127: [empty commit message] cannam@127: cannam@127: commit caf4303b1448b64d7f82cf9ba36eee8071674421 cannam@127: Author: athena cannam@127: Date: Wed Nov 15 18:00:12 2006 -0500 cannam@127: cannam@127: Initial port to Cell Broadband Engine. cannam@127: [empty commit message] cannam@127: cannam@127: commit 45eebf6ef925ca51e2749ea8658cfb39216fd5b7 cannam@127: Author: athena cannam@127: Date: Wed Mar 14 10:19:53 2007 -0400 cannam@127: cannam@127: Remove Codesourcery contributions from commercial tarball. cannam@127: [empty commit message] cannam@127: cannam@127: commit 71e740a2b810c009c637addb3f87bba3338fa0d2 cannam@127: Author: athena cannam@127: Date: Wed Mar 14 08:59:18 2007 -0400 cannam@127: cannam@127: Added FFTW_WISDOM_ONLY, at the request of Phil Dumont. cannam@127: [empty commit message] cannam@127: cannam@127: commit 8c4485fd3fffb1cfd1aacddfecb58250b5b69607 cannam@127: Author: stevenj cannam@127: Date: Tue Mar 13 00:32:05 2007 -0400 cannam@127: cannam@127: fixed potential MPI deadlock if timer misbehaves cannam@127: [empty commit message] cannam@127: cannam@127: commit 193dbead568fc6582fce99e2b1824f7aac2c66b1 cannam@127: Author: stevenj cannam@127: Date: Mon Mar 12 23:31:52 2007 -0400 cannam@127: cannam@127: more work on MPI documentation cannam@127: [empty commit message] cannam@127: cannam@127: commit 4374a330a301a85267faf67eb71833daeeeefa72 cannam@127: Author: stevenj cannam@127: Date: Tue Feb 27 13:48:43 2007 -0500 cannam@127: cannam@127: index cannam@127: [empty commit message] cannam@127: cannam@127: commit 8dd26fb6008dec917db3ff3f34bbd437b21ba12f cannam@127: Author: stevenj cannam@127: Date: Tue Feb 27 13:46:45 2007 -0500 cannam@127: cannam@127: rename "new-data execute" to "new-array execute", since of course you do not need a new array to have new data cannam@127: [empty commit message] cannam@127: cannam@127: commit 89ebde1693423d225eb9a50b56dc0a5703d30384 cannam@127: Author: stevenj cannam@127: Date: Tue Feb 27 13:43:55 2007 -0500 cannam@127: cannam@127: consistency with manual (guru execute -> new-data execute) cannam@127: [empty commit message] cannam@127: cannam@127: commit 8cce0d9c67bf193b5cf177483ad0500e71a18a84 cannam@127: Author: stevenj cannam@127: Date: Tue Feb 27 13:42:24 2007 -0500 cannam@127: cannam@127: texinfo fixes; renamed "guru execute" section to "new-data execute", since previously it seemed to lead to endless confusion with the guru planner API cannam@127: [empty commit message] cannam@127: cannam@127: commit 7188f6250c91692fb25976542298900e557d092a cannam@127: Author: stevenj cannam@127: Date: Mon Feb 26 18:57:11 2007 -0500 cannam@127: cannam@127: consistently use n0/n1/.. everywhere instead of nx/ny/... (for consistency with d-dimensional case n[0], n[1], ...) ... first start at MPI documentation cannam@127: [empty commit message] cannam@127: cannam@127: commit 385b92bc1fa159e0423f02059cd15c93b7444c92 cannam@127: Author: athena cannam@127: Date: Sat Mar 10 18:48:05 2007 -0500 cannam@127: cannam@127: Changed C++-style comment into K&R cannam@127: [empty commit message] cannam@127: cannam@127: commit 32f8fc24e66030c4e5fdc42b9ec503c50d163435 cannam@127: Author: athena cannam@127: Date: Sat Mar 10 18:47:12 2007 -0500 cannam@127: cannam@127: Forgot to add file cannam@127: [empty commit message] cannam@127: cannam@127: commit acf05fd6f7275e013b16abcfafbc2db7437145f1 cannam@127: Author: athena cannam@127: Date: Sat Mar 10 18:44:39 2007 -0500 cannam@127: cannam@127: Note removal of K7 support. cannam@127: [empty commit message] cannam@127: cannam@127: commit e768b9aeeddd3937eacf72bf4bcd1fe6b67681b8 cannam@127: Author: athena cannam@127: Date: Sat Mar 10 18:41:52 2007 -0500 cannam@127: cannam@127: Updated manual for MIPS PS cannam@127: [empty commit message] cannam@127: cannam@127: commit 56c7d29b2740a24da19b5c022569e60a9bf1abaa cannam@127: Author: athena cannam@127: Date: Sat Mar 10 18:37:07 2007 -0500 cannam@127: cannam@127: Adopted MIPS_PS patches from Codesourcery. cannam@127: [empty commit message] cannam@127: cannam@127: commit 117c18b54c2398c74c00d3f375e60e490cfd0a55 cannam@127: Author: athena cannam@127: Date: Sun Feb 25 11:34:51 2007 -0500 cannam@127: cannam@127: Incorrect initialization of win32 semaphores cannam@127: [empty commit message] cannam@127: cannam@127: commit 835fb99c05fc32b63c000aaa65fa8f098d66d1a1 cannam@127: Author: stevenj cannam@127: Date: Tue Jan 30 11:43:09 2007 -0500 cannam@127: cannam@127: win32 fixes (I think, still untested) cannam@127: [empty commit message] cannam@127: cannam@127: commit 22544bfee3ccdd6810c7f7b9552eb8ec67f58562 cannam@127: Author: stevenj cannam@127: Date: Fri Jan 19 17:31:47 2007 -0500 cannam@127: cannam@127: message-size heuristic in tranpose-recurse cannam@127: [empty commit message] cannam@127: cannam@127: commit c3f9a60853f7d021b8e2e84aca81167fe0742499 cannam@127: Author: athena cannam@127: Date: Tue Jan 30 08:53:55 2007 -0500 cannam@127: cannam@127: Threading layer for Win32, completely untested. cannam@127: [empty commit message] cannam@127: cannam@127: commit e9103c2fa36af2816f233d39aa4b4e6aad4bccd2 cannam@127: Author: athena cannam@127: Date: Mon Jan 29 14:26:30 2007 -0500 cannam@127: cannam@127: Check for EINTR after sem_wait(), as suggested by Chip Salzenberg. cannam@127: [empty commit message] cannam@127: cannam@127: commit 719f223ad7ec385208d4d10171374f44d6dcbfa9 cannam@127: Author: athena cannam@127: Date: Mon Jan 22 13:58:23 2007 -0500 cannam@127: cannam@127: Force vector recursion by means of a separate function pointer. I need this for Cell. cannam@127: [empty commit message] cannam@127: cannam@127: commit 44d62282fe44de7c794ce22ec8a5a3120e71d57f cannam@127: Author: athena cannam@127: Date: Mon Jan 22 09:28:35 2007 -0500 cannam@127: cannam@127: Merge multiplications by twiddle with multiplications by i for faster r2c transforms. cannam@127: [empty commit message] cannam@127: cannam@127: commit 37defea1b213e2cb3e5f73fc481e34551ca72e59 cannam@127: Author: athena cannam@127: Date: Sun Jan 21 19:02:44 2007 -0500 cannam@127: cannam@127: Disabled vector recursion, too messy. cannam@127: [empty commit message] cannam@127: cannam@127: commit f6f7ab5fd044a6ed0b9803c8ea10e176c37137dd cannam@127: Author: athena cannam@127: Date: Sun Jan 21 14:23:35 2007 -0500 cannam@127: cannam@127: Changed heuristics for vector recursion. cannam@127: cannam@127: As in fftw-3.1, NO_VRECURSE disables vector recursion. As an cannam@127: exception, however, vector recursion is allowed when the predicate cannam@127: VRECURSE_ANYWAYP is true. We need some form of vector recursion to cannam@127: obtain decent plans on Cell, and this solution captures the common cannam@127: cases without increasing planning time too much. cannam@127: cannam@127: commit 3612cb7be2fce875627ec720c48a70f9204b42f6 cannam@127: Author: athena cannam@127: Date: Sun Dec 17 22:31:17 2006 -0500 cannam@127: cannam@127: fixed hc2c for vector-recursion branch cannam@127: [empty commit message] cannam@127: cannam@127: commit af9505fcc532b01fb7d7d4e4df0793f1d58bbedd cannam@127: Author: athena cannam@127: Date: Tue Dec 5 12:52:36 2006 -0500 cannam@127: cannam@127: switch to default vector recursion cannam@127: [empty commit message] cannam@127: cannam@127: commit d49ea8d1b8d123219d25c7279a06f0146ff0020b cannam@127: Author: athena cannam@127: Date: Sat Jan 20 23:37:33 2007 -0500 cannam@127: cannam@127: Smarter algorithm for selection of nbuf. cannam@127: [empty commit message] cannam@127: cannam@127: commit 983a3c8a18351c2aa89b096d17419c8ecc8ee4eb cannam@127: Author: athena cannam@127: Date: Sat Jan 20 22:15:33 2007 -0500 cannam@127: cannam@127: Increased buffer sizes according to Moore's law. cannam@127: [empty commit message] cannam@127: cannam@127: commit 3063d37e369e9b607af9a1870c3e9c67966969d3 cannam@127: Author: stevenj cannam@127: Date: Fri Jan 19 16:02:00 2007 -0500 cannam@127: cannam@127: fix another MPI synchronization bug -- several more places where cost_hook must be called to synchronized process timings (sigh) cannam@127: [empty commit message] cannam@127: cannam@127: commit ed26acb97814e71ca8961385f95d136fb532d3e5 cannam@127: Author: athena cannam@127: Date: Fri Jan 19 12:08:07 2007 -0500 cannam@127: cannam@127: Set havewisdom=0 when calling forget_wisdom() in the test program. cannam@127: [empty commit message] cannam@127: cannam@127: commit 9c8fc20de720a1f8588230add1f732504a489797 cannam@127: Author: stevenj cannam@127: Date: Fri Jan 19 10:29:56 2007 -0500 cannam@127: cannam@127: remove redundant check cannam@127: [empty commit message] cannam@127: cannam@127: commit ea709110aaac1eac97acdc9e6d6dccb1a319f491 cannam@127: Author: stevenj cannam@127: Date: Thu Jan 18 22:37:59 2007 -0500 cannam@127: cannam@127: fixed potential (unlikely) bug in wisdom import (triggered when importing impatient wisdom after creating more patient plans, but apparently only for nonstandard configure.c configurations) cannam@127: [empty commit message] cannam@127: cannam@127: commit c30ae9a7d180707e86d8a42ce607c6e7717b49e6 cannam@127: Author: stevenj cannam@127: Date: Thu Jan 18 21:50:14 2007 -0500 cannam@127: cannam@127: added functions to gather/broadcast wisdom for MPI cannam@127: [empty commit message] cannam@127: cannam@127: commit a87ad4116c7bf1ac3e28709b2dc7a3f942beba34 cannam@127: Author: stevenj cannam@127: Date: Thu Jan 11 18:33:17 2007 -0500 cannam@127: cannam@127: whoops, another int/INT bug cannam@127: [empty commit message] cannam@127: cannam@127: commit 51ddf455e30f2f8448b94dc40b8a19a8f296067b cannam@127: Author: stevenj cannam@127: Date: Thu Jan 11 17:42:24 2007 -0500 cannam@127: cannam@127: whoops, fixed bug in transpose-recurse for r != m cannam@127: [empty commit message] cannam@127: cannam@127: commit d2550926efaedd18154f03ae20b464f57ebbc71f cannam@127: Author: stevenj cannam@127: Date: Thu Jan 11 17:25:36 2007 -0500 cannam@127: cannam@127: canonicalize mpi-transposed flags by setting TRANSPOSED_IN/OUT where possible cannam@127: [empty commit message] cannam@127: cannam@127: commit 0129b3159f5ffd78b1d5e8c99a80e5aac5ae1743 cannam@127: Author: stevenj cannam@127: Date: Thu Jan 11 17:16:24 2007 -0500 cannam@127: cannam@127: replace transpose-radix2 with much more general transpose-recurse solver cannam@127: [empty commit message] cannam@127: cannam@127: commit b5399f6884419e5aac9bd45b2f99a55c722dbae6 cannam@127: Author: stevenj cannam@127: Date: Wed Jan 10 20:23:48 2007 -0500 cannam@127: cannam@127: rename transpose-inplace to transpose-pairwise, as the algorithm is not restricted to inplace operation cannam@127: [empty commit message] cannam@127: cannam@127: commit 1db83491ac2308011e874a0e14867ab82285ca87 cannam@127: Author: stevenj cannam@127: Date: Wed Jan 10 14:39:08 2007 -0500 cannam@127: cannam@127: whoops, some int/INT bugs cannam@127: [empty commit message] cannam@127: cannam@127: commit 7c54c7285fedadb55997fa5032a86721a5d73c00 cannam@127: Author: stevenj cannam@127: Date: Tue Jan 9 18:50:07 2007 -0500 cannam@127: cannam@127: fix FAQ Makefile for vpath builds cannam@127: [empty commit message] cannam@127: cannam@127: commit 1f9ce0c767890a637491a26dc6d671cb48d899e1 cannam@127: Author: athena cannam@127: Date: Tue Jan 9 20:22:11 2007 -0500 cannam@127: cannam@127: Missing ``static'' keyword. cannam@127: [empty commit message] cannam@127: cannam@127: commit 5719264a71b3d5a725179d6c6f38fc9844f614c7 cannam@127: Author: athena cannam@127: Date: Tue Jan 9 20:13:18 2007 -0500 cannam@127: cannam@127: Minor cleanup. cannam@127: [empty commit message] cannam@127: cannam@127: commit 31f23769946e8cbd668eae280cf6fa5e0d731cc6 cannam@127: Author: stevenj cannam@127: Date: Tue Jan 9 00:04:03 2007 -0500 cannam@127: cannam@127: interleave twiddle mults with DFTs (should we use dftw?) cannam@127: [empty commit message] cannam@127: cannam@127: commit 5ee274d2bfff3ccfa48faf75d4c5ba4254b6403a cannam@127: Author: stevenj cannam@127: Date: Mon Jan 8 18:35:41 2007 -0500 cannam@127: cannam@127: simplified (and somewhat sped up) dft-rank1 by exploiting dft-rank1-bigvec cannam@127: [empty commit message] cannam@127: cannam@127: commit 6603c476a81bd7d9a84eeec0106ce87ea7af55eb cannam@127: Author: stevenj cannam@127: Date: Sun Jan 7 00:31:31 2007 -0500 cannam@127: cannam@127: rearranged TRANSPOSED format, numerous speedups cannam@127: cannam@127: Split the TRANSPOSED and non-TRANSPOSED rank-geq2 solvers, and changed cannam@127: the DFT TRANSPOSED format to be more like fftw2 (both globally and cannam@127: locally transposed). In general, more emphasis on arranging the data cannam@127: contiguously for the DFTs, and more flexibility in intermediate cannam@127: transposed formats. Also disable NO_SLOW when planning transposes, cannam@127: since otherwise non-square in-place transposes gratuitously put the cannam@127: planner in SLOW mode. cannam@127: cannam@127: Currently, dft-rank1-bigvec has 5 variants (or 10, if DESTROY_INPUT). cannam@127: It looks like only 2 of these are commonly used, so I should probably cannam@127: add some UGLY tags once I do more benchmarking. cannam@127: cannam@127: commit 8efa4e83812fc0d52b20291b0ae6b6d863d873b4 cannam@127: Author: stevenj cannam@127: Date: Thu Jan 4 19:13:17 2007 -0500 cannam@127: cannam@127: add bench_cost_postprocess to prevent deadlocks in mpi-bench cannam@127: [empty commit message] cannam@127: cannam@127: commit ad8fbe7775bfe2a214cefd3759493f11d3330532 cannam@127: Author: stevenj cannam@127: Date: Thu Jan 4 16:46:29 2007 -0500 cannam@127: cannam@127: whoops cannam@127: [empty commit message] cannam@127: cannam@127: commit f1d13c4b532737e65ce9f8cdb058875fed16aac7 cannam@127: Author: stevenj cannam@127: Date: Wed Jan 3 14:23:42 2007 -0500 cannam@127: cannam@127: pass proper pointer types as arguments, so that ACX_PTHREAD still works with C++ and -Werror (thanks to Ewald Arnold for the suggestion) cannam@127: [empty commit message] cannam@127: cannam@127: commit 85662c73ef1053f67e55830adb20a7660c7f546e cannam@127: Author: athena cannam@127: Date: Mon Jan 1 19:30:43 2007 -0500 cannam@127: cannam@127: Renamed [io]vs => [io]vs_by_nbuf, which is more appropriate and would have saved me 30mins debugging. cannam@127: [empty commit message] cannam@127: cannam@127: commit 011d6fa311a3126c66527f22f76a55acababb0f3 cannam@127: Author: stevenj cannam@127: Date: Mon Jan 1 18:52:38 2007 -0500 cannam@127: cannam@127: add --with-g77-wrappers option & always include g77 wrappers on GNU systems and/or with gfortran cannam@127: cannam@127: Upcoming GNU/Linux distros will most likely switch to configuring FFTW cannam@127: with gfortran by default, since g77 isn't even included with recent gcc cannam@127: versions. However, we still want to include g77-compatible wrappers in cannam@127: this case (two underscores) in addition to gfortran wrappers (one cannam@127: underscore) lest we silently break binary compatibility and provoke cannam@127: lots of annoying emails. cannam@127: cannam@127: commit fbb0f99f47d4c09c87cd81573b3532809b44fee1 cannam@127: Author: stevenj cannam@127: Date: Mon Jan 1 16:48:36 2007 -0500 cannam@127: cannam@127: use AC_HELP_STRING for --disable-fortran cannam@127: [empty commit message] cannam@127: cannam@127: commit 4c6880164b7e43be156bd10825038bc5fe83b9b3 cannam@127: Author: stevenj cannam@127: Date: Mon Jan 1 15:56:12 2007 -0500 cannam@127: cannam@127: terminology cannam@127: [empty commit message] cannam@127: cannam@127: commit ae75dc0d2a5fb8286ebadc5fa70a1ff2e17ed7ba cannam@127: Author: athena cannam@127: Date: Sat Dec 30 16:18:35 2006 -0500 cannam@127: cannam@127: Free buffers before calling cldrest. cannam@127: [empty commit message] cannam@127: cannam@127: commit 053f45629c9af2bc608086640e6684ef1e76bf0b cannam@127: Author: athena cannam@127: Date: Fri Dec 29 10:52:15 2006 -0500 cannam@127: cannam@127: Removed obsolete code. cannam@127: [empty commit message] cannam@127: cannam@127: commit e6ffd09841ff145db9112e4fca774cc3454e1170 cannam@127: Author: athena cannam@127: Date: Thu Dec 28 21:37:48 2006 -0500 cannam@127: cannam@127: Attempt to work-around old gcc bugs in a more efficient fashion that does not lose performance on newer gcc's. cannam@127: [empty commit message] cannam@127: cannam@127: commit b1d16645f02bf03f9934c6acfe86d41705734cf6 cannam@127: Author: athena cannam@127: Date: Thu Dec 28 16:10:33 2006 -0500 cannam@127: cannam@127: Make sure that the speed() input is zero even in paranoid mode. cannam@127: [empty commit message] cannam@127: cannam@127: commit bd281a3248526dd660f3cc5db5662a38af6aca70 cannam@127: Author: athena cannam@127: Date: Thu Dec 28 11:41:46 2006 -0500 cannam@127: cannam@127: cld0 and cldm problems must be tainted because they are used in a v-loop. cannam@127: [empty commit message] cannam@127: cannam@127: commit 762203e35e50a636ebbe34f1bb4a9a72dbdfceae cannam@127: Author: athena cannam@127: Date: Wed Dec 27 17:17:45 2006 -0500 cannam@127: cannam@127: Run paranoid-check in patient mode. cannam@127: [empty commit message] cannam@127: cannam@127: commit 13fcf5a8a0073c72a967a6b5c21009dc09b0e63e cannam@127: Author: athena cannam@127: Date: Wed Dec 27 10:51:42 2006 -0500 cannam@127: cannam@127: Fixed incorrect initialization to zero. cannam@127: [empty commit message] cannam@127: cannam@127: commit 957a6a68a80da76a90adfd5c50e6570a0102a174 cannam@127: Author: athena cannam@127: Date: Wed Dec 27 09:33:02 2006 -0500 cannam@127: cannam@127: Fixed wrong TAINT() cannam@127: [empty commit message] cannam@127: cannam@127: commit 131e00d9d05b13400d93ba18bddd02ee53db276a cannam@127: Author: athena cannam@127: Date: Tue Dec 26 22:50:28 2006 -0500 cannam@127: cannam@127: Grrrrr... cannam@127: [empty commit message] cannam@127: cannam@127: commit e30b1ccf6d58829fdae52ee072c605802f728761 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 22:48:44 2006 -0500 cannam@127: cannam@127: Give up trying to verify rdft2 when vrank=-infinity. cannam@127: [empty commit message] cannam@127: cannam@127: commit 50b70cffda5f011216fd43162aa1b6b69f7ef912 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 22:31:38 2006 -0500 cannam@127: cannam@127: typo cannam@127: [empty commit message] cannam@127: cannam@127: commit 4cdf2b4a16da17795eadebf1a1215e1060f1c573 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 21:54:53 2006 -0500 cannam@127: cannam@127: Correctly verify rdft2 when vrank = -infinity. cannam@127: [empty commit message] cannam@127: cannam@127: commit a93befe83478f18c088fc474973185660c237154 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 21:25:02 2006 -0500 cannam@127: cannam@127: rdft/buffered2.c now generates rdft2 subproblems, not rdft. cannam@127: cannam@127: The old rdft2->rdft reduction is now in rdft/rdft2-rdft.c cannam@127: and still does way too much. cannam@127: cannam@127: commit 994d04b97146dcfb849bc7d83136402cb1a0a070 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 14:03:27 2006 -0500 cannam@127: cannam@127: Buffer the input in hc2r problems, as opposed to the output. cannam@127: [empty commit message] cannam@127: cannam@127: commit 47f2f5a1335a6cc49ed95c73655fa08a19958606 cannam@127: Author: athena cannam@127: Date: Tue Dec 26 10:02:59 2006 -0500 cannam@127: cannam@127: streamlined buffered solvers cannam@127: [empty commit message] cannam@127: cannam@127: commit add79ce24b3c20348a098ca15c7431ce95835a54 cannam@127: Author: athena cannam@127: Date: Mon Dec 25 16:08:22 2006 -0500 cannam@127: cannam@127: c++ compatibility cannam@127: [empty commit message] cannam@127: cannam@127: commit 7caedf2e6196972d2a5c4745ff5badc6856c1b29 cannam@127: Author: athena cannam@127: Date: Sun Dec 24 20:27:23 2006 -0500 cannam@127: cannam@127: Gratuitous renaming of directories and files since the old naming was becoming too inconsistent for my taste. cannam@127: [empty commit message] cannam@127: cannam@127: commit ab5397b31582ea9372345a7868a165a340a8aefb cannam@127: Author: athena cannam@127: Date: Sun Dec 24 20:11:50 2006 -0500 cannam@127: cannam@127: Fixed another dftw bug (sigh) cannam@127: [empty commit message] cannam@127: cannam@127: commit aa1c7cfcbfb0b6bedd716d02d2b84ea86432f03b cannam@127: Author: athena cannam@127: Date: Sun Dec 24 11:48:01 2006 -0500 cannam@127: cannam@127: Removed debugging leftovers. cannam@127: [empty commit message] cannam@127: cannam@127: commit 401d4cd84bec4348694725e268ccb78c5cbbb71c cannam@127: Author: athena cannam@127: Date: Sun Dec 24 11:34:17 2006 -0500 cannam@127: cannam@127: Moved dftw-generic* to new dftw protocol. cannam@127: [empty commit message] cannam@127: cannam@127: commit 91d2ba6242b230cf1195cf9c8e7d61f2ee226a28 cannam@127: Author: athena cannam@127: Date: Sun Dec 24 09:37:19 2006 -0500 cannam@127: cannam@127: Oops cannam@127: [empty commit message] cannam@127: cannam@127: commit ef28c24515ebda3edb9fcb0e67682b3e608f7e67 cannam@127: Author: athena cannam@127: Date: Sun Dec 24 09:31:46 2006 -0500 cannam@127: cannam@127: Fixed wrong verification of rank-1 rdft2 cannam@127: [empty commit message] cannam@127: cannam@127: commit a0f01cb9aaaa491d23686acf3ccedc38c47cd8bb cannam@127: Author: athena cannam@127: Date: Sat Dec 23 20:11:29 2006 -0500 cannam@127: cannam@127: minor tweaks cannam@127: [empty commit message] cannam@127: cannam@127: commit de3d507c8f454e02a382e0cbf7d9453f04621021 cannam@127: Author: athena cannam@127: Date: Sat Dec 23 19:50:03 2006 -0500 cannam@127: cannam@127: Removed obsolete items. cannam@127: [empty commit message] cannam@127: cannam@127: commit 67ca9cb9258b4fb320d26445040b6fae0e450594 cannam@127: Author: athena cannam@127: Date: Sat Dec 23 17:56:37 2006 -0500 cannam@127: cannam@127: Modified the problem_dftw invocation protocol. cannam@127: cannam@127: apply() now requires pointers to the beginning of the full array. cannam@127: Each thread processes a slice mb <= m < me. This protocol is cannam@127: consistent with the one used in hc2hc, where there is no other choice. cannam@127: cannam@127: commit 378686a490f47d469eedfb5383cb46f6500cf835 cannam@127: Author: athena cannam@127: Date: Sat Dec 23 16:18:25 2006 -0500 cannam@127: cannam@127: typo cannam@127: [empty commit message] cannam@127: cannam@127: commit a15a3176f43a12770d66407d77b3fc138f278a53 cannam@127: Author: athena cannam@127: Date: Sat Dec 23 16:06:56 2006 -0500 cannam@127: cannam@127: changed hc2hc twiddle storage to be the same as hc2c cannam@127: [empty commit message] cannam@127: cannam@127: commit a4d048b7080396fae83e41bd64c4740ba0ab9f7a cannam@127: Author: athena cannam@127: Date: Sat Dec 23 15:16:36 2006 -0500 cannam@127: cannam@127: Allowed extra_iter in dftw-direct. Rationalized twiddle factors in hc2c. cannam@127: [empty commit message] cannam@127: cannam@127: commit 4c0bf02653b7e58b218d47e0cf01d719edc1d015 cannam@127: Author: athena cannam@127: Date: Sat Dec 23 10:37:11 2006 -0500 cannam@127: cannam@127: Implemented unmentionable hack to use 4-way SIMD with an odd number of cannam@127: iterations. cannam@127: cannam@127: commit 992d3ce4a54640d5af4d942ef17eb880f56ec36e cannam@127: Author: athena cannam@127: Date: Fri Dec 22 22:13:30 2006 -0500 cannam@127: cannam@127: altivec support for new codelets cannam@127: [empty commit message] cannam@127: cannam@127: commit e0908cfd6fe22ae0544576667bec649b71dcb922 cannam@127: Author: athena cannam@127: Date: Fri Dec 22 19:09:15 2006 -0500 cannam@127: cannam@127: fixed incorrect computation of W cannam@127: [empty commit message] cannam@127: cannam@127: commit df8bd57748278b92e0975a66062bd71bb5ac2e8d cannam@127: Author: athena cannam@127: Date: Fri Dec 22 18:51:22 2006 -0500 cannam@127: cannam@127: Implemented 4-way simd hc2cdftv cannam@127: cannam@127: Also eliminated the twiddle_shift hack. A zillion changes dictated cannam@127: by this choice, which was in turn necessary for the hc2cdftv thing cannam@127: to work. cannam@127: cannam@127: commit 7bac8d1f715f737bfed8742521fe60d5dec6b963 cannam@127: Author: athena cannam@127: Date: Fri Dec 22 08:45:46 2006 -0500 cannam@127: cannam@127: Fixed verification of rdft2 problems with new format. cannam@127: [empty commit message] cannam@127: cannam@127: commit c18f29a56027a5e08aa164530d10ff55c1950170 cannam@127: Author: athena cannam@127: Date: Fri Dec 22 00:05:59 2006 -0500 cannam@127: cannam@127: Added file cannam@127: [empty commit message] cannam@127: cannam@127: commit aafef0ef88f37d8b0e63a31afec168dad67a29f4 cannam@127: Author: athena cannam@127: Date: Fri Dec 22 00:02:50 2006 -0500 cannam@127: cannam@127: Hmm, previous commit did not work cannam@127: [empty commit message] cannam@127: cannam@127: commit 21c810018e7f4993ebadf9a05682f3bd0a6d2c8f cannam@127: Author: athena cannam@127: Date: Thu Dec 21 23:58:33 2006 -0500 cannam@127: cannam@127: Added SIMD r2cdft codelets. cannam@127: [empty commit message] cannam@127: cannam@127: commit fe2f5075f1443d522b445b31027cacb32e8add18 cannam@127: Author: athena cannam@127: Date: Thu Dec 21 21:19:21 2006 -0500 cannam@127: cannam@127: Bug in buffering, grrr... cannam@127: [empty commit message] cannam@127: cannam@127: commit 32f34cf494321ef860f20924b84df527d63e0ce0 cannam@127: Author: athena cannam@127: Date: Thu Dec 21 20:58:14 2006 -0500 cannam@127: cannam@127: Oops, memory leak. cannam@127: [empty commit message] cannam@127: cannam@127: commit 179cfb2f9ddf45916458e2dfdd0e0adbcf762044 cannam@127: Author: athena cannam@127: Date: Thu Dec 21 17:12:31 2006 -0500 cannam@127: cannam@127: minor changes, cleanup. cannam@127: [empty commit message] cannam@127: cannam@127: commit 2505062c2e9ded3822b01c123e82033d37968917 cannam@127: Author: athena cannam@127: Date: Wed Dec 20 22:09:28 2006 -0500 cannam@127: cannam@127: Unified hc2hc-direct, hc2hc-directbuf. Cleanup. cannam@127: [empty commit message] cannam@127: cannam@127: commit 5f80c36202acc746148c007e394a4260457f1f60 cannam@127: Author: athena cannam@127: Date: Wed Dec 20 17:55:56 2006 -0500 cannam@127: cannam@127: removed obsolete rdft2-radix2 cannam@127: [empty commit message] cannam@127: cannam@127: commit 18d4fddebb71049478d41152af043e33ed90d014 cannam@127: Author: athena cannam@127: Date: Wed Dec 20 17:51:20 2006 -0500 cannam@127: cannam@127: implemented reduction rdft2->dft cannam@127: [empty commit message] cannam@127: cannam@127: commit f84c7e67e6c77dc3b8ef4c0703277aa884852ab0 cannam@127: Author: athena cannam@127: Date: Wed Dec 20 09:29:39 2006 -0500 cannam@127: cannam@127: Implemented buffered direct-r2c, direct-hc2c. cannam@127: Also, removed some old cruft: cannam@127: cannam@127: * okp() functions were never used and a pain to maintain---now they cannam@127: are gone. cannam@127: cannam@127: * ``m'' in hc2hc and hc2c codelets is now the number of iterations, cannam@127: not the ``logical'' m. cannam@127: cannam@127: commit de904f19b230a114ead0b9580646689ab8519a29 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 17:07:04 2006 -0500 cannam@127: cannam@127: added memcpy-loop rank0 solver (it makes a 5-20% difference for transposes of large tuples) cannam@127: [empty commit message] cannam@127: cannam@127: commit d91736d4442e92910eabaa0e923d0cda833213cc cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 16:15:54 2006 -0500 cannam@127: cannam@127: new variable to disable libbench2's problem allocation during speed benchmarking (to benchmark MPI transforms where the array does not fit into the memory of a single process) cannam@127: [empty commit message] cannam@127: cannam@127: commit 8c1f9aabc4cd60f5509b287de2850c2767d07fd8 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 14:55:08 2006 -0500 cannam@127: cannam@127: allow transpose-inplace to use input as scratch for DESTROY_INPUT plans (to avoid non-square in-place transpositions) ... on supersgj, the planner often prefers transpose-inplace to transpose-alltoall in this case (apparently MPI_Alltoall in LAM MPI isn't that great) cannam@127: [empty commit message] cannam@127: cannam@127: commit 762d58ccde3a29468b9b522c8426ba6f48f0e74b cannam@127: Author: athena cannam@127: Date: Tue Dec 19 17:07:14 2006 -0500 cannam@127: cannam@127: For some reason HB2 codelets were not generated. cannam@127: [empty commit message] cannam@127: cannam@127: commit 928be47a01cfc332b729fd60775949d699d60795 cannam@127: Author: athena cannam@127: Date: Tue Dec 19 15:12:39 2006 -0500 cannam@127: cannam@127: split rdft/direct.c into direct-r2r and direct-r2c, since the file was getting out of control. cannam@127: [empty commit message] cannam@127: cannam@127: commit cf38c33836313129b7e98c192434dae261777810 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 02:59:35 2006 -0500 cannam@127: cannam@127: added dft-rank1 solver - MPI now supports 1d complex DFTs! cannam@127: [empty commit message] cannam@127: cannam@127: commit 061b341f302122d768db24c7aab043ade2e2dbb8 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 01:27:20 2006 -0500 cannam@127: cannam@127: fftw_flops must call cost_hook directly; iestimate_cost always uses COST_MAX cannam@127: [empty commit message] cannam@127: cannam@127: commit ebf61714b949775e7004b86b828112ae82b69726 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 01:16:54 2006 -0500 cannam@127: cannam@127: fftw_flops and fftw_estimate_cost must now be called from every process, to prevent deadlocks in the MPI code (since they sum/max the cost over all processes) cannam@127: [empty commit message] cannam@127: cannam@127: commit bea9d02f1cc9edd02ad6d30a11795bc11ff64d1a cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 00:55:34 2006 -0500 cannam@127: cannam@127: whoops, typo in assert cannam@127: [empty commit message] cannam@127: cannam@127: commit aee20bd2d616611547ef7826e5d30bf033994736 cannam@127: Author: stevenj cannam@127: Date: Tue Dec 19 00:51:07 2006 -0500 cannam@127: cannam@127: remove multiplication by FFT_SIGN cannam@127: [empty commit message] cannam@127: cannam@127: commit 8583a34cfe3ae51cd835c45d9035d80a0f944b52 cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 22:51:45 2006 -0500 cannam@127: cannam@127: need to synchronize ESTIMATE costs in MPI planner, and sum ESTIMATE costs for flop reporting: generalize measure_hook to cost_hook(..., {COST_SUM, COST_MAX}) cannam@127: [empty commit message] cannam@127: cannam@127: commit 97ebd8fa0c58fd24345e3386b28f1c7abba8cb8d cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 15:36:15 2006 -0500 cannam@127: cannam@127: previous patch slowed down transpose-alltoall when TRANSPOSED_IN and DESTROY_INPUT; now allow planner to choose old behavior in this case cannam@127: [empty commit message] cannam@127: cannam@127: commit 045a9c7e5b7ac5d91eb7567f34a2b4e307adeeeb cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 15:15:39 2006 -0500 cannam@127: cannam@127: transpose-alltoall doesn't require input to be destroyed if TRANSPOSED_IN is set cannam@127: [empty commit message] cannam@127: cannam@127: commit 43ab77408bc3d76457a0e89ad02aec84f4949bf3 cannam@127: Author: athena cannam@127: Date: Mon Dec 18 17:41:25 2006 -0500 cannam@127: cannam@127: Added t2-style hc2c codelets, fixed typos. cannam@127: [empty commit message] cannam@127: cannam@127: commit 7e431baa8e5da11432db111e201a4da9b19d6f49 cannam@127: Author: athena cannam@127: Date: Mon Dec 18 16:02:42 2006 -0500 cannam@127: cannam@127: Renamed certain variables to avoid calling an output stride `is'. cannam@127: [empty commit message] cannam@127: cannam@127: commit 141dcad10c248a45577b80c26d1e396530597d3b cannam@127: Author: athena cannam@127: Date: Mon Dec 18 15:54:43 2006 -0500 cannam@127: cannam@127: Oops---wrong stride cannam@127: [empty commit message] cannam@127: cannam@127: commit 23386506c8294fc1d61bc7cfcedb0bebc4e8fe60 cannam@127: Author: athena cannam@127: Date: Mon Dec 18 14:59:16 2006 -0500 cannam@127: cannam@127: Oops, forgot to add file cannam@127: [empty commit message] cannam@127: cannam@127: commit 8e69f0617fe6f90d09c23d4ce8f125f1822eb363 cannam@127: Author: athena cannam@127: Date: Mon Dec 18 14:17:02 2006 -0500 cannam@127: cannam@127: Renamed r2hc/hc2r codelets to r2c cannam@127: After the recent changes, r2hc/hc2r codelets became rdft2 cannam@127: problems, so I renamed them accordingly to r2cf/r2cb. cannam@127: Codelet parameters are now a real array and a complex array, instead cannam@127: of an input array and an output array, and forward and backward cannam@127: codelets have the same type, which removes some clutter from the rdft cannam@127: code. cannam@127: cannam@127: commit 91b8d21aa599744cf6f9bb2141bcd4193fcdb957 cannam@127: Author: athena cannam@127: Date: Mon Dec 18 10:48:07 2006 -0500 cannam@127: cannam@127: Implemented backward radix-2k rdft2. cannam@127: [empty commit message] cannam@127: cannam@127: commit c91a0bf1ed85466bcb46d2f55128399855c90f06 cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 13:56:09 2006 -0500 cannam@127: cannam@127: move extract_reim into kernel, since it is used by internal MPI stuff and not just in the API code any more cannam@127: [empty commit message] cannam@127: cannam@127: commit ab0e79e7a1319598924b4d434f1a0ce57bc58a43 cannam@127: Author: athena cannam@127: Date: Mon Dec 18 08:40:14 2006 -0500 cannam@127: cannam@127: Do not check r1==cr unless rnk>0 cannam@127: [empty commit message] cannam@127: cannam@127: commit 4a203a4d39cb5b02d1f6d83f2f525a6a6c0cf065 cannam@127: Author: athena cannam@127: Date: Sun Dec 17 21:03:50 2006 -0500 cannam@127: cannam@127: Implemented radix-2k RDFT2, forward only for now cannam@127: [empty commit message] cannam@127: cannam@127: commit 6d86c9dd27b6fbfb45faf91980565df40ec8d825 cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 01:23:45 2006 -0500 cannam@127: cannam@127: separate TRANSPOSED/SCRAMBLED flags internally (this is required so that dft-rank1-bigvec and the future dft-rank1 won't have incompatible SCRAMBLED formats) cannam@127: [empty commit message] cannam@127: cannam@127: commit 7920d86807a6fe9829cca1cb4e633ab3156c8b38 cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 01:02:27 2006 -0500 cannam@127: cannam@127: ops_add -> ops_add2 where possible, to shrink code cannam@127: [empty commit message] cannam@127: cannam@127: commit 81d5eddab81d4c0265863e1da6302b63fb1a9a80 cannam@127: Author: stevenj cannam@127: Date: Mon Dec 18 00:43:02 2006 -0500 cannam@127: cannam@127: added dft-rank1-bigvec solver (easy case for 1d parallel transforms) cannam@127: [empty commit message] cannam@127: cannam@127: commit bf7a77840dcbe0b1d5a8a1d7568877f093299e69 cannam@127: Author: stevenj cannam@127: Date: Sun Dec 17 20:42:21 2006 -0500 cannam@127: cannam@127: rewrote MPI stuff to use dtensor data structure cannam@127: cannam@127: A dtensor is an ordered tuple of triplets (n, ib, ob) giving the size of cannam@127: a dimension (n) and its input and output block sizes of a distributed cannam@127: row-major multi-dimensional array. An MPI DFT (etc.) is now specified cannam@127: in terms of dtensors, which provide a much more flexible data layout. cannam@127: cannam@127: For example, we can now describe multidimensional block distributions, cannam@127: which are important if the number of processors is greater than the cannam@127: size of any given dimension. Currently, we only have solvers for cannam@127: 1d slab distributions, and this is all that is supported in the basic cannam@127: and advanced APIs. The guru API allows one to specify more general cannam@127: distributions, however, which will be useful when/if we have solvers cannam@127: for this case. cannam@127: cannam@127: We now also don't need a TRANSPOSED flag, at least internally, since cannam@127: TRANSPOSED multi-dimensional DFT plans just correspond to dtensors cannam@127: where the input and output block distributions are different. cannam@127: cannam@127: Other changes include the use of the XM(foo) macro for X(mpi_foo). cannam@127: cannam@127: commit c9e4b997dd8cd38b753a8c40fb0004ab07124ca7 cannam@127: Author: athena cannam@127: Date: Fri Dec 15 16:01:23 2006 -0500 cannam@127: cannam@127: Distinguished mutexes from semaphores. cannam@127: The distinction is useful because the linux implementation of cannam@127: sem_post() in unnecessarily slow when semaphores are used for mutual cannam@127: exclusion. This change made spinlocks messier to implement, so I cannam@127: excised them. cannam@127: cannam@127: commit 352252ac74f50d01ae2c996f0072533f84c9a043 cannam@127: Author: athena cannam@127: Date: Fri Dec 15 12:46:11 2006 -0500 cannam@127: cannam@127: Use posix semaphores where available. cannam@127: Paranoid declaration of all shared variables as ``volatile''. Paranoid cannam@127: initialization of all shared variables within locks. cannam@127: cannam@127: commit 3364aa89b74e5e6a3514c653dc4ef4ebeebddae4 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Dec 13 20:12:13 2006 -0500 cannam@127: cannam@127: paranoia cannam@127: [empty commit message] cannam@127: cannam@127: commit 0be1cd7e754cf3b9550bcf6f60c2b31643d4512a cannam@127: Author: stevenj cannam@127: Date: Tue Dec 12 17:28:13 2006 -0500 cannam@127: cannam@127: punt on detecting unsolvable rdft2 problems; make r==iio rdft2 problems unsolvable, since it doesn't look like we've consistently checked for this case and it's not clear why we would want to support it (it was also not documented in the manual) cannam@127: [empty commit message] cannam@127: cannam@127: commit 8f3194d212eeb8c2382a60a4db59ff1cf935faef cannam@127: Author: athena cannam@127: Date: Sat Dec 9 12:14:13 2006 -0500 cannam@127: cannam@127: Obey stupid const rules cannam@127: [empty commit message] cannam@127: cannam@127: commit 1f7938759fd6c3a52293ffeffea1961692a22e72 cannam@127: Author: stevenj cannam@127: Date: Fri Dec 8 18:21:50 2006 -0500 cannam@127: cannam@127: added unsolvable check for rdft2 problem cannam@127: cannam@127: An in-place rdft2 problem is ill-formed if the real data, including the cannam@127: extra "padding" elements, do not coincide with the complex data. cannam@127: cannam@127: CHANGE: the new code considers all in-place split r2c and c2r problems cannam@127: to be ill-formed. Previously, these could be done, but only if the cannam@127: entire multi-dimensional array fit into the buffer, which is kind of cannam@127: stupid. I'm not sure it's worth it to even try to support the cannam@127: split in-place r2c case. cannam@127: cannam@127: commit e5241fedc1b2a9be448809abfc8b812e07598801 cannam@127: Author: stevenj cannam@127: Date: Fri Dec 8 13:47:53 2006 -0500 cannam@127: cannam@127: check in-placeness after joining taints cannam@127: [empty commit message] cannam@127: cannam@127: commit b7a9db50bbf343d80e6b5d0cbaae7cd4d8653f1e cannam@127: Author: athena cannam@127: Date: Fri Dec 8 13:43:44 2006 -0500 cannam@127: cannam@127: Grrr... paranoid-check was not testing in exhaustive mode cannam@127: [empty commit message] cannam@127: cannam@127: commit d34bae7a5b89e8298450b98759be43f22e7c99d3 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 8 10:00:30 2006 -0500 cannam@127: cannam@127: Implemented PROBLEM_UNSOLVABLE. cannam@127: In-place DFT and RDFT problems with inconsistent I/O strides are cannam@127: now unsolvable, and we don't check for them any longer in solvers. cannam@127: cannam@127: While I was at it, declared all problem pointers to be ``const'' cannam@127: for extra safety. cannam@127: cannam@127: commit db6e8b81d60e8cbe9f49ac3035a5151759afc88b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Dec 7 20:13:46 2006 -0500 cannam@127: cannam@127: Avoid qsort'ing one element. cannam@127: [empty commit message] cannam@127: cannam@127: commit a6d29bce88258799bb4bc6ee27c36aa5ccf4cbcd cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Dec 7 18:25:47 2006 -0500 cannam@127: cannam@127: In-place vrank>=1 is now applicable only if the problem is really in-place. cannam@127: [empty commit message] cannam@127: cannam@127: commit f0e0bda2dacfef167a5387c26d0c5631b59aaa0c cannam@127: Author: athena cannam@127: Date: Tue Dec 5 12:21:38 2006 -0500 cannam@127: cannam@127: unused variable cannam@127: [empty commit message] cannam@127: cannam@127: commit 07dc6817f4991f8690c6b6952bc3879c4712a624 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 3 19:16:33 2006 -0500 cannam@127: cannam@127: Removed CVS $Id$ everywhere, since darcs does not update them. cannam@127: [empty commit message] cannam@127: cannam@127: commit 2cf2355d9987a09710a6b6b345cf232abf8c1a8b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 3 16:11:17 2006 -0500 cannam@127: cannam@127: generalized dftw to encompass q codelets. As a side effect, q codelets are now threaded. cannam@127: [empty commit message] cannam@127: cannam@127: commit 942c7b34d673282b52aacbb51237b38f3ffef3f5 cannam@127: Author: stevenj cannam@127: Date: Sat Nov 25 16:34:38 2006 -0500 cannam@127: cannam@127: add missing __declspec attribute to threads API functions when compiling for Windows (thanks to Robert O. Morris for the bug report) cannam@127: [empty commit message] cannam@127: cannam@127: commit 8c4b9a9a79e7fdbbd7eab18f4aac4ac94c8ce2cb cannam@127: Author: stevenj cannam@127: Date: Mon Nov 20 17:39:20 2006 -0500 cannam@127: cannam@127: add AC_SUBST to AX_OPENMP, thanks to Sebastien Maret for the suggestion cannam@127: [empty commit message] cannam@127: cannam@127: commit 02d141e6157d606dce0bf146248958313ea19466 cannam@127: Author: stevenj cannam@127: Date: Mon Oct 16 23:02:29 2006 -0400 cannam@127: cannam@127: not gcc bug for MIPS (thanks to Jonathan Day) cannam@127: [empty commit message] cannam@127: cannam@127: commit 87fc8a66ef9cf8161a4cf23c0580f91ec9e86e25 cannam@127: Author: stevenj cannam@127: Date: Sat Sep 23 17:52:36 2006 -0400 cannam@127: cannam@127: in maintainer/debug mode, don't modify CFLAGS if they were explicitly set (-pedantic seems to cause problems with LAM's mpicc, so I need a way to override) cannam@127: [empty commit message] cannam@127: cannam@127: commit d1ebd06376db08bd5afecbe45d6059f60f7cd09d cannam@127: Author: athena cannam@127: Date: Wed Nov 22 18:14:47 2006 -0500 cannam@127: cannam@127: Removed obsolete comment. cannam@127: [empty commit message] cannam@127: cannam@127: commit b7bdd19e3b444a9c0ce68991739644a0fd4d9166 cannam@127: Author: athena cannam@127: Date: Sun Nov 19 11:21:44 2006 -0500 cannam@127: cannam@127: Use p->v when comparing TW_FULL fields. cannam@127: [empty commit message] cannam@127: cannam@127: commit fd6481c30c72bf15c3316dd7db6664e5b801160e cannam@127: Author: athena cannam@127: Date: Thu Nov 16 14:49:05 2006 -0500 cannam@127: cannam@127: removed useless definition cannam@127: [empty commit message] cannam@127: cannam@127: commit d161dc974c19cf43addd4b6cb516ae6b770827e7 cannam@127: Author: athena cannam@127: Date: Mon Nov 13 09:18:32 2006 -0500 cannam@127: cannam@127: paranoid avoidance of integer overflows cannam@127: [empty commit message] cannam@127: cannam@127: commit e5a1cce0ead5ae9d73c2c38c48e66c3bf059a874 cannam@127: Author: athena cannam@127: Date: Mon Nov 13 09:00:11 2006 -0500 cannam@127: cannam@127: avoid potential overflows in cycle counters cannam@127: At the suggestion of Alex Cichowski, convert all ticks cannam@127: to double before operating on them, to avoid potential cannam@127: signed/unsigned confusion and integer overflow. cannam@127: cannam@127: commit 98e4c9b9751d26d0adecc56634cc77e5689357bc cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Nov 5 09:00:52 2006 -0500 cannam@127: cannam@127: Removed unused struct field cannam@127: [empty commit message] cannam@127: cannam@127: commit 303349e158e3fdf0231790fe32a6831aa671f895 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Nov 4 09:43:13 2006 -0500 cannam@127: cannam@127: use pthread condition variables instead of semaphores cannam@127: Condition variables are more likely to be portable everywhere, and cannam@127: somehow they appear to introduce less overhead at least on my cannam@127: linux box. cannam@127: cannam@127: commit eb7fb1efca70242568f0e74266ea88e8c9a45eff cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Oct 31 20:45:24 2006 -0500 cannam@127: cannam@127: Imprecise help message. cannam@127: [empty commit message] cannam@127: cannam@127: commit 08bdb758c515972281738ca7567e38d3aeb05cf4 cannam@127: Author: athena cannam@127: Date: Mon Oct 30 20:13:35 2006 -0500 cannam@127: cannam@127: Experimental implementation of spinlocks. cannam@127: This patch implements spinlocks via a semi-portable hack, and adds cannam@127: the -ospinlocks option to the bench program so that we can play with them. cannam@127: cannam@127: commit fe239f5afbec9a3868f4330849cfe3d6dccb54b0 cannam@127: Author: athena cannam@127: Date: Thu Oct 26 22:29:18 2006 -0400 cannam@127: cannam@127: Updated manual for new openmp configure options. cannam@127: [empty commit message] cannam@127: cannam@127: commit 9809db57ce3f7e945e7bf04105a3fce74cdf7f47 cannam@127: Author: athena cannam@127: Date: Thu Oct 26 21:52:39 2006 -0400 cannam@127: cannam@127: Added back openmp. cannam@127: Rationalized threads naming conventions: cannam@127: cannam@127: * threads explicitly managed by us are enabled by --enable-threads, cannam@127: predicated on HAVE_THREADS, etc. cannam@127: * openmp is enabled by --enable-openmp, predicated on HAVE_OPENMP, etc. cannam@127: * SMP denotes either THREADS or OPENMP. cannam@127: cannam@127: commit 22cd21b038b6dd972444d5b00a6ebd00a932aa7f cannam@127: Author: athena cannam@127: Date: Mon Oct 23 20:14:31 2006 -0400 cannam@127: cannam@127: different thread protocols cannam@127: [empty commit message] cannam@127: cannam@127: commit f61da0e9c9db3c22dfd61c108f5f65bc4afe78bd cannam@127: Author: athena cannam@127: Date: Sun Oct 22 14:49:32 2006 -0400 cannam@127: cannam@127: fix memory leak cannam@127: Added pthread_attr_destroy to avoid memory leak. cannam@127: cannam@127: commit 6f6532928b29c6ac7599424c9dc834c41fc7fcf0 cannam@127: Author: athena cannam@127: Date: Sun Oct 22 14:23:30 2006 -0400 cannam@127: cannam@127: Experimental new pthread implementation that recycles threads. cannam@127: [empty commit message] cannam@127: cannam@127: commit e8c76bbab164fd6ad784bcfd9ec0fe5f002bfb43 cannam@127: Author: athena cannam@127: Date: Tue Oct 24 23:28:10 2006 -0400 cannam@127: cannam@127: switched buddies cannam@127: Switched order of buddies in rdft2 rank-geq2 for consistency cannam@127: with analogous dft and rdft solvers. Furthermore, this change reduces cannam@127: the MEASURE planning time for rank == 3. cannam@127: cannam@127: commit 3e13c85d1651dfc698143631f47a45b4c3947d12 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Oct 21 10:10:00 2006 -0400 cannam@127: cannam@127: Typo. cannam@127: [empty commit message] cannam@127: cannam@127: commit 7151038f7642b5ca050afb037fd91719a6f733b8 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Oct 21 09:56:50 2006 -0400 cannam@127: cannam@127: Out of place is the default. Thanks to Kirk Kern for pointing this out. cannam@127: [empty commit message] cannam@127: cannam@127: commit 281e20106cf076681392cb66050e11d2ac758dd2 cannam@127: Author: stevenj cannam@127: Date: Fri Sep 29 01:36:11 2006 -0400 cannam@127: cannam@127: rename "test" to "tst", since a user (Igor Levicki) reports that "test" is a reserved words in some x86 assemblers cannam@127: [empty commit message] cannam@127: cannam@127: commit 859e712b9c8051f6a1fb5a6500472042f194712f cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 26 09:01:08 2006 -0400 cannam@127: cannam@127: Stylistic change. cannam@127: [empty commit message] cannam@127: cannam@127: commit d024e575f98dc4a3452b9e0d5115a6650a7bf951 cannam@127: Author: athena cannam@127: Date: Tue Sep 26 08:45:37 2006 -0400 cannam@127: cannam@127: Do not set tmin=1e10, since a large FFT may take longer than that. cannam@127: [empty commit message] cannam@127: cannam@127: commit f31fffd32e9497675200784973ee8420ef3d60db cannam@127: Author: athena cannam@127: Date: Sat Sep 23 22:07:10 2006 -0400 cannam@127: cannam@127: Disable certain gcc optimizations. cannam@127: When PRECOMPUTE_ARRAY_INDICES is #define'd, array indices have the cannam@127: form array[stride[k]] for compile-time constant k. Apparently new cannam@127: gcc's copy stride[k] onto the stack before the codelet loop, which is cannam@127: an idiotic optimization if ever there was one. This patch confuses cannam@127: gcc enough to prevent this optimization. cannam@127: cannam@127: commit 0cc1f20ac1bbadb2e82d6465941755a9754d69d7 cannam@127: Author: stevenj cannam@127: Date: Sat Sep 23 13:02:58 2006 -0400 cannam@127: cannam@127: re-enable TOMS algorithm - it is the best for large vector lengths, since for such sizes the cache line is not an issue and the bookkeeping overhead is negligible cannam@127: [empty commit message] cannam@127: cannam@127: commit 458c7ee2e058d2cbe6fc90d23780b59bb29fbb52 cannam@127: Author: stevenj cannam@127: Date: Thu Sep 21 15:40:15 2006 -0400 cannam@127: cannam@127: add measure_hook so that MPI can synchronize timing measurements (otherwise different processors might end up with different MPI plans, yikes!) cannam@127: [empty commit message] cannam@127: cannam@127: commit 9fa6c37e2e09ea6bd226bde9a62f95d7c66f33bb cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 21:26:19 2006 -0400 cannam@127: cannam@127: added O(p log p) transpose algorithm (radix 2) cannam@127: [empty commit message] cannam@127: cannam@127: commit 1316689f96089a3b53799a3733d15add7c2f267c cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 21:05:09 2006 -0400 cannam@127: cannam@127: comments cannam@127: [empty commit message] cannam@127: cannam@127: commit c4b69d82b2c7de2fa2a963e27e3e498544c80262 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 19:54:58 2006 -0400 cannam@127: cannam@127: whoops cannam@127: [empty commit message] cannam@127: cannam@127: commit 19265283619d376581c5815adab98a2b51fb0cb2 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 19:49:01 2006 -0400 cannam@127: cannam@127: synchronize planning so that if one process fails to create a plan then all of them do. cannam@127: [empty commit message] cannam@127: cannam@127: commit e1b970b70562528b08d665ad2f7a17ee31f23e47 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 18:17:38 2006 -0400 cannam@127: cannam@127: call MPI_Alltoall instead of MPI_Alltoallv for equal-blocks case, in case MPI implementation has special optimizations for the common case of equal sizes cannam@127: [empty commit message] cannam@127: cannam@127: commit 3adcd54e6a54928afc2c58b7aab4bc75664be1d1 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 12:07:35 2006 -0400 cannam@127: cannam@127: whoops cannam@127: [empty commit message] cannam@127: cannam@127: commit 31b25a3bccaa0a8290dab5010199da8a5fec619d cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 02:20:06 2006 -0400 cannam@127: cannam@127: typo in comment cannam@127: [empty commit message] cannam@127: cannam@127: commit 519395bdb6ed85a2f18ba95bc82a68b2d42ddaae cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 01:58:55 2006 -0400 cannam@127: cannam@127: more filename simplifications cannam@127: [empty commit message] cannam@127: cannam@127: commit 2870207d79654158ecf7ae1d68d37382be5e39da cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 01:49:52 2006 -0400 cannam@127: cannam@127: canonicalize file names (hyphens, not underscores) cannam@127: [empty commit message] cannam@127: cannam@127: commit 41c88ac7e2cc2df2ff894c7223d55a41800b4c98 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 01:34:07 2006 -0400 cannam@127: cannam@127: add mpi-dft-serial cannam@127: [empty commit message] cannam@127: cannam@127: commit cbf1beef8656c2b0bc205c4787e6986cf32b405a cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 01:21:47 2006 -0400 cannam@127: cannam@127: silence warnings cannam@127: [empty commit message] cannam@127: cannam@127: commit ff7df52317b96acc1966b2d5920e46a3e368bded cannam@127: Author: stevenj cannam@127: Date: Tue Sep 19 00:31:59 2006 -0400 cannam@127: cannam@127: make "t" problem semantics match FFTW_MPI_TRANSPOSED cannam@127: [empty commit message] cannam@127: cannam@127: commit 694244ed8ede7153eca565e43ff553a26db39b8f cannam@127: Author: stevenj cannam@127: Date: Mon Sep 18 23:50:43 2006 -0400 cannam@127: cannam@127: whoops, fixed backwards mpi_dft cannam@127: [empty commit message] cannam@127: cannam@127: commit a36a49ee7709c1f3010f32039814f655e502850a cannam@127: Author: stevenj cannam@127: Date: Mon Sep 18 22:26:31 2006 -0400 cannam@127: cannam@127: initial stab at rank-geq2 mpi-dft; seems to be mostly working cannam@127: [empty commit message] cannam@127: cannam@127: commit c8e0a65f63c0d8eb6a148cee255c5aca2ff4c68b cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 13:41:32 2006 -0400 cannam@127: cannam@127: support SCRAMBLED_OUT in alltoall transpose cannam@127: [empty commit message] cannam@127: cannam@127: commit 72887de15eec06aeb7426d6a7cc527fc171821b0 cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 12:34:30 2006 -0400 cannam@127: cannam@127: skeleton of future support for block-cyclic cannam@127: [empty commit message] cannam@127: cannam@127: commit 62b562d1df02409e10395385348f56318e46a2e7 cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 12:11:19 2006 -0400 cannam@127: cannam@127: test program now checks scrambled in/out via -obflag=28/29 cannam@127: [empty commit message] cannam@127: cannam@127: commit b3ef0c11ca26e02875c29eb154f1a5b9bf386ad5 cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 11:58:36 2006 -0400 cannam@127: cannam@127: added -obflag to make it easier to set high-order bits cannam@127: [empty commit message] cannam@127: cannam@127: commit a4cbe985fe508661ccf587b5331e62b0a6526289 cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 01:30:51 2006 -0400 cannam@127: cannam@127: use proper child plans for 2nd transpose in transpose_alltoall; implement opcount in transpose_inplace cannam@127: [empty commit message] cannam@127: cannam@127: commit ab2dd6cfd9903487e3ac3cf9401c9a1f35de4862 cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 01:08:01 2006 -0400 cannam@127: cannam@127: fix in test program for transposes of vectors -- transpose routines seem to completely work now (except for scrambled in/out, which is untested) cannam@127: [empty commit message] cannam@127: cannam@127: commit aa5eecbf266020f2e2788bba862c13f6575d1ce8 cannam@127: Author: stevenj cannam@127: Date: Sun Sep 17 01:01:16 2006 -0400 cannam@127: cannam@127: fixed bug in transpose_alltoall for unequal blocks cannam@127: [empty commit message] cannam@127: cannam@127: commit 202f232a54abc2cb04e0a3d0d32bdb727c306c4e cannam@127: Author: stevenj cannam@127: Date: Sat Sep 16 15:29:46 2006 -0400 cannam@127: cannam@127: correctly handle cld2rest cannam@127: [empty commit message] cannam@127: cannam@127: commit 36668c0dd144f2bc2f6b2bf1f10eb1677593b9c6 cannam@127: Author: stevenj cannam@127: Date: Sat Sep 16 15:29:31 2006 -0400 cannam@127: cannam@127: some debugging code and other fixes cannam@127: [empty commit message] cannam@127: cannam@127: commit 66dcf1f5c673fd16b2f0f88988c4aaf388eeaf27 cannam@127: Author: stevenj cannam@127: Date: Sat Sep 16 14:54:30 2006 -0400 cannam@127: cannam@127: whoops, forgot to check in mpi_bench.c file cannam@127: [empty commit message] cannam@127: cannam@127: commit 3d96f316225934ef4485bcc2432314b89292914b cannam@127: Author: stevenj cannam@127: Date: Sat Sep 16 14:54:02 2006 -0400 cannam@127: cannam@127: added bench_exit routine so that it can be overridden (by MPI_Abort) if needed cannam@127: [empty commit message] cannam@127: cannam@127: commit b14337a969f6fee88bda25464c7ef7c0e56b5c00 cannam@127: Author: stevenj cannam@127: Date: Sat Sep 16 14:52:56 2006 -0400 cannam@127: cannam@127: bug fix in mpi_transpose_inplace for case where some processors are idle cannam@127: [empty commit message] cannam@127: cannam@127: commit 43fd42786f54710bffe85528beae2fff76e4a58e cannam@127: Author: stevenj cannam@127: Date: Fri Sep 15 18:47:13 2006 -0400 cannam@127: cannam@127: allow vecloop for sz->rnk==0 in exceptional (SLOW) cases, e.g. it is necessary for loops of non-square transposes (otherwise e.g. ik1v5:200:200x10:20:1x20:1:10 planning fails) cannam@127: [empty commit message] cannam@127: cannam@127: commit 121eaa69908a7b465f21f3529f74e983a63801ad cannam@127: Author: stevenj cannam@127: Date: Thu Sep 14 23:36:48 2006 -0400 cannam@127: cannam@127: first pass at working mpi_bench test program; transpose seems to work iff dimensions are divisible by #processors cannam@127: [empty commit message] cannam@127: cannam@127: commit 51101a902b4fdaef585e1d9e975238100951601c cannam@127: Author: stevenj cannam@127: Date: Wed Sep 13 17:28:07 2006 -0400 cannam@127: cannam@127: whoops cannam@127: [empty commit message] cannam@127: cannam@127: commit 7986cd7f00327db5f156e8d4d1458456f309e37a cannam@127: Author: stevenj cannam@127: Date: Tue Sep 12 22:27:03 2006 -0400 cannam@127: cannam@127: initial stub for mpi_bench cannam@127: [empty commit message] cannam@127: cannam@127: commit 279ca0155c7cb9dcd9bb9c75149a24bb1f44ba50 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 12 21:54:31 2006 -0400 cannam@127: cannam@127: do no output at all if verbose < 0 (for use with MPI, where we only want output from process 0) cannam@127: [empty commit message] cannam@127: cannam@127: commit 481f3838af04ae3db7aee15094ecf748f71d03da cannam@127: Author: stevenj cannam@127: Date: Tue Sep 12 21:39:15 2006 -0400 cannam@127: cannam@127: whoops cannam@127: [empty commit message] cannam@127: cannam@127: commit 792aaa1acca61e89b5605cbed49e9dd86bfbc2b1 cannam@127: Author: stevenj cannam@127: Date: Tue Sep 12 21:31:40 2006 -0400 cannam@127: cannam@127: split bench.c into bench.c and fftw_bench_common.c so that we can re-use some of the code in the MPI test program cannam@127: [empty commit message] cannam@127: cannam@127: commit 3b3b95ad0b1ab373687a9df59cdf2ec4bcdd502b cannam@127: Author: stevenj cannam@127: Date: Tue Sep 12 21:00:36 2006 -0400 cannam@127: cannam@127: MPI stuff at least compiles now cannam@127: [empty commit message] cannam@127: cannam@127: commit f5092f54bc2b8ee7289a2fb5148fc5315cbb2ee8 cannam@127: Author: stevenj cannam@127: Date: Mon Sep 11 22:26:36 2006 -0400 cannam@127: cannam@127: initial (nonfunctional) start at MPI support (similar to FFTW 2.x in spirit, but mostly rewritten) cannam@127: [empty commit message] cannam@127: cannam@127: commit 64d68fafe0c7f4433aec4a0925ce5972c33c78b4 cannam@127: Author: stevenj cannam@127: Date: Mon Sep 11 22:25:38 2006 -0400 cannam@127: cannam@127: make X(plan_awake) work for NULL argument to reduce code size cannam@127: [empty commit message] cannam@127: cannam@127: commit ed05c503c219544b0fe91af61db02d9cbb4027b5 cannam@127: Author: stevenj cannam@127: Date: Fri Sep 15 23:47:08 2006 -0400 cannam@127: cannam@127: -mt should go before -mthreads to avoid spurious warnings on HPUX (thanks to Peter O'Gorman for the bug report) cannam@127: [empty commit message] cannam@127: cannam@127: commit d737c7b3eca4ff8d7f372273f114dfd4e765b70c cannam@127: Author: stevenj cannam@127: Date: Mon Sep 11 13:53:44 2006 -0400 cannam@127: cannam@127: Fortran init_threads wrapper didn't return result; thanks to Markus Wetzstein for the bug report cannam@127: [empty commit message] cannam@127: cannam@127: commit ba5664a7958d533904b9251a4bfaa56b0f338a8a cannam@127: Author: stevenj cannam@127: Date: Thu Sep 7 18:43:55 2006 -0400 cannam@127: cannam@127: make sure wrappers are included even if Fortran compiler was not detected (unless --disable-fortran was specified explicitly) ... this was supposed to be done before, but the definition was in the wrong place, grr cannam@127: [empty commit message] cannam@127: cannam@127: commit 188c9dde71d0bc56ba30a052b82d02b6676f20ed cannam@127: Author: stevenj cannam@127: Date: Thu Aug 31 19:33:29 2006 -0400 cannam@127: cannam@127: I'm sick of answering this question about non-deterministic results cannam@127: [empty commit message] cannam@127: cannam@127: commit 336fb6116c43aa5559392ea2d0759606efd6f275 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 22 21:27:29 2006 -0400 cannam@127: cannam@127: Add --tag=CC flag to libtool. cannam@127: cannam@127: This change is consistent with the libtool invocation in the latest cannam@127: automake, and is required to compile with (some version of) xlc. cannam@127: cannam@127: commit d98d86f9115ca1a836e92d8df8e061f98f329032 cannam@127: Author: athena cannam@127: Date: Mon Aug 21 21:40:36 2006 -0400 cannam@127: cannam@127: avoid ``fma'' because it is defined in c99. cannam@127: [empty commit message] cannam@127: cannam@127: commit 841eb8db14a22936ba8ef81f439f42cb2411073d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 20 11:40:53 2006 -0400 cannam@127: cannam@127: Obey -standalone flag. cannam@127: [empty commit message] cannam@127: cannam@127: commit f270abac4732fe5f77708bef5f0d0cdc599bdb61 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 19 13:34:27 2006 -0400 cannam@127: cannam@127: obey -standalone when generating simd codelets cannam@127: [empty commit message] cannam@127: cannam@127: commit f573bbe2aaafabfbb21daf7da62972b8b071167f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 19 13:33:43 2006 -0400 cannam@127: cannam@127: removed obsolete athfft cannam@127: [empty commit message] cannam@127: cannam@127: commit 84e5b7792da92198e101b168d10710f0b81df5e8 cannam@127: Author: stevenj cannam@127: Date: Thu Aug 17 21:50:50 2006 -0400 cannam@127: cannam@127: updated citation to Proc. IEEE paper cannam@127: [empty commit message] cannam@127: cannam@127: commit 5ca3a79e05b95a688c21e7cb37a1ef7fa42a7f04 cannam@127: Author: stevenj cannam@127: Date: Thu Aug 17 21:47:05 2006 -0400 cannam@127: cannam@127: use darcs changes --summary to make nice changelog; emacs fill-region hack is obsolete cannam@127: [empty commit message] cannam@127: cannam@127: commit b80ff1b7affc5ba9c62bde0b06a548c3baf7c615 cannam@127: Author: athena cannam@127: Date: Mon Aug 14 17:53:19 2006 -0400 cannam@127: cannam@127: removed timer calibration cannam@127: Timer calibration seems not to work any longer on recent processors--- cannam@127: too much noise. I have remove it completely. cannam@127: cannam@127: commit 8986b3fa943f3f424a2f75541f8627a86af31a0a cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 14 10:47:15 2006 -0400 cannam@127: cannam@127: removed k7 cannam@127: Removed obsolete k7 support. cannam@127: cannam@127: commit 820835bfa680e9a0193435bfbcaf21923df9e7fc cannam@127: Author: athena cannam@127: Date: Sun Aug 13 11:02:11 2006 -0400 cannam@127: cannam@127: Use darcs instead of cvs. cannam@127: [empty commit message] cannam@127: cannam@127: commit 818c52da26a5d0781db8d9b45d4026403fb7e922 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 19 08:52:15 2006 -0400 cannam@127: cannam@127: Treat a the string "-" as a nonoption. cannam@127: cannam@127: commit 4e8a814e90696ee38898bfb5f079ac9bb6b614c2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 4 17:10:47 2006 -0400 cannam@127: cannam@127: comment out pkginclude dir for now cannam@127: cannam@127: commit 1bc4dd79b8cc59be7b18676f338c78013da54dab cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 3 20:51:08 2006 -0400 cannam@127: cannam@127: make sure CCAS = CC to avoid libtool confusion cannam@127: cannam@127: commit ebddd6bce119dec0b9a970a6d6194131321bdc5e cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jun 23 04:07:31 2006 -0400 cannam@127: cannam@127: install x77.h guru.h guru64.h in pkgincludedir cannam@127: cannam@127: commit e272fe53d7d822aa7d5ce03277f40c87aa843eef cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jun 23 04:03:42 2006 -0400 cannam@127: cannam@127: whitespace cannam@127: cannam@127: commit d6d23fdac18d0d01e363ff60bdba1285be017d0c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jun 23 02:33:45 2006 -0400 cannam@127: cannam@127: support cycle counter with xlc on Linux/ppc cannam@127: cannam@127: commit 677dd906902cf9dd2215c576a8f9d9e6755cc7cd cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 20 08:16:08 2006 -0400 cannam@127: cannam@127: Stylistic change. cannam@127: cannam@127: commit bb6bed2564fdec63eb8439031bc45caf8436b378 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jun 20 02:20:34 2006 -0400 cannam@127: cannam@127: bump date cannam@127: cannam@127: commit 6944a35c403fdcbf6b7b46f1aa9df9288991efca cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jun 20 02:20:06 2006 -0400 cannam@127: cannam@127: correct bug reported by Andrew Salamon ... --enable-portable-binary was cannam@127: ignored (or rather, treated unpredictably) due to typo, grrr cannam@127: cannam@127: commit ad98ebc35798f8713ac299ebe9ce74ca9fefe2f1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jun 1 20:30:06 2006 -0400 cannam@127: cannam@127: install 'internal' header files into includedir/fftw3/, includedir/fftw3f/, etcetera....this will make it easier to write external libraries that plug into FFTW internals, e.g. to add new solvers cannam@127: cannam@127: commit 4ce51f61d823524e8bebc4bc92ad2b17b6e7b53a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 29 23:59:19 2006 -0400 cannam@127: cannam@127: bug fix, thanks to James Donald for the bug report (only affects experimental semaphore stuff) cannam@127: cannam@127: commit ca9e38be107c761af7cd66a3ce9f0cfe93e9c069 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 29 23:58:16 2006 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 7e4b4be5e1bcdd9706a3ded5e2f59010ff751401 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 29 21:02:50 2006 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 5af69a3ec3b932c0d7e3e2dfdbcbff2aa067c5bf cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 27 19:36:15 2006 -0400 cannam@127: cannam@127: version bump cannam@127: cannam@127: commit 27dd43e42fa0b4ccea275b2143a9056f42f8c7f9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 27 14:54:47 2006 -0400 cannam@127: cannam@127: only check for xlc_r/cc_r if we are not using gcc cannam@127: cannam@127: commit c222c025be6649da84164ba5d2334fdcf0b3ac0b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 26 15:00:38 2006 -0400 cannam@127: cannam@127: use ptrdiff_t (it's C89 and standard C++, hooray) cannam@127: cannam@127: commit c3450d7f654ac2adf06bbbe9687f99cf1c6641b5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 26 12:59:33 2006 -0400 cannam@127: cannam@127: version bump cannam@127: cannam@127: commit 0be4f57c071dc97314660a66f4d46eee4ac143e9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 26 12:57:32 2006 -0400 cannam@127: cannam@127: noted 64-bit guru API cannam@127: cannam@127: commit 90455678a81def7a9aa3bc14f17047deb714271b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 26 12:53:09 2006 -0400 cannam@127: cannam@127: note that newer versions of VC++ support long long cannam@127: cannam@127: commit efddf05184fe6977af120842d10faf89399f14e0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 26 12:46:09 2006 -0400 cannam@127: cannam@127: try harder to get a portable 64-bit type cannam@127: cannam@127: commit 1d34caa16af08ff47fd75006c7576242e4643d17 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 25 22:04:18 2006 -0400 cannam@127: cannam@127: added draft guru64 API cannam@127: cannam@127: commit f987e828891ddd69efa3c664c68d231c1d3fc460 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 22 16:41:44 2006 -0400 cannam@127: cannam@127: added FIXME note cannam@127: cannam@127: commit 245cd07427cd24c953e4f1eea383790c7f557701 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 22 16:40:30 2006 -0400 cannam@127: cannam@127: check for xlc_r in addition to cc_r; thanks to Guy Moebs for the bug report cannam@127: cannam@127: commit 8a76c773855a145883608d47ca0aaa369e3ec408 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 21 12:35:25 2006 -0400 cannam@127: cannam@127: added note about gcc 4.0.1 on MacOS/Intel cannam@127: cannam@127: commit 9bb0ec78947a8597e0642379e7348e6b1c03af0b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 20 23:08:42 2006 -0400 cannam@127: cannam@127: added code for Core Duo; thanks to Eric Branlund cannam@127: cannam@127: commit d7a2e4a3ad51c01ec1bffbbadad602bb643da270 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 20 20:21:03 2006 -0400 cannam@127: cannam@127: fixed failure for -fPIC or for gcc-4 on Apple Intel machines; thanks to cannam@127: Eric Branlund for the bug report cannam@127: cannam@127: commit 3cb3cea549b4e8e0f9a16a1952eae4b4d8be1189 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 11 20:00:31 2006 -0400 cannam@127: cannam@127: Use -maltivec when checking for altivec.h. cannam@127: cannam@127: commit e2fb474c726118343e25059e2e1e8d2da6a21f62 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Apr 3 15:52:44 2006 -0400 cannam@127: cannam@127: note planner overwriting input in planner-flags reference cannam@127: cannam@127: commit 89a78d79d22078ee258d43d581cee6aaa3ba1d80 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 28 09:05:26 2006 -0500 cannam@127: cannam@127: FAQ entry about --enable-k7 in 64-bit mode. cannam@127: cannam@127: commit bfc115831ce70cd5cbf96fc005710862cf10bef6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 27 23:41:05 2006 -0500 cannam@127: cannam@127: sprintf -> snprintf, to avoid (harmless) complaints by users/compilers cannam@127: cannam@127: commit a2e4f6bfa281ed8b11c3a42e4cf32570e6a5c4d9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 27 23:30:22 2006 -0500 cannam@127: cannam@127: silence compiler warning cannam@127: cannam@127: commit 2c39d368d18c97bb079456491d60bb9a0c4c4342 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 17 09:20:10 2006 -0500 cannam@127: cannam@127: Remove dft/codelets/inplace, add simd/nonportable to list of cannam@127: directories to be compiled on non-unix systems. cannam@127: cannam@127: commit 01fa8ec4e8e6bd7a560437afe4ce4e37c13e0806 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 4 16:17:56 2006 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 6687db156af27c4ba2a4ddab66b6aa0a951b1a35 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 4 16:13:08 2006 -0500 cannam@127: cannam@127: note that we align the stack ourselves if necessary, with gcc and icc cannam@127: cannam@127: commit eee84dd2a9317a44c05e2f4dc2c05ff42709a973 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 4 16:08:16 2006 -0500 cannam@127: cannam@127: clearer distinction between static and automatic storage in C cannam@127: cannam@127: commit bc1aba15a1ddb5cd37b8088ea70f81ea6093e8d2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 25 20:27:01 2006 -0500 cannam@127: cannam@127: rm unused var cannam@127: cannam@127: commit d93efe4d5783cc4d5791894d58524c93644d5cb2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 25 17:30:28 2006 -0500 cannam@127: cannam@127: Improved usage of goto (Dijkstra miserere nostri) cannam@127: cannam@127: commit 36a203c3ada1b6257109162fee8dc563da9c4bc2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 25 14:19:15 2006 -0500 cannam@127: cannam@127: boilerplate cannam@127: cannam@127: commit 4041499e9299726d5840a0d119af094517810bda cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 25 14:14:40 2006 -0500 cannam@127: cannam@127: update for upcoming 3.1.1 cannam@127: cannam@127: commit 579c413f3bb5e24ac92d433aa17a063b1f11f8a1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 25 13:57:34 2006 -0500 cannam@127: cannam@127: replace obsolete IMPATIENT with MEASURE cannam@127: cannam@127: commit e0e594ba308e101ba93aacdceabdf0a35b4b0221 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 25 13:52:25 2006 -0500 cannam@127: cannam@127: corrected comment cannam@127: cannam@127: commit 5c1e2c07d1d8e21c219853b35212ba7373b35b45 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 25 10:19:26 2006 -0500 cannam@127: cannam@127: -v does not take an argument. cannam@127: cannam@127: commit dca8aaed07eadc0d1db6fe19b4a86d00ff7a328b cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 25 10:17:18 2006 -0500 cannam@127: cannam@127: Obey the unix convention that -ab = -a -b cannam@127: cannam@127: commit 95450e7e4d3ed287b4ff36d6ccd1250023cc06a2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 24 23:13:49 2006 -0500 cannam@127: cannam@127: minor fixes (return error on unrecognized option) cannam@127: cannam@127: commit af67fa909fc8ad31f5163b26da1693b1f9a61649 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 24 22:46:12 2006 -0500 cannam@127: cannam@127: ugh cannam@127: cannam@127: commit fec17358e3fbfb5e049933495db198312f9e10f9 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 24 21:42:56 2006 -0500 cannam@127: cannam@127: require exact match for long options. cannam@127: cannam@127: commit 5538e310cb61df6e3b5bd880ae604d86ec8f6121 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 24 21:38:02 2006 -0500 cannam@127: cannam@127: better fix cannam@127: cannam@127: commit 6f17dbe2db5e098604a35c03a7a1514040ea47d6 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 24 21:37:06 2006 -0500 cannam@127: cannam@127: Fix cannam@127: cannam@127: commit f10cae7e4cf944a6ef5928afbbaead482a4692d3 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 24 21:25:48 2006 -0500 cannam@127: cannam@127: nothing cannam@127: cannam@127: commit 01a4d4b5c204ef2c6d0afc7402a72481ad4a6c3e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 20 17:37:21 2006 -0500 cannam@127: cannam@127: rm transpose-indirect-inplace solver, which was buggy cannam@127: cannam@127: commit baa641d48341281eb3a9d4d9792f4482042836b4 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 15 08:43:05 2006 -0500 cannam@127: cannam@127: Comment fix. cannam@127: cannam@127: commit 207d1eae51bef5d4d14c7c670d6d16dadd8c8edf cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 15 08:18:41 2006 -0500 cannam@127: cannam@127: Cycle counter for Visual C++ x86-64, courtesy of Dirk Michaelis cannam@127: cannam@127: commit 0aefc1a066f619ed6f5b54791b00ab0acffe2901 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 14 19:17:30 2006 -0500 cannam@127: cannam@127: rfftwnd.png is in builddir cannam@127: cannam@127: commit 1799b5fa46fef72b8f116de92605ea4b8118a3db cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 14 19:03:27 2006 -0500 cannam@127: cannam@127: fixed typo: --enable-portable-binary, not --with cannam@127: cannam@127: commit f75d618a590c61c3bdba28ad0155f327670e231f cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 13 07:59:06 2006 -0500 cannam@127: cannam@127: estimator tweaks. cannam@127: cannam@127: commit 1c0cc8d7bb3570ce31320d87bbe35eb5b03ef38d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 12 20:43:39 2006 -0500 cannam@127: cannam@127: sse/sse2 support for t3?v codelets cannam@127: cannam@127: commit b1116627def6398d97dc443ba0f9bba1e2989f86 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 12 20:39:22 2006 -0500 cannam@127: cannam@127: Use CEXP instead of SIN/COS. cannam@127: cannam@127: commit dd361f593b5e85eeea97f2a2c048d3a8dce9d7e5 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 12 20:12:10 2006 -0500 cannam@127: cannam@127: bug in randomized cse eliminator. cannam@127: cannam@127: commit 9d329a9d010b44b728449d0f566eaa9356682a82 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 12 18:34:12 2006 -0500 cannam@127: cannam@127: Added support for t2-style simd codelets. This is altivec only for cannam@127: now; sse/sse2 don't even compile yet. cannam@127: cannam@127: commit b187b797d17ccda6efb853eb93b27aa42474a02a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 12 15:30:27 2006 -0500 cannam@127: cannam@127: Added support for t2-style simd split-complex codelets. cannam@127: cannam@127: commit 78281302f4fd1a0d42f0a0baf64cb364076aedb3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 10 18:21:28 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 948abfe0c2c7db954b7de51ff8da674dab3258ee cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 10 18:19:46 2006 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit e24cb9776a48a5cb0673e4ee8d75d142fcf2b117 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 10 18:00:35 2006 -0500 cannam@127: cannam@127: windows DLL stuff for Fortran interface cannam@127: cannam@127: commit fd7272f2e39eddd4491666e0bfe1e5c69c1eb04c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 10 09:48:52 2006 -0500 cannam@127: cannam@127: Bumped version to 3.1.1 cannam@127: cannam@127: commit 81a965e9d51e8e647f2eeaa12a973f7ef5188314 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 10 09:18:39 2006 -0500 cannam@127: cannam@127: Precompute array indices on x86-64. Speeds up Pentium IV and makes no cannam@127: appreciable difference on AMD. cannam@127: cannam@127: commit 5dbfa49ad81db0d3dca7e419507654fc0adc63fe cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 7 22:01:36 2006 -0500 cannam@127: cannam@127: Check whether the processor supports CPUID before issuing the cannam@127: instruction. (Grrr...) Code contributed by Eric J. Korpela. cannam@127: cannam@127: commit ce017677182a7662b7b1db85f32c6a8f34773703 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 7 21:36:47 2006 -0500 cannam@127: cannam@127: icc supports x86_64 these days. cannam@127: cannam@127: commit a7f132f06de9d343ee68b436c089bd37e6b7fc17 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 5 18:19:55 2006 -0500 cannam@127: cannam@127: Paranoia. cannam@127: cannam@127: commit 8645d5236a621db86ff7094b0e1a3e2946abc9fc cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 30 15:27:53 2006 -0500 cannam@127: cannam@127: whoops, fixed assert (y <= x) cannam@127: cannam@127: commit 13864d94f5c655cee3914be9d8751e184f86b8c9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 30 15:26:22 2006 -0500 cannam@127: cannam@127: note that safe_mulmod requires {x,y} < p (or at least < 2p), and added cannam@127: assert cannam@127: cannam@127: commit 849af348d142662e71fc4f3efe2866907e3bc745 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 30 11:09:32 2006 -0500 cannam@127: cannam@127: fixed aix/xlc lossage cannam@127: cannam@127: commit 106ee57674a134e1f876b6b6b77accd3a3b7a5f8 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 29 20:42:51 2006 -0500 cannam@127: cannam@127: In the impuse test, normalize the impulse so that the impulse and the cannam@127: random vectors have roughly the same L2 norm. This change reduces the cannam@127: number of bits that we lose because of floating-point cancellation, so cannam@127: that we can focus on the bits that we lose because of bugs. cannam@127: cannam@127: commit 45098b30a5e91b87bb97474de6ff2c16fd7373e3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 29 20:37:47 2006 -0500 cannam@127: cannam@127: Compute omega in trigreal precision, as opposed to R. cannam@127: cannam@127: commit 2c5480453a0fd877bdec040eb421b975eb2c63f4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 27 19:16:22 2006 -0500 cannam@127: cannam@127: add --with-combined-threads option as workaround to Windows inability to build shared libs with dependencies cannam@127: cannam@127: commit 68fde0a7351209d643634dfc19367da685c7e455 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 27 17:20:45 2006 -0500 cannam@127: cannam@127: libfftw3_threads should *not* used -no-undefined because, in fact, it is not true -- this library depends on -lfftw3, and is not self-contained cannam@127: cannam@127: commit ba85fd54aba2401c937c7acbff52a7c557956f68 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 26 22:04:34 2006 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 42feb604758692ce9936076f37e10c0f4098d46d cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 26 21:10:50 2006 -0500 cannam@127: cannam@127: Added paranoid stack alignment when awaking plans. While I was at it, cannam@127: removed obsolete, redundant AWAKE macro. cannam@127: cannam@127: commit 6b9831ddefcd83bf50aeafd90a6aa1effb44183e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 26 20:54:39 2006 -0500 cannam@127: cannam@127: Updated for 3.1. cannam@127: cannam@127: commit 4f2fadf55d8ba8d714bc96fb2236dfa981e3d244 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 26 19:15:12 2006 -0500 cannam@127: cannam@127: ditched one alignment check and noted that we should eliminate the rest as well cannam@127: cannam@127: commit 7c89983f07d925a997e5c293f8cdd5fbe577e3fb cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 25 23:02:19 2006 -0500 cannam@127: cannam@127: alignment hack cannam@127: cannam@127: commit 34f414ddf79840e5a7a9122c98e97bb2a09ecbd7 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 25 22:05:11 2006 -0500 cannam@127: cannam@127: detect pentium M cannam@127: cannam@127: commit 4f5853890a7ad01e763186bd03e44a5b20e5ef4a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 25 18:42:58 2006 -0500 cannam@127: cannam@127: don't trust host_cpu if it claims we are on i386/i486, and call cpuid anyway (if it fails we use no arch flag). This is needed on FreeBSD cannam@127: cannam@127: commit 16caea410e6cc85276555146cf41c370534074b6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 25 18:00:04 2006 -0500 cannam@127: cannam@127: suggest --with-our-malloc16 in error message cannam@127: cannam@127: commit 7574f2ae7e872c4f05f34b73057069b57fb2df01 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 24 19:53:34 2006 -0500 cannam@127: cannam@127: ditto for -no-gcc cannam@127: cannam@127: commit 04b5cc720e50ebe4cd2360425e79d6767356288a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 24 19:51:08 2006 -0500 cannam@127: cannam@127: flags required for successfull compilation should be added even if the cannam@127: user overrides CFLAGS cannam@127: cannam@127: commit 623ce195f6fe58d67f3bc8928ecc173f753e55db cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 24 18:43:59 2006 -0500 cannam@127: cannam@127: upcoming gcc OpenMP support uses -fopenmp cannam@127: cannam@127: commit abec9a1443aa235af2e2cfbd86f636599bcfea5e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 24 18:26:59 2006 -0500 cannam@127: cannam@127: note that PGI uses -mp as well cannam@127: cannam@127: commit 883f0f18de0f8528fe77129192f521b1a77bfece cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 23 15:31:24 2006 -0500 cannam@127: cannam@127: my best guess at how to fix the microsoft crap du jour cannam@127: cannam@127: commit b457e9e371cf5e13bb818868495b38dce7ccdcd6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 23 14:05:14 2006 -0500 cannam@127: cannam@127: use -Masmkeyword for PGI cycle counter, grr cannam@127: cannam@127: commit 91a65bac96c2fc134d11c8d551eb410de6d18bc2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 22 18:09:06 2006 -0500 cannam@127: cannam@127: Bumped version number to 3.1. cannam@127: cannam@127: commit 1745639a638440b17bfcea324c068679f655df24 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 21 10:03:59 2006 -0500 cannam@127: cannam@127: Report that --enable-k7 is incompatible with --enable-shared. cannam@127: cannam@127: commit 840da056365df79ea63fc3d5a21b1ab5a13707e9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 21 09:17:54 2006 -0500 cannam@127: cannam@127: Do not use empty libraries in LIBADD, since otherwise the linker fails cannam@127: on Solaris. cannam@127: cannam@127: commit 4228f20154f81216ab4ddae092d7661bb8af1652 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 18 10:47:59 2006 -0500 cannam@127: cannam@127: warn end-users away from this file cannam@127: cannam@127: commit b6e0f0a6eac2561efe417f2cfc0eb8686196a385 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 16:16:42 2006 -0500 cannam@127: cannam@127: Gcc sucks. cannam@127: cannam@127: commit 8560506aa44b2740ea378c83c403373dfce2a662 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 11:48:55 2006 -0500 cannam@127: cannam@127: Disabled checks that may turn out to be too paranoid. cannam@127: cannam@127: commit 782888694f5690298d87cc67cf9963f97aabc412 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 10:35:03 2006 -0500 cannam@127: cannam@127: Some paranoid checks. cannam@127: cannam@127: commit c93e48fa31d081994b9e3b11cca9f1ab25bdf6a1 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 09:31:08 2006 -0500 cannam@127: cannam@127: Flush stdout after printing. cannam@127: cannam@127: commit 8a84f237ca9d96babf1f4edeecb181c47cd74dbe cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 08:28:18 2006 -0500 cannam@127: cannam@127: Run the leak detector in all cases, not just when verbose > 2. cannam@127: cannam@127: commit 884a08a129046af3f84ce0fc138f385976f5a5a9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 08:11:41 2006 -0500 cannam@127: cannam@127: Eliminate calls to pow(), rint(). cannam@127: cannam@127: commit 370ddffe8e4854e4826b1ff4ea14c617d1eea504 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 17 00:45:06 2006 -0500 cannam@127: cannam@127: put # in first column, for stylistic consistency cannam@127: cannam@127: commit ede00270785b328279288ada254a11f7314bcd6c cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 17 00:17:27 2006 -0500 cannam@127: cannam@127: Made timeout part of impatience flags, in order to improve the cannam@127: usability of wisdom. Also, fixed bogus error recovery logic in cannam@127: planner.c:imprt(). cannam@127: cannam@127: commit 693f01973548254de258d7efa4217cabea005e79 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 16 23:03:34 2006 -0500 cannam@127: cannam@127: make timelimit < 0 .eq. FFTW_NO_TIMELIMIT cannam@127: cannam@127: commit 5af63c169becdefc68db3b4f2df8e788b9867c98 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 16 21:52:01 2006 -0500 cannam@127: cannam@127: Eliminated the FFTW_TIMELIMIT flag in favor of this simpler logic: cannam@127: fftw_set_timelimit(0) disables time limit. cannam@127: fftw_set_timelimit(X), X>0 sets the time limit to X. cannam@127: cannam@127: commit 8a9d6dd6b442050ad202a6f7154926d145e359b1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 16 08:38:04 2006 -0500 cannam@127: cannam@127: Force the use of the estimator when wisdom fails because of md5 cannam@127: collisions, otherwise the planner takes forever. cannam@127: cannam@127: commit 7c6a1a3f7e16df4dca8f78ee994d0488278977fb cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 21:30:31 2006 -0500 cannam@127: cannam@127: Ranted about how broken gcc-4 is. cannam@127: cannam@127: commit 383c1374f7af522dfcfe363c508d7fb630e83746 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 15 19:59:38 2006 -0500 cannam@127: cannam@127: change fftw_timelimit global var to fftw_set_timelimit(double) function, for simpler usage with shared libraries and for consistency with e.g. set_numthreads cannam@127: cannam@127: commit 584641592c2c273e233b919c8e68e1dbb840d72f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 19:32:27 2006 -0500 cannam@127: cannam@127: Minor tweaks. cannam@127: cannam@127: commit f8fd8093b44aca863601612ae0b4818e91cca853 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 16:32:54 2006 -0500 cannam@127: cannam@127: tweaks to make sure that time_n() is always called from the same stack position. cannam@127: cannam@127: commit 1a5445769d483d86df5d2de6e41f4c5e9515a4f1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 16:09:53 2006 -0500 cannam@127: cannam@127: Major simplification of the timer calibration logic. Also, use an FFT cannam@127: as a unit of work instead of the old pointer chasing, because God cannam@127: knows how pointer chasing interacts with the idiotic cache-hit cannam@127: speculation on the Pentium IV. cannam@127: cannam@127: commit 1838fc3c1290495355ce10791c9a9f376dd7522d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 15:12:08 2006 -0500 cannam@127: cannam@127: Fixed broken aligment checks when sizeof(R)==12. cannam@127: cannam@127: commit 17b67db5fa051c8eef9c962abfb698b51d11f303 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 10:36:40 2006 -0500 cannam@127: cannam@127: Manual unrolling of loop. cannam@127: cannam@127: commit 7465e7b2ab1c48d06a50189a8545af2b0d98fdda cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 15 10:12:55 2006 -0500 cannam@127: cannam@127: Various improvements to timer calibration routines. cannam@127: cannam@127: commit 049684aa3e3f0411c535a71b0f4adc0ecff7327d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 14 22:16:09 2006 -0500 cannam@127: cannam@127: cygwin defines __CYGWIN__, not __WIN32__ etc. cannam@127: cannam@127: commit c418027ffb0304f446af729a93415df506d093f3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 14 20:40:12 2006 -0500 cannam@127: cannam@127: fixed confusion between libbench and user timers cannam@127: cannam@127: commit 64c27bbd6553e2d8e2e988456890e3e31266b89e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 14 12:32:44 2006 -0500 cannam@127: cannam@127: update cannam@127: cannam@127: commit 1b00d512abfe62699264bc556e84a50e89c9a377 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 14 10:24:11 2006 -0500 cannam@127: cannam@127: Comment. cannam@127: cannam@127: commit 7fa69534cca1e07e3c1260151ed8dee5e5a645cc cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 14 10:19:28 2006 -0500 cannam@127: cannam@127: Workaround gcc bug. cannam@127: cannam@127: commit 686f1af1d0f1c63d99f2891a47c5de2b459b92d2 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 13 19:13:18 2006 -0500 cannam@127: cannam@127: Switched to -beta2. cannam@127: cannam@127: commit 9f370230780a4f3c03643f6c35f4114fada8c1e7 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 12 22:21:57 2006 -0500 cannam@127: cannam@127: Fixed technically correct but highly obfuscated use of the enum tag cannam@127: R2HC as a null pointer. cannam@127: cannam@127: commit 38965981e1187e5d0574e129690f3e02b4bc1cae cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 12 19:25:20 2006 -0500 cannam@127: cannam@127: --enable-unsafe-mulmod is obsolete cannam@127: cannam@127: commit 3f29e7d2e1993de8b7a9759bc879955cb8ae569d cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 12 19:23:18 2006 -0500 cannam@127: cannam@127: More thoughts. cannam@127: cannam@127: commit 84082b78ddfaf5133e49453cc3a62c3d3dde9c04 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 12 19:17:57 2006 -0500 cannam@127: cannam@127: Removed loop unrolling because it slows things down on at least one cannam@127: powerpc and it generates clumsy x86 code. cannam@127: cannam@127: commit 51caa62b55dcdb8e1aeb9da2d10a40874cef875a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 12 19:17:35 2006 -0500 cannam@127: cannam@127: tweaks cannam@127: cannam@127: commit e29d0b0a8ceacbe19adba501d8e5799c7647bb87 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 12 15:55:52 2006 -0500 cannam@127: cannam@127: MacOSX x86 ABI specifies that the stack is kept 16-byte aligned cannam@127: cannam@127: commit 487e03a1ee35bc63877b6ec2c2e410da5f3dd4d5 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 12 12:46:49 2006 -0500 cannam@127: cannam@127: ``ret'' is a reserved word in the evil empire. cannam@127: cannam@127: commit ef109b1d0703dbf67144c7ef5afe9ee4dd0ef489 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 12 08:31:43 2006 -0500 cannam@127: cannam@127: Changed ret => result because ret ``is a reserved word'' in the evil cannam@127: empire. cannam@127: cannam@127: commit 7dfbcb39afd28daaa10eba6e7909e0e8e3dd56bc cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 11 19:30:42 2006 -0500 cannam@127: cannam@127: Workaround Visual c++ lossage. cannam@127: cannam@127: commit 93876be963fec88768744d04a2c027a4c14f49f0 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 11 19:26:16 2006 -0500 cannam@127: cannam@127: Workaround visual c++ lossage. cannam@127: cannam@127: commit b2e9544d09112da7db08f07f268e2ed3ad707634 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 11 19:10:52 2006 -0500 cannam@127: cannam@127: isprint() is guaranteed to work for unsigned char + EOF only. cannam@127: cannam@127: commit 5b926765db935776483660d88b2ce02dca54081e cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 11 13:47:49 2006 -0500 cannam@127: cannam@127: rm obsolete fixme cannam@127: cannam@127: commit 9237b1a5063d1190e4a8a79d924599a240706756 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 11 13:38:46 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 72c1af743cd4da543e142aa9d51b600f47811378 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 11 13:32:26 2006 -0500 cannam@127: cannam@127: fix comment cannam@127: cannam@127: commit 47b608a52f08027e1429325bd1639ee4f176aea2 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 11 12:27:05 2006 -0500 cannam@127: cannam@127: Paranoid use of K(x) for all constants x, to avoid runtime double->float conversions on sufficiently stupid compilers. cannam@127: cannam@127: commit fd9ac529906a8db6d171aa999e4a848b495a8fde cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 20:10:38 2006 -0500 cannam@127: cannam@127: Workaround to gcc nonsense. cannam@127: cannam@127: commit 90aaf565346f372e580fb899f9212558ff87d0d2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 18:44:28 2006 -0500 cannam@127: cannam@127: bug fix: infinite loop in transpose-cut planning cannam@127: cannam@127: commit 5cd8a9482a90f25c76df01e4f8ea4b2a3386c449 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 18:12:14 2006 -0500 cannam@127: cannam@127: clarified comment cannam@127: cannam@127: commit 4c5e2af8af80c06734bac116adefdc9e346caa47 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 18:10:32 2006 -0500 cannam@127: cannam@127: more Windows decorations cannam@127: cannam@127: commit c0bb01fdec0e5c0d9636187641f43690cdb771e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 17:57:45 2006 -0500 cannam@127: cannam@127: added FIXME comment cannam@127: cannam@127: commit 82eb98885049d8d0b69490915a39614c17d5263c cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 17:52:07 2006 -0500 cannam@127: cannam@127: 'make clean' should not delete codlist.c since it is included in the dist tarball cannam@127: cannam@127: commit 7fece302306db2d854caf017c680e29eadf79cb7 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 17:50:12 2006 -0500 cannam@127: cannam@127: Change threshold for ``large'' Cooley-Tukey to 256K from 64K, since it cannam@127: seems to benefit the Pentium IV with sse and the planning cost is not cannam@127: too horrible. cannam@127: cannam@127: commit 46c94f013ca0ab45344996479ff3059a6b835241 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 17:45:11 2006 -0500 cannam@127: cannam@127: more missing Windows DLL decorations cannam@127: cannam@127: commit 67d487e555c8d365aa9530173dba788656f2d91b cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 17:41:28 2006 -0500 cannam@127: cannam@127: remove unused var cannam@127: cannam@127: commit 3356ac92a38eea7582b9712a6cef2067dd9ccf28 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 14:00:50 2006 -0500 cannam@127: cannam@127: allow compiler threads, if enabled, to take precedence over explicit threads cannam@127: cannam@127: commit d2c3905718e0dbe3bb6e67befc3a2f4c63badbeb cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 12:30:09 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 755b3ecdb747b9b6f1dbc52036cf1d08f724596d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 12:21:56 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 22db6a0e68da61729239444ff941e84f8de1b336 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 09:13:20 2006 -0500 cannam@127: cannam@127: Fixed comment typo. cannam@127: cannam@127: commit 219609390fc443e6defd5f4940aa36e059b0e6c6 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 08:59:22 2006 -0500 cannam@127: cannam@127: Rearranged timeout checks so as to eliminate one of them. cannam@127: cannam@127: commit 5d22885da57a28d4ce96128650dba99a3ea76481 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 08:56:55 2006 -0500 cannam@127: cannam@127: Converted residual CK() -> A(). cannam@127: cannam@127: commit d82a20e3e3a4d47aebbd2ce4350da4976ba32652 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 08:36:13 2006 -0500 cannam@127: cannam@127: Maintain the invariant TIMED_OUT ==> NEED_TIMEOUT_CHECK. cannam@127: cannam@127: commit 08f674254d16c7770944dc9e7c0eaa3579f333bb cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 08:24:41 2006 -0500 cannam@127: cannam@127: silence some 64-bit warnings cannam@127: cannam@127: commit 37aaadd4f3be4c0a5f03fffae1df96e82e8064c9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 10 07:58:48 2006 -0500 cannam@127: cannam@127: Assertions. cannam@127: cannam@127: commit 3ee7cd888752144ff48442480446982dcbf3bba3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 00:14:00 2006 -0500 cannam@127: cannam@127: some condensing cannam@127: cannam@127: commit 667419d3ac72c2cc43df2d10f704111a40320338 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 10 00:03:32 2006 -0500 cannam@127: cannam@127: eliminate X(seconds) in favor of X(elapsed_since), in paranoia of clock wrap cannam@127: cannam@127: commit f696d1fe43a496c64fcf0daaa83060ac70c789fb cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 23:21:21 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 95280e070545a42bbd407c52877e6f8b48d778d6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 23:21:06 2006 -0500 cannam@127: cannam@127: hmm, a bit more pessimistic about clock wrapping cannam@127: cannam@127: commit 47d7479fd14d1baf3102a699e72eb7158fede8bd cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 23:20:26 2006 -0500 cannam@127: cannam@127: Revert to md5uint = unsigned int whenever possible, so as to cannam@127: avoid wasting space for unsigned long on 64-bit machines. cannam@127: cannam@127: commit 887d8a089ff5b925f88a198bf1b1cc7ddf61392a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 23:12:27 2006 -0500 cannam@127: cannam@127: note why clock() wrap should not be a concern cannam@127: cannam@127: commit 2991a94ba406fa1f245b62216e463a778f646bbd cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 22:57:16 2006 -0500 cannam@127: cannam@127: bugfix in recent timeout changes - check for case where last solver times out cannam@127: cannam@127: commit 3a0c958aa20c64dab514ec5afe74531e933ac77a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 22:40:26 2006 -0500 cannam@127: cannam@127: started changes list from beta cannam@127: cannam@127: commit 63922f26968d87550c4fcfd47f41225d056e4977 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 22:34:13 2006 -0500 cannam@127: cannam@127: Paranoia. cannam@127: cannam@127: commit 5bf3d3dd9c6ab2be2e450cd03cabb775bb3db4ca cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 22:27:37 2006 -0500 cannam@127: cannam@127: Paranoid assertions. cannam@127: cannam@127: commit 967f0848ad28ffe357a4758d477c5826075a4bac cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 22:13:32 2006 -0500 cannam@127: cannam@127: Added FIXME comment stating the 64-bit uncleaniness of cannam@127: fftw_tensor_to_bench_tensor(). cannam@127: cannam@127: commit 0a2228df28268ba4855063849942199ed2c86d31 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 22:06:05 2006 -0500 cannam@127: cannam@127: Another 64-bit bug. cannam@127: cannam@127: commit dc20e0d303713ae5664b91eb4762b4a0f5cf0623 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 21:54:07 2006 -0500 cannam@127: cannam@127: more Windows DLL nonsense cannam@127: cannam@127: commit d884e3edc5ca24864e92470966ed04aeaccab8f9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 21:18:25 2006 -0500 cannam@127: cannam@127: some additional dllexport tags required to build the test program, due to internal stuff called by hook.c cannam@127: cannam@127: commit c0fc6ffb403456e03f5e8dc425182e6607c1cd2a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:31:15 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit a85549d03edbe4bee47b0248fac7d76d9cdfeb06 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:30:19 2006 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 1f72b4d52a499bd63cd52ec259ae2585c6df2b66 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:20:28 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit fb0eb0e86fcfffb4c1b5ba17cc520e27914c4ff4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:16:50 2006 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit 29de1846aefcd05cb14e0dd286a8374a4a9b18e8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:12:23 2006 -0500 cannam@127: cannam@127: define FFTW_DLL if DLL_EXPORT (defined by libtool) is supplied cannam@127: cannam@127: commit b99426a4b113d6c8017a6feeb22d89b8a32211f7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:05:11 2006 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 5bb395fcc8798c697d5158b9242d4d914922d9db cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 20:00:47 2006 -0500 cannam@127: cannam@127: another stab at Windows DLL mess cannam@127: cannam@127: commit 9453c5ed5d0c160deb3aef127870b7d65c26d8d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 19:23:42 2006 -0500 cannam@127: cannam@127: 64-bit clean SIMD header file. I missed those because sparse cannam@127: does not know vector types. Grrr... cannam@127: cannam@127: commit a27e044b39f52abb6066e070b1a3492b6be2e155 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 19:08:36 2006 -0500 cannam@127: cannam@127: this option is called AC_DISABLE_SHARED in the documentation cannam@127: cannam@127: commit 25cd95982a2acda3a3d6220728768164d6d9c890 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 17:34:13 2006 -0500 cannam@127: cannam@127: fixed --with-gcc-arch to work when cross-compiling cannam@127: cannam@127: commit c0b9d3122ba267c448b98b0ede12bcf27b9b4e02 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 9 12:04:04 2006 -0500 cannam@127: cannam@127: Moved the timeout check back into the search loop, sicut erat in cannam@127: principio. This gives us a precise control over the timeout. To cannam@127: avoid the overhead of X(seconds)(), only call X(seconds)() if some cannam@127: time measurement was taken since the last call to X(seconds)(). cannam@127: cannam@127: commit 8c4448e2b69fb02b70e85405bf58a77ec4c13de9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 9 00:07:40 2006 -0500 cannam@127: cannam@127: comments cannam@127: cannam@127: commit 3cd770cab6fac7657b7cd55d6d98f3f516a20fb3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 23:58:23 2006 -0500 cannam@127: cannam@127: generalized transpose-cut routine to be able to call transpose-gcd recursivly; TOMS follow-the-cycles algorithm now seems to be completely superseded cannam@127: cannam@127: commit 7ce8a67fabd9ed925a7aee905fa50c658ed2fd20 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 20:53:18 2006 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit ee3cbdc7ad5dd5bfcb2f1f3df6b5ace55f121f32 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 20:52:16 2006 -0500 cannam@127: cannam@127: ignore errors from setscope -- POSIX standard does not require PTHREAD_SCOPE_SYSTEM to be supported, and PTHREAD_SCOPE_PROCESS is usually okay in that case cannam@127: cannam@127: commit 0ee88684468fc0dae5fd08cc684b8c174d885dd9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 15:58:40 2006 -0500 cannam@127: cannam@127: added TODO comment cannam@127: cannam@127: commit 839a6d3192f804e3bc018419b90d18aa82d00292 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 15:39:28 2006 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit d3fdf3fcd234dbb12aea0ab1029db2f121356f1e cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 8 14:44:23 2006 -0500 cannam@127: cannam@127: Boasted ``much faster altivec performance''. cannam@127: cannam@127: commit 6d85298a5a4c987cf192ae2df25673a8250d265d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 8 11:44:52 2006 -0500 cannam@127: cannam@127: Added a new pass to the generator to schedule for the pipeline cannam@127: latency. (This schedule modifies the ``optimal'' cache-oblivious cannam@127: schedule and hence it uses more registers.) cannam@127: cannam@127: This pass is currently: cannam@127: cannam@127: * disabled for non-fma code, under the assumption that this will cannam@127: run on a register-starved fma. cannam@127: cannam@127: * enabled for non-simd fma code, under the assumption that this will cannam@127: run on a processor with 32 or more FP registers. The latency of 4 cannam@127: is conservative and does not introduce too much register pressure. cannam@127: cannam@127: * enabled for simd fma code, under the assumption that this will run cannam@127: on altivec. The latency of 8 seems to produce the best results. cannam@127: cannam@127: commit 1e7e0cd308f67033c681c0ae52836283f874fe51 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 03:13:53 2006 -0500 cannam@127: cannam@127: fixed estimator for vrank3-transpose cannam@127: cannam@127: commit 13dd2e84c1d331d07eaaef76bb78110dbe941446 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 02:02:11 2006 -0500 cannam@127: cannam@127: more detail on VC++ workaround cannam@127: cannam@127: commit 699008e51d100801bb19d99b2dbe595b1e33c445 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 00:19:19 2006 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 2f842c52fbd7cac9b7564045378e1d649af6dbbf cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 8 00:16:20 2006 -0500 cannam@127: cannam@127: screw it, just use planner for all sub-transposes in vrank3-transpose (still just use memcpy for contiguous copies, though) cannam@127: cannam@127: commit e6908d5d37b50dff661acfecd3687d1a9fd3300c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 23:13:45 2006 -0500 cannam@127: cannam@127: add an assert cannam@127: cannam@127: commit 000c5f8a4f6f83143f85268a03709d12ba1e896e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 21:57:34 2006 -0500 cannam@127: cannam@127: vrank3-transpose now uses planner to decide whether to use cpy2d, cpy2d_tiled, etc. cannam@127: cannam@127: commit f9db072d1270330e0fde90db33e71576d4a2e141 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 20:57:16 2006 -0500 cannam@127: cannam@127: too annoying to have isqrt unexpectedly fail for n==0 cannam@127: cannam@127: commit 9c8847c394cdd9bdd3d02a127a2497e09bab2d28 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 17:49:37 2006 -0500 cannam@127: cannam@127: clarifications cannam@127: cannam@127: commit 9fdeaf83ff81ca6931d74b65f8477f95fcfae323 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 16:39:20 2006 -0500 cannam@127: cannam@127: comment fix cannam@127: cannam@127: commit ad7b11b215b77bad24047e811e8bdaa2ee320edb cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 15:16:22 2006 -0500 cannam@127: cannam@127: more faq updates cannam@127: cannam@127: commit 8bc87da1486f5f1a451cc418a345bb12b95479fc cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 7 15:12:16 2006 -0500 cannam@127: cannam@127: enable fma on hppa, update FAQ entry cannam@127: cannam@127: commit 479aa905ff4136d48a86ef8ea28e46c06c07ee79 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 7 14:06:31 2006 -0500 cannam@127: cannam@127: Accomodate different semantics of 'const' in C and C++ cannam@127: cannam@127: commit cf0d153fd10cf6e894520f58c2ce1e6259b683c9 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 6 23:40:53 2006 -0500 cannam@127: cannam@127: Altivec is called VMX in IBM land. cannam@127: cannam@127: commit a46734a158edbc1e170c0e043d64fb3a320c8d80 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 6 23:40:16 2006 -0500 cannam@127: cannam@127: Noted faster altivec support. cannam@127: cannam@127: commit 4e7329c580102980a2862964df1474c403d59f9d cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 6 21:49:10 2006 -0500 cannam@127: cannam@127: updated icc flag detection cannam@127: cannam@127: commit 2de66ca6567360268fa4f1653c787903471a2ab7 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 6 10:01:50 2006 -0500 cannam@127: cannam@127: Note ``memoize triggen''. cannam@127: cannam@127: commit c19609ea4726f8e842db68cbf15f2ee94abdf33d cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 6 09:36:51 2006 -0500 cannam@127: cannam@127: Use --enable-threads to generate dependencies in the threads/ directory. cannam@127: cannam@127: commit 7538d17a7e277e5f3099b285f85944ee81df6a7c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 6 09:26:29 2006 -0500 cannam@127: cannam@127: Workaround to icc #defining __GNUC__. cannam@127: cannam@127: commit 3623ea4c4e5649470d360af6c89410b22da9b9ef cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 6 09:21:19 2006 -0500 cannam@127: cannam@127: Switched name to 3.1-beta1. cannam@127: cannam@127: commit 5022d2e2f5e385f82c9b298f958a6935de39233c cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 23:08:44 2006 -0500 cannam@127: cannam@127: More thoughts. cannam@127: cannam@127: commit d6262891e97139b27fdb2ca73addf122be568d17 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 22:30:51 2006 -0500 cannam@127: cannam@127: Note wish that (block_size % 4) == 0. cannam@127: cannam@127: commit d6779fe4008a3ff1b5341cc82946e24a6e0cf418 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 22:19:09 2006 -0500 cannam@127: cannam@127: Check alignment of mstart, mcount in SIMD codelets. cannam@127: cannam@127: commit 3d4fc920479d90ecc75a2256c6306c148d2a7bd8 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 21:56:19 2006 -0500 cannam@127: cannam@127: Enable threads at bootstrap time, so I get the compiler warnings that cannam@127: I would otherwise ignore. cannam@127: cannam@127: commit 90f3ef0fb9b081f29eae1e1923e94ea3bb29d7ba cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 18:23:15 2006 -0500 cannam@127: cannam@127: made compilable by c++ cannam@127: cannam@127: commit b68d5ed7c28299cf92764bff3ab8b8f06ec1cf00 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 17:39:02 2006 -0500 cannam@127: cannam@127: FIXED: incorrect twiddle_shift() cannam@127: cannam@127: commit b56739cdd0ea335b6ca48c8dd34103316cc43785 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 16:01:51 2006 -0500 cannam@127: cannam@127: Replaced remnants of awake flag with the new enum wakefulness type. cannam@127: cannam@127: commit 8871d572d270aa76dea86073fc11362c6d516c9a cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 11:20:59 2006 -0500 cannam@127: cannam@127: Oops---there is no need to find a free slot. cannam@127: cannam@127: commit fedf131be6c553e13212c16f7a8f474a0e61fed6 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 09:41:58 2006 -0500 cannam@127: cannam@127: Assertions. cannam@127: cannam@127: commit ff66bb4a211ea2640f833ae48bedb1b34a0b47f2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 09:29:55 2006 -0500 cannam@127: cannam@127: Commented the hash table lookup algorithm. cannam@127: cannam@127: commit 4bafb30ddfc85ff74bb758a23532ce60bb621d19 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 5 09:12:00 2006 -0500 cannam@127: cannam@127: Fixed infinite loop in hashtable lookup/insert. Grrr... cannam@127: cannam@127: commit 02a5374038e878b9e0cfe88ee88b0389bf20a255 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 22:04:28 2006 -0500 cannam@127: cannam@127: updated copyright years to 2006 cannam@127: cannam@127: commit b2d48f50aa87d2b9e5f57c6c04959b7ce0984732 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 21:57:23 2006 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 24baeff279c41dbe00c5fd1b13844175e8f70cfe cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 21:52:18 2006 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 490a044a9e2b2f599506ef415c3f87c2b64ba83a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 21:51:40 2006 -0500 cannam@127: cannam@127: more updates for recent pentia/amd cannam@127: cannam@127: commit 21fc6cf5d45450edd194c6d83d328dd7c27c8142 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 4 20:57:47 2006 -0500 cannam@127: cannam@127: Pruned TODO. cannam@127: cannam@127: commit 96c862a6929365a5a78a2196cd72c5037082c5d8 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 4 20:43:41 2006 -0500 cannam@127: cannam@127: Prototype of problem_destroy() cannam@127: cannam@127: commit 700b7dcd5331fe4317b214d64086771a404814ef cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 20:43:13 2006 -0500 cannam@127: cannam@127: rm obsoleted TODOs cannam@127: cannam@127: commit f722e923cd823d4501bc8c3a730fbc09d2c26e06 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 4 20:37:24 2006 -0500 cannam@127: cannam@127: Fallback to 970 if neither -mcpu=power5 nor -mcpu=power4 are supported. cannam@127: cannam@127: commit b5823feffb1b189d536e5c562959969c247a61c3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 20:29:07 2006 -0500 cannam@127: cannam@127: NEWS updates, clarifications, and reorganization cannam@127: cannam@127: commit fffa543ce9d6cb43d2c09bf401c029b5f6830356 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 4 19:54:41 2006 -0500 cannam@127: cannam@127: remove some compiler warnings, add an assert check, make estimator work properly for nop plans cannam@127: cannam@127: commit 3c4889a04995ac9f01ffdb3c4dd0ddc4ef42dc53 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 3 19:34:04 2006 -0500 cannam@127: cannam@127: Two big changes: cannam@127: cannam@127: 1) revised the twiddle generation machinery, to avoid generating cannam@127: twiddles when measuring, and to use a faster O(sqrt(N)) table cannam@127: when this entails no loss of precision. cannam@127: cannam@127: 2) implemented new ALLOW_PRUNING estimator hack. cannam@127: cannam@127: commit 30e3e40e0439f7109a75c063ebb0544bbe68a0c7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 22:08:29 2005 -0500 cannam@127: cannam@127: Estimator tweaks, mostly to favor generic over rader for small n. cannam@127: cannam@127: commit 2e0e06d43cef1259a6fdda21744c8fa71960ea69 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 17:55:47 2005 -0500 cannam@127: cannam@127: Grrr... missing break statement in switch. cannam@127: cannam@127: commit 12348cb25f94416b730862ea4d0a5e85eb2c98b2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 16:08:50 2005 -0500 cannam@127: cannam@127: Swapped fields TW and OPS in struct ct_desc_s, to make k7 asm cannam@127: code insensitive to -malign-double. For consistency, changed cannam@127: struct hc2hc_desc_s in the same way. cannam@127: cannam@127: commit 33a820de9270d537b4079f08fe258a969c410632 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 16:00:42 2005 -0500 cannam@127: cannam@127: Wrong check for infeasible slvndx in imprt(). cannam@127: cannam@127: commit 4b5008a48fbfaf95504f2816b980f971d6678326 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 15:56:59 2005 -0500 cannam@127: cannam@127: Removed obsolete function invoke_solver_if_correct_kind(). cannam@127: cannam@127: commit e1959cade352dd407f5c1c87cf37580ef60f6eb3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 14:22:12 2005 -0500 cannam@127: cannam@127: Faster implementation of safe_mulmod(), avoiding divisions altogether. cannam@127: Works for 0 <= p <= INT_MAX. cannam@127: cannam@127: commit f827b89e687419b19b7133b64651c3a2f10de064 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 24 12:05:54 2005 -0500 cannam@127: cannam@127: FFTW_ALLOW_LARGE_GENERIC must belong to flags->l, it cannot be cannam@127: overridden by fftw. cannam@127: cannam@127: commit 5dbe4dcaa75797cb76e09e4349b526993fb435b2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Dec 23 20:46:24 2005 -0500 cannam@127: cannam@127: no more need for limits.h, add some explanatory comments cannam@127: cannam@127: commit 1dba2396d5d50261e6c82014e279b4ac035120f2 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 23 17:50:25 2005 -0500 cannam@127: cannam@127: Paranoia. cannam@127: cannam@127: commit a09014d7cc40be154096f5b14b0b136985ac39fb cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 23 17:40:41 2005 -0500 cannam@127: cannam@127: Fixed subtle bug involving overflow of the slvndx field in flags_t. cannam@127: cannam@127: commit 1a5304605e6f104eb147f96a5bc76dad55ad9dbf cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 23 16:33:56 2005 -0500 cannam@127: cannam@127: Note 64-bit clean. cannam@127: cannam@127: commit 7d6e177477acee44216776a7afff2306b58eb963 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 23 15:34:32 2005 -0500 cannam@127: cannam@127: Threads are now 64-bit clean cannam@127: cannam@127: commit 208ba330fb9eaaa58a138350dc9f9e965b95bd2c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 23 13:00:31 2005 -0500 cannam@127: cannam@127: Restored the old numbering TW_NEXT=3 etc, because the k7 code depends cannam@127: on it. cannam@127: cannam@127: commit 5a7e2e7cbedf9021d8b278afdd9762f3fe0cc697 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Dec 23 11:58:00 2005 -0500 cannam@127: cannam@127: Portable implementation of MULMOD() and safe_mulmod(). cannam@127: Removed all unnecessary AC_CHECK_SIZEOF() from configure.ac. cannam@127: cannam@127: commit e515294ed8f991b8efb4dc7a0891c16562783679 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Dec 22 11:12:29 2005 -0500 cannam@127: cannam@127: Inline the loop body in r2r codelets like we do everywhere else. cannam@127: cannam@127: commit 94210bafc8387499f631cdd6187ab293943261a2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Dec 22 10:48:53 2005 -0500 cannam@127: cannam@127: Oops. cannam@127: cannam@127: commit 2dcf5d5b1908062b236d6aa2fba93b28937e9488 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Dec 22 10:25:15 2005 -0500 cannam@127: cannam@127: Renamed X(sin_and_cos)() to X(cexp)(). cannam@127: cannam@127: commit de2f6ff5df500a8d15c1cb36f620d277994ec098 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Dec 21 22:49:58 2005 -0500 cannam@127: cannam@127: Somewhat faster generation of twiddle factors. cannam@127: cannam@127: commit 2bda3ba8833c53949694b05f2518b57b2cda80a3 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Dec 20 23:50:01 2005 -0500 cannam@127: cannam@127: tweaks cannam@127: cannam@127: commit 86c8779bcf89bca6fad1812b716a0171b7ab0f91 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Dec 20 22:29:19 2005 -0500 cannam@127: cannam@127: Sped up planner, esp. in estimate mode. The planner now classifies cannam@127: all solvers into DFT, RDFT, and RDFT2, and it only invokes solvers cannam@127: appropriate for the problem being planned. Because we have several cannam@127: hundred solvers, the overhead of calling irrelevant solvers is cannam@127: significant, and this modification mitigates the issue somewhat. cannam@127: cannam@127: commit 98ea24afbd44d88617f25cd467def39b934cbed5 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Dec 19 22:04:00 2005 -0500 cannam@127: cannam@127: Eliminated all calls to sprintf() in favor of own routines, so as not cannam@127: to force users to link stdio and the associated locale/pthreads crap. cannam@127: cannam@127: commit 112a5e19c813a918315e26a80ed9e1f427aa59c3 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Dec 19 21:27:25 2005 -0500 cannam@127: cannam@127: Implemented routine to print INT, removing the need for c99's cannam@127: %td format. cannam@127: cannam@127: commit 5c20f07423e4661b32498afa8071e1f6dacd47c7 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Dec 19 12:06:33 2005 -0500 cannam@127: cannam@127: info->n is size_t cannam@127: cannam@127: commit 6ae75f3b9b700352da7e3ad728d49d988f80e864 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 18 18:15:04 2005 -0500 cannam@127: cannam@127: Explicit casts in front of pointer difference in printf() context, cannam@127: just in case INT != ptrdiff_t. cannam@127: cannam@127: commit 25abe60b6b82d9cab328fbfc8dc17f33ffd6803a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 18 16:52:38 2005 -0500 cannam@127: cannam@127: Forgot to add %D to print.c cannam@127: cannam@127: commit 7e07750df2164e8f8c88185b8857c527f145b444 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 18 16:43:26 2005 -0500 cannam@127: cannam@127: Use %D as format character for type INT. cannam@127: cannam@127: commit 1bf67aff56a4e6b2f0fc41cb8b66e9b09d4b2ea0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 18 15:14:03 2005 -0500 cannam@127: cannam@127: Changed type of an_int_guaranteed_to_be_zero. Changed name as well. cannam@127: cannam@127: commit 602b07fee7f1fbb86b429e682fbce4a4f886e0d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Dec 18 14:41:31 2005 -0500 cannam@127: cannam@127: converted %o -> INT cannam@127: cannam@127: commit e99c67870f4d09190598610fc7c1bd5df8e4515e cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Dec 17 20:28:50 2005 -0500 cannam@127: cannam@127: Major 64-bit cleanup. cannam@127: cannam@127: commit 3cd29a6839b31e093a5c715d6deb2867eafb1b15 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Dec 7 22:39:01 2005 -0500 cannam@127: cannam@127: PGI x86-64 cycle counter, courtesy Cristiano Calonaci cannam@127: cannam@127: commit 7b830d38cb785513bde604f14a3253e171a75e0c cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Dec 5 21:25:57 2005 -0500 cannam@127: cannam@127: Must insert into hash table when wisdom_state == WISDOM_ONLY, cannam@127: otherwise wisdom does not work. cannam@127: cannam@127: commit 9cfa064f6635afd41f01788e5a16a7a56babfca0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Oct 8 18:08:44 2005 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 7fd8f4a4ff768b59317a318d3d83ac0726609868 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Oct 2 11:49:13 2005 -0400 cannam@127: cannam@127: Paranoia: made planner robust against MD5 collisions. cannam@127: cannam@127: commit 55004ef918346e933b7d46aa529fc76258c0b673 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 22:33:18 2005 -0400 cannam@127: cannam@127: Note that --enable-3dnow is unsupported. cannam@127: cannam@127: commit 317d36cb4265710fe5ccbf3518f15f7f24c076cb cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 22:31:04 2005 -0400 cannam@127: cannam@127: * Removed --enable-3dnow support. cannam@127: cannam@127: * SIMD support for split complex arrays. cannam@127: cannam@127: commit 2f87ee31a3c1a416b983aee2ad2441b0624f6839 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 22:28:41 2005 -0400 cannam@127: cannam@127: Removed --enabled-3dnow, since it is becoming useless as the world cannam@127: moves to x86-64, and it is a pain to maintain. (We should probably cannam@127: remove the k7 stuff as well.) cannam@127: cannam@127: commit e5a5da39405e5960f93478937fea04c98feabf49 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 21:59:16 2005 -0400 cannam@127: cannam@127: Missing BEGIN_SIMD(), END_SIMD() statements. cannam@127: cannam@127: commit 7898dae11c979e9b069616b3d922b09b23b8750f cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 12:16:08 2005 -0400 cannam@127: cannam@127: Tweaks cannam@127: cannam@127: commit 3bc850803f4f000f1c979a3576bdd066c37eaafe cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 10:04:32 2005 -0400 cannam@127: cannam@127: Fixed wrong opcount for simd codelets. cannam@127: cannam@127: commit 2c35b6d0d3217976f3597d04403cfac7a4f7da57 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 27 09:25:50 2005 -0400 cannam@127: cannam@127: Fixed wrong opcount for simd codelets. cannam@127: cannam@127: commit 27aa07803ba692bbdbc563607e6531222bb56488 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 26 22:58:19 2005 -0400 cannam@127: cannam@127: fixed flop counts cannam@127: cannam@127: commit 97b8e6bc0d2daddf10da0eb41c94e8e8c4e92bf1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 26 22:34:40 2005 -0400 cannam@127: cannam@127: Silence warnings cannam@127: cannam@127: commit 804b1a4d34edaba87c4aa0f6f7fe3f173bb926f8 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 26 20:52:36 2005 -0400 cannam@127: cannam@127: Implemented split-complex SIMD codelets cannam@127: cannam@127: commit 4c34b9513f4003ec04ebc836dd009d15d4f913f1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 25 22:25:35 2005 -0400 cannam@127: cannam@127: Generalized the ``store pairs'' trick (now called ``store multiple''). cannam@127: cannam@127: commit c8eb4f532fe1b280cd93313eab57b1e51cd6d4cf cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 25 18:58:20 2005 -0400 cannam@127: cannam@127: Silence some warnings. cannam@127: cannam@127: commit 7ecbbeacf952a07cbc1a338fa9bdc9612d99b7bb cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 24 12:37:16 2005 -0400 cannam@127: cannam@127: Removed obsolete cruft cannam@127: cannam@127: commit 9a8a94ca3fb2d0ee33268ae8527f65260631d958 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 19 22:55:19 2005 -0400 cannam@127: cannam@127: Re-enabled check for because OSX requires it. cannam@127: cannam@127: commit 2525a542b0277af07f89f45a3e68c2ac022d4189 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 11 11:03:03 2005 -0400 cannam@127: cannam@127: Check for sizeof(unsigned int) unconditionally, because the cannam@127: result is used by ifftw.h. cannam@127: cannam@127: commit 5750c658cabc6d64ab0f9817312b2399d75f4041 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 11 10:59:40 2005 -0400 cannam@127: cannam@127: Higher size limit for t2 codelets. cannam@127: cannam@127: commit c5134ff6de3bfe5306428398c14cb7dcc9a09afe cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 11 10:50:37 2005 -0400 cannam@127: cannam@127: Heuristic: do not use t2 simd codelets for N>1024. cannam@127: cannam@127: commit 8c4b74a02763d61fd64c98f01fd2658bf80fbc68 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 5 22:22:50 2005 -0400 cannam@127: cannam@127: Larger tolerance in timer calibration routine. cannam@127: cannam@127: commit ed07b941c3be22c7f19c569bd29230c683783b47 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 5 16:03:33 2005 -0400 cannam@127: cannam@127: #include unconditionally. (There is no point in checking.) cannam@127: cannam@127: commit f03e0aced4c470b2b24d8d5abb94be526833a2b1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 5 15:23:27 2005 -0400 cannam@127: cannam@127: Removed SSE and SSE2 asm because it was bitrotting. Use the Intel cannam@127: API instead, which seems to be supported by gcc >= 3.3. cannam@127: Moved files that require -msse, -msse2 to new directory. cannam@127: cannam@127: commit a12a85c774d25cb85391f200a8e6d62da2572cce cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 5 12:56:28 2005 -0400 cannam@127: cannam@127: Parse cputypes of the form 7447A,altivecsupported cannam@127: cannam@127: commit 1d5a7d722689e83fdcccae9edae36ec276b68241 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 5 12:52:30 2005 -0400 cannam@127: cannam@127: Distinguish powerpc 7400 from the 7450, which has a different cannam@127: pipeline. cannam@127: cannam@127: commit b363c2bb7fe126fe80afcd974a463349e63a48a6 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 5 12:46:00 2005 -0400 cannam@127: cannam@127: Paranoia: define RIGHT_CPU unconditionally. cannam@127: cannam@127: commit 558789684b3fa4435a4fab4d86769f2a5ee53b57 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 11 20:56:41 2005 -0400 cannam@127: cannam@127: Removed obsolete name fftw-wisdom2c. cannam@127: cannam@127: commit d73fb7f9d84bc1acccdf9c8f7f2b71e10b3d7854 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 11 20:55:59 2005 -0400 cannam@127: cannam@127: Avoid creation of temporary files---use cpp magic instead. cannam@127: This fix solves a security bug and avoids nonportable tempfile cannam@127: creation hacks. cannam@127: cannam@127: commit a74941c286a12d9a008c3b89ba558cfab82587af cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 5 10:03:02 2005 -0400 cannam@127: cannam@127: Workaround for with gcc-3.3 altivec bug. cannam@127: cannam@127: commit 259f7d688fec2615a29b1aeb22321568cdcc4bc4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jun 15 21:36:46 2005 -0400 cannam@127: cannam@127: solaris fix: check -pthreads first since gcc does not like -pthread but chokes due to stubbed libc (grr) cannam@127: cannam@127: commit 261b7c0fcfaa8c8e6a34d06b051c4355bcac60b1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jun 3 17:19:56 2005 -0400 cannam@127: cannam@127: note that VC++ bug was fixed in 2005 cannam@127: cannam@127: commit 14832d8b25d4091667d3f0e5c8fd8fa1c14f8ce1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 30 16:30:45 2005 -0400 cannam@127: cannam@127: generalized ax_cc_vendor to ax_compiler_vendor cannam@127: cannam@127: commit b13949fd1df86e14fcd73495557bea7532b49b8c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 30 15:55:07 2005 -0400 cannam@127: cannam@127: updated message cannam@127: cannam@127: commit ead701adfc138233d26e86258f0daa8041a41d37 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 30 15:45:14 2005 -0400 cannam@127: cannam@127: update for new AC archive format cannam@127: cannam@127: commit 56c34ca4db1ff26982040ff00e1cb549653ab720 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 23 23:12:22 2005 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit c04871b2f43fe56cd9e921b4864a26ad354cf3f5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 23 18:17:38 2005 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit c4afbfd4ef5235b1b88715bac592b8f091d76d13 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 23 18:13:08 2005 -0400 cannam@127: cannam@127: more notes cannam@127: cannam@127: commit 1cf10c2f758f89da2c0f8bd68f0a8c974e93f33c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 23:37:08 2005 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 568dac7da89c3fe5dbab61ff28e2aa6dc52ca71f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 22:37:50 2005 -0400 cannam@127: cannam@127: note icc 8.x annoyance cannam@127: cannam@127: commit 1b1f5c242db3f55c2dfadb248a9fb292981c5e6b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 22:36:04 2005 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit f66bc7b513029ac91ec983bb3279f3c0dec3468c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 22:35:34 2005 -0400 cannam@127: cannam@127: note gcc 3.4.[0123] bug, which is fixed in gcc 3.4.4 cannam@127: cannam@127: commit 0f2a7eb61a2bcf44583bd41245ad55c7e78eb70f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 22:21:26 2005 -0400 cannam@127: cannam@127: added automatic detection of icc architecture flag cannam@127: cannam@127: commit 7b90a23bc9ceeeb03131b4774aa0ff5d04e91c63 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 21:47:19 2005 -0400 cannam@127: cannam@127: add -no-gcc to icc flags...even if it is Intel's fault, I'm sick of dealing with bug reports about this cannam@127: cannam@127: commit ff0439a0bc1dc149d302630cb96062fc7fb053f1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 21:40:59 2005 -0400 cannam@127: cannam@127: added @cindex portability cannam@127: cannam@127: commit e18637fa933a8a75ef831024c4c966d6a2dff76b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 21:34:10 2005 -0400 cannam@127: cannam@127: note --without-gcc-arch cannam@127: cannam@127: commit 7131ee53a750ff084f05b97c67e34a39e1a7011c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 20:54:54 2005 -0400 cannam@127: cannam@127: bsd ppc detection; some odd 603 types cannam@127: cannam@127: commit 7f439b2ab6289af0e08134c659480f9589b13387 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun May 22 11:53:20 2005 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 32419ec5a48e285cbcbee2f0a4c49e628fcf6ccb cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 21 20:34:52 2005 -0400 cannam@127: cannam@127: ensure no spaces in cputype cannam@127: cannam@127: commit 7a6288d8a7617720cb8c46fc9152a31c7dab793a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 21 20:31:41 2005 -0400 cannam@127: cannam@127: nevermind cannam@127: cannam@127: commit b9bac647b7039e381615e0faac27fc3a8de06eb4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 21 20:30:08 2005 -0400 cannam@127: cannam@127: more bsd stuff cannam@127: cannam@127: commit f1c985e46f8c17122e47ece0e9696258638be1f1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 21 20:28:40 2005 -0400 cannam@127: cannam@127: added BSD cpu detection for SPARC and better super/hypersparc detection cannam@127: cannam@127: commit e35c028649be9cc1568401e9e39eb2e19d1cda3b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 21 20:22:11 2005 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit a0582b1056c2562cd639c18f2827fc124dd79fa6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 20 19:40:09 2005 -0400 cannam@127: cannam@127: "alternate" == "alternative" is US-centric cannam@127: cannam@127: commit 333d9eb5086ed1afa77719e9f24142a8bd5dada9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 20 19:36:26 2005 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit e2d0b93f5de6abb830a0d28324399d4689850b09 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 20 01:28:34 2005 -0400 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit f8a4a4af8c47ae8e572e5f169c0eeb0720eb7473 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 17 18:56:46 2005 -0400 cannam@127: cannam@127: print out estimate-planner time from can_do in verbose>2 mode cannam@127: cannam@127: commit e1bbc2ce6ff2b094ad3549a5140d6acd0218b7d8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 9 00:47:19 2005 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 2e2b68117557549932c89d24586be1852a189462 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 5 23:47:55 2005 -0400 cannam@127: cannam@127: fixes for building Windows DLLs with Cygwin; thanks in part to Stephane Fillod cannam@127: cannam@127: commit bb8fc9fb4dda639b9f0b1f13ef448e39d71a4b39 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 22 19:47:43 2005 -0400 cannam@127: cannam@127: -ffast-math seems to produce code that is either about the same speed or slightly faster (gcc 3.3 and 4.0, x86) cannam@127: cannam@127: commit 2f7b1f2707810c171bb85b330c99a94196a257d0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 22 19:18:23 2005 -0400 cannam@127: cannam@127: power5 fallback to power4 sched for older gcc's cannam@127: cannam@127: commit 169cba437dfb6f553bb1a8e2a404ca2bf74a5b56 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 22 19:14:53 2005 -0400 cannam@127: cannam@127: check for power5 cannam@127: cannam@127: commit 1978d7cd087b7e6e93133c7b4aa2c612f664203d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 19 21:55:13 2005 -0400 cannam@127: cannam@127: Removed clause #3 cannam@127: cannam@127: commit 3c385073178a321cc4108d4b88f121276b5d0020 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 19 21:44:57 2005 -0400 cannam@127: cannam@127: license clarification cannam@127: cannam@127: commit ab865d9025afbb6c923e94956c3e7ebdd64ef75d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 19 21:42:51 2005 -0400 cannam@127: cannam@127: Changed license of fftw3.h to X11. cannam@127: cannam@127: commit d851f36c4ff5e1febbc2ed47cb08eba3f8dbaf19 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Apr 11 13:15:12 2005 -0400 cannam@127: cannam@127: delete fixed-input code cannam@127: cannam@127: commit cc673385bfc98894c37272241fcb6135756d2c14 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Apr 10 16:33:24 2005 -0400 cannam@127: cannam@127: joned L-U-planner branch cannam@127: cannam@127: commit d4b2b38d4a6b40919a6229bb574ecd49884ad58f cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 7 23:15:02 2005 -0400 cannam@127: cannam@127: ref cannam@127: cannam@127: commit 8895af84fb9e4970420b21451977fde49072c2b9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 7 00:11:13 2005 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 6dbfe38e27a7f4a5090917f8b53a03e334a40881 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 6 22:06:21 2005 -0400 cannam@127: cannam@127: added (optional) new split-radix algorithm, enabled with -newsplit; also new -standalone option to omit desc; also -unitary, -normalization, and -normsqr options to generate r2r codelets with various normalization (to match lit. in DCT-II, use: -unitary -normsqr 2) cannam@127: cannam@127: commit 5e1deadac7dbe4d60d493b86f66b37474388b11e cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 25 08:59:43 2005 -0500 cannam@127: cannam@127: Moved timeout check outside the search loop, because X(seconds) is cannam@127: expensive. cannam@127: cannam@127: commit 094cbe955f1ad43c143f7781eb524ede71d164bc cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 20 18:35:53 2005 -0500 cannam@127: cannam@127: Enable vector recursion for in-place problems, otherwise cannam@127: dftw-genericbuf works only in PATIENT mode. cannam@127: cannam@127: commit 14a9b596a784705637abb9cd5a47595ed2a4bcbd cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 20 17:53:58 2005 -0500 cannam@127: cannam@127: oops cannam@127: cannam@127: commit 7ea889cca28101323df5287b988ee6bd96c531a0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 20 17:49:13 2005 -0500 cannam@127: cannam@127: make solver UGLY for small N cannam@127: cannam@127: commit a4abcfa708787e3e18b32fc37506992215578c4b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 20 17:16:37 2005 -0500 cannam@127: cannam@127: new dftw-genericbuf solver cannam@127: cannam@127: commit 70997fbe34952f59b14245e68e5fd4614d13c3ac cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 20 16:12:44 2005 -0500 cannam@127: cannam@127: new dftw-genericbuf solver cannam@127: cannam@127: commit 3d40d10cca6f0fb8ed0e327ae23d569829a43768 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 17 21:48:19 2005 -0500 cannam@127: cannam@127: Hmm... what was I thinking? cannam@127: cannam@127: commit b27eff441bd1e24148569ed9ee02c05c08b46ea4 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 17 19:20:54 2005 -0500 cannam@127: cannam@127: Workaround for a MSVC bug. cannam@127: cannam@127: commit 433960d78aef7dc12c5611baa3213b4db99f99cc cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 17 08:18:39 2005 -0500 cannam@127: cannam@127: Workaround for a MSVC bug that was reported by Eddie Yee. cannam@127: cannam@127: commit 0c4f3dfe86c936003eed705208a100c11a5bcce6 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 15 13:25:53 2005 -0500 cannam@127: cannam@127: try both contiguous input and contiguous output when in doubt cannam@127: cannam@127: commit 155f07c46c6589d374f886a8ed86f985a64642e3 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 15 08:44:41 2005 -0500 cannam@127: cannam@127: Added genfft flag -precompute-twiddles which moves the computation of cannam@127: the twiddle factors before the main schedule. This flag produces cannam@127: smaller code everywhere, and slightly faster code on powerpc. cannam@127: I observe no speed difference on x86. cannam@127: cannam@127: commit 5cc6165f9756f2faeab137eed5f8c25ebac08773 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 14 21:43:53 2005 -0500 cannam@127: cannam@127: sp cannam@127: cannam@127: commit 255c6db9915f31c3b323cee61a7900999c7b4cfe cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 14 21:43:05 2005 -0500 cannam@127: cannam@127: whoops, spelling error (thanks to Steve Eddins for bug report) cannam@127: cannam@127: commit 556965536b7671795bc6e4ef86edfffe75b2ffd9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 12 15:03:47 2005 -0500 cannam@127: cannam@127: Do not approximate pcost = vl * child->pcost unless child is guaranteed cannam@127: not to be a simple codelet. cannam@127: cannam@127: commit a5282a50ce6211585a443fa099e2fa6e47450ceb cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 9 20:00:02 2005 -0500 cannam@127: cannam@127: Relaxed applicability conditions. cannam@127: cannam@127: commit 2496640b61c0ac594325d4fa68e3729873c004bd cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 9 00:05:47 2005 -0500 cannam@127: cannam@127: Minor optimization cannam@127: cannam@127: commit 81c49148f9fb58b0c541b2636b37bd8422a458b1 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 8 22:14:02 2005 -0500 cannam@127: cannam@127: Interpret K to mean *1024. Similarly for M. cannam@127: cannam@127: commit b94f2eb04282f6c3c511944e3767a8895a19ef77 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 8 20:44:25 2005 -0500 cannam@127: cannam@127: Hmm... somehow some previous commit got lost. cannam@127: cannam@127: commit 55b8abdbbc3a2bbb26f005735bd9d121634c4055 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 8 20:30:42 2005 -0500 cannam@127: cannam@127: Paranoia cannam@127: cannam@127: commit 752db4c71fd1a447d9ed1699ed0382e042d4f89c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 7 14:30:01 2005 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 6c18ecea25e2a9f685131b49c7365fc35b8c4c7c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 7 14:29:43 2005 -0500 cannam@127: cannam@127: move fftw-specific HP/UX tweak into configure.ac cannam@127: cannam@127: commit 3916e3b25257834172ce4eb126a2d745b8943123 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 7 14:19:24 2005 -0500 cannam@127: cannam@127: ax_cc_family -> ax_cc_vendor (vendor names are easier to remember), add checks for many new compilers, use in ax_cc_maxopt cannam@127: cannam@127: commit a0ad3ef6add8118e82611c08b4c252ec8346efea cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 6 21:36:05 2005 -0500 cannam@127: cannam@127: Count FMA as one flop in estimator when HAVE_FMA cannam@127: cannam@127: commit 10a57b3a5a428bab777ec22f4eb83203498a743a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 6 19:16:06 2005 -0500 cannam@127: cannam@127: Do not try radix-2 generic. cannam@127: cannam@127: commit e38ef2e30e6f41fb2301acf208ff7f9b775de0ac cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 6 13:04:23 2005 -0500 cannam@127: cannam@127: Use -O3 for xlc now that we use -O for CODELET_OPTIM cannam@127: cannam@127: commit be3c47c96bcc1ef146a296202b53db7a457b3230 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 6 13:02:41 2005 -0500 cannam@127: cannam@127: New AX_CC_FAMILY macro, that detects the compiler based on symbols cannam@127: that it defines (as opposed to the name of the compiler). cannam@127: We need to start use this strategy everywhere else. cannam@127: cannam@127: commit 562882d5c889b0bce256013a056ce07f55c27dfb cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 6 11:33:15 2005 -0500 cannam@127: cannam@127: Runtime checks to guarantee small strides. cannam@127: cannam@127: commit 1fcf24126783752b3ab8f35f480a0e5d0fa90aab cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 5 20:09:25 2005 -0500 cannam@127: cannam@127: Reduced the search space for rank-0 transforms cannam@127: cannam@127: commit 77cbffe7c30bbac4d294cd2c7321163054732418 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 4 17:50:29 2005 -0500 cannam@127: cannam@127: little assert cannam@127: cannam@127: commit 495b9d7617c0167346817c4d5620fe80ee2d1194 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 1 09:19:16 2005 -0500 cannam@127: cannam@127: Implemented directbuf, enabled for now. cannam@127: cannam@127: commit 1869b027f29cef23f101026dee512744fba87eaa cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 28 22:21:14 2005 -0500 cannam@127: cannam@127: Unified dftw-direct, dftw-directbuf in an attempt to tame code cannam@127: growth cannam@127: cannam@127: commit 82fce69cd912d2a58b86a5699c04d2eea3b9a536 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 27 13:51:24 2005 -0500 cannam@127: cannam@127: fixed copyright cannam@127: cannam@127: commit 7d1a5530230d76d105f3ed4aeebdf4f708ed0e8a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 26 22:21:03 2005 -0500 cannam@127: cannam@127: silence warnings cannam@127: cannam@127: commit 753ab3b636f099eedb841e643898aed3e8c5c817 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 26 22:19:16 2005 -0500 cannam@127: cannam@127: oops cannam@127: cannam@127: commit a64fecb2ccd2670c6b37d40d70558d553e4cb17d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 26 21:28:39 2005 -0500 cannam@127: cannam@127: Tweaking while thinking about a higher-rank transposer (bitreverser) cannam@127: cannam@127: commit 9c7a7d3c45be7ca132fdece876ebea7eb053fad7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 26 20:06:49 2005 -0500 cannam@127: cannam@127: Transposed the buffer, and skewed it. This allows for contiguous cannam@127: copy operations, and the codelet should not incur associativity cannam@127: conflicts if the buffer is large. cannam@127: cannam@127: commit 521fa92ebcf99b32b35cb4c26b304f42a2812e22 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 26 18:14:11 2005 -0500 cannam@127: cannam@127: make tensor_max_index more reasonable (take maximum of input and output cannam@127: max indices, computed separately) cannam@127: cannam@127: commit c6c2bcbb2b8c8f3b1da7d5465e4bee93905c8d32 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 26 10:04:30 2005 -0500 cannam@127: cannam@127: Use cpy2d instead of cpy2d_tiled, because vl may be too large. cannam@127: cannam@127: commit 269e71f3db6c3d1bcf8dc77e25983dcc9989d5f7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 26 00:31:52 2005 -0500 cannam@127: cannam@127: Fixed old bug that was introduced with yesterday's changes. cannam@127: cannam@127: commit e769a1735dd71165677025498471db8a41271198 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 25 21:54:23 2005 -0500 cannam@127: cannam@127: ``Interesting'' switch statement. cannam@127: cannam@127: commit 7e729390b41355c7abf6c2a3901dec6cb40c4c23 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 25 12:29:54 2005 -0500 cannam@127: cannam@127: Disabled -reorder-loads -reorder-stores, since they seem to do cannam@127: nothing. cannam@127: cannam@127: commit 4350026ea3252e1dbc25b1539941ee79b3cb6124 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 25 12:19:10 2005 -0500 cannam@127: cannam@127: Because of the recent changes to kernel/pickdim.c, splitrnk=0 is no cannam@127: longer equivalent to splitrnk=1 for rnk < 4, where the latter is the cannam@127: FFTW2 behavior. For small rnk, however, I observe the planner to pretty cannam@127: consistently choose the FFTW2 behavior (splitrnk=1), despite its not cannam@127: being asymptotically optimal in the cache oblivious sense. So, make cannam@127: splitrnk=1 instead of splitrnk=0 the default in FFTW_MEASURE and cannam@127: FFTW_ESTIMATE modes (rnk > 3 is pretty rare in practice anyway). cannam@127: cannam@127: commit 3bfeb642d11098a707ca70b7332077b6472917d6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 25 00:33:27 2005 -0500 cannam@127: cannam@127: tweak cannam@127: cannam@127: commit 24560b26faac0a352c23e15c892c38a762bbb453 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 25 00:29:09 2005 -0500 cannam@127: cannam@127: slight relaxation cannam@127: cannam@127: commit cadf7b9d5561d14d8042ad3b051f7f95a010cb1f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 25 00:21:00 2005 -0500 cannam@127: cannam@127: cruft cannam@127: cannam@127: commit 42d46a1c8af18b951c978ee2cf1cc57ca106929f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 25 00:03:14 2005 -0500 cannam@127: cannam@127: added experimental indirect-transpose solver: when transforming the columns of the matrix, allow us to do a transpose to make the DFTs contiguous cannam@127: cannam@127: commit eec7f69ff78e1b95f1bdd09a2f96b3be5cf1b407 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 23:04:58 2005 -0500 cannam@127: cannam@127: check for abort() cannam@127: cannam@127: commit e1d0f900a4e4444b4ef0fa230de11da87a48a192 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 23:04:43 2005 -0500 cannam@127: cannam@127: call abort() on failed assertion cannam@127: cannam@127: commit 4d8aee345fa2da4b2383722a482d245d38288dad cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 21:17:23 2005 -0500 cannam@127: cannam@127: Forgot to change X(isqrt) -> isqrt_maybe cannam@127: cannam@127: commit 47e79fca2a795dcd96ecf59852cdc53bc883f9d1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 20:18:59 2005 -0500 cannam@127: cannam@127: require finite_rnk cannam@127: cannam@127: commit 7e29047649fc202d7061c007ce3ba8a3962ed38c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 20:07:38 2005 -0500 cannam@127: cannam@127: #ifdef HAVE_STRING_H must come after rdft.h so that we get config.h cannam@127: cannam@127: commit d0b93533d99e69f85e2aaf759989f652311206ac cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 18:59:40 2005 -0500 cannam@127: cannam@127: Implemented reordering of loads and stores so that the real and cannam@127: imaginary part are loaded/stored together. This should improve cannam@127: out-of-cache performance in the presence of associativity conflicts, cannam@127: and maybe worsen in-cache performance because of worse scheduling. cannam@127: Enabled for now, for experimental purposes. cannam@127: cannam@127: commit 827ad1c139031037135765c5600dcf05b58030e4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 18:10:49 2005 -0500 cannam@127: cannam@127: fix comment cannam@127: cannam@127: commit 35e5d61fd3b5f769ea631e357ac6f55002f74f96 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 18:10:23 2005 -0500 cannam@127: cannam@127: better message cannam@127: cannam@127: commit d2c6d9c9d37a6ea058c48c7445fbaca7089a6489 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 18:08:36 2005 -0500 cannam@127: cannam@127: use gcc version > 3.0 as fallback in check for alignment bug cannam@127: cannam@127: commit 9efbf189a95137e78b39f48e223e66df384eb89c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 24 18:02:31 2005 -0500 cannam@127: cannam@127: don't use -malign-double unconditionally (it is only available on x86) cannam@127: cannam@127: commit 858b560880b60856698a28728dd44964d456b7cf cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 12:03:30 2005 -0500 cannam@127: cannam@127: Subtler selection of tilesz. cannam@127: cannam@127: commit c44a6cff160e0ecd38f2a4f56bff4e34ddda2b59 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 11:52:25 2005 -0500 cannam@127: cannam@127: Call cpy2d_tiledbuf, not cpy2d_tiled. cannam@127: cannam@127: commit 826a2387489dd9efde0ed09afc92e91e50a6d578 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 11:29:28 2005 -0500 cannam@127: cannam@127: buffer sizes were wrong :-( cannam@127: cannam@127: commit fdabdfc4ef5010ed7965168b1ab583c296db3637 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 11:19:01 2005 -0500 cannam@127: cannam@127: Single function for computing tile size. Eliminate spurious assertions. cannam@127: cannam@127: commit add19c2d3c32f843ff951cc227dc4ce1221fafb6 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 10:00:02 2005 -0500 cannam@127: cannam@127: Do tiling recursively. cannam@127: cannam@127: commit 203fc5647fea6fe99f2d23cc43a24eeea47aee49 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 24 09:40:30 2005 -0500 cannam@127: cannam@127: Reworked tiled transposes; provide tiling with and without buffering. cannam@127: I can't believe that one has to waste his life with this @#$%. cannam@127: cannam@127: commit c92a1fc69c9315d97f71a3070003d37923ac02b8 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 23 22:21:19 2005 -0500 cannam@127: cannam@127: Clarified logic. I am not sure why the code was so confusing to begin cannam@127: with. The computation of *dp in the which_dim == 0 case was also cannam@127: wrong, returning e.g. *dp == -1 if sz->rnk == 1. cannam@127: cannam@127: commit 44692fa46d7313f08a624ec68bd421e282fa139f cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 23 22:00:15 2005 -0500 cannam@127: cannam@127: Enable aggressive inlining in codelets only, to avoid code bloat. cannam@127: cannam@127: commit e94240f1731b33ff9ad18ffe4c14a08a7d66d65a cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 23 21:51:50 2005 -0500 cannam@127: cannam@127: Removed cache-oblivious copy/transpose algorithms in favor of cannam@127: explicitly blocked algorithms. The cache-oblivious algorithms fail if cannam@127: there are associativity conflicts, in which case buffering is cannam@127: necessary, as per Carter and Gatlin. Once you set the buffer size, cannam@127: there is no point whatsoever to do the algorithm recursively, and you cannam@127: may as well use blocking. cannam@127: cannam@127: commit 77aeedee308c8b7bce0ff4c36986f715ced6748c cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 23 18:46:12 2005 -0500 cannam@127: cannam@127: --disable-fortran now differs from --enable-fortran that fails cannam@127: cannam@127: commit 3cb3e167e76d53336c1307cecb6b1eb975bdda61 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 23 18:42:21 2005 -0500 cannam@127: cannam@127: comment tweak cannam@127: cannam@127: commit e0f881c48bd199f098eaa764fb17982cf1435475 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 23 18:41:14 2005 -0500 cannam@127: cannam@127: If a Fortran compiler was not detected, just make our best guess at cannam@127: what wrappers to use...I'm sick of dealing with user complaints from cannam@127: cases where wrapper detection fails for whatever reason. cannam@127: cannam@127: commit aa2c11cd3b47c6352d13b8f869f858082bb7a52a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 23 18:10:40 2005 -0500 cannam@127: cannam@127: fflush(stdout) after print_plan, in case F77 doesn't cannam@127: cannam@127: commit 76bdaf349e332587c7b5b4ae1fe55f4d3c0cc92d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 22:54:42 2005 -0500 cannam@127: cannam@127: --enable-sse is necessary after all, to generate all dependencies cannam@127: correctly. cannam@127: cannam@127: commit 5844ac653fc5e937e4f2939d8a73dcc282657fd2 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 22:32:06 2005 -0500 cannam@127: cannam@127: Put cpy2d_pair into its own file, so that I can experiment with cannam@127: buffering of nontwiddle codelets. cannam@127: cannam@127: commit e7d485c4f71be2a762c91d4d7e96a321afdfe858 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 20:07:11 2005 -0500 cannam@127: cannam@127: Copy rfftwnd.png from ${srcdir}, not $PWD cannam@127: cannam@127: commit 0c56019ec6dc8f3c778b628a8a0b6094cd8a31d0 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 17:08:48 2005 -0500 cannam@127: cannam@127: Do not bother memcpy-ing complex numbers. cannam@127: cannam@127: commit 6accb53a30744a5793b451670a70afb371cceeff cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 16:20:46 2005 -0500 cannam@127: cannam@127: Tighther layout of buffers. I am not sure it matters, but just in case... cannam@127: cannam@127: commit 0f5938fa6bcc89ad947656aa949a89feb73b7c77 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 10:13:02 2005 -0500 cannam@127: cannam@127: Usec cpy1d for rank-0 copies cannam@127: cannam@127: commit 24a0b716253a1914882d738969bc8b101b70380f cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 22 10:06:13 2005 -0500 cannam@127: cannam@127: Implemented in-place transposes with buffering. Moved cannam@127: copy/transposition routines into own files, so that we can reuse them cannam@127: from multiple places. TODO: merge vrank3-transpose.c with rank0.c, or cannam@127: rename vrank3-transpose.c to rank0-fancy.c or something like that; cannam@127: decide whether square in-place transposes should be in rank0.c or cannam@127: vrank3-transpose.c; apply FIXME's in vrank3-transpose.c. cannam@127: cannam@127: commit 52f669f4280a8ad0834f201919290dc382898a4c cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 21 23:29:52 2005 -0500 cannam@127: cannam@127: Indentation should be printed after newline, not at the beginning cannam@127: of print() cannam@127: cannam@127: commit decdf03722050f50fba24b8152927c2327109e16 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 21 10:07:24 2005 -0500 cannam@127: cannam@127: generalized in anticipation of more complicated solvers. cannam@127: cannam@127: commit 2a7b91a46dd814576f0dbfa54f17d38380bd35f0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 20 22:18:59 2005 -0500 cannam@127: cannam@127: Implemented buffered recursive transpose cannam@127: cannam@127: commit 4ce9d94def9d52633bb76b107aba65caa8c4fcf4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 20 18:27:29 2005 -0500 cannam@127: cannam@127: Fixed comment cannam@127: cannam@127: commit ac7a99027ee51e48f6be6dadcf00eb593d6017d9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 20 18:22:15 2005 -0500 cannam@127: cannam@127: grand unification of rank0 solvers cannam@127: cannam@127: commit 20af4f6724d7080f17a83aae996a6fd00e08ae7b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 20 15:35:24 2005 -0500 cannam@127: cannam@127: manual tail-recursion optimization cannam@127: cannam@127: commit e834b974175d946c82b66c99c7bf18593f85cd8c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 17:57:44 2005 -0500 cannam@127: cannam@127: implemented check for transpositions cannam@127: cannam@127: commit 6f6c5d224ae74b757b7013102ab25c018d7f9a30 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 17:28:43 2005 -0500 cannam@127: cannam@127: Previous fix was wrong for rdft2 problems. cannam@127: cannam@127: commit 6bd660a504ef0345ea0f55db133690f9de7218ec cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 17:23:36 2005 -0500 cannam@127: cannam@127: vecsz->rnk must be finite for this solver to apply. cannam@127: cannam@127: commit 05d2a86385b2655cca135d882688ff493eccaa22 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 17:15:19 2005 -0500 cannam@127: cannam@127: unified the various simple'' transposers cannam@127: cannam@127: commit e67ffc01608a1ebeedd99bb1390ff0ad58e33c0c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 16:55:29 2005 -0500 cannam@127: cannam@127: Fixed stupid bug in rec_transpose_swap. Fixed stupid verifier that did not catch the bug. cannam@127: cannam@127: commit 49f3542f8f1ee7aa2bc7ddb12ded96d4b330b452 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 15:24:03 2005 -0500 cannam@127: cannam@127: Minor cleanup of transposition routines. cannam@127: cannam@127: commit 770952578791d8ac1394ba8e19890fce2779ad67 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 19 09:31:14 2005 -0500 cannam@127: cannam@127: Make the batch size B=Theta(r) instead of B=Theta(1) in buffered cannam@127: twiddle solvers. Theory: for cache line size L, we want B = Omega(L) cannam@127: to utilize the cache line fully. We also want B*r =O(Z), where Z is cannam@127: the size of the cache. It is safe to assume that Z = Theta(L^2): cannam@127: cache designers will tend to make L as large as they can get away cannam@127: with, because they don't have to program the machines that they build, cannam@127: and Z < Theta(L^2) will screw up the little matrix transposition cannam@127: benchmarks that they use to design the cache. Hence, B=Theta(r) is cannam@127: the right number. cannam@127: cannam@127: commit 0fc1650f8f411bc3fd1b6019b33d8e67d54b43a3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 18 23:47:22 2005 -0500 cannam@127: cannam@127: for --enable-portable-binary, only try -mcpu=$arch and -m$arch on x86, cannam@127: since these generate non-portable code on every other target (and cannam@127: some other targets, like Alpha, don't support -mtune=$arch). cannam@127: cannam@127: commit 77be37a9825edf45432db688f9b6e307fc779320 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 17 21:15:42 2005 -0500 cannam@127: cannam@127: gcc/aix defines _POWER, not __powerpc__ like the rest of the world cannam@127: does. cannam@127: cannam@127: commit da4852a84de13f2ed74462052a1081a8517fac9c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 16 22:30:27 2005 -0500 cannam@127: cannam@127: enable fma for ia64, since it seems to help with the hpux compiler. cannam@127: cannam@127: commit e9b2b83177aabb8ff8d42f4b239e9eda1fbd10bf cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 16 21:47:48 2005 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 9f01f364832d025554f5912bd4f71c3c0b972d5c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 16 15:27:18 2005 -0500 cannam@127: cannam@127: Fixes for darwin cannam@127: cannam@127: commit ff3f2d0d66afc832a1ec7f70d14e6d1520e40858 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 16 14:27:42 2005 -0500 cannam@127: cannam@127: Made the correctness of the code more obvious. cannam@127: cannam@127: commit 0eaea796c7d8dfc833c38cc2485c68004bcb9d4c cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 16 12:30:29 2005 -0500 cannam@127: cannam@127: s/with-portable-binary/enable-portable-binary/ to be GNUlly correct; I'm sticking with --with-gcc-arch=arch, however, as --enable-gcc-arch=arch has the wrong connotations for me cannam@127: cannam@127: commit 1f54539fae28f217a239c3dbc5c66a31784dbcd9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 16 11:44:48 2005 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 743d6f8aa35cf29485b805e657e72afb83e401cf cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 16 11:23:38 2005 -0500 cannam@127: cannam@127: bless wisdom with patience used to create it cannam@127: cannam@127: commit 741a55c0cb7529ae5ce8b1b3a01375a3f176a5e0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 16 11:18:56 2005 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit ab2c1f6788b6309abe08b585fa21ac7254e02f07 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 16 10:50:28 2005 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit aa37add40de415143b25c5c3fa09d3d212af9ec2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 15 23:53:53 2005 -0500 cannam@127: cannam@127: added 'timed' planner option cannam@127: cannam@127: commit 79f70936e6e19cb09dafb45f8ead8d9fff715111 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 15 23:08:29 2005 -0500 cannam@127: cannam@127: Do not use SIMD_CFLAGS. The theory is that if taint.c is unsafe cannam@127: with SIMD_CFLAGS, then all files in this directory are as well. cannam@127: Conversely, if these files require SIMD_CFLAGS because they include cannam@127: "simd.h", then taint.c requires SIMD_CFLAGS as well, and thus we need cannam@127: some other hack. cannam@127: cannam@127: commit f9e6da507bcacf5aa503ce42e7cd73c0c501cbe2 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 15 22:49:05 2005 -0500 cannam@127: cannam@127: Do not override CFLAGS in Makefile.am. cannam@127: cannam@127: commit 932e8f656a8a592700a3ca153c416e3e1504d278 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 15 10:30:12 2005 -0500 cannam@127: cannam@127: Allow users to build long double version even if sizeof(long double) cannam@127: == sizeof(double) cannam@127: cannam@127: commit b35d88cdc10b06342c2c39a8d2012a71875aecf6 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 14 19:55:38 2005 -0500 cannam@127: cannam@127: Updated for 3.1 cannam@127: cannam@127: commit 3c20661d7ca87a19ec855d94791bd24a3202e30d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 14 19:07:14 2005 -0500 cannam@127: cannam@127: Oops, version.h is no longer used cannam@127: cannam@127: commit 485e6dbbea69f8e6438ec11fdb265cbe3b786464 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 14 18:51:05 2005 -0500 cannam@127: cannam@127: unified fma and non-fma versions cannam@127: cannam@127: commit 800ea93e6f610aa9a7c15f1e9e7ed779dedefcfa cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 14 14:12:09 2005 -0500 cannam@127: cannam@127: forgot to remove inplace/Makefile from configure.ac cannam@127: cannam@127: commit 48bfe71f273d592eb0010911c4df16e12df1b9b4 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 14 12:08:52 2005 -0500 cannam@127: cannam@127: Merged dft/codelets/inplace with the main dft/codelets/standard cannam@127: directory. This step makes dft codelets consistent with the rest cannam@127: of the naming conventions, and will simplify the eventual merge cannam@127: of fma and non-fma codelets. cannam@127: cannam@127: commit 1f70ee8f508d17b3cb0b694d838c71d4b411d740 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 14 11:16:15 2005 -0500 cannam@127: cannam@127: inline altivec constants, since gcc seems to generate better code this way. cannam@127: cannam@127: commit 454930e2baceefbda8523cfbc103db0061604799 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 13 18:17:32 2005 -0500 cannam@127: cannam@127: group altivec constants into a single array, for faster access cannam@127: cannam@127: commit 6cfc3df81b5b843ac0641d7aff61b76d29f82a63 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 13 18:15:37 2005 -0500 cannam@127: cannam@127: code cleanup cannam@127: cannam@127: commit e8d683e0260b327eeedec8e25249bfd8c81cdda9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 13 10:29:32 2005 -0500 cannam@127: cannam@127: removed some unused stuff cannam@127: cannam@127: commit d495f6e14d219a63d1ed2a3e77e2c526e185a82c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 12 22:04:40 2005 -0500 cannam@127: cannam@127: New twiddle scheme for altivec, 3dnow cannam@127: cannam@127: commit 510cdba23c47b1838f8a027da5680ad9ff21dcf3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 12 20:17:35 2005 -0500 cannam@127: cannam@127: Implemented new twiddle scheme for sse2 cannam@127: cannam@127: commit fd74e1eb06f6460dc3f0d8b6c5504fc005f98806 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 12 19:57:46 2005 -0500 cannam@127: cannam@127: Implemented experimental t2* codelets, which store twiddle factors cannam@127: in a more convenient format, at the expense of twice the storage. cannam@127: Currently only SSE works; I have to port SSE2, altivec, etc. to the cannam@127: new scheme. After this, we will decide whether these codelets cannam@127: are worth the price. cannam@127: cannam@127: commit 9ba2ad18ff0a5c9a683120d7737cc6d343b83246 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 11 08:07:12 2005 -0500 cannam@127: cannam@127: Forgot to define SIMD_STRIDE_OKPAIR cannam@127: cannam@127: commit 24aa1c39dc04c158a5275310b779bec639962a38 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 10 22:20:00 2005 -0500 cannam@127: cannam@127: fixed sse2, 3dnow, and altivec, as promised cannam@127: cannam@127: commit fa8ee16c80d02c0a0a19391f9aa5897b37ac004b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 10 21:47:40 2005 -0500 cannam@127: cannam@127: Generate n2?v_* codelets in such a way that we may or may not cannam@127: pair stores, depending on which mode happens to work best on cannam@127: a particular SIMD implementation. sse2, 3dnow, and altivec cannam@127: are currently broken---will fix soon. cannam@127: cannam@127: commit 8a141e0f8570683466ef4cf2aa4e8027d7ea698e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 10 08:53:22 2005 -0500 cannam@127: cannam@127: instantiate altivec constants only once cannam@127: cannam@127: commit b23eef5ad62b650caafba583fae089d173718eac cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Feb 10 06:37:56 2005 -0500 cannam@127: cannam@127: Fixed alignment checks for new SIMD scheme cannam@127: cannam@127: commit bf8b613b6a4299e8fcc3b36c1c0ec6c61ae944d6 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 9 21:35:01 2005 -0500 cannam@127: cannam@127: Change n2?v_* codelets to store pairs of vectors, with implicit cannam@127: 2x2 transposition. Works for 2-way SIMD as well. Tested with sse cannam@127: and sse2. I haven't tried altivec yet, but I observed a huge cannam@127: speedup when I transformed one codelet by hand. cannam@127: cannam@127: commit b45f5e7af8fe63c291238eded48cff440ad1f4b9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 8 21:28:38 2005 -0500 cannam@127: cannam@127: Resurrected old DIF codelets for experimental purposes. They cannam@127: are disabled for now, but I am keeping the setup around for cannam@127: future reference. cannam@127: cannam@127: commit 2b2271e7df0c994e8ed02a49304a2ef279c084d2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 8 20:10:19 2005 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit c06695785e699d90aab66ce15e718ccab31f42bc cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 8 19:37:09 2005 -0500 cannam@127: cannam@127: clarifications, document --with-portable-binary and --with-gcc-arch cannam@127: cannam@127: commit 4658829ef2505ec43aab6986fdc4778314c3e0bf cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 8 19:23:41 2005 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 44be70997db3875b83dfe5dee436014717bdf235 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 8 01:36:22 2005 -0500 cannam@127: cannam@127: more change comments cannam@127: cannam@127: commit b7802bbb738b279d8d061756f90f03caecd0767a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 8 00:41:38 2005 -0500 cannam@127: cannam@127: fma is definitely beneficial on Itanium with the HP/UX compiler cannam@127: cannam@127: commit 95f76ca2081a043388616e815c0364bc6ffde166 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 7 22:58:47 2005 -0500 cannam@127: cannam@127: Silence warnings. cannam@127: cannam@127: commit fe63ebfa96d081c7d45183e96a8d904d3dcfd226 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 7 22:55:49 2005 -0500 cannam@127: cannam@127: when we compile our own getopt, change symbol names to avoid conflicts (e.g. avoid build failure on MacOS X with --enable-shared) cannam@127: cannam@127: commit 151717343ac9ebd9197dfa0065de4176fa9d0894 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 7 22:36:42 2005 -0500 cannam@127: cannam@127: grr, more bugfixes for in-place case cannam@127: cannam@127: commit ca853db7099972e3b3840be7d1d3ee1abff00d04 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 7 22:29:35 2005 -0500 cannam@127: cannam@127: removed relics of FRANZ mode cannam@127: cannam@127: commit b5015c430276d969565a9b6fe816a55556f8d6f7 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 7 18:48:36 2005 -0500 cannam@127: cannam@127: Somehow xlc does not like ``vector int dummy;'' cannam@127: cannam@127: commit e8ba7b5c1885c85755dd33973ec8d2c5305f41e9 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 7 13:59:47 2005 -0500 cannam@127: cannam@127: There is no need to enable sse to make the distribution. This might cannam@127: have been true in the past but not anymore. cannam@127: cannam@127: commit fea3ce788e0bd8cfd350e05d05c418e90b27ec63 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 7 13:55:17 2005 -0500 cannam@127: cannam@127: Oops---included fortran file in C sources cannam@127: cannam@127: commit 2f4c935bb52c2e34940f4ad58ea6fd26ba30740f cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 7 13:42:45 2005 -0500 cannam@127: cannam@127: Set version string at ``make dist'' time, not at ``configure'' time, cannam@127: so we know whether a user is using the fma version or not. cannam@127: cannam@127: commit fcd17cfa8271300c8a41d87c9abd4968502ebaca cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 6 17:00:33 2005 -0500 cannam@127: cannam@127: Removed useless files cannam@127: cannam@127: commit 2707963bd735e791f7f5b8200c8c9d4f155bc4f8 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 6 16:59:39 2005 -0500 cannam@127: cannam@127: Different (simpler?) way to prevent the compiler from optimizing loop cannam@127: inductive variables. We now explicitly corrupt stride variables by cannam@127: xor-ing them with another variable that happens to be zero (but the cannam@127: compiler does not know it). In this way, the compiler does not cannam@127: attempt to extract a zillion loop indices from codelets, which would cannam@127: overflow the register set. Set the -fno-loop-optimize flag to further cannam@127: help the process. cannam@127: cannam@127: Consequences: removed m* codelets. Smaller library size. Slightly cannam@127: faster code with gcc/powerpc (including altivec). Much faster code cannam@127: with xlc/powerpc. No changes for gcc/pentium. Maybe slightly faster cannam@127: with icc/pentium. cannam@127: cannam@127: commit 1e222893c8c84f35b16a63384ad1239e471ce684 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 5 18:51:08 2005 -0500 cannam@127: cannam@127: paranoia about in-place rodft00 plans cannam@127: cannam@127: commit 1d442744933c7161e86dd825d65aeb3d0c640e53 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 5 18:39:55 2005 -0500 cannam@127: cannam@127: don't believe pcost when using the estimator...there is no point, and cannam@127: it screws up estimator hacks to prefer in-codelet loops to vecloops cannam@127: cannam@127: commit 9ad39d1cad4ef56e0c29fc64a12a76e2e6195c52 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 5 18:34:25 2005 -0500 cannam@127: cannam@127: Reduced optimization level from -O3 to -O for xlc, since -O generates cannam@127: faster code. cannam@127: cannam@127: commit 91fa9ff722538be49b29c22a3174cef3fdce9c25 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 5 16:26:58 2005 -0500 cannam@127: cannam@127: whoops, only applicable to redft00/rodft00 plans cannam@127: cannam@127: commit 6591b1e69eec3c3d11199ec3f84c341aa8e754db cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 5 16:22:39 2005 -0500 cannam@127: cannam@127: fixed in-place operation, and don't create size-0 sub-plans cannam@127: cannam@127: commit f01834e572803db476083af9b0a0906b951ac9d0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 4 11:30:30 2005 -0500 cannam@127: cannam@127: Autodetect altivec on linux. This code works with gcc-3.4 and cannam@127: -maltivec, with or without -mabi=altivec. The code *should* work with cannam@127: gcc-3.3 without -mabi=altivec. However, disabling -mabi=altivec on cannam@127: gcc-3.4 produces much worse code (I don't know why). cannam@127: cannam@127: commit 2ac42677bbf31c868ad589a378f93887163910c1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 28 00:04:58 2005 -0500 cannam@127: cannam@127: update reference cannam@127: cannam@127: commit 2f3db335dab469a165ed2d9a4f19435371ef9590 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 27 15:48:28 2005 -0500 cannam@127: cannam@127: note that DCT-II/III are often called the'' DCT/DCT cannam@127: cannam@127: commit a2480b0a7742cb4792f0a17ef54fcfa47bf9299f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 21 14:42:04 2005 -0500 cannam@127: cannam@127: added MSVC++ for ia64 (based on information at http://www.intel.com/cd/ids/developer/asmo-na/eng/19949.htm?prn=Y) cannam@127: cannam@127: commit fa86c1be03f8a3ac77ad8f17a4a0db76b8a08d04 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 21 14:22:50 2005 -0500 cannam@127: cannam@127: vc++ defines _M_AMD64 on x86-64, apparently cannam@127: cannam@127: commit 905e261576a2333fdc356609f6f6533740716663 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 18 22:30:27 2005 -0500 cannam@127: cannam@127: avoid gratuitous breakage with -Werror, requested by Simon Perreault cannam@127: cannam@127: commit 6fb09d4fad8df9be7c5cadda330234fbcf6bdecd cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 17 18:54:55 2005 -0500 cannam@127: cannam@127: comment typo cannam@127: cannam@127: commit 2f9aac9cff6654101febb130659eab9345b58783 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 16:56:23 2005 -0500 cannam@127: cannam@127: bumped shared-lib revision# cannam@127: cannam@127: commit c793a51d3eafa054b132ebbc6095810261ac56b6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 16:35:42 2005 -0500 cannam@127: cannam@127: add X(estimate_cost) to get estimator cost, and print from bench, to aid in tweaking estimator cannam@127: cannam@127: commit ef81def3aef05a8e513d2c28f9eba162af22020b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 14:57:56 2005 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 1b90ee6f155399994c4234601dfdce43c854555e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 12:57:07 2005 -0500 cannam@127: cannam@127: formatting fix cannam@127: cannam@127: commit 2abab58ebcf1286120285091b31ff706fa81cf81 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 12:31:28 2005 -0500 cannam@127: cannam@127: tweaks cannam@127: cannam@127: commit 044466122b66a254d87c396cbf0b17039543fd13 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 12:03:24 2005 -0500 cannam@127: cannam@127: use less buffer space cannam@127: cannam@127: commit 3e78c0361397476b699825b883be3d32331e8439 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 15 01:41:58 2005 -0500 cannam@127: cannam@127: added split-radix-based dct/dst I for odd n cannam@127: cannam@127: commit d994d2ded5077bfb54d19ee5c062e607b73ce73a cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 14 21:50:08 2005 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit cf8ef77af5eddfdda0d6c952ae0ae1955890bca4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 14 21:49:55 2005 -0500 cannam@127: cannam@127: warn silly users who confuse CVS id with FFTW version cannam@127: cannam@127: commit e7ab0f25025fb3be5f73408419e51a2fcf54f031 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 14 16:57:36 2005 -0500 cannam@127: cannam@127: get sparc cpu type on solaris as well as with linux cannam@127: cannam@127: commit e82ef68d349c8df79cb772c944164b79b7f2c77a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 13 19:21:58 2005 -0500 cannam@127: cannam@127: detect prescott mobile (f37) cannam@127: cannam@127: commit 3622c28434b7292df2153c577f8262a2974fd6ce cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 13 18:09:52 2005 -0500 cannam@127: cannam@127: use cpuid for x86_64 as well as i[56]86 cannam@127: cannam@127: commit ba6d8352bbd435da164d15a693e824711bcd86ce cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 13 17:59:55 2005 -0500 cannam@127: cannam@127: update with x86info 1.7 and other sources (identify k8, nocona, etc), handle nonzero leading bytes in eax cannam@127: cannam@127: commit 92d9e4b244a2689bc7fb64105d20c874d09f9cca cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 13 16:30:33 2005 -0500 cannam@127: cannam@127: compactified check for JOINABLE; use AC_DEFINE_UNQUOTED instead of AC_DEFINE for PTHREAD_CREATE_JOINABLE (thanks to Oliver Niekrenz for the bug report) cannam@127: cannam@127: commit 5440f786f094cdfb2b624e1e9050ba74a06ad780 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 12 12:22:13 2005 -0500 cannam@127: cannam@127: The scheduler hack was incorrect because it swapped instructions cannam@127: of the form A = *B and *B = C. Fixed. cannam@127: cannam@127: commit 124a19a9d293ffa06f8b50519fc1e53ced2ca1ab cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 11 22:13:24 2005 -0500 cannam@127: cannam@127: Quote expressions such as ``if test $FOO = yes'' when $FOO may be cannam@127: empty. Also, $GCC is set to either ``yes'' or empty, never to ``no''. cannam@127: cannam@127: commit d52e4f122a2b71ab9272261bfec25931b8d9cd5b cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 11 19:30:47 2005 -0500 cannam@127: cannam@127: Hmm---somehow the previous commit did not work. cannam@127: cannam@127: commit ca5f6331f2b2432591707b129dc343705209e482 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 11 16:54:45 2005 -0500 cannam@127: cannam@127: Fixed various gcc-related problems on powerpc: cannam@127: - gcc-3.4 becomes totally confused by expressions like cannam@127: vec_add(a, vec_add(b, vec_add(c, ...))) cannam@127: The compiler uses gigabytes of memory and then crashes, presumably cannam@127: because of the exponential-time search problem involved in typing the cannam@127: above expression (since vec_add can take either ints or floats). cannam@127: I changed VADD and similar macros to be inline functions, thus cannam@127: constraining the type system. cannam@127: cannam@127: - New flags cannam@127: --param inline-unit-growth=1000 --param large-function-growth=1000 cannam@127: to work around limitations of the gcc-3.4 inliner. cannam@127: cannam@127: commit 43a34b10c3383ccf1277216826b8201c3a0f3276 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 21:27:24 2005 -0500 cannam@127: cannam@127: Check for HAVE_ALTIVEC_H cannam@127: cannam@127: commit 558d64554efbc303c104513b4f6243d2178335a8 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 21:09:30 2005 -0500 cannam@127: cannam@127: Remove support for altivec using gcc builtins, since these keep cannam@127: changing across gcc versions. These changes work on gcc-3.4/linux; I cannam@127: haven't tried MacOS X yet. (The altivec ``spec'' differs between cannam@127: Motorola/Apple and gcc, grrr...) cannam@127: cannam@127: commit d9289c88276c6c878bd61c454049052d420013fa cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 18:57:30 2005 -0500 cannam@127: cannam@127: Stylistic changes cannam@127: cannam@127: commit 34b131fad38155ebb215614d02fba749c35c10ed cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 17:34:41 2005 -0500 cannam@127: cannam@127: Changed incorrect ugliness condition. cannam@127: cannam@127: commit 79acbd2e7f6820db37155925cdb1411a808b5bb4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 10 16:09:43 2005 -0500 cannam@127: cannam@127: note x86info version number that was used, to make it easier to update cannam@127: the cpuid for changes in later versions cannam@127: cannam@127: commit 114d644618ae24c93784d908c47981dfbd32719e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 15:00:51 2005 -0500 cannam@127: cannam@127: Make dft-r2hc non-UGLY for rank-0 problems cannam@127: cannam@127: commit 1a81406a03ef105ec8c188ee2f77dd605c25d422 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 14:50:23 2005 -0500 cannam@127: cannam@127: Do not use -mcpu=970 on power4 processors, because power4 does cannam@127: not have altivec. cannam@127: cannam@127: commit 393ce48d0f58f8f7788198d59fa203e19f36db69 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 14:48:47 2005 -0500 cannam@127: cannam@127: Note gcc-3.4 problem with inlining. cannam@127: cannam@127: commit c25eb53aa1676a746b9243a7463a62e3d753fb0f cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 13:51:08 2005 -0500 cannam@127: cannam@127: Oops, forgot to remove ``static'' from the declaration of noninlinable cannam@127: functions. cannam@127: cannam@127: commit 1447d501267177b4d3f4b5160a7e3b4fc16e7aba cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 10 12:31:26 2005 -0500 cannam@127: cannam@127: Recognize power4. Use ``head -n COUNT'' instead of obsolete ``head cannam@127: -COUNT'' (which fails on gentoo). cannam@127: cannam@127: commit 3de5bb754f1b9eb2514402c3b542a3735009f223 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 22:12:16 2005 -0500 cannam@127: cannam@127: Remind to add FAQ entry concerning gcc-3.4.[1-3] crashes. cannam@127: cannam@127: commit 669ca8a3c4968477bf695ebc2961279779e0ec37 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 9 21:53:08 2005 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit dcaa702e5c8d172b42b79a0c8ae14a1c8525f0a3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 9 21:48:02 2005 -0500 cannam@127: cannam@127: support checking for major.minor.patchlevel cannam@127: cannam@127: commit 584fa85e1d7ca47d71b72c14f7dab1ac448048ec cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 21:40:18 2005 -0500 cannam@127: cannam@127: Revert CODELET_OPTIM to -O on IA32, which is faster than -O2. cannam@127: cannam@127: commit ec5ec6cbc0d0325a26eda54206f7f17253b39bae cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 20:30:12 2005 -0500 cannam@127: cannam@127: /bin/sh allows no spaces in assignments. cannam@127: cannam@127: commit 2b5a7ef73ab8bd55c32f63badf3120d1c4a62a28 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 20:05:55 2005 -0500 cannam@127: cannam@127: Make non-inlinable functions external, so that gcc becomes confused cannam@127: and does not try to inline them. cannam@127: cannam@127: commit 321304bb50c85a0d5353f7bed5116d33b865dc4b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 13:44:25 2005 -0500 cannam@127: cannam@127: Add -fno-web to CFLAGS, because -fweb destroys FMAs. cannam@127: cannam@127: commit 2f4f3044ed140d5b0edf1cf7415e0c0035392b40 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 10:31:47 2005 -0500 cannam@127: cannam@127: Allow -mcpu=970 besides -mcpu=G5 cannam@127: cannam@127: commit e00f75f258a1b31526633b408804ed3c231cef68 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 10:26:20 2005 -0500 cannam@127: cannam@127: configure was not using -fno-schedule-insns :-( cannam@127: cannam@127: commit e86fb1669da7d88ee98278e686d078ed205237c3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 08:52:40 2005 -0500 cannam@127: cannam@127: In mkplan() and elsewhere, use solver index instead of solver cannam@127: *pointer*, which looks marginally clearer. cannam@127: cannam@127: commit 446a3894d345237cabc59f659d5a2186c1f26554 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 9 08:15:36 2005 -0500 cannam@127: cannam@127: Split planner hash table into two tables, for blessed and unblessed cannam@127: solutions respectively. Now an unblessed solution never overwrites a cannam@127: blessed solution, thus avoiding wisdom leakage by construction. cannam@127: Further, forget() is now a O(1) operation, which speeds up the cannam@127: estimator when the wisdom table is large. cannam@127: cannam@127: commit ee5380a2af1b55803ff5d64557ff5b9a2005b54b cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 8 21:19:45 2005 -0500 cannam@127: cannam@127: New TODO idea. cannam@127: cannam@127: commit 8bf4164bfd1d6aef62dac0e09eb5c5ef712ed8f4 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 6 11:02:29 2005 -0500 cannam@127: cannam@127: Split search() into two routines to make the UGLY/NO_UGLY logic cannam@127: obvious. cannam@127: cannam@127: commit 1f170904d9848a43935bbd9a7c95d0249fa39138 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Dec 17 16:08:54 2004 -0500 cannam@127: cannam@127: push/pop 64-bit registers on ia64; thanks to Orion Poplawski for the fix cannam@127: cannam@127: commit c53a0b8fa44ee6e63d41cdf2e4eb12589981f43d cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Dec 9 21:41:09 2004 -0500 cannam@127: cannam@127: patch from FreeBSD ports - FreeBSD does not have memalign, but its cannam@127: malloc is 16-byte aligned cannam@127: cannam@127: commit 31b763b9455632deddfb6425b630c4ce458b444e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Nov 23 17:06:47 2004 -0500 cannam@127: cannam@127: don't compile taint.c with SIMD_CFLAGS (fixed Debian bug #259612) cannam@127: cannam@127: commit 5f505f2c11b292e769afc7de1e1fbb9bb75d1495 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Nov 18 11:37:32 2004 -0500 cannam@127: cannam@127: revert incorrect change -- codlist.c should be rebuilt, but it is built in the build directory and not in the source directory cannam@127: cannam@127: commit 247e871cccf86dee2fa5543473c76373e5c46b34 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Nov 17 22:53:53 2004 -0500 cannam@127: cannam@127: $(CODLIST) should be rebuilt only if Makefile.am changes, or cannam@127: alternatively only in maintainer mode, to prevent stomping in the cannam@127: source directory during user builds. (Thanks to Grant Cook for the cannam@127: bug report.) cannam@127: cannam@127: commit 7b6e452ba1709033b19a1056184ef5e7865773c3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Nov 13 13:43:01 2004 -0500 cannam@127: cannam@127: corrected #ifdef for icc/ia64, thanks to Matt Boman cannam@127: cannam@127: commit 80176573959dd2f034b41ab5d38c541281a5987a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Nov 13 13:34:55 2004 -0500 cannam@127: cannam@127: spelling correction (Larsen, not Larson) cannam@127: cannam@127: commit 4e72b0ba4a2ee4245a1c996aabcea979753ded6e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Nov 8 22:12:39 2004 -0500 cannam@127: cannam@127: use standard withval cannam@127: cannam@127: commit 38a050f2474601bd6fc7f1e9faca33e8656f0a63 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Nov 8 22:09:16 2004 -0500 cannam@127: cannam@127: match doc cannam@127: cannam@127: commit 8d34c77d933aba00013d63875fb0a8cfdb5c5058 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Nov 8 22:00:34 2004 -0500 cannam@127: cannam@127: formatting cannam@127: cannam@127: commit f354a059a0559c7816da1f1bfcbf30fef2965584 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Nov 8 21:59:33 2004 -0500 cannam@127: cannam@127: make sure OPENMP_CFLAGS environment variable is used correctly cannam@127: cannam@127: commit caffdb38e0d057c260d21dcd45fee9d04ba48520 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Nov 8 21:46:50 2004 -0500 cannam@127: cannam@127: replace ax_check_cc_flags with more generic ax_check_compiler_flags cannam@127: cannam@127: commit bc44b190250c3a55ddc841fdb85623efef8a1d04 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Nov 8 17:49:42 2004 -0500 cannam@127: cannam@127: separate macro for OpenMP test cannam@127: cannam@127: commit 7bdd20309c710d7f29cb11cd2a130a2a453252ca cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Nov 5 16:24:22 2004 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit ba62ab6d94914626b1bb5c4fa59d239a92f5789a cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Oct 29 00:48:13 2004 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit bbe80b4b34e5e86fb09b40b44a0f686b07bbd17b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Oct 28 00:09:38 2004 -0400 cannam@127: cannam@127: better guessing of sparc type on Linux cannam@127: cannam@127: commit 93d85f0ab3a391bf35f1eb8c51e0d693736fa416 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 27 13:44:08 2004 -0400 cannam@127: cannam@127: note default cannam@127: cannam@127: commit 78065724b3f4e1170788d4d75cc1c1e318663b06 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 27 13:41:57 2004 -0400 cannam@127: cannam@127: tweak cannam@127: cannam@127: commit e43858fa862ad22519805870bef8be66593db88c cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 27 13:34:25 2004 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit af53c27b20c589cc956cc567f7a85d05e5f9996d cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 27 13:31:10 2004 -0400 cannam@127: cannam@127: whoops, m4 is EXTRA_DIST, not SUBDIR, since it doesn't have a Makefile cannam@127: cannam@127: commit fc7444822d899746b1c4e68cb06847ce95ff12b7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 27 13:16:57 2004 -0400 cannam@127: cannam@127: silence warnings cannam@127: cannam@127: commit be281108e1c825de4313ece30b12fd918273b1a9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 27 13:14:22 2004 -0400 cannam@127: cannam@127: clean up m4 macros; try to detect correct gcc -march flag on x86; new --with-portable-binary, --with-gcc-arch= flags; use -O2 for codelets with gcc 3.4 to work around bug cannam@127: cannam@127: commit 9403174ddea85728f959287755950e43901c2d39 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Oct 26 16:46:14 2004 -0400 cannam@127: cannam@127: rename cexp -> mcexp to avoid conflict with C99 builtin cannam@127: cannam@127: commit d581a67939f4d7c95a0b07b3a4952d35e44bb17c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Oct 25 16:58:23 2004 -0400 cannam@127: cannam@127: use basename , w/o args, for compiler-name comparisons; also detect Compaq ccc on alpha-linus cannam@127: cannam@127: commit a1d9fccd5bdda57ae410ba0ce15367e987d64f73 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Oct 24 22:05:10 2004 -0400 cannam@127: cannam@127: note recent icc problems cannam@127: cannam@127: commit dfddc484065adab609af43ba17821394c23dc5cd cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Oct 24 02:10:12 2004 -0400 cannam@127: cannam@127: whoops, disable semaphores again (for now) cannam@127: cannam@127: commit a2dad5feeb7e13cc8d93adb55bb59ed0431341be cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Oct 24 02:04:58 2004 -0400 cannam@127: cannam@127: POSIX semaphores are *not* the same as SYSV semaphores cannam@127: cannam@127: commit 64a5d0fd73897b1f811382dc5238209dfe9672be cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Oct 24 01:18:14 2004 -0400 cannam@127: cannam@127: re-implement threaded stuff; dftw now takes parameters to indicate a portion of m loop cannam@127: cannam@127: commit 99fecf91b80dfe5aabdd4b3d69cc71639de2c483 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Oct 21 20:44:51 2004 -0400 cannam@127: cannam@127: more C++ notes cannam@127: cannam@127: commit 77e885e9ca5d60b2b34f126b21cce95382cafc59 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Oct 14 09:50:38 2004 -0400 cannam@127: cannam@127: note bug report for VC++ 6.0 from Dale Dickerhoof cannam@127: cannam@127: commit e5523dbd23cd0d4beff0d5b53ca76a275b7e5b1f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Oct 1 16:06:59 2004 -0400 cannam@127: cannam@127: fmt cannam@127: cannam@127: commit 8e9f882720c1fc5f2c7c3b168a8f48608af95057 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Oct 1 15:59:17 2004 -0400 cannam@127: cannam@127: comment typo cannam@127: cannam@127: commit 689ac491bc35a2728b1ae0ccc6e1698f84a04f4f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Oct 1 15:48:09 2004 -0400 cannam@127: cannam@127: bug fix -- ishift/oshift only apply to execution of child plan cannam@127: cannam@127: commit 6438e86b96980ae10958e4483acf04e80573c1dd cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 30 21:12:47 2004 -0400 cannam@127: cannam@127: New planner that tries never to lose wisdom. cannam@127: cannam@127: commit 28f9e28b2b6d2c7d5969c93cdf3c460f6fd895a2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 30 13:36:43 2004 -0400 cannam@127: cannam@127: Nested comment was triggering a warning. cannam@127: cannam@127: commit 7f1f6a5fe5723ce3079588306a98c43289f6df32 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Sep 10 15:20:07 2004 -0400 cannam@127: cannam@127: system "root" under dgjpp is /dev/env/DJDIR, not /dev/env/DJGPP, cannam@127: according to djgpp's libc.info; patch confirmed with J. M. Guerrero cannam@127: cannam@127: commit 354611ae36fd7494d3f90789fa33d6b26febeec2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Sep 8 18:50:03 2004 -0400 cannam@127: cannam@127: some minor portability fixes for djgpp; thanks to Juan Manuel Guerrero for the patch cannam@127: cannam@127: commit 133be56f2adeb3f4ab3c394a03da4254f758eacf cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 19 12:41:23 2004 -0400 cannam@127: cannam@127: pointer to tutorial for quick start cannam@127: cannam@127: commit 6a23ed45415cae1a9825953e80dc99ceee5d185a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 19 12:39:50 2004 -0400 cannam@127: cannam@127: point users to manual cannam@127: cannam@127: commit b759a1ca992dee63a97a67ba2beddde782dba6c7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 7 13:42:22 2004 -0400 cannam@127: cannam@127: minor typo cannam@127: cannam@127: commit 693ed3bc9f47a262b3502ad06b42be41f68ee47f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 18 18:54:18 2004 -0400 cannam@127: cannam@127: use __DECCXX for Compaq cxx, not Linux-specific symbol cannam@127: cannam@127: commit df4ddeeaad67144bd7d6f855f690cf06907f1d56 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 16 13:55:25 2004 -0400 cannam@127: cannam@127: patch by John Bowman to make cycle counter work with DEC cxx under Linux cannam@127: cannam@127: commit fd9cd11e5b8806245d6b5522fdef29b1626eda0f cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jun 30 00:45:10 2004 -0400 cannam@127: cannam@127: updated pruned FFT discussion, with link to further details on www.fftw.org/pruned.html cannam@127: cannam@127: commit 243e4dafca54e62e83d796c176d4af2ce00690b1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jun 14 20:08:27 2004 -0400 cannam@127: cannam@127: darwin is based on freebsd cannam@127: cannam@127: commit fa86af755d34199fa6ddf2a1e40dbedb9898f5bb cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jun 3 14:23:41 2004 -0400 cannam@127: cannam@127: in --with-windows-f77-mangling, add lowercase + single underscore for Intel compilers, etc. (thanks to David Gomez for the bug report) cannam@127: cannam@127: commit 3f13a0eb176fe03d5937ef282b9defa42c258876 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 7 00:46:07 2004 -0400 cannam@127: cannam@127: whoops, extra alignment check cannam@127: cannam@127: commit f0e8345ba78e99831a1589192d6fc3b2e1e41e38 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 7 00:16:49 2004 -0400 cannam@127: cannam@127: disable most 2-float-as-double copying, add alignment check in one remaining place cannam@127: cannam@127: commit a12d8b846381396d54acbc1748e53cbc0c09baac cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 6 13:49:13 2004 -0400 cannam@127: cannam@127: make sure it is clear that real-even/odd refers to symmetry, not size cannam@127: cannam@127: commit 920197fd649070eadef659b39572b155a8b0c36c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Apr 5 20:18:29 2004 -0400 cannam@127: cannam@127: optimization cannam@127: cannam@127: commit c0d199f22910faaf1f4850900185c161a585f96b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 2 21:31:00 2004 -0500 cannam@127: cannam@127: separate cutoff for ugliness...these cutoffs are still not ideal cannam@127: cannam@127: commit e1920963de856b058811b84764d848947cab454f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 2 21:30:17 2004 -0500 cannam@127: cannam@127: transpose.c is gone cannam@127: cannam@127: commit a115ba2703fa6d7cfb8e1453904bf94cd9c25b7d cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 2 21:18:27 2004 -0500 cannam@127: cannam@127: move all rank0 transforms to rdft cannam@127: cannam@127: commit 444b8f48586e952b107d4ee2ad58c56e357e5fbd cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 2 20:35:35 2004 -0500 cannam@127: cannam@127: enable fp-moves/us comparison of rank-0 transforms cannam@127: cannam@127: commit 2cc4d9f4818fb41d6aa0c1be4224eb25a94b3ac7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 1 16:13:22 2004 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit b0ee7083fd7d7ff73366c8011fbf43d675380d8c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 1 15:25:30 2004 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 50854b83979e79e4a0a2f6e90404ca553d0a3d33 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 31 18:11:02 2004 -0500 cannam@127: cannam@127: sort tensor dims by stride absolute values, not strides cannam@127: cannam@127: commit 39cd8178427b5a70d5fa503a14c663c6a4f96edf cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 30 20:22:50 2004 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 9b5e15aa7f7e515b9faa505be708a14d568ceb1b cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 30 19:44:54 2004 -0500 cannam@127: cannam@127: added improved transpose algorithm for N x M where |N-M| is small cannam@127: cannam@127: commit 2db4ea7db1f587486546e244ea42930e51275806 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 30 19:41:14 2004 -0500 cannam@127: cannam@127: check to make sure SIMD matches precision, and make sure user doesn't select both SSE and SSE2 cannam@127: cannam@127: commit 8995d09da3bc5d0fd7daf6f6ad295fccd9e94893 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 28 09:26:38 2004 -0500 cannam@127: cannam@127: Implemented hc2hc-generic hc2r. cannam@127: cannam@127: commit 581a83475a46b89a73b8a7fb3f2dccb140f72629 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 25 11:19:25 2004 -0500 cannam@127: cannam@127: Inverted loop for stride-1 access. cannam@127: cannam@127: commit b0d68fa533d26d6bebf433c43c4dfee7b99a3701 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 25 11:18:49 2004 -0500 cannam@127: cannam@127: Swapped j <-> k for consistency cannam@127: cannam@127: commit dc715359aa2bb496a60ae650612b42f0cdf998dc cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 23 12:08:07 2004 -0500 cannam@127: cannam@127: Require that R be odd cannam@127: cannam@127: commit 36e2199cf602f511e50a4bbc56e472d79c935e8f cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 23 11:49:01 2004 -0500 cannam@127: cannam@127: Implemented hc2hc-generic (DIT only for now). cannam@127: cannam@127: commit bc377e92e0d11be803dc1a3deb60f05a82799f85 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 22 14:43:16 2004 -0500 cannam@127: cannam@127: Relax equality of twiddle description, since the `i' field cannam@127: is not used by TW_FULL or TW_HALF. cannam@127: cannam@127: commit ede9d975b188649b84cca9bf24c5f7feab3653c4 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 22 13:22:44 2004 -0500 cannam@127: cannam@127: Do not allocate tw_instr's on the stack. Thus, the ``consistency check'' cannam@127: in twiddle.c becomes wrong. cannam@127: cannam@127: commit 19b8fbca72260c622266cd93466267c9dfb57cc3 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 22 13:21:28 2004 -0500 cannam@127: cannam@127: Fixed incorrect malloc()/free() logic. cannam@127: cannam@127: commit 050be8cad10f411ab6ca025f59e5cffc3f7bf42d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 22 09:04:37 2004 -0500 cannam@127: cannam@127: Silence warnings cannam@127: cannam@127: commit ae20d94938c08cb65f257fec653a9e3b1961a77b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 22 09:02:55 2004 -0500 cannam@127: cannam@127: Separate file for hc2hc common routines cannam@127: cannam@127: commit e35b856a11108d5bdf61855976cddd7e8e7a84e9 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 22 08:23:56 2004 -0500 cannam@127: cannam@127: (re)Implemented buffered hc2hc. Slight simplification of cannam@127: twiddle-factors management. cannam@127: cannam@127: commit de8ff3b06710f0dda76007150592239d4aa7565c cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 19:53:05 2004 -0500 cannam@127: cannam@127: Incremented libtool revision number before we forget. cannam@127: cannam@127: commit 5004b2e13de1b9b8635441bba800e8f6b850900a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 19:25:56 2004 -0500 cannam@127: cannam@127: Fixed opcnt cannam@127: cannam@127: commit af360d8473ebdda79f57a3fa6bd3bbb2b7b041a2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 17:56:15 2004 -0500 cannam@127: cannam@127: Renamed files. These solvers are not really cooley-tukey. cannam@127: cannam@127: commit c6c735fb857127becb133e21c37544052b985806 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 12:38:45 2004 -0500 cannam@127: cannam@127: Started moving rdft/ to the new cooley-tukey ontology cannam@127: cannam@127: commit 5df5843950df1fb50697f28d983ff0a9b8d5c5b9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 10:59:42 2004 -0500 cannam@127: cannam@127: Plans in ct-*.c are subtypes of plan_dftw, not plan_dft cannam@127: cannam@127: commit 5a4eb1dc842c864c311f175e9f97dde3f42dba2f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 10:38:18 2004 -0500 cannam@127: cannam@127: Slight simplification cannam@127: cannam@127: commit 23b338208a4b752b307a0c6ff8d03f4e3f3c077f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 21 10:20:06 2004 -0500 cannam@127: cannam@127: Minor simplification cannam@127: cannam@127: commit 71c684955ce3dfb91065a561e806edc213b2a1ae cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 20 08:43:57 2004 -0500 cannam@127: cannam@127: Workarounds for icc-8.0 nonsense. cannam@127: cannam@127: commit 446cbae42c628ae2ba7e6f63f4771355a10b5e0f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 7 07:56:08 2004 -0500 cannam@127: cannam@127: FFTW_FORWARD is not technically an ``option''. cannam@127: cannam@127: commit 150af2bf6e6d380dc31ebffdcb79961e64d47f97 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 24 12:17:06 2004 -0500 cannam@127: cannam@127: Alejandro requested that his name be removed from @author cannam@127: cannam@127: commit 6948af91f0140722c52246a2b09faaeb7e664d99 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 23 17:42:56 2004 -0500 cannam@127: cannam@127: GNU Pth emulation library check cannam@127: cannam@127: commit b28089821d98c117e9688fdb7c65b4bfc0645345 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 21 17:51:13 2004 -0500 cannam@127: cannam@127: calling can-do calls the estimating-planner, which creates wisdom that we don't want ...we should be able to do all of the documented problems, anyway cannam@127: cannam@127: commit 262bd966d1d8394d17b2dc1ae7b76446b9300323 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 21 17:46:06 2004 -0500 cannam@127: cannam@127: don't forget_wisdom because of side effects cannam@127: cannam@127: commit b5c61a6821de885155d1b960a1d8b50a5464bc3d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 21 17:42:47 2004 -0500 cannam@127: cannam@127: forget wisdom from can_do cannam@127: cannam@127: commit 51442d30f17617100834e2fb27cbe7df79b3d61b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 19 14:11:14 2004 -0500 cannam@127: cannam@127: parenthesization cannam@127: cannam@127: commit 63bf06148e526cb5c90550fbd7b53a40fc73f2d4 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 13 07:20:31 2004 -0500 cannam@127: cannam@127: Split malloc into kernel_malloc and API malloc cannam@127: cannam@127: commit 26fb1d12ecc37fda0f9760386b1f59a87e193e01 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 12 15:42:20 2004 -0500 cannam@127: cannam@127: X(malloc) must be extern "C" cannam@127: cannam@127: commit bb95c42e188e35ab4f22703978e5da7ba796eaff cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 12 15:41:44 2004 -0500 cannam@127: cannam@127: satsify C++ compiler cannam@127: cannam@127: commit 5560fa42b5df5500dfd63303262c412f308ceb76 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 5 20:39:14 2004 -0500 cannam@127: cannam@127: with the new flags, fma is definitely beneficial on PA-RISC with HP/UX cc cannam@127: cannam@127: commit 795e5b6919fec5bbdf1fb9cff3be1db5f63ddeee cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 5 19:52:17 2004 -0500 cannam@127: cannam@127: grr, Ofaster etcetera are not supported under older versions of the compiler. Note that +Ofltacc *disables* fp-reordering optimizations (which are enabled by +Oall). +Optrs_ansi is the older version of the aliasing stuff cannam@127: cannam@127: commit db287e0973e7d6ef19261b5a96979dff3b339b9a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 5 19:26:01 2004 -0500 cannam@127: cannam@127: +Otype_safety=ansi on hpux cannam@127: cannam@127: commit c98916ab1aa18fd95e8e43584e6eb618015de573 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 5 19:22:34 2004 -0500 cannam@127: cannam@127: just use +Ofaster on hpux (+O3 +Onolimit +Olibcalls +Ofltacc=relaxed -Wl,+mergeseg) cannam@127: cannam@127: commit 4e3bf163dca0615df17146ee2e18481d0a20a9e6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 30 14:17:15 2004 -0500 cannam@127: cannam@127: check for win32 threads for mingw32; thanks to Alessio Massaro cannam@127: cannam@127: commit 86652c99050dcc4f52d17974597bdbf56a5998de cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 29 15:23:33 2004 -0500 cannam@127: cannam@127: added missing 'static', thanks to Alessio Massaro cannam@127: cannam@127: commit 96566e4ddd0a9d00b23ad7c8ad04240cce7bfac1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 9 16:36:48 2004 -0500 cannam@127: cannam@127: print more like bluestein cannam@127: cannam@127: commit 9ce57c8bf150fdf7d8177b0252abd7721d82d28b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 9 15:45:22 2004 -0500 cannam@127: cannam@127: fixed op count for R2HC_ONLY_CONV cannam@127: cannam@127: commit 1d28fc43969f2824c21efd2d1f2ce5a365dc0a07 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 9 15:41:50 2004 -0500 cannam@127: cannam@127: include DESTROY_INPUT in buffered flags for in-place...otherwise in-place hc2r uses rdft-dhtcvs diff cannam@127: cannam@127: commit bf6f542cb4702cf9d2c9346254cf09f0bce0e032 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 9 15:41:09 2004 -0500 cannam@127: cannam@127: resurrected R2HC_ONLY_CONV option to share plans and save on planning time cannam@127: cannam@127: commit 04d01b659718c6e66f9b14ad925fc5c630c7c1e8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 9 14:47:00 2004 -0500 cannam@127: cannam@127: precompute folding for cyclic convolution cannam@127: cannam@127: commit 41947ea5b1fd5203da848d5afe3ecfed87f90d91 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 7 16:48:39 2004 -0500 cannam@127: cannam@127: minor cannam@127: cannam@127: commit 766e29f31c88457b71f94bc7607e81a39b445fdd cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 7 16:48:25 2004 -0500 cannam@127: cannam@127: note reports of successful compilation on Windows cannam@127: cannam@127: commit a3b6ef73f675682810957a7770b13c5ede3c75a3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 7 14:16:16 2004 -0500 cannam@127: cannam@127: citation year cannam@127: cannam@127: commit c6ff6592cd7196994610014cb3da0caafa4354df cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 01:07:36 2004 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 326cb17c5ee1b2d61f66a81bf90c012128148add cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 01:07:08 2004 -0500 cannam@127: cannam@127: comment fix cannam@127: cannam@127: commit 16c7ff8cc5de0d74ad68a628e80e9ac8ede5e918 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 01:06:57 2004 -0500 cannam@127: cannam@127: fixed naming cruft cannam@127: cannam@127: commit 3c33d645c75a685c1b95f12be64b272dd01fb621 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 00:56:16 2004 -0500 cannam@127: cannam@127: space cannam@127: cannam@127: commit daf2625a4becbd9aac13358827eef8bbf2ef115e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 00:55:53 2004 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 8bc4eaa51d81346cd07706771fc58bb8767bd428 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 00:54:07 2004 -0500 cannam@127: cannam@127: moved assert cannam@127: cannam@127: commit 679d41c77182afe28023a32c667fb6771f7df10d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 00:49:16 2004 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 0ead6d3532ee317a8e8e99391655c0ef3b13bfd3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 00:41:06 2004 -0500 cannam@127: cannam@127: delete old R2HC_ONLY_CONV hack, now defunct cannam@127: cannam@127: commit 4f45958b21e10e4b0ad128e4a36b3b1c456e7a22 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 6 00:32:58 2004 -0500 cannam@127: cannam@127: added padded real rader cannam@127: cannam@127: commit 2743f45c6e17d1cc72cce590faa7475c03e8ea74 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 5 22:56:58 2004 -0500 cannam@127: cannam@127: removed unused var cannam@127: cannam@127: commit 57cf035f36b08fae46144c4e03a52b3fb7a65f51 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 5 21:20:29 2004 -0500 cannam@127: cannam@127: handle both FFT_SIGN values cannam@127: cannam@127: commit 0f0e531863a2392a747c6b540a966b48b77debaa cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 2 06:07:51 2004 -0500 cannam@127: cannam@127: Oops: d->ros ==> d->ios cannam@127: cannam@127: commit 6ce67e298050d47829243bf692910eeef126601f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 2 06:05:10 2004 -0500 cannam@127: cannam@127: Oops: d->ris should have been d->iis cannam@127: cannam@127: commit 3ddc923840694e141348bb8f48eb6bc9272b607b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 1 16:00:07 2004 -0500 cannam@127: cannam@127: Removed rdft rader cooley-tukey, to be superseded by a generic cannam@127: reduction of rdft twiddle problems to dft + pre/post processing cannam@127: cannam@127: commit 3f82980635418e49dad204ff327021a8adf4bcfb cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 1 15:44:09 2004 -0500 cannam@127: cannam@127: In anticipation of the upcoming revision of rdft, removed rdft generic cannam@127: dit/dif cooley-tukey, in favor of generic rh2c and hc2r solvers. cannam@127: Cleaned up stuff that became unused after this change, such as cannam@127: TW_GENERIC. cannam@127: cannam@127: commit f7546dadff20223e87c698a9c3e8bcdb8496547e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 1 12:59:30 2004 -0500 cannam@127: cannam@127: Removed useless file cannam@127: cannam@127: commit 822bd0498b7fe0a45a001af73cd4b317f33b0230 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Dec 26 13:54:00 2003 -0500 cannam@127: cannam@127: whoops, don't call AC_F77_DUMMY_MAIN if no Fortran compiler is found; thanks to Charles Radley for the bug report. cannam@127: cannam@127: commit 6d8fa3754568aeb1979cbd6d1f6b91c90f524989 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Dec 19 13:58:05 2003 -0500 cannam@127: cannam@127: guess good flags for Solaris/intel, suggested by J. Gregory Wright cannam@127: cannam@127: commit e393cf5533fdff4834a269b4d163641553f9532f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Dec 5 19:55:13 2003 -0500 cannam@127: cannam@127: blah cannam@127: cannam@127: commit bcb1ecc806de458e9744b90452468221ab65d36d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Nov 30 06:59:41 2003 -0500 cannam@127: cannam@127: DIF generic solver was destroying the input. cannam@127: cannam@127: commit 1580db9c3c75c57928058fc81faa981295c7b6ca cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Nov 29 19:28:39 2003 -0500 cannam@127: cannam@127: Fixed bug that caused HC2R transforms to destroy the input in cannam@127: certain cases, even if the user specified FFTW_PRESERVE_INPUT. cannam@127: cannam@127: commit 24f8af52f2a239ba51cd03e37e4c1c74befdc2f2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Nov 29 16:49:01 2003 -0500 cannam@127: cannam@127: Implemented swap_io hack for r2r verifier. cannam@127: cannam@127: commit e15bf89a0e15d64e294ea23deb4f3de422e467b6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Nov 20 22:00:53 2003 -0500 cannam@127: cannam@127: citation cannam@127: cannam@127: commit f8afd813d4a9af4549ccca07ea7604c12c081761 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Nov 14 20:57:55 2003 -0500 cannam@127: cannam@127: Trying to get ``make paranoid-check'' to work. (Still broken.) cannam@127: cannam@127: commit 74399102001e6b633c21ca16f1b4369f63facab1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Nov 14 20:05:54 2003 -0500 cannam@127: cannam@127: fixes for input-preservation tests cannam@127: cannam@127: commit 29fc95192096b664dc7f23bfc6d3530f5880c66f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Nov 14 19:19:31 2003 -0500 cannam@127: cannam@127: Assume FFTW_PRESERVE_INPUT unless either the `d' flag is given in the cannam@127: problem, or the problem is multidimensional c2r (which fftw3 cannot cannam@127: without destroying the input). With this change, we can at least test cannam@127: that FFTW_PRESERVE_INPUT works in the c2r 1d case. cannam@127: cannam@127: commit 5a2907cf7121cc9e824150f654d83c9ff984aa92 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Nov 14 19:14:40 2003 -0500 cannam@127: cannam@127: apply should copy back input for input-preservation check cannam@127: cannam@127: commit e5b287efdbd1a909467ac69e49f09c708462ea6d cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Nov 14 19:01:36 2003 -0500 cannam@127: cannam@127: Undone previous bogus changes cannam@127: cannam@127: commit 55075f65e5f23bb045b5a076bd559d508e15fab8 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Nov 14 18:27:12 2003 -0500 cannam@127: cannam@127: Check dr[fb] in addition to r[fb] cannam@127: cannam@127: commit 74d5a2653d744d48adc748e8ebfaafdad0198cb7 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Nov 14 17:33:44 2003 -0500 cannam@127: cannam@127: Fixed conditions under which the rank-geq2-rdft2 solver is applicable. cannam@127: cannam@127: The old solver was not applicable for out-of-place problems cannam@127: unless DESTROY_INPUT. This is bogus. As long as the subsolvers cannam@127: honor !DESTROY_INPUT, the solver is always applicable. cannam@127: cannam@127: Changed semantics of test program, so that PRESERVE_INPUT is always cannam@127: true unless the problem specifies destroy_input explicitly. Without cannam@127: this change, there is no way to test the new solver. cannam@127: cannam@127: commit edcc72abc5fafc3147bfb8b802cea42d249c711b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Oct 30 15:10:42 2003 -0500 cannam@127: cannam@127: added AIX OpenMP (-qsmp=omp) support; thanks to Greg Bauer cannam@127: cannam@127: commit 8dffe5112def767dc95ffe8c722d009bd2d3a5aa cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Oct 30 10:11:39 2003 -0500 cannam@127: cannam@127: G5 CFLAGS cannam@127: cannam@127: commit 3cdf00d461370ae110601bf4612b31601a9b7100 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Oct 24 04:17:39 2003 -0400 cannam@127: cannam@127: western FAQ cannam@127: cannam@127: commit 8ae00fe74ec955caecad22123ab716b908fb595a cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Oct 23 11:34:11 2003 -0400 cannam@127: cannam@127: Oops. cannam@127: cannam@127: commit f7b3b4aae8910a377ab2c391da7855bad4c6875a cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Oct 23 11:28:28 2003 -0400 cannam@127: cannam@127: Autodetect altivec cannam@127: cannam@127: commit 7458c1cb849f028a73209b18da00a2b63da0d861 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 22 01:14:10 2003 -0400 cannam@127: cannam@127: MinGW gets confused by a single / cannam@127: cannam@127: commit 3b8090b60a50893cb21b3e7442c5d8eec86c756e cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Oct 17 10:46:41 2003 -0400 cannam@127: cannam@127: Paranoid portability fix cannam@127: cannam@127: commit 82175d1a4b25bd246759a5a9499a50037b51bc94 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Oct 16 11:07:46 2003 -0400 cannam@127: cannam@127: size -> length, which should make clear that we are not talking cannam@127: about arbitrary precision. cannam@127: cannam@127: commit b5f6b9f86f41b0429dd03c64101bc2cbd4a0b261 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 15 15:01:40 2003 -0400 cannam@127: cannam@127: pruned transforms are a FAQ cannam@127: cannam@127: commit 736d76c64078a558714dc87e31bc158117a274d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 8 23:54:17 2003 -0400 cannam@127: cannam@127: NO_SEARCH has already been mapped to FFTW_WISDOM_ONLY cannam@127: cannam@127: commit 02a8d7f029570cfb78c03215109a391d74f22ebd cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 8 23:53:19 2003 -0400 cannam@127: cannam@127: newline cannam@127: cannam@127: commit 4177857895703cbd233c7b10e80cb95b8e8c25c5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 20:27:32 2003 -0400 cannam@127: cannam@127: fix cannam@127: cannam@127: commit 694836051975049f0d2981df1372a23be9d9bc1b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 20:24:39 2003 -0400 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit 0b20096c7a3afe66a13dc3efd84ba9e535748248 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 17:43:57 2003 -0400 cannam@127: cannam@127: minor fix cannam@127: cannam@127: commit 39ef965f44a5ba2245f12c7e1b182032746d6dda cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 17:42:30 2003 -0400 cannam@127: cannam@127: grammar cannam@127: cannam@127: commit f7d34b13ca2adef07ab32520d97f05c403d34bd4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 17:29:04 2003 -0400 cannam@127: cannam@127: html output fix cannam@127: cannam@127: commit 4b0c92ff25547e2f4fef16bf2089b6525407fc9e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 17:22:48 2003 -0400 cannam@127: cannam@127: mentioned sqrt(2) factors for DCT/DST cannam@127: cannam@127: commit 512c0e8650df0efa89bc8fea862a005f6f58f2cc cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 27 17:07:18 2003 -0400 cannam@127: cannam@127: FFTW_WISDOM_ONLY flag (undocumented for now), suggested by Phil Dumont cannam@127: cannam@127: commit 54b4afc9f006bc10ce0423f09625b91af30d9dc3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 23 23:36:19 2003 -0400 cannam@127: cannam@127: removed UpTime code cannam@127: cannam@127: commit 26c7e51b101004fe0cc9c2a5f90c732fd7bcf2b2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 23 23:27:29 2003 -0400 cannam@127: cannam@127: updated documentation for mach_absolute_time cannam@127: cannam@127: commit 12a80b367661367374d32deeb29e01fd75e311e3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 23 23:25:52 2003 -0400 cannam@127: cannam@127: use mach_absolute_time on MacOS/Darwin, as a fallback; don't bother checking for UpTime since it requires extra libs cannam@127: cannam@127: commit fa1787b57c9fb3539af76bd43c35ce224da7dace cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 23 22:59:29 2003 -0400 cannam@127: cannam@127: support Apple UpTime function for asm-less xlc, grrr... cannam@127: cannam@127: commit 13e7c9ab7273a625f9b21015a75eff8ef163d468 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 23 15:42:29 2003 -0400 cannam@127: cannam@127: additional paranoia for xlc etc. cannam@127: cannam@127: commit 2da4a3de8376303c716cd7ee5b3b47b8f759983b cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 22 15:28:56 2003 -0400 cannam@127: cannam@127: work around _Complex_I weirdness in xlc, reported by Greg Allen cannam@127: cannam@127: commit 3533775b3de13e397feb15794631f1d2fedd1e98 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Sep 5 18:03:11 2003 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit ef0a55daee823f7ae55367f4304e32abf20c0d40 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 5 13:11:40 2003 -0400 cannam@127: cannam@127: New script that produces commercial version. cannam@127: cannam@127: commit 4c4873b3aface20d5444216fcb5ea79221e5a289 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 5 07:27:06 2003 -0400 cannam@127: cannam@127: Noted that VC++ is buggy. Noted that we know nothing about Windows. cannam@127: Noted that the sky is blue as well. cannam@127: cannam@127: commit a3d172eb32ea6f17849604998b55db355d60cc26 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 2 09:04:19 2003 -0400 cannam@127: cannam@127: Noted that certain arrays are no longer used after the planner has cannam@127: completed. cannam@127: cannam@127: commit c68f62d97627bb869061a9433c4f4605f4dc8fd4 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 26 08:22:38 2003 -0400 cannam@127: cannam@127: Typo cannam@127: cannam@127: commit b95fbd832374dacc4e8c9fb21b8c4085b33b3460 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 25 21:27:43 2003 -0400 cannam@127: cannam@127: New item cannam@127: cannam@127: commit b2a02ef7112f03f73e25ccb67227265306b69b6c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 21 17:36:08 2003 -0400 cannam@127: cannam@127: try creating output file before planning (thanks to Phil Dumont for the suggestion) cannam@127: cannam@127: commit 766c3757fc99565ef8a14a9f3d5729740b1e7182 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 19 10:08:07 2003 -0400 cannam@127: cannam@127: Clarified fftw_cleanup() cannam@127: cannam@127: commit 8eecb544f6ac10687ec258d3162a5c7508c6156d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 16 03:13:41 2003 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 443c1d796f44813179c18d5b34c2836bd441a1e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 28 18:01:13 2003 -0400 cannam@127: cannam@127: use time() instead of clock() (FIXME: what to do for non-POSIX systems?) ...thanks to JP Sugarbroad and James A. Treacy for the bug report cannam@127: cannam@127: commit 9de40445c11cba14fa6eb7ae8e06d2792ebb2262 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 24 18:58:10 2003 -0400 cannam@127: cannam@127: Need __volatile__ in sparc cycle counter. This is why the debian cannam@127: port hangs. cannam@127: cannam@127: commit 3da4cd5a3334e2b8415224657c5f5d9ce17eef12 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 20 16:02:43 2003 -0400 cannam@127: cannam@127: merged 3.0.1 notes cannam@127: cannam@127: commit cb6949db0271a00b15369efbd59d5c74af6d8c0c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 13 20:57:34 2003 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 906832308c9b7ba4ce3e55a3efa502ab55e6a676 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 10 11:48:50 2003 -0400 cannam@127: cannam@127: Dealing with constants in a way that seems to confuse gcc less. cannam@127: cannam@127: commit e823de6c6b1433905ed2851aadcd1e11bf9b81b9 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 9 17:39:23 2003 -0400 cannam@127: cannam@127: Enabled scheduler hack for FMA, where it seems to help. cannam@127: cannam@127: commit cff00fb9b000446f13c060876536184a03873ca5 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 9 12:57:38 2003 -0400 cannam@127: cannam@127: Hmm---the new scheduler seems make things worse for gcc/x86, better cannam@127: for gcc/ppc, and about the same for icc/x86. Disabled for now. cannam@127: cannam@127: commit ca5556ba3956352649cd6d7342d16ffa660db23f cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 9 08:09:53 2003 -0400 cannam@127: cannam@127: New scheduling pass that keeps ``x = a + b'' and ``y = a - b'' close cannam@127: together. This property was no longer automatic for the dags cannam@127: generated in SIMD mode. cannam@127: cannam@127: I cannot measure any speed difference due to this change. However, cannam@127: the change is justified by a minimal-screwup argument. Moreover, the cannam@127: sse2 fftw library is now 1% smaller than it was before. cannam@127: cannam@127: commit 82ab8c034224f1364c148e9f725c90e6e16f721d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 8 20:42:22 2003 -0400 cannam@127: cannam@127: -(FNMS()) => FMS() cannam@127: cannam@127: commit 471cc543a0805c19f676664861e5dc9d1fb5f1d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 6 13:53:23 2003 -0400 cannam@127: cannam@127: added more convenient target name cannam@127: cannam@127: commit 18303fef32c580b6c7dc03c4e8bfe30ca2a92724 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 5 13:30:10 2003 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit c30db69dc73fc965683cd147c948c3dba1f6aad7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 5 13:19:36 2003 -0400 cannam@127: cannam@127: Consistent naming cannam@127: cannam@127: commit da3b10c5645b3bb482b26cb42d0821dda213b994 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 5 13:05:51 2003 -0400 cannam@127: cannam@127: Got rid of problemw. cannam@127: cannam@127: commit 21c3f87f755cadd9ef25945fe33448eeea7cb511 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 4 06:56:26 2003 -0400 cannam@127: cannam@127: Increase TIME_MIN on intel only cannam@127: cannam@127: commit c749315d331fac4826ab1754a84f2a66e00197df cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 4 06:36:02 2003 -0400 cannam@127: cannam@127: A little hack to get more consistent scheduling. cannam@127: cannam@127: commit 1af463bd5b8ae6bd696ecb6a86d8b1952aaf6fd0 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 3 16:47:42 2003 -0400 cannam@127: cannam@127: New experimental scheduler (currently disabled). cannam@127: cannam@127: The old scheduler is ``optimal'' in the sense that it minimizes cannam@127: register pressure. The only way to reduce register pressure is to cannam@127: schedule dependent instructions as closely as possible, so as to cannam@127: minimize the life time of registers. This strategy maximizes the cannam@127: number of pipeline stalls, however. With enough registers and short cannam@127: enough pipelines, this tradeoff is fine. This is no longer the case cannam@127: for the devilish pipeline of the Pentium IV or (probably) the PowerPC cannam@127: 970. cannam@127: cannam@127: The new scheduler switches to a ``list scheduler'' for dags smaller cannam@127: than a specified size. The list scheduler executes a butterfly left cannam@127: to right one column at the time. This amounts to the best possible cannam@127: pipeline utilization, and the worst possible register pressure. cannam@127: cannam@127: The ``specified size'' defaults to 0, i.e., no change from fftw2 and cannam@127: fftw-3.0. It seems like a value of 7--10 produces the best results cannam@127: for Pentium IV (probably screwing the G3/G4 powerpcs and sparc, but I cannam@127: haven't tried.) As time goes by, we may want to increase this number cannam@127: to favor newer processors over older processors. cannam@127: cannam@127: commit b4dc4ef39f37d036b9120f6e273549fd7bbaaa2a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jun 25 17:43:59 2003 -0400 cannam@127: cannam@127: remove non-portable use of tempfile; thanks to Nicolas Decoster for the patch cannam@127: cannam@127: commit 9630b97551ac80d0b1a8ecc26722ab98149a2be6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jun 25 17:14:03 2003 -0400 cannam@127: cannam@127: increase stupid HP preprocessor limits cannam@127: cannam@127: commit a31916f0f41eb0a1398f38341da28991919123f0 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 19 15:21:52 2003 -0400 cannam@127: cannam@127: Distribute gen_mdct.ml cannam@127: cannam@127: commit 367373d615cb66b335d9586544ead2f9d424effc cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 11 06:55:21 2003 -0400 cannam@127: cannam@127: Cleared int/ptrdiff_t confusions cannam@127: cannam@127: commit f1e0319906231b0e8d5675cc4ab24db00e560dc0 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 10 22:15:42 2003 -0400 cannam@127: cannam@127: Cleared int/ptrdiff_t confusion cannam@127: cannam@127: commit 48d9ab9a8d3ddcb8fc21d4c8e90adfcbf7e4e200 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 8 09:52:57 2003 -0400 cannam@127: cannam@127: Increased TIME_MIN. This seems to produce more reliable plans cannam@127: on Pentium IV. cannam@127: cannam@127: commit 3ba082c0f617d0c03cde783d3d7eba4392d13397 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 7 21:43:00 2003 -0400 cannam@127: cannam@127: Removed relic -trivial-stores, which dates back to Franz's early cannam@127: experiments. Speed improved on SSE2, both with gcc and icc. cannam@127: cannam@127: commit 25a3b0e594d1b7fbe6b87c322ae34470bfdcccba cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jun 5 22:29:52 2003 -0400 cannam@127: cannam@127: fix direntry cannam@127: cannam@127: commit 4ceff26934f86ada8712eabca96511462e7e8eaf cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jun 5 13:41:34 2003 -0400 cannam@127: cannam@127: added imdct cannam@127: cannam@127: commit 1f23163e190c844ee3b2caf4564245cfbddd1c5b cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 4 19:54:38 2003 -0400 cannam@127: cannam@127: Collect pattern (a * b) +- (c * d) in generic-arith, because this cannam@127: operation can usually be computed with one rounding in fixed-point cannam@127: (and it possibly exposes a FMA instruction) cannam@127: cannam@127: commit 10d5f543e3ab8e1f2acb3f0bdeef0b196f22d6e0 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 4 15:11:29 2003 -0400 cannam@127: cannam@127: Generic-arithmetic unparser cannam@127: cannam@127: commit d705a296a095ba4947adbd5e745b1a5ccf39f04a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 1 09:05:30 2003 -0400 cannam@127: cannam@127: Oops---randomized CSE was using the same random numbers cannam@127: over and over cannam@127: cannam@127: commit fed2aa57dc95873ed83371338d16b9667c84c6f1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 1 07:01:17 2003 -0400 cannam@127: cannam@127: Paranoia. cannam@127: cannam@127: commit 4905e1d1498f044f37a6efd2b168c9bfac54a3de cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 1 07:00:54 2003 -0400 cannam@127: cannam@127: Use relative error instead of absolute error, to avoid problems cannam@127: when normalization factors are used. cannam@127: cannam@127: commit 967eecb8b745332dff25a3610a617bd73897abd6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 31 22:11:28 2003 -0400 cannam@127: cannam@127: slight opt cannam@127: cannam@127: commit 6186c7e4fbaec8d816ce12a74c211890ad8c7fe9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 31 22:10:45 2003 -0400 cannam@127: cannam@127: slight optimization cannam@127: cannam@127: commit f547dd3851ad0f88d4b79ac5b32af9ecbe727e4e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 31 20:43:31 2003 -0400 cannam@127: cannam@127: *W is const cannam@127: cannam@127: commit 9d9e1ec4a2623c32dd10304f84d3d3854b17b938 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat May 31 20:41:15 2003 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 4269fae3e68941b115c0f8855ad3bb58e18c6f89 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 29 21:31:31 2003 -0400 cannam@127: cannam@127: added experimental MDCT cannam@127: cannam@127: commit 3028a550c6acc0fb367dbe26af192714715a09e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 28 22:01:37 2003 -0400 cannam@127: cannam@127: altivec (fma) needs simd codlist.c too cannam@127: cannam@127: commit 083e7c5e6bf88cffc3938cb6225551ec5af0f869 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 28 22:00:49 2003 -0400 cannam@127: cannam@127: make sure we include SIMD codlist.c for non-Unix folks cannam@127: cannam@127: commit 08028f9b0217ea164f25bcbf01824df23bb7245e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 27 20:31:25 2003 -0400 cannam@127: cannam@127: noted howmany_rank == 0 is a single transform cannam@127: cannam@127: commit 46b08adb7f478a6f5d51e1fefa6ab99891d3391d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 27 20:02:31 2003 -0400 cannam@127: cannam@127: further stride clarification cannam@127: cannam@127: commit 1264855f26ebf31705c1a6cf6067c3e996c71114 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 26 10:21:22 2003 -0400 cannam@127: cannam@127: Removed transposed dftw problems. cannam@127: cannam@127: I now consider transposed dftw a Bad Idea, since it does not cannam@127: apply to the case that it was originally meant for (speed up four-step) cannam@127: and it complicates the implementation of the other thing I want to try cannam@127: (dftw m-slices). cannam@127: cannam@127: commit c198c3ed318d9864a22877795f7bac4f3eaf4d8b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 26 07:22:59 2003 -0400 cannam@127: cannam@127: Obsolete comment cannam@127: cannam@127: commit bb96207501d8ca498bde638f46932c2bdd4bec66 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 24 15:00:53 2003 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit a3733f2596b8d6d3972edfd942ce61296f2051d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 24 07:20:35 2003 -0400 cannam@127: cannam@127: Oops---wrong test NO_UGLYP instead of !NO_UGLYP cannam@127: cannam@127: commit d8575658ab032ccf0b6553c4a84af8510d0cae55 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 24 07:05:34 2003 -0400 cannam@127: cannam@127: Implemented radix r, where n=r^2 * p cannam@127: cannam@127: commit c2ca438ef110583287ce3c0e8527d4d382ccedde cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 21 01:54:32 2003 -0400 cannam@127: cannam@127: xlc seems to properly use fma as well cannam@127: cannam@127: commit 3bf5cca5763573337c68978f05125e6e3eb4fe3d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 20 23:07:43 2003 -0400 cannam@127: cannam@127: print warning if there is no cycle counter cannam@127: cannam@127: commit 00cef912cbb09cf8a1080a3544d45d6bacc44222 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 20 17:32:04 2003 -0400 cannam@127: cannam@127: updated Funda reference cannam@127: cannam@127: commit a331d8a0841d1ccb7c314dbb98749bd77d69e709 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 19 20:12:36 2003 -0400 cannam@127: cannam@127: const cannam@127: cannam@127: commit 200b3519c0ee6bd32babf52b352b8f8385da45d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 19 15:41:09 2003 -0400 cannam@127: cannam@127: Implemented generic dif square transposed (q-style) solver. cannam@127: cannam@127: commit 8c6184ff5904082729018b5c36926b5bc479db14 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 19 07:00:36 2003 -0400 cannam@127: cannam@127: applicable() is now a property of the solver (in anticipation of cannam@127: transposed solvers) cannam@127: cannam@127: commit 75102fd59a69589a2b23faeab596f43e7bcdc46c cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 19 06:33:40 2003 -0400 cannam@127: cannam@127: Slight cleanup cannam@127: cannam@127: commit fb49407efc496838a8f21a6cdfa52ad602b1cdfa cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 18 13:05:51 2003 -0400 cannam@127: cannam@127: Nothing, really cannam@127: cannam@127: commit df97ba383cdf4cc6a473e4e1ad041b11b8f965af cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 18 09:05:20 2003 -0400 cannam@127: cannam@127: Moved vector loop inside bytwiddle(), in anticipation of cannam@127: a q-style dftw-dit transposed solver. cannam@127: cannam@127: commit ee4edcdb0d229ae56167c04d4ffc36ee84d32361 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 18 08:52:02 2003 -0400 cannam@127: cannam@127: Fixed flops count cannam@127: cannam@127: commit dccb90a9de4bd56cb5dad7340b27302bf2c6d245 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 18 08:47:20 2003 -0400 cannam@127: cannam@127: style cannam@127: cannam@127: commit b83dec51bb0e588e9bc8fd4aff422b3e8b20e8ca cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 18 07:16:34 2003 -0400 cannam@127: cannam@127: Faster inner loop. cannam@127: cannam@127: commit c7bd9e9fc6d14ae3a28012fa0425a14330746572 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 17 08:02:38 2003 -0400 cannam@127: cannam@127: Print vector length cannam@127: cannam@127: commit 74d6f9435c81aaafcf08927ebec58dd49ecea003 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 17 07:55:33 2003 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 668b7018dd6ccd5a7802f71b29eb4e155801ff86 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 17 07:50:35 2003 -0400 cannam@127: cannam@127: Allow vl > 1 cannam@127: cannam@127: commit 1919840986e87d611bccc5a1ffbaa469bccaa94a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat May 17 07:01:42 2003 -0400 cannam@127: cannam@127: Radix can be derived from problem---no need to pre-specify it. cannam@127: cannam@127: commit a25c7ceebb0e65bb9b03bd7c30e3b97a29063185 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 22:50:50 2003 -0400 cannam@127: cannam@127: fixed comment cannam@127: cannam@127: commit e12663ca110ee43071e2344163161cd2122312c4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 22:48:20 2003 -0400 cannam@127: cannam@127: whoops, gcd should be static cannam@127: cannam@127: commit 96c37858e236a75341bc45eca6e8a9a6180c61c8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 22:40:32 2003 -0400 cannam@127: cannam@127: more unrolling cannam@127: cannam@127: commit ce888f64f2b19c2b1d0a2306c0f5a0bfb282e2cf cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 20:03:48 2003 -0400 cannam@127: cannam@127: Hack to avoid infinite recursion. cannam@127: cannam@127: commit 3be61e5d6f8b5c517feed2417902fc4fc8205180 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 19:52:43 2003 -0400 cannam@127: cannam@127: consistency cannam@127: cannam@127: commit 10ef200843d878ba4e88a6e59c53ec89b48cba11 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 19:45:15 2003 -0400 cannam@127: cannam@127: Wrong comment. cannam@127: cannam@127: commit 9ad0f5ae45b2265e1307381d14a436eb43998c8d cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 19:45:03 2003 -0400 cannam@127: cannam@127: Style. cannam@127: cannam@127: commit bc609cdbd63dfd90154f4f0b07a0fb6c9c1f7354 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 18:35:27 2003 -0400 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit 3c931e88733a24d99d4639165f16f073ff25b35e cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 18:33:45 2003 -0400 cannam@127: cannam@127: added allzero FAQ cannam@127: cannam@127: commit 4e3c1c97240d8dc8b8595f179994e9999e2a7b98 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 18:22:45 2003 -0400 cannam@127: cannam@127: simplification: instead of cldb, just use cldf with inputs/output values swapped cannam@127: cannam@127: commit 734444b8d502c323a04fd2cc6e6fc37d53cd4b04 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 15:47:17 2003 -0400 cannam@127: cannam@127: Allow more general transform sizes. cannam@127: cannam@127: commit 77d47783b6445a14416a351ffb88c70348bb7ae1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 14:22:37 2003 -0400 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit 7bdfe7acb85d3e5c4081537789ee95c49d852e0c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri May 16 14:22:05 2003 -0400 cannam@127: cannam@127: MS has __int64 type, not long long (grr) cannam@127: cannam@127: commit 6778b57c92f2c729955fec383ee1ec817559b35f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 13:34:16 2003 -0400 cannam@127: cannam@127: Fixed printout cannam@127: cannam@127: commit 26cad01414a0e9ec0f5809c3fafdcd0ebeca8eee cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 13:23:00 2003 -0400 cannam@127: cannam@127: Fixed flop count cannam@127: cannam@127: commit 6663b7b12eed820ccad08bbe8a7b13d4e7e96baf cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 13:02:06 2003 -0400 cannam@127: cannam@127: New bluestein solver cannam@127: cannam@127: commit a3444150521abc3e4522064bfbce259979d143dd cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 09:51:05 2003 -0400 cannam@127: cannam@127: Implemented generic radix. cannam@127: cannam@127: commit 105268f348df67714aaff107d5a4817a8ea9069e cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 08:19:38 2003 -0400 cannam@127: cannam@127: Removed conditional branch from inner loop in generic.c cannam@127: cannam@127: commit 44179d747df49fe429ae0108b108e1f28b71023c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 07:48:28 2003 -0400 cannam@127: cannam@127: Simplified indexing cannam@127: cannam@127: commit 6a89bb8523df8e372f221f4ccdd6fa9e75120ec1 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 06:53:56 2003 -0400 cannam@127: cannam@127: Better still. cannam@127: cannam@127: commit bc7126aa1f6bca65323f8d367629a9e6ddb18310 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 06:24:31 2003 -0400 cannam@127: cannam@127: Further improvement of generic solver cannam@127: cannam@127: commit 29931919e62a0633afc7fdfe2738fba18419e30c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 05:57:07 2003 -0400 cannam@127: cannam@127: Cleanup cannam@127: cannam@127: commit 516d81975ebfabe254800e61a072976ce2493792 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 05:42:57 2003 -0400 cannam@127: cannam@127: Cleanup cannam@127: cannam@127: commit 9c40e9a8ab92ae76b96b1e57a51a8a46cd4202aa cannam@127: Author: Matteo Frigo cannam@127: Date: Fri May 16 05:31:40 2003 -0400 cannam@127: cannam@127: Generic now only works for odd sized. Added check. cannam@127: cannam@127: commit a4abb5b3c869ca3d4c1b572793c4128935461f2e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 21:53:25 2003 -0400 cannam@127: cannam@127: Increased GENERIC_MIN_BAD because of new algorithm. cannam@127: cannam@127: commit 60bf38f4720ea99600008e1ad0772e3871cc975d cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 21:40:27 2003 -0400 cannam@127: cannam@127: Much, much better. cannam@127: cannam@127: commit bd0ae8b86cd6e44542f040b89670e8559cb4daef cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 21:25:00 2003 -0400 cannam@127: cannam@127: Still trying to understand why rdft-generic-dit is faster cannam@127: then dft-generic... cannam@127: cannam@127: commit 1ace458103964bbd9cc763efde55b6c64543e072 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 21:04:33 2003 -0400 cannam@127: cannam@127: Nothing, really cannam@127: cannam@127: commit e68561ce796750faf2eb70606053723da3a3651b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 20:59:45 2003 -0400 cannam@127: cannam@127: Never be clever for the sake of being clever. cannam@127: cannam@127: commit 3480d0dea4ccb5fa65ea2c8950aea4821cea6e3a cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 20:58:06 2003 -0400 cannam@127: cannam@127: Simplified. generic-dit is gone. The solver is now out-of-place cannam@127: only---buffering is done by the buffered solver. cannam@127: cannam@127: commit 21161d67e447696eda8fb463a6e629fdb9b9a286 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 19:18:18 2003 -0400 cannam@127: cannam@127: rader-dit is gone. cannam@127: cannam@127: commit 99baac8e16f51413086aa8b35ff5894b43ddce25 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 19:13:03 2003 -0400 cannam@127: cannam@127: Cast cannam@127: cannam@127: commit 9bc90955564668ef3b897434df873ea7a9e987b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu May 15 19:09:07 2003 -0400 cannam@127: cannam@127: Introduced twiddle problem ``dftw''. Changed most other things cannam@127: to deal with this change. cannam@127: cannam@127: commit 57d761eab36c018f98849a04c82df5fdc61db498 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 15 18:47:18 2003 -0400 cannam@127: cannam@127: whoops, X(safe_mulmod) not fftw_safe_mulmod cannam@127: cannam@127: commit ad0f04aa3eead44c4f82e436a20241a8d76fdfba cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 15 16:53:16 2003 -0400 cannam@127: cannam@127: add VC++ versions of asm cannam@127: cannam@127: commit 4e67675d7f66ce57718045e8ddf3769ba44f378f cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 15 15:03:06 2003 -0400 cannam@127: cannam@127: VC++ reportedly supports the intel intrinsics, but requires __inline instead of __inline__ cannam@127: cannam@127: commit 8d72a4d25a56b9b8c9e918cb462ae7f8429fce9c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu May 15 14:32:06 2003 -0400 cannam@127: cannam@127: precompute array indices with VC++ cannam@127: cannam@127: commit 7a8ca44fd207d2de8947e6d7dad9b6122d5eacf4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 14 21:57:39 2003 -0400 cannam@127: cannam@127: added doc note cannam@127: cannam@127: commit e21c7e8fbc7073874cec2c052a810cb8c8bafb6c cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 14 19:45:54 2003 -0400 cannam@127: cannam@127: autodetect windows cannam@127: cannam@127: commit 9e15f067241ae365258fdae039f13855799566b5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 14 15:08:49 2003 -0400 cannam@127: cannam@127: don't bother with #ifdef HAVE_CONFIG_H, since non-Unix users always forget to define it cannam@127: cannam@127: commit 8fd89cce099546af6bb94f83b4e84bc46609708a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 13 16:58:07 2003 -0400 cannam@127: cannam@127: VC++ uses __inline cannam@127: cannam@127: commit 02aaa87cb911b0d6d67ec4f11932357f702aa75f cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 13 14:51:26 2003 -0400 cannam@127: cannam@127: added leak question cannam@127: cannam@127: commit 64d02f177161f96e87c02cde6015ceff42ac0bfe cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 12 18:26:51 2003 -0400 cannam@127: cannam@127: LARGE_INTEGER needs windows.h (supposedly, there is some problem converting _itnt64 to double...damn MS and their nonstandard types) cannam@127: cannam@127: commit 342ab9adfdb7bf9a5936f1c1d3f0820fa729ab2e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 12 18:22:16 2003 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit a8cb5339f580f0f4ee0ff02c58f91036b7fef886 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 12 17:16:19 2003 -0400 cannam@127: cannam@127: added 256x256 to canonical list cannam@127: cannam@127: commit 446260f6a6a84986c7eec6e1b3c0eea0f66f759d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon May 12 07:02:06 2003 -0400 cannam@127: cannam@127: Oops... cannam@127: cannam@127: commit 924714d15f5ed063b1fa8a40a3bcd2ebe406e572 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 11 11:04:46 2003 -0400 cannam@127: cannam@127: Unrolled loops, changed cutoff cannam@127: cannam@127: commit a8e681a4dd1d8cbae25becdb745926efde43cf3c cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 11 10:20:04 2003 -0400 cannam@127: cannam@127: Do not multiply strides by 2 twice. cannam@127: cannam@127: commit cba6f4731943edfe50f4bacd9de28d0551593f43 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 7 21:09:43 2003 -0400 cannam@127: cannam@127: added 'make smallcheck' cannam@127: cannam@127: commit 8f61201655a4d04b402d2cec34acb86e89cbe35f cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 7 20:46:10 2003 -0400 cannam@127: cannam@127: --without-cycle-counter becomes --with-slow-timer, updated docs cannam@127: cannam@127: commit e77df17a33148f0cd531fcf1bcf371af0b69ced6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 7 18:05:29 2003 -0400 cannam@127: cannam@127: remove duplicate -openmp check; Sun requires -xopenmp cannam@127: cannam@127: commit 34594ee9cdef76091eff1164f9518e92bf0b855b cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed May 7 17:59:23 2003 -0400 cannam@127: cannam@127: fixed compilation under Sun C++ cannam@127: cannam@127: commit fe5788275ebad911c952662c50694c2f296ae4b2 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed May 7 14:24:46 2003 -0400 cannam@127: cannam@127: Use estimator if cycle counter is unavailable, regardless cannam@127: of the FFTW_MEASURE/ESTIMATE setting. cannam@127: cannam@127: commit 7d2473af5f98cac96951e94c009b96f1d96dcea3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 6 23:15:34 2003 -0400 cannam@127: cannam@127: _WIN32 (not __WIN32__) is always defined cannam@127: cannam@127: commit a9d4840add6bac6ef0f0f194868861615f52d91c cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 6 23:11:52 2003 -0400 cannam@127: cannam@127: minor cleanup cannam@127: cannam@127: commit 33d6678021b7c966f0490729f9839282d88b7e68 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 6 22:50:07 2003 -0400 cannam@127: cannam@127: tentative VC++ stuff, some consolidation cannam@127: cannam@127: commit 4b2c8b1fdf43a442999e9309b1abb5f525b2a5e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue May 6 12:17:56 2003 -0400 cannam@127: cannam@127: made cycle.h more self-contained cannam@127: cannam@127: commit 17b78ccd6e3bf378453a85f671f02c82eb45ced4 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue May 6 08:30:39 2003 -0400 cannam@127: cannam@127: Use ``%'' flag to denote commutative operations. cannam@127: cannam@127: commit b98342890cbb95b8f450d6da8e22637b455fccc0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 5 20:42:30 2003 -0400 cannam@127: cannam@127: MIT license, brief documentation cannam@127: cannam@127: commit 95f79870876aa95354fecf59c0da025edb8982f9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon May 5 20:31:16 2003 -0400 cannam@127: cannam@127: whoops, forgot f77_wisdom.f cannam@127: cannam@127: commit 5cdde47077a07f4aa39487741662e868f5cf11df cannam@127: Author: Matteo Frigo cannam@127: Date: Sun May 4 19:37:09 2003 -0400 cannam@127: cannam@127: Improved speed of accuracy test. cannam@127: cannam@127: commit ff00ccc1b3a83962c18e91ab12740ca63bbb6081 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 29 11:45:34 2003 -0400 cannam@127: cannam@127: s390 cycle counter cannam@127: cannam@127: commit 2ab4e6e2c3bfdd6882bad8bca36fdc105f742847 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Apr 26 12:26:15 2003 -0400 cannam@127: cannam@127: forgot r2r directory cannam@127: cannam@127: commit 990abcc219adbf0759807b9b2c20d80f639dd940 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 25 20:52:23 2003 -0400 cannam@127: cannam@127: delete unused files, since they don't compile any more cannam@127: cannam@127: commit a35c9fd7e587b8301131f8e530387dad7e62fc00 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 24 06:37:41 2003 -0400 cannam@127: cannam@127: Better gcc code generation cannam@127: cannam@127: commit 462d92a21265012fd4fb89326da6bccd69f05406 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 23 15:30:50 2003 -0400 cannam@127: cannam@127: ccc is the Compaq C compiler on Linux/alpha cannam@127: cannam@127: commit a97d5f6b3b451179f501157bfe6fafde3481ea0d cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 23 00:06:03 2003 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit a53aa0afb9f63f64ee6235e07fd99014f6da32fb cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 19 09:18:25 2003 -0400 cannam@127: cannam@127: ia64 cycle counter with intel compiler. cannam@127: cannam@127: commit 20978a6bcaad9c07b4c969eae56ab29ae092e2bb cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 18:27:30 2003 -0400 cannam@127: cannam@127: More gcc bugs. Sigh. cannam@127: cannam@127: commit 1a9f1a74c640a09efbabff5043cc7074b6bfefe0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 18:01:49 2003 -0400 cannam@127: cannam@127: touch ChangeLog to observe GNU standards cannam@127: cannam@127: commit 4241f03ed18cc0acf61072c70b9c38b1c7dc7c31 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 18:01:12 2003 -0400 cannam@127: cannam@127: We now build ChangeLog automatically at distribution time cannam@127: cannam@127: commit 1797417a3e517b1f9b9931e49797dff29ae760da cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 18:00:17 2003 -0400 cannam@127: cannam@127: Automatic ChangeLog hackery cannam@127: cannam@127: commit d76cd97496030b79d3450f2c6de88da7b4458bd6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 18 13:25:26 2003 -0400 cannam@127: cannam@127: plural cannam@127: cannam@127: commit 6c59528dcbd6ce3676a0852bad2d909e68fdeeac cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 18 13:25:01 2003 -0400 cannam@127: cannam@127: updated cannam@127: cannam@127: commit b37ba8f3518f8be4c0c25dd86f102c9e64527104 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 12:59:41 2003 -0400 cannam@127: cannam@127: Updated cannam@127: cannam@127: commit 61ac8832cb1bd238132fe475c699f12e11232781 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 18 11:48:39 2003 -0400 cannam@127: cannam@127: a -> an cannam@127: cannam@127: commit 287a97f4366aeb55745345972896f06e378dcacf cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 18 11:47:56 2003 -0400 cannam@127: cannam@127: hyphen cannam@127: cannam@127: commit 8791b19a3135636c10dd4e99695799a6ac315870 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 18 11:47:28 2003 -0400 cannam@127: cannam@127: comma cannam@127: cannam@127: commit ad823e8abbe446885056a5fc05555e295b7a174f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 18 11:46:59 2003 -0400 cannam@127: cannam@127: minor cannam@127: cannam@127: commit 7a09c0d18816d7d1c2ed89a0dfbd62843d3442eb cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 10:37:31 2003 -0400 cannam@127: cannam@127: Updated cannam@127: cannam@127: commit 13381c1ba5e8af34a8bb03710aa7741a8040cc07 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 10:14:59 2003 -0400 cannam@127: cannam@127: New script that builds the distributions cannam@127: cannam@127: commit 459fa38ea6735a02e430cb0ecbca2323e07cef57 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 08:51:07 2003 -0400 cannam@127: cannam@127: Oops again cannam@127: cannam@127: commit e57ee39dc083cd4461a548cded7fdc45e57fc74f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 08:39:05 2003 -0400 cannam@127: cannam@127: Oops, forgot -sign 1 cannam@127: cannam@127: commit 396b6bc876a10a52ce9dc68230e2eb25af544f29 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 18 08:28:25 2003 -0400 cannam@127: cannam@127: Reorganization of simd codelets cannam@127: cannam@127: commit fd7cb9b51d78aabe5b32969758bf472ca0d563ae cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 21:21:45 2003 -0400 cannam@127: cannam@127: k7 assembly was not updated after conversion of opcnt from cannam@127: int to double cannam@127: cannam@127: commit 3bf64e2b77975db0ec3c2ad2232fa9dfceeae35f cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 19:15:53 2003 -0400 cannam@127: cannam@127: Capital `X' looks bad in all-lowercase plans cannam@127: cannam@127: commit 448802e951a73d406bacc449b5d7eb0ece3bfaf6 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 18:53:29 2003 -0400 cannam@127: cannam@127: Removed redundant inline/noinline codelets cannam@127: cannam@127: commit b72b1f8cdf96f12c6776efdb456cdd6e2eff00b3 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 15:25:50 2003 -0400 cannam@127: cannam@127: New noinline cannam@127: Noinline real codelets cannam@127: cannam@127: commit 4f5ff427ae9e39bda6b17cdc61b9ecfad075f322 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 17 15:23:03 2003 -0400 cannam@127: cannam@127: more ideas cannam@127: cannam@127: commit e549828bdf29f4ba5b352f8d54e1d8fb65e86b0b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 13:18:45 2003 -0400 cannam@127: cannam@127: Removed duplicate rules. cannam@127: cannam@127: commit 29889996e0d7e2beec6c4afaf8d06936f0bc8e7a cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 10:51:09 2003 -0400 cannam@127: cannam@127: acx_pthread.m4 was not distributed cannam@127: cannam@127: commit 4bcda610088022777266a9d4723e3108871a2382 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 07:21:17 2003 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 92ba4bfacade8ab1d4dae0e256cda08e8b22eb3c cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 07:07:19 2003 -0400 cannam@127: cannam@127: Both inlined and non-inlined notw codelets. cannam@127: cannam@127: commit bc5fcf6d524989489b08f05b59fd2660b4331765 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 06:44:21 2003 -0400 cannam@127: cannam@127: Initial experiment with both inlined and non-inlined simd codelets. cannam@127: Both are included for now. cannam@127: cannam@127: commit 5586bdcd9d791a373355fae20e4df01e8b51ef32 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 17 05:57:36 2003 -0400 cannam@127: cannam@127: --enable-fma to build FMA distribution cannam@127: cannam@127: commit 6719f26d34cd9bd0dce5a3d279a06b113cd774a7 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 16 17:21:53 2003 -0400 cannam@127: cannam@127: Inline SIMD nontwiddle codelets cannam@127: cannam@127: commit 5db4d2ae77f5600008ce54b88e93a0e7fbcc649b cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 16 16:18:29 2003 -0400 cannam@127: cannam@127: Pathetic attempt at saving a couple of registers... cannam@127: cannam@127: commit 7267a94d763a0380970f1b07ee84aad71f138c8f cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 16 15:51:27 2003 -0400 cannam@127: cannam@127: for (i = 0; i < m; ++i) ==> for (i = m; i > 0; --i) cannam@127: No proof of evidence that this is any faster, but just in case... cannam@127: cannam@127: commit 607d75d99f7b0c558d5664cc18b0c6a070d3aa02 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 15 15:03:20 2003 -0400 cannam@127: cannam@127: added hack to make sure that codelet loops are preferred to vecloop solvers in the estimator cannam@127: cannam@127: commit a772926574ae28c851b95b9eea8d22d0b244e25e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 15 14:53:44 2003 -0400 cannam@127: cannam@127: use double for flops cannam@127: cannam@127: commit 84c191f258a2ee1fde8b39e4a567f48bc84d273a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 15 14:51:50 2003 -0400 cannam@127: cannam@127: metrowerks reportedly supports gcc assembly extensions on ppc cannam@127: cannam@127: commit d83a8e3d6b322b293fb8b048bb46c1048faf430b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Apr 14 15:00:50 2003 -0400 cannam@127: cannam@127: foo_CFLAGS generates some automake junk that breaks the build cannam@127: on Redhat 7.3. Screw it. cannam@127: cannam@127: commit 54128f6b2dd2ea009736debbb0c2eee43c4a0ade cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Apr 14 12:22:59 2003 -0400 cannam@127: cannam@127: Carefully check return status cannam@127: cannam@127: commit 8935dbb4a98ff9a7780aa1ce7a98c656b7284d79 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Apr 13 16:46:12 2003 -0400 cannam@127: cannam@127: Removed annoying -FMA() expressions. cannam@127: cannam@127: commit 8ce45c13aca6d081f84f8dcb84a973383d3e5ee8 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 12 14:32:22 2003 -0400 cannam@127: cannam@127: Major fma hackery cannam@127: cannam@127: commit 4a294df6343206e437a3f24ed268a71c9cd9edb0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 12 14:25:43 2003 -0400 cannam@127: cannam@127: Slight cleanup cannam@127: cannam@127: commit bda29baca81b8098e25fb1d61d3aa7b882f33ef5 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 12 10:04:51 2003 -0400 cannam@127: cannam@127: Updated version number cannam@127: cannam@127: commit 80a70d1b1e8e1d7e336301a3f65b26d153ba15a7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 12 08:03:07 2003 -0400 cannam@127: cannam@127: Damn autoconf cannam@127: cannam@127: commit 27c1334aec2b8b3bbc7e679761d0ddab53212487 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 12 07:54:20 2003 -0400 cannam@127: cannam@127: Recognize all 74xx processors cannam@127: cannam@127: commit 7b1c87d4184dace0460075da82295c1fc90e523d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 12 07:35:17 2003 -0400 cannam@127: cannam@127: Detect 7400 processor. cannam@127: cannam@127: commit ea19ce217870bff790c8e91865228328a08cd769 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 11 20:42:11 2003 -0400 cannam@127: cannam@127: No need to check for gcc-2.95 cannam@127: cannam@127: commit fb6560399e6c1dc7382fc48209545f4494f87e65 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 11 16:14:39 2003 -0400 cannam@127: cannam@127: removed duplicate cannam@127: cannam@127: commit 114c5faeddef91d05efc3af062e58f21879cd77d cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 11 08:45:37 2003 -0400 cannam@127: cannam@127: mflops ==> ``mflops'' cannam@127: cannam@127: commit be09e2c160458ffa571c8c207546d77dd86f1022 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 11 07:00:53 2003 -0400 cannam@127: cannam@127: Print setup time as well cannam@127: cannam@127: commit a55b833ea382a9dcf478517c83d68d044db4dcaa cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 10 15:36:18 2003 -0400 cannam@127: cannam@127: Enforce pointer equality for in-place problems. cannam@127: cannam@127: commit a920de28e41596cc862a7f7bcc14c27c68920b53 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 9 17:47:54 2003 -0400 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 080c6f9a24e129da6a216d44e581ea9c5012df83 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 9 14:53:38 2003 -0400 cannam@127: cannam@127: cross-ref fftw-wisdom man page cannam@127: cannam@127: commit 920b026d6c76cc8b23b877e8238cb6cd9f71e7af cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 9 10:13:00 2003 -0400 cannam@127: cannam@127: Undone previous change, committed by mistake. cannam@127: cannam@127: commit c636a6d32d90fda78f15b6e5b0060feeea45b47c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 9 10:12:24 2003 -0400 cannam@127: cannam@127: Quick and dirty README for bench cannam@127: cannam@127: commit e542ee15fcc1eaa62b3c1ea6a58cc3bc468ff599 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 9 08:50:25 2003 -0400 cannam@127: cannam@127: Consider additional command-line arguments as problems to be cannam@127: benchmarked. cannam@127: cannam@127: commit 2196cad41694a580854e0f04991d3055d9e365e3 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 9 08:44:13 2003 -0400 cannam@127: cannam@127: Default report format is now human-readable. Removed cannam@127: unnecessary complexity in benchmark reporting. cannam@127: cannam@127: commit 1ecc5f59fb401c7ecdeb54596088ac86ede3639a cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 9 06:10:40 2003 -0400 cannam@127: cannam@127: Updated for new interleaved/split api. cannam@127: cannam@127: commit 8d3536f2476441b3f7754a1d5298102d6cd14ac2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 9 03:01:03 2003 -0400 cannam@127: cannam@127: updated citation cannam@127: cannam@127: commit aaf6b0e66924b89fb8c927bbcc3bc655512aab13 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 8 19:35:59 2003 -0400 cannam@127: cannam@127: Time for beta3 cannam@127: cannam@127: commit f5c162435d5334990a7c6f3421191592a831bdd7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 8 17:40:59 2003 -0400 cannam@127: cannam@127: whoops, added cannam@127: cannam@127: commit 14108da14b0738a1c8b036a901971104d5374ef7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 8 17:33:47 2003 -0400 cannam@127: cannam@127: more comparison of different R*DFT types cannam@127: cannam@127: commit 75bbef9a658de085d8c21952d597dbca9e8fa722 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 8 16:48:08 2003 -0400 cannam@127: cannam@127: comments cannam@127: cannam@127: commit 0ba9318b95b11d2cb5470ebcf73a1819d31caafb cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 8 16:19:39 2003 -0400 cannam@127: cannam@127: more accurate DCT-I and DST-I, at the expense of up to a factor of 2 in speed and memory cannam@127: cannam@127: commit 404912f3cee76dd7a6b3928798304c64c55129bf cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 8 05:38:09 2003 -0400 cannam@127: cannam@127: Workaround gcc/sparc bug cannam@127: cannam@127: commit f34586737c6b234a5e2316dddbbe425331d4a5b3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 8 01:34:12 2003 -0400 cannam@127: cannam@127: rumors cannam@127: cannam@127: commit 78f9ac7bedec0c22cb7fc5fccdbb2adc8b8c4f3b cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Apr 7 18:54:11 2003 -0400 cannam@127: cannam@127: added rdft2 paranoid mode cannam@127: cannam@127: commit 55b03e7ef8ab680294bbbe73059cb32cef82ac65 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Apr 7 18:47:37 2003 -0400 cannam@127: cannam@127: added paranoid mode for r2r cannam@127: cannam@127: commit 8361bbcd9549f5c5819b31ce44c2e733e185f1a5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Apr 7 15:10:08 2003 -0400 cannam@127: cannam@127: whoops, sincos is predefined on some systems cannam@127: cannam@127: commit 0e1c9cf517e0d4010af5a35d10d141b2d74f7d0e cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 5 16:50:57 2003 -0500 cannam@127: cannam@127: bp->destroy_input was not initialized cannam@127: cannam@127: commit 7b13a4d8dfbf05ed699e7a1a533e7edf9355e2bc cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 5 09:29:11 2003 -0500 cannam@127: cannam@127: Asserted correctness conditions for tainted pointers. cannam@127: cannam@127: (For now, use CK() while we test. They should be changed into cannam@127: A() at some point.) cannam@127: cannam@127: commit 6b16dfb3e11fcfa65d2064af8ee1c66f0e0ba2a2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 5 08:18:23 2003 -0500 cannam@127: cannam@127: Untaint pointers before zero'ing arrays and before hashing cannam@127: cannam@127: commit 1495e7c627b502d282f4fb290357d6fe573e12dd cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 5 07:11:56 2003 -0500 cannam@127: cannam@127: Alignment check did not work with icc, which seems to be cannam@127: confused by the fact that the variable is not used. cannam@127: cannam@127: commit e013a83a5d1fb9e98a7d035baa3e4c5a89f768f1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 5 06:41:20 2003 -0500 cannam@127: cannam@127: More paranoid paranoid-check cannam@127: cannam@127: commit 377dd60fc68c62595e87882b3a5caede183a1251 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Apr 5 06:19:25 2003 -0500 cannam@127: cannam@127: 0 == x & 7 parses as (0 == x) & 7, which is wrong cannam@127: cannam@127: commit 109ea5550e130de0017dcdf2665c0872ff776e98 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 21:35:49 2003 -0500 cannam@127: cannam@127: alignment checks cannam@127: cannam@127: commit 97c940cb1b38b75289d99b2206d77a13db97f6fb cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 21:04:14 2003 -0500 cannam@127: cannam@127: prevent infinite loops in exhaustive planning cannam@127: cannam@127: commit 20a0d16cc3bf2a26e0e5b2ed4f1691c617ab676e cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 20:58:20 2003 -0500 cannam@127: cannam@127: split/unsplit guru interface cannam@127: cannam@127: commit ec77ade6672c46fde379e806bb25f26569eaa874 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 20:39:55 2003 -0500 cannam@127: cannam@127: Need UNTAINT in verifier too. cannam@127: cannam@127: commit 2fd0ede87b74bd6a6b325910eb68ce55ee69ae66 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 19:36:46 2003 -0500 cannam@127: cannam@127: Forgot #if HAVE_SIMD cannam@127: cannam@127: commit 0509bf08e01662eb716c8038093caab6bbbe867e cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 19:30:37 2003 -0500 cannam@127: cannam@127: Keep track of two separate taint bits cannam@127: cannam@127: commit dc9c49340a92349a47d46befc4f85937b95911c5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 19:16:32 2003 -0500 cannam@127: cannam@127: added NO_SIMD problem flag, made UNALIGNED an API issue (taints input pointers) cannam@127: cannam@127: commit 1b8c8e9e88fa0f39226f2a8a853a07719d4faf40 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 18:14:14 2003 -0500 cannam@127: cannam@127: bugfix in buffered: wrong pointers passed for cldrest; also use TAINT instead of UNALIGNED in buffered2 cannam@127: cannam@127: commit 99fccbefe65b6e39c9b4e9be40e647facb900b4f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 17:19:51 2003 -0500 cannam@127: cannam@127: Reverted previous change, committed accidentally cannam@127: cannam@127: commit db841c316cbd202532fd5b428396e5f4e9c74cf0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 17:18:39 2003 -0500 cannam@127: cannam@127: What was I thinking? cannam@127: cannam@127: commit 6235b967442a2150ad1e7100ae39070ecbee0ca9 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 17:18:21 2003 -0500 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 4d690f88fdc36870e18b359db01ee23dbc005d72 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 16:48:32 2003 -0500 cannam@127: cannam@127: added --enable-debug-alignment cannam@127: cannam@127: commit 8890a79f285088b2b04ca1c2db939a582ac0328e cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Apr 4 16:29:43 2003 -0500 cannam@127: cannam@127: X(taint) prototype, define corresponding function only if HAVE_SIMD cannam@127: cannam@127: commit 3f29be3cc8a63846e725e496ae01474af84ab9fc cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 16:15:53 2003 -0500 cannam@127: cannam@127: Initial checkin of tained pointers cannam@127: cannam@127: commit faad01bdd384c083438df8ef016b8a18804cb72a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Apr 4 13:12:58 2003 -0500 cannam@127: cannam@127: More conservative preservation of alignment cannam@127: cannam@127: commit ac40b45c34f80bd09d25405935c3722528595a97 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 3 23:16:27 2003 -0500 cannam@127: cannam@127: plan/execute with aligned stack cannam@127: cannam@127: commit 978b7f409d31dde15736857998dada802a3ef49c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 3 15:40:01 2003 -0500 cannam@127: cannam@127: whoops, missed FFTW_MEASURE in fftw3.f cannam@127: cannam@127: commit 3274f607dea4bbf61b89f1d09703054007b28776 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 3 13:44:46 2003 -0500 cannam@127: cannam@127: use WITH_ALIGNED_STACK for experimental semaphore stuff, too cannam@127: cannam@127: commit 345df91b8bab3216268ca697850bc00767799265 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 3 09:04:23 2003 -0500 cannam@127: cannam@127: Removed old file cannam@127: cannam@127: commit a36ad0e14ceb92c16cdc3bec19938c53c69f79a5 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Apr 3 07:50:43 2003 -0500 cannam@127: cannam@127: Improved stack-alignment hack cannam@127: cannam@127: commit c4f4e2d7d432203a0f99e3a50b29168c3d653a83 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 3 02:37:57 2003 -0500 cannam@127: cannam@127: use aligned stack for experimental semaphores, too cannam@127: cannam@127: commit cea2d48a884c03d448b9c688a192081e3e984983 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 3 02:17:58 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit c3bdcb8375c4c5181c0c642b0331d2a7268757f8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Apr 3 01:58:32 2003 -0500 cannam@127: cannam@127: fix(?) for SIMD thread problems cannam@127: cannam@127: commit 13dde386673933410cafa316f241cdc6544ecd65 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 20:33:12 2003 -0500 cannam@127: cannam@127: noted n=1 REDFT01 case cannam@127: cannam@127: commit 4a2c5556d9c6d080f3c3fee8c87d0aee50c12531 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 20:32:07 2003 -0500 cannam@127: cannam@127: note about n=2 REDFT00 formula cannam@127: cannam@127: commit ac5fe8c3ecce10f2e79f84279ce6e406db3891e8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 20:30:10 2003 -0500 cannam@127: cannam@127: note about undefined REDFT00 cannam@127: cannam@127: commit 4761b3e61b5cf393deeacf6eba73d9f3a35e2d12 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 20:18:03 2003 -0500 cannam@127: cannam@127: noted n=1 RODFT01 case cannam@127: cannam@127: commit 00cd3721f0f757f691e62c836aff445fece4a9ef cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 20:14:07 2003 -0500 cannam@127: cannam@127: corrected definitions cannam@127: cannam@127: commit 20545fe3112a9aa8bd9529129f24586a66f39f9d cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 19:43:59 2003 -0500 cannam@127: cannam@127: added REODFT_KINDP, fixed nontrivial test for R2HC11 and HC2R11 (not that we support these yet anyway) cannam@127: cannam@127: commit dcd456710f59aea75abb5a4b62ad7b8c8592c28f cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 19:16:54 2003 -0500 cannam@127: cannam@127: size 2 hc2r and dht are equivalent to r2hc cannam@127: cannam@127: commit dd3db55bba543cc4db74f3760716a251892089a9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Apr 2 15:09:08 2003 -0500 cannam@127: cannam@127: noted overwriting in upgrading section cannam@127: cannam@127: commit 4752fd3dcd81d75371bc667be6ab701ee36a24d3 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Apr 2 05:25:56 2003 -0500 cannam@127: cannam@127: Moved with_aligned_stack to its own file cannam@127: cannam@127: commit 821f37e9c6396afa7dcf22eae25e2ddb56f16218 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 1 21:11:31 2003 -0500 cannam@127: cannam@127: Fixed comments cannam@127: cannam@127: commit 44b77936443c9dcbba1ccf21d3e90c2426a46e01 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 1 20:57:39 2003 -0500 cannam@127: cannam@127: Alignment hacks cannam@127: cannam@127: commit 52974f9347f673ccfc5eca68ed2af2e39f0ae148 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 1 14:26:48 2003 -0500 cannam@127: cannam@127: phew, no, previous version was okay cannam@127: cannam@127: commit f599fa6d8cb159f0e636411e51f0bd07feca296e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 1 14:26:15 2003 -0500 cannam@127: cannam@127: whoops, crap cannam@127: cannam@127: commit da939ebd27d69c1e3693ebf71f81060e816af54e cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Apr 1 08:01:06 2003 -0500 cannam@127: cannam@127: support sse2 in forthcoming gcc-3.3 cannam@127: cannam@127: commit bad66cbc1963d1beecba1205ff4d528026003427 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 1 01:17:15 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit bde4d633afdc37f663c31f7aa2c4b3f8673e9607 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 1 01:16:46 2003 -0500 cannam@127: cannam@127: noted ac_check_headers cannam@127: cannam@127: commit 0e70968689aed47b11b44eb15752b97e21534366 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 1 01:11:31 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 716a92cca66059e083cc6dc764db18de707a6318 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Apr 1 01:06:53 2003 -0500 cannam@127: cannam@127: documented autoconf tests, so that cycle.h can be distributed separately cannam@127: cannam@127: commit a081cb59d2fbd65042f4a1cec68ec04698a03594 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 22:12:02 2003 -0500 cannam@127: cannam@127: IRIX is all-caps cannam@127: cannam@127: commit c21fee75eade0b3c38780e252bb0dbe24383d2cf cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 22:11:42 2003 -0500 cannam@127: cannam@127: noted Irix fix cannam@127: cannam@127: commit 0521214bc9e4b224ee18c31c165c8971d5d09fc6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 22:10:33 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit e42bd5b20e10661a6cd8228c3b19ee7d8f1a1602 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 22:04:35 2003 -0500 cannam@127: cannam@127: use ithreads_init so as not to confuse fftw 2 users cannam@127: cannam@127: commit 95c74b4b2e3ef14b347ae7b50fdb455d6a7aa719 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 22:00:42 2003 -0500 cannam@127: cannam@127: IRIX lossage cannam@127: cannam@127: commit 746ced9c6d3a5d53c7b95090cbb99fcfd0b07344 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 21:19:20 2003 -0500 cannam@127: cannam@127: check for -openmp (icc) among the OpenMP flags (TODO: make this a cannam@127: separate macro, with a loop instead of repeated checks) cannam@127: cannam@127: commit 589adf0e340eafbabdd43f5beacae6740e9e64a0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 17:12:19 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit 4242c9c8bf63111190cbcccd162a224af036e5af cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 17:01:16 2003 -0500 cannam@127: cannam@127: More liberal test for solaris CC cannam@127: cannam@127: commit d5928079a514ffaba5eedc2cd5ce76eb2dd9fa9b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 15:13:33 2003 -0500 cannam@127: cannam@127: Allow x86-64 simd cannam@127: cannam@127: commit b7a2252e112c67968e6695c7ef13e375a04d23d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 15:13:21 2003 -0500 cannam@127: cannam@127: Added x86-64 timer code cannam@127: cannam@127: commit 7f0d1b516cd025f72f304fdeb210c563b94bff31 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 13:10:54 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 914e74201ca244b441f7f971d2f05aced6aa405c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 13:07:19 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 202febc5ba6f89ad6e834d4e36a01caf4bb5fde2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 31 13:05:27 2003 -0500 cannam@127: cannam@127: colon cannam@127: cannam@127: commit 2deea3231269a3e4bcbdfa9498ad253ad1b26a48 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 07:20:20 2003 -0500 cannam@127: cannam@127: Reorganized compiler bugs section (which is growing out of control) cannam@127: cannam@127: commit b4bb5597a0d941eeefe4ec01208c139d37e9fce2 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 07:15:20 2003 -0500 cannam@127: cannam@127: solaris gcc bug appears to be also in 2.95.2 cannam@127: cannam@127: commit efb7874ecc58a7e086abf8428d481a6e19c4e0d7 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 07:13:45 2003 -0500 cannam@127: cannam@127: Workaround works---there is another gcc/sparc bug elsehwere cannam@127: cannam@127: commit 8ab897ec05303f83b56d2e349c3dba59da173ef2 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 07:08:56 2003 -0500 cannam@127: cannam@127: Grrr, workaround does not work. cannam@127: cannam@127: commit 32707cc1247ff03834c3d37fceb57f53e268da65 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 31 07:02:23 2003 -0500 cannam@127: cannam@127: ADDMOD is now function, which seems to avoid gcc bugs. cannam@127: cannam@127: commit c4c605027021db9b801e3e2695c802ed6e1bc44a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 30 16:40:26 2003 -0500 cannam@127: cannam@127: Workaround sparc gcc bug cannam@127: cannam@127: commit b77fba4459439cf3d969088c9edb010ab151a893 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 30 15:51:59 2003 -0500 cannam@127: cannam@127: note cannam@127: cannam@127: commit 91e398fb21a1c46fac8174a5f2faf0b79548e188 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 30 15:34:57 2003 -0500 cannam@127: cannam@127: make non-square UGLY, for now cannam@127: cannam@127: commit 4233309534b8e309bce0dafeeff64c29ac9f4b1c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 30 15:33:57 2003 -0500 cannam@127: cannam@127: added -o amnesia to forget_wisdom before each plan cannam@127: cannam@127: commit 055907acc9ab6486266e2601f13e76e768bd990f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 30 09:41:27 2003 -0500 cannam@127: cannam@127: Report setup time in benchmark cannam@127: cannam@127: commit 6a49d54d587cc678c9a4063e3ed620c998d2602e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 20:21:15 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 1ccc921a0398eb08789ac928e28840a524100587 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 19:11:10 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit ee2cf222eef51c2ee38a761765c58ed6a2faa35e cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 29 18:46:16 2003 -0500 cannam@127: cannam@127: More relaxed definition of UGLYness cannam@127: cannam@127: commit 2afbef3c1cd1edca0168bc5341dac85de41790ba cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 15:28:01 2003 -0500 cannam@127: cannam@127: no more cvs id strings in header files...I'm tired of having to rebuild everything after a commit cannam@127: cannam@127: commit 6922449e07c11f405107d7e5fc63d7dfb0379b5d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 15:22:28 2003 -0500 cannam@127: cannam@127: rdft2 stride unification cannam@127: cannam@127: commit 383f9ebcd63e13f756a57f0801b6bdc4080f4887 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 14:38:23 2003 -0500 cannam@127: cannam@127: preserve in-place-ness cannam@127: cannam@127: commit 4989fd02c94baef2f163547b88f643fcd1172a72 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 14:23:31 2003 -0500 cannam@127: cannam@127: make nowisdom the default cannam@127: cannam@127: commit a2f08dcbee1508f40df179ca67ed0ddcefd66f37 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 29 14:13:18 2003 -0500 cannam@127: cannam@127: --verbose in paranoid-check produces too much output. Make it quiet. cannam@127: cannam@127: commit 90cdd14a2b342236ae6e8367d94ad2e29ecadd76 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 13:45:13 2003 -0500 cannam@127: cannam@127: fixed transpose bugs...need to check ri-ii before deciding whether Ntuple fits cannam@127: cannam@127: commit 71fc37fa553e50623f56a5fb21731833cb2d9dfd cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 29 08:10:40 2003 -0500 cannam@127: cannam@127: try more 2^k cannam@127: cannam@127: commit d4e0d59380ec69df5a4250ebd0f62f002c964e19 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 29 08:05:41 2003 -0500 cannam@127: cannam@127: MIN_ALIGNMENT was defined after being used, causing crash in sse2. cannam@127: cannam@127: commit 101331222a4ff1189042a5997260a7e171ae1136 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 03:07:34 2003 -0500 cannam@127: cannam@127: real transposes are currently unused, and are not needed for MPI code either cannam@127: cannam@127: commit 55b24758612593bac4f6b7065d32b33b815eb81f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 29 02:58:39 2003 -0500 cannam@127: cannam@127: added general transpose cannam@127: cannam@127: commit c111a90447eb6c3c1a0058a93e97557beaaf1605 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 22:49:04 2003 -0500 cannam@127: cannam@127: added transposition option cannam@127: cannam@127: commit 90fb8971b2092e9a1fa97b10065683ba8af9247c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 22:09:22 2003 -0500 cannam@127: cannam@127: yikes, fixed incorrect applicability of transpose plans cannam@127: cannam@127: commit 5776651de7f7152e07630b99ee8445fb004131a1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 22:06:14 2003 -0500 cannam@127: cannam@127: in the future, we might want to allow sz->rnk == 0, vecsz->rnk arbitrary to be converted to r2hc (the apply function already should work for this case)...disabled for now, though cannam@127: cannam@127: commit db6988d7af647595db1ef218c039bb2755070b59 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 19:12:08 2003 -0500 cannam@127: cannam@127: use most_unaligned in rdft2 cannam@127: cannam@127: commit bf69a12d650bc9daee88f41bd0a04bf1abe664c8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 19:11:47 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit b79360114562af8636d8c3da2898cc7ed7df7b98 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 19:00:21 2003 -0500 cannam@127: cannam@127: output message when checks pass cannam@127: cannam@127: commit 14afb1d94a6d7eb23f853cd2097814989148a9e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 17:21:47 2003 -0500 cannam@127: cannam@127: added ifndef alloca around alloca stuff cannam@127: cannam@127: commit 6cccb2f2fed1d79204cbbb7e1ee44685bf2ed300 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 28 13:45:50 2003 -0500 cannam@127: cannam@127: Proper alignment in rader cannam@127: cannam@127: commit 643528ab599946750ef668ce19266fe5a0bab5c1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 12:43:23 2003 -0500 cannam@127: cannam@127: whitespace cannam@127: cannam@127: commit 8c9af83603806d8d769f21b1222dfe717068f7c6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 12:41:39 2003 -0500 cannam@127: cannam@127: whoops, alloca stuff inside HAVE_ALLOCA cannam@127: cannam@127: commit 4f4ed55f3679a721e23cf9cb61e9180646f0f176 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 12:35:21 2003 -0500 cannam@127: cannam@127: make check can afford to be a little bigger cannam@127: cannam@127: commit e7db3e5ee6c6cbb0c24626dc09c00e23e46e70e7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 12:31:32 2003 -0500 cannam@127: cannam@127: use same alloca macrology as configure script cannam@127: cannam@127: commit 3c6ec07b6659b60cfb8e77365e75c872ccacd66d cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 03:05:15 2003 -0500 cannam@127: cannam@127: fallback is no longer needed for mingw cannam@127: cannam@127: commit cbc91a4cae1193e9e27ef5885e1bb37c548c191b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 02:58:45 2003 -0500 cannam@127: cannam@127: alloca fallback for gcc cannam@127: cannam@127: commit 37a6e5be53d9273006dc360b8dafe2e3e53356e5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 28 02:49:59 2003 -0500 cannam@127: cannam@127: _alloca was added for MinGW, but it causes problems there cannam@127: cannam@127: commit 37adf3eddeb59f98c6d4e8888ddb4208b10fb42a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 22:06:07 2003 -0500 cannam@127: cannam@127: fixed most_unaligned for split format cannam@127: cannam@127: commit fae5ff2e94558a024ef43a1cd4470f5c68b4de17 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 19:01:58 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 5a2216ff945775dcd769967d2a58125b51c4b3c0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 19:00:20 2003 -0500 cannam@127: cannam@127: added pkg-config cannam@127: cannam@127: commit 669fc84978762faee9e8d48a7b852eca22ee4303 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 15:59:01 2003 -0500 cannam@127: cannam@127: fixed asserts cannam@127: cannam@127: commit 8108cd595625c0e28683a556df95de39588c7fb4 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 27 15:49:53 2003 -0500 cannam@127: cannam@127: Do not adjust r/i pointers separately. cannam@127: cannam@127: commit d9b6e6ea20bc01290efaabb99405a0b10a3dadca cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 27 15:17:40 2003 -0500 cannam@127: cannam@127: iForgot to add files cannam@127: cannam@127: commit 9c79c521f6c57b91f21a90731610214d9dafca8e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 27 15:10:41 2003 -0500 cannam@127: cannam@127: Specialized n simd codelets for unit vector stride. cannam@127: cannam@127: commit 7e309fd9c2284d234e3932b6d3a2d5bbcc44c9e0 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 27 08:22:03 2003 -0500 cannam@127: cannam@127: Changed version number to beta2 cannam@127: cannam@127: commit 513db4fd67e83952d5e510cf7a1eb23fbd6ef2bb cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 27 06:37:07 2003 -0500 cannam@127: cannam@127: Changed alignment requirements for n1 simd codelets. Changed cannam@127: mechanism for detecting lack of alignment. cannam@127: cannam@127: commit be8495756a69c610211f28e3f9a7ff20016eb901 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Mar 27 04:25:06 2003 -0500 cannam@127: cannam@127: Oops, wrong place for hook cannam@127: cannam@127: commit 575731d46f96f87a35f3a1a37dba70cc153728df cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 02:37:52 2003 -0500 cannam@127: cannam@127: added comments to codelet makefiles, to aid people wanting to generate their own code cannam@127: cannam@127: commit 59245164a590789dd1ed892f910bc43a346b791b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:42:27 2003 -0500 cannam@127: cannam@127: Matteo is also a copyright holder cannam@127: cannam@127: commit c558091f6b4b0f37175f86a623e2f2376da9c01a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:41:08 2003 -0500 cannam@127: cannam@127: FORTRAN is officially Fortran, these days cannam@127: cannam@127: commit d0b28f4043bbc0aae200dd359e7ad52da98bc903 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:40:32 2003 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit 46bd3b7c143c2291021d10121a74d23936e3ccdd cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:40:14 2003 -0500 cannam@127: cannam@127: don't use "wrapper" cannam@127: cannam@127: commit 9a8eba97546a52e4070f6910534c849009d027d8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:37:53 2003 -0500 cannam@127: cannam@127: plural cannam@127: cannam@127: commit bc26c4cd5feb70158f734130ef3415ce557e207b cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:35:32 2003 -0500 cannam@127: cannam@127: grammar cannam@127: cannam@127: commit 246a46ccdfc59616fdc3234ee8f773f54b9b5260 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 27 01:33:35 2003 -0500 cannam@127: cannam@127: better phrasing cannam@127: cannam@127: commit 8d4fae80f2b4558c3ec62f108316187e7dad2b84 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 22:47:58 2003 -0500 cannam@127: cannam@127: stddef.h should not be needed anymore for this file cannam@127: cannam@127: commit 1080fb42895231d251238b34f4af9458ee7329ec cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 22:13:48 2003 -0500 cannam@127: cannam@127: added comments for Franz mode cannam@127: cannam@127: commit e995cc9d9c287c7681b4e8ff9e97dc57e5dcbdb4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 22:11:58 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit cff23bce9ddaeddc3338be9cfcbc3cd3d57f2370 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 22:08:51 2003 -0500 cannam@127: cannam@127: commented on FRANZ codelets cannam@127: cannam@127: commit 0b33d349e4eab5f3fce6b0873cf0366e83e51d59 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 22:06:45 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 8467b9db21a1c618e7ccaf958299f9edc31bbb42 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 21:51:15 2003 -0500 cannam@127: cannam@127: disable DIF codelets, since they are never used (apparently) except cannam@127: for some non-power-of-two sizes...improve support for the latter by cannam@127: adding size 3, 5, and 6 q^2 codelets. cannam@127: cannam@127: commit 100f8e1667a8ffdc4ad997bbe4346603e7da122b cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 20:07:11 2003 -0500 cannam@127: cannam@127: DHT has no forward/backward cannam@127: cannam@127: commit 9f5c7271cdd393f08d42a71669c9d3d1686ab641 cannam@127: Author: fftw cannam@127: Date: Wed Mar 26 19:46:12 2003 -0500 cannam@127: cannam@127: added hacky way to use an arbitrary flag cannam@127: cannam@127: commit 59d54e87e6bb971ba93e6b371aad0c3ee5d88d11 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 26 19:44:31 2003 -0500 cannam@127: cannam@127: Better place to install hook cannam@127: cannam@127: commit b35aa5670a5cf242f215c8281c3c09097c3c740a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 26 19:40:28 2003 -0500 cannam@127: cannam@127: noted that the user should run make check if they think FFTW has a bug cannam@127: cannam@127: commit dd17b391f48608fdfe190c514eb865ff891689b9 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 26 17:31:16 2003 -0500 cannam@127: cannam@127: Oops, what am I thinking cannam@127: cannam@127: commit 47c3588218fccd048fb32989c007dc693f402abc cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 26 17:23:56 2003 -0500 cannam@127: cannam@127: Grrr.... fixed bug in estimator cannam@127: cannam@127: commit a3f8ee308a4b9b1f83d1031991c9f8fdc55b3bc2 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 26 17:16:19 2003 -0500 cannam@127: cannam@127: Oops---the flop count was right. The estimator is broken elsewhere. cannam@127: cannam@127: commit f2103b394847d39a74d720c5dc18b9f3139fc257 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 26 14:28:41 2003 -0500 cannam@127: cannam@127: Fixed SIMD estimator cannam@127: cannam@127: commit 442a17b47519435071b0c7373c83cc50f5e4b826 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 26 07:45:03 2003 -0500 cannam@127: cannam@127: Added twidsq simd codelets cannam@127: cannam@127: commit 9647b9a35046476b0697bb196f6ad80a1c81b763 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 23:33:03 2003 -0500 cannam@127: cannam@127: gensrc -> genfft cannam@127: cannam@127: commit e668b629605e9193d33403c9c87be52a7d08d134 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 23:32:16 2003 -0500 cannam@127: cannam@127: newline cannam@127: cannam@127: commit 76eeb4a83c788c638126d62924bbedb833573028 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 25 19:17:08 2003 -0500 cannam@127: cannam@127: Noted need to add dif simd codelets cannam@127: cannam@127: commit 350bf8c788a6f8a0ec21b7b004ce7a83c163f511 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 13:03:47 2003 -0500 cannam@127: cannam@127: noted shift cannam@127: cannam@127: commit a5fa31a29076ae51d870e2db210b7f51aa46adbb cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 13:02:47 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit 276ff68eb803fb179adefc146b05b4f616fd226f cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 12:46:44 2003 -0500 cannam@127: cannam@127: need make after bootstrap cannam@127: cannam@127: commit f48787b41d83d8f21ec8ce19b275eaedf5316484 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 12:31:49 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit 7143220a87f7444e90964aadccece0c31bf3830b cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 12:30:56 2003 -0500 cannam@127: cannam@127: libtool is also needed cannam@127: cannam@127: commit 212581eeb5c4011118653b3d8fe433b774bbcd1b cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 12:29:52 2003 -0500 cannam@127: cannam@127: added code generator introduction cannam@127: cannam@127: commit e22b4de0a314136783316cc1acbbc7bf97ca105c cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 11:51:49 2003 -0500 cannam@127: cannam@127: added support for REDFT/RODFT/DHT direct codelets cannam@127: cannam@127: commit cc149df36c0ddc161d91558da702572cd01f99c6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 25 11:29:29 2003 -0500 cannam@127: cannam@127: noted ARM bug; thanks to Jay Treacy cannam@127: cannam@127: commit e313a7fb2e0c1c2524eaed8926b25055a38fb957 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 25 07:55:54 2003 -0500 cannam@127: cannam@127: bugfix from Stefan cannam@127: cannam@127: commit dc62fc48ad26abb231c697a5a18b5f7ca64ab6fe cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 24 15:59:08 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit 878030bb1ea7efd3b1e1dab02601732fd5c90c36 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 24 15:58:44 2003 -0500 cannam@127: cannam@127: caveat cannam@127: cannam@127: commit 776dd0aac7bb400bce14f59781f664062d7b4117 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 24 15:58:04 2003 -0500 cannam@127: cannam@127: warning about DHT cannam@127: cannam@127: commit 2bd26e46c0f7e3622be81d9922f0089923143c65 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 24 08:34:14 2003 -0500 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 597693dba60d0535d890bbb5f161c3a01830a1ea cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 24 08:13:15 2003 -0500 cannam@127: cannam@127: Regression test for p4fftwgel cannam@127: cannam@127: commit 92603541cee018def425427f93dcb3739ab7c0f1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 24 03:09:06 2003 -0500 cannam@127: cannam@127: make check is faster, old tests are in make bigcheck cannam@127: cannam@127: commit 518b188f2a8a30b7cbc2a5c34b335940afa54530 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 22 00:41:21 2003 -0500 cannam@127: cannam@127: note cannam@127: cannam@127: commit 82b8e611b2d1a65af695db0ddf0cf306a5804886 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 22 00:40:05 2003 -0500 cannam@127: cannam@127: whoops, line wrapping cannam@127: cannam@127: commit b9e7ade930fd2e2de8105a28ff7a8f32a799237f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 21 15:10:00 2003 -0500 cannam@127: cannam@127: Franz-mode codelets even without SIMD. (disabled) cannam@127: cannam@127: commit bd548cc599b6178d2e1bdbc2c6abc08f276ae386 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 21 09:09:30 2003 -0500 cannam@127: cannam@127: Bug is in netbsd-1.6, not 1.5 cannam@127: cannam@127: commit 1d1b6b166a1164c2499c4a7e5f9bd9b69f3cf5c3 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 21 07:45:48 2003 -0500 cannam@127: cannam@127: const cast, should placate c++ compilers. cannam@127: cannam@127: commit 677ff57df2415f59cc701368e26dd23d1c6ec956 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 20 18:49:49 2003 -0500 cannam@127: cannam@127: added FAQ on why plans are array-specific cannam@127: cannam@127: commit f0c1a0a7c03bfb68f4559001c2b652aa7a601c0d cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 20 16:12:56 2003 -0500 cannam@127: cannam@127: comment fix cannam@127: cannam@127: commit 3ac192669e4bbb596cc30adb429179fa58f11387 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 20 16:12:15 2003 -0500 cannam@127: cannam@127: noted comparison to NR cannam@127: cannam@127: commit 3f8a990d38ae5f796daa261636120dcb936acb2a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 19 20:13:16 2003 -0500 cannam@127: cannam@127: whoops, C99 complex didn't work if complex is a macro (as it is with glibc); thanks to Keh-Cheng Chu for the bug report cannam@127: cannam@127: commit faab1981e35c6596ac99e9c1e8379c77d92155fe cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 19 16:52:54 2003 -0500 cannam@127: cannam@127: noted in help that --enable-k7 enables 3dnow, and that --enable-3dnow is only a fallback cannam@127: cannam@127: commit 297a4fd9785fe05d2149abf128413bd363fa2dbc cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 19 15:09:52 2003 -0500 cannam@127: cannam@127: New gcc bug. html.refs was not in repository/distribution. cannam@127: cannam@127: commit 8a81ec059d2c1c567c69a45f77d76b8f242c8836 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 19 10:09:16 2003 -0500 cannam@127: cannam@127: Don't write wisdom if you don't have it. cannam@127: cannam@127: commit 4bc446d3e46e7a8c2f084d425e193e58f9ff76ec cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 18 15:44:41 2003 -0500 cannam@127: cannam@127: Added index entries for DHT. Similarly for DCT, DST cannam@127: cannam@127: commit 4a72bfaf1f333116de1e5e0a154bc87d17c9c234 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 18 14:50:04 2003 -0500 cannam@127: cannam@127: execute should not go through C api, for efficiency cannam@127: cannam@127: commit 22f933b01c30e0f68f46f8a73e474a1e8a893360 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Mar 18 06:14:51 2003 -0500 cannam@127: cannam@127: Renamed FFTW_IODIM, FFTW_R2R_KIND cannam@127: cannam@127: commit e57a38d55f979644a5fecd702c1d4bd105b1eac3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 18 00:30:17 2003 -0500 cannam@127: cannam@127: added rfftwnd.eps to dist, so that transfig is not required for people trying to build other formats (e.g. ps); thanks to Brian Gough for the bug report cannam@127: cannam@127: commit f5713b796921f5e1cfded3ce96e33c6df0d09a8c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 17 15:17:59 2003 -0500 cannam@127: cannam@127: pointer to upgrading section from tutorial cannam@127: cannam@127: commit 7e222b6349b8a2bceeb8703d347715fb763efadd cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 17 14:44:40 2003 -0500 cannam@127: cannam@127: make print_plan and fprint_plan, so that the former can be more easily called from other languages cannam@127: cannam@127: commit f358b64955871f01d87a42a05275f5f0cb5094e5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 17 14:19:10 2003 -0500 cannam@127: cannam@127: whoops, forgot to change equation image links to .png cannam@127: cannam@127: commit ea32f5a93a88f6ddec9185886bbbea43cf8ed067 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 17 04:15:50 2003 -0500 cannam@127: cannam@127: fixed c++ linkage problems cannam@127: cannam@127: commit de7c276d4b4ab36471c8dcb639d3c522d2cbe7cc cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 17 03:25:17 2003 -0500 cannam@127: cannam@127: Removed ``const'', otherwise c++ link fails cannam@127: cannam@127: commit f3bf675c6d0003e3087d634aab2ef34a6745dcb9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 20:24:31 2003 -0500 cannam@127: cannam@127: fixed C++ annoyances: void* casts, and global variables are static by default(?!?) cannam@127: cannam@127: commit 45e54b3f9a8c0b5942cc21c0b2d2f19682d3a7c0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 15:29:11 2003 -0500 cannam@127: cannam@127: ranlib bug is in binutils cannam@127: cannam@127: commit a17b7eb5a896ea6d7ca3f73fd7251bfc76de500d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 15:26:42 2003 -0500 cannam@127: cannam@127: ranlib Irix bug cannam@127: cannam@127: commit f482abd59b1c2afed27292d79bef782b935d0c51 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 15:13:35 2003 -0500 cannam@127: cannam@127: start with random tests cannam@127: cannam@127: commit a216647a57733c53d3407957caaaf759ed0dd700 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 15:00:04 2003 -0500 cannam@127: cannam@127: silenced some compiler warnings, eliminated unused variables, and fixed Makefile.am for f77funcs.h cannam@127: cannam@127: commit 2b581243067955d0e82eb7cf487def793b8f66b6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 14:55:13 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 058c4751ed4a98a52e6a878c78335f4997c60294 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 14:28:22 2003 -0500 cannam@127: cannam@127: 3dnow is float cannam@127: cannam@127: commit 0b50f9ef3433e59f0b9cc7983652a8fe3c361fdf cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 14:27:45 2003 -0500 cannam@127: cannam@127: fixed k7 docs cannam@127: cannam@127: commit 9179a17f1ba8db6733ccb2dbe541aa3f5f59727e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 14:19:10 2003 -0500 cannam@127: cannam@127: SGI compilers now support inline cannam@127: cannam@127: commit 353d40e964502f46aba99f094c08fd610cb9fc2c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 14:18:32 2003 -0500 cannam@127: cannam@127: cruft cannam@127: cannam@127: commit 455c3aa4c3e0cc6d5404c78ef12ed70b8751da45 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 14:15:47 2003 -0500 cannam@127: cannam@127: texinfo doesn't like commas in nodes cannam@127: cannam@127: commit 323b6d34cebb6be520075efaf4eeef0a369a6635 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:52:04 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 7762fe2f89dc86791560cad9326ece6fbdbceaf7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:47:44 2003 -0500 cannam@127: cannam@127: f77funcs.c -> f77funcs.h so that people don't try to compile it cannam@127: cannam@127: commit acd3f5b16c0b85acfad30bb086199cc65bc6b326 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:46:11 2003 -0500 cannam@127: cannam@127: minor changes cannam@127: cannam@127: commit 8d4f8a05ac24ce13ba6adea137099c22c6f5362b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:39:24 2003 -0500 cannam@127: cannam@127: updated compiler bug list cannam@127: cannam@127: commit dc84fdefd84cac3bd6ecf521f48ca6cab0ae2b0e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:39:07 2003 -0500 cannam@127: cannam@127: noted how to set CC cannam@127: cannam@127: commit d71b55ff07b10fe5ee5dc24799511bdbb0b3f772 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:01:01 2003 -0500 cannam@127: cannam@127: TODONE cannam@127: cannam@127: commit 454b2a79327b2582f18024204a6ab683d97f9f41 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 13:00:42 2003 -0500 cannam@127: cannam@127: yikes, bugfix cannam@127: cannam@127: commit e741c61f2ab8b259c217e9e25adbcece21a6be4b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 16 10:26:28 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit f0073024ddb3bb621a4c71fcc7ddb575adf42871 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 16 09:24:19 2003 -0500 cannam@127: cannam@127: Report SIMD extensions in version string cannam@127: cannam@127: commit 0b40f7e79a8110bd4d2215f9d81a3d100f1e9ecc cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 18:56:11 2003 -0500 cannam@127: cannam@127: more verbose output cannam@127: cannam@127: commit 1310aa1ef6043afa44bc6c8bcc2d7b3bae66190c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 17:41:25 2003 -0500 cannam@127: cannam@127: a couple of additional non-Unix instructions cannam@127: cannam@127: commit 12cb13aafd73275762b5f2c098c436457b8f9be9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 17:15:26 2003 -0500 cannam@127: cannam@127: hyphen cannam@127: cannam@127: commit 7aea3d41ed7a9dde86b14f410caf606a05f15fd5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 17:12:29 2003 -0500 cannam@127: cannam@127: softened cannam@127: cannam@127: commit 2a251916b17e7380f33bf556d666781828819789 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 17:09:44 2003 -0500 cannam@127: cannam@127: added FAQ, used PNGs cannam@127: cannam@127: commit d3669c90789fbfcc99404a8fbd8d90540fae6c52 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 15:29:43 2003 -0500 cannam@127: cannam@127: great copyright update cannam@127: cannam@127: commit 1b82fbfbe632120cba76c9c6107bd3e1abbe4547 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 15:14:02 2003 -0500 cannam@127: cannam@127: threads in make check cannam@127: cannam@127: commit a7ebafd6aec670afd0a9d5165893abf7d7413870 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 15:11:24 2003 -0500 cannam@127: cannam@127: fixed const warnings cannam@127: cannam@127: commit b72d4726555aa5ef40e612f712eaa2190324c89e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 15:08:25 2003 -0500 cannam@127: cannam@127: make sure spawn_loop size > 1 (it has to be at least > 0 lest we crash, but > 1 is an optimization) cannam@127: cannam@127: commit 8f82cc0405e8d264d1a201e4b65d0e82e5822834 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 15 14:00:17 2003 -0500 cannam@127: cannam@127: hpux seems to want machine/sys/inline.h as opposed to cannam@127: machine/inline.h. cannam@127: cannam@127: commit 195978c28fbdd1b1ead25d381c9c6af6f71a74fb cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 13:36:56 2003 -0500 cannam@127: cannam@127: Sourceforge is really SourceForge.net, and is run by VA cannam@127: cannam@127: commit 93eaa99ca18255b538bd37c4742ff87898a9350c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 13:34:05 2003 -0500 cannam@127: cannam@127: comma cannam@127: cannam@127: commit aa16c88c1efdf9283884a6f3c28bda36d54c1cb9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 13:31:42 2003 -0500 cannam@127: cannam@127: fixed AMD company name cannam@127: cannam@127: commit fa4887fa3ddccb2e53b50158d92f8cb9da3223f2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 13:29:41 2003 -0500 cannam@127: cannam@127: minor changes cannam@127: cannam@127: commit 689f73454e57451cc4ceca48e6c9b3856550cc3f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 13:13:55 2003 -0500 cannam@127: cannam@127: more emitter->read_char renaming cannam@127: cannam@127: commit 469d7370865e70079d60fc5d2144c477847ff50b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 13:08:45 2003 -0500 cannam@127: cannam@127: more wisdom docs, noted wisdom utilities cannam@127: cannam@127: commit 69c2e6ee0d6523c9181828e9d918d00390f1b07f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 11:41:32 2003 -0500 cannam@127: cannam@127: compound adjectives are hyphenated cannam@127: cannam@127: commit 1c816b975a4d35c3296bceb2700bc665c2838788 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 11:40:30 2003 -0500 cannam@127: cannam@127: fftw does support another type of packed array via r2r cannam@127: cannam@127: commit 4510d672da97fc9273a574d9cad23f807c811192 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 11:29:12 2003 -0500 cannam@127: cannam@127: write_char/read_char for export/import functions cannam@127: cannam@127: commit 5d042765f68d22c08849f8120b432d637364a95a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 15 11:19:19 2003 -0500 cannam@127: cannam@127: comments cannam@127: cannam@127: commit ec4d319ec4d855dd2e5c3521429d77dcba1deffa cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 15 10:08:26 2003 -0500 cannam@127: cannam@127: Enabled randomized-cse cannam@127: cannam@127: commit 85619e6f972e3105691588bba210448ad468726f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 15 09:47:49 2003 -0500 cannam@127: cannam@127: Changed to 3.0-beta1 cannam@127: cannam@127: commit 6c58169a5ef565ec595054c8a1a3644a119575ad cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 15 09:07:31 2003 -0500 cannam@127: cannam@127: First complete draft cannam@127: cannam@127: commit e014222e1611b0fda35eb4e81010d764371f645a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 15 08:37:52 2003 -0500 cannam@127: cannam@127: EMITTER is a misnomer cannam@127: cannam@127: commit b4e71cdebd8e08a8e4cb6e4e021c9839b0240220 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Mar 15 05:50:50 2003 -0500 cannam@127: cannam@127: Revision, wisdom tutorial, acks. cannam@127: cannam@127: commit 304d6a33a960a6867e345b7a2391f580de183901 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 22:59:04 2003 -0500 cannam@127: cannam@127: noted OpenMP cannam@127: cannam@127: commit f0132ff87cdec8cce3eec22776267630ce5d52a8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 22:38:49 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit d32e3536671b2ddf95fd19eefd595903f53369c9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 22:38:30 2003 -0500 cannam@127: cannam@127: comments cannam@127: cannam@127: commit 758a708f03680fe53ce46466e344370a9537adfc cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 22:38:05 2003 -0500 cannam@127: cannam@127: reformatting cannam@127: cannam@127: commit f31a618619a119ba5df49807d225f5fef53e2acc cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 22:26:28 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 47acccb2c662f75a8b9b082032072bfa154f13e5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 22:11:23 2003 -0500 cannam@127: cannam@127: some threads fixes, and added experimental semaphore (pre-thread-spawning) and Linux spinlock support cannam@127: cannam@127: commit b3f95134caa95e434d418ab40f2bb57c07521a33 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 20:50:46 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit caedcb4f9b8df5449616654ec8782156a2e63e7f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 18:23:03 2003 -0500 cannam@127: cannam@127: added note that FFTW_PATIENT will disable threads if they are not beneficial cannam@127: cannam@127: commit 34677912b28d4342f4ac1f84e27ee248d2c9ca71 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 18:20:44 2003 -0500 cannam@127: cannam@127: made fftw_cleanup* more restrictive, in that we don't want to cannam@127: guarantee that previously created plans will still work (they won't, cannam@127: in the case of threaded plans and fftw_cleanup_threads), and there is cannam@127: no reason to provide such a guarantee anyway. cannam@127: cannam@127: commit 4311c764859ea3a4a45fbb507ff0e131d12a5d44 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 17:23:13 2003 -0500 cannam@127: cannam@127: Moved version.c from kernel/ into api/ cannam@127: cannam@127: commit b79acfd84c9dc9bf6ce933ef72af7aafa01623e4 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 17:19:50 2003 -0500 cannam@127: cannam@127: icc-7.0 requires -openmp cannam@127: cannam@127: commit 36f49567ecc9ec71ab72b760ee70ceb688f51f4c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 14:47:52 2003 -0500 cannam@127: cannam@127: Ensure that one can do make dist given the distribution cannam@127: cannam@127: commit 266bb8c14f0aa494b54fcaf1fd0b517c646d5618 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 14:38:11 2003 -0500 cannam@127: cannam@127: Dist fftw3.pdf, not fftw.pdf cannam@127: cannam@127: commit a79801bc40a8ba8ba6f7b27f78aebb9426010b5a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 14:36:25 2003 -0500 cannam@127: cannam@127: Support -onthreads=%d cannam@127: cannam@127: commit 84c91507e6f7f6a050cc8651c7ee8c017d5d1b2f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 14:34:21 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit a8ef843faf74d0384c6ee1320b456f6aae56c5b2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 14:33:27 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 3e4f6ed2ad2fda1dbaa2bb444f81cbf116ab1931 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 12:32:18 2003 -0500 cannam@127: cannam@127: fftw_real is gone cannam@127: cannam@127: commit b55295b022d814a869b207fea2dbbb79c5091525 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 14 12:26:04 2003 -0500 cannam@127: cannam@127: typos cannam@127: cannam@127: commit 469579587defd8532f362c0ca4a2935532bae16a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 06:21:43 2003 -0500 cannam@127: cannam@127: More BENCH_DOC strings cannam@127: cannam@127: commit 18f0d31d803f348a8494ac190b4b9ff8d9be7a97 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 05:58:53 2003 -0500 cannam@127: cannam@127: Fixed xref's cannam@127: cannam@127: commit d39f035994e443ebbc933eae51b3d9116bc50bb4 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Mar 14 05:38:26 2003 -0500 cannam@127: cannam@127: Revised manual (esp. intro and tutorial), fixed texinfo hackery cannam@127: for figures. cannam@127: cannam@127: commit f0cf0419996f46abb0bdf85068d67c1f88435a87 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 02:42:33 2003 -0500 cannam@127: cannam@127: redirect users from guru execute to advanced interface, if possible cannam@127: cannam@127: commit d30d60239f8f57975f53876649f04f04458b8d90 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 02:35:22 2003 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit cc3b4e3f2fd1880b0a9ced57de8bc592ac868aab cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 02:28:51 2003 -0500 cannam@127: cannam@127: use correct heading level cannam@127: cannam@127: commit 16e33bb6e9eba6c6ac3a3b5e88192f0937cbc79a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 02:24:37 2003 -0500 cannam@127: cannam@127: html generation cannam@127: cannam@127: commit 8ea08e261cef0528db1c181268c6aabca6c52e50 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 01:44:00 2003 -0500 cannam@127: cannam@127: added equation GIFs cannam@127: cannam@127: commit 6b511ad0e8551382fb008d5f7d9d6db7c923f5d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 01:43:27 2003 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit b223dbcdf2607d546dcde4593dfeb29740b5a2c3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 01:26:46 2003 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit 8e6421b39b31952d4cde709e9a7dc68146eeac77 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 01:25:12 2003 -0500 cannam@127: cannam@127: added multi-dimensional transform definitions cannam@127: cannam@127: commit da7ac31fa42d9b594d9a458bc86b31e326d2631b cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 00:14:03 2003 -0500 cannam@127: cannam@127: slight changes cannam@127: cannam@127: commit 4fa36533cd5df28fb24a7cd7678c4ff3a2b8e1f7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 12 00:06:34 2003 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 93fdbbd4434ff6db48765645e2af3eb2031caece cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 11 23:50:43 2003 -0500 cannam@127: cannam@127: added 1d version of What FFTW Really Computes cannam@127: cannam@127: commit 989a15455a04e193bd71a2fe4b1daea5649d0f2d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 11 21:17:54 2003 -0500 cannam@127: cannam@127: note in upgrading section about FFTW_PATIENT cannam@127: cannam@127: commit f94fc8414c8477ad076f17bed5a1bffe87557ea9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 11 15:18:39 2003 -0500 cannam@127: cannam@127: added cycle-counter section cannam@127: cannam@127: commit 32e58f9ac101c22551198abe31c5021196f69f0e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 11 14:53:44 2003 -0500 cannam@127: cannam@127: more ideas cannam@127: cannam@127: commit 54102c10c10da11afcf1dac0451ce4a1e064be8c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 10 17:41:35 2003 -0500 cannam@127: cannam@127: noted that indirect should probably be merged with rank-geq2, to make a rank-split solver cannam@127: cannam@127: commit e93a7d1eda3519a9467a0d1a7af57a176aae195c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 03:01:52 2003 -0500 cannam@127: cannam@127: added non-Unix installation instructions cannam@127: cannam@127: commit 910a5988b2529e4ebd33372540c9db14626a3e8c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:30:59 2003 -0500 cannam@127: cannam@127: also talk about stack alignment with SSE/SSE2 cannam@127: cannam@127: commit 620f6439ff6d382e7f79fba9735243ffbc4e98d6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:24:07 2003 -0500 cannam@127: cannam@127: made warning more dire cannam@127: cannam@127: commit 6c49e3a0d90853a504b55ee2bb9e67e6961334c6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:13:25 2003 -0500 cannam@127: cannam@127: fix cannam@127: cannam@127: commit 076cf960691702683f560140c3c90932f531c802 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:09:55 2003 -0500 cannam@127: cannam@127: number cannam@127: cannam@127: commit abe3e1b3e2ac5d1ce15dd74544550011079f056c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:09:08 2003 -0500 cannam@127: cannam@127: fix cannam@127: cannam@127: commit a43149065f2c521c8ce705f9ac0eeb519899ec2b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:08:01 2003 -0500 cannam@127: cannam@127: minor cannam@127: cannam@127: commit cb19343373774be75d78469cbcd3ac4f0f4a903a cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 02:04:45 2003 -0500 cannam@127: cannam@127: minor fix cannam@127: cannam@127: commit d962180e504c71e46dc5b2f71d2304c254fcdace cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 01:58:15 2003 -0500 cannam@127: cannam@127: cross-ref cannam@127: cannam@127: commit 35ef1ce130da4c0389a2f7cef5eaab36dbd614ae cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 01:57:31 2003 -0500 cannam@127: cannam@127: minor cannam@127: cannam@127: commit 28fe03b9f79a6a80be8cc0d02cfc87e090f408d5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 01:53:28 2003 -0500 cannam@127: cannam@127: more installation manual cannam@127: cannam@127: commit 650bf3b91d1fe392906f9aa25faed1707244f4f4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 00:43:40 2003 -0500 cannam@127: cannam@127: GNU-lly correct cannam@127: cannam@127: commit aec18000f9851e8985d704ee50f49ea4d17f324e cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 00:38:48 2003 -0500 cannam@127: cannam@127: started installation section cannam@127: cannam@127: commit f7bf8016fa681c46c51385297a58d6dae611862f cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 00:25:02 2003 -0500 cannam@127: cannam@127: added --without-cycle-counter option as a last resort cannam@127: cannam@127: commit e97d01d48d003b290d6d2da7dc53cea35c90357d cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Mar 7 00:07:12 2003 -0500 cannam@127: cannam@127: macros with () arguments were only standardized in C99, and we don't need them anyway cannam@127: cannam@127: commit 459a56abf74ce71af7c63047b31d39f7befefbb9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 23:10:41 2003 -0500 cannam@127: cannam@127: wording cannam@127: cannam@127: commit 20d77f4e2a461fab512a8b5cd0ccd301d42f3673 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 23:03:03 2003 -0500 cannam@127: cannam@127: parallelism cannam@127: cannam@127: commit 881feeb3f3d4813a30da4baf5d71b8af8ca72d23 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 23:01:47 2003 -0500 cannam@127: cannam@127: additions to upgrading chapter cannam@127: cannam@127: commit 59f6ac21ab762470d0d4740130fa2131cc3f684e cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 22:39:36 2003 -0500 cannam@127: cannam@127: noted additional humility of FFTW 3 wisdom cannam@127: cannam@127: commit cf933ec73d68c2839a79d1fc53ba1198dc63fe39 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 22:32:44 2003 -0500 cannam@127: cannam@127: renaming cannam@127: cannam@127: commit 724b52700268a45264d168aaf7a63977a16af8bb cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 22:31:00 2003 -0500 cannam@127: cannam@127: added placeholder for wisdom reference cannam@127: cannam@127: commit c286ee068195c75e012cdf36534aa5f4154b394f cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 22:29:38 2003 -0500 cannam@127: cannam@127: wrote upgrading chapter cannam@127: cannam@127: commit 35c5a163f14e561b45a226dece35564f5773ce69 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 18:01:10 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit de53b4abb5481ee319ffcfc4e4b215861d814ed6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 18:00:43 2003 -0500 cannam@127: cannam@127: placeholder for upgrade chapter cannam@127: cannam@127: commit a8a06d66b81a785625077d6de4fb8699ee4c718f cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 13:47:49 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit a128a59973d9f74fa491a56fc22b374ad69a5ebc cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 13:36:38 2003 -0500 cannam@127: cannam@127: strengthed warning about time cannam@127: cannam@127: commit 271819893ab4e7634f8cee294f9c68612ff811f2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 13:35:42 2003 -0500 cannam@127: cannam@127: noted -t in example cannam@127: cannam@127: commit 08b64e3b15f7ad163677a348ba8d0a1a62720b07 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 13:21:03 2003 -0500 cannam@127: cannam@127: pay attention to WINDOWS_F77_MANGLING cannam@127: cannam@127: commit 5428bbf998b549e46c06f6f3e2ed9ff435304631 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:52:30 2003 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit 1462402c458e7a21360fcde1e6a5e9a023987747 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:51:02 2003 -0500 cannam@127: cannam@127: index cannam@127: cannam@127: commit 3cfc6a120672eeb46fca1300ba357ef6bff2b1cc cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:50:38 2003 -0500 cannam@127: cannam@127: documented C++ usage cannam@127: cannam@127: commit 675b0233f6e57d4aa15fe422acb4c156e2c3692a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:25:32 2003 -0500 cannam@127: cannam@127: got rid of overfull hbox TeX warnings cannam@127: cannam@127: commit a5a689c09a184e7f361240b46f8a74cd5c0bea78 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:20:38 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 9bdfa427108e546c8fd707d8bde9151b5cacd81d cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:20:13 2003 -0500 cannam@127: cannam@127: noted fftw_iodim split for Fortran guru interface cannam@127: cannam@127: commit ba02448b7f27ddbff45651477c0ca5ea4d28b7bd cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Mar 6 02:14:21 2003 -0500 cannam@127: cannam@127: added guru reference cannam@127: cannam@127: commit db7990c25a72ecb1a1acddfa63bdd8c38fdaeedf cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 5 22:56:05 2003 -0500 cannam@127: cannam@127: minor cannam@127: cannam@127: commit 58778ac5172128991fd8e88d4461004a03763596 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 5 22:45:31 2003 -0500 cannam@127: cannam@127: use @r{...} for comment text in code examples cannam@127: cannam@127: commit bd4b0411a2a7a9485f83d430455ff5d1571019f8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 5 13:14:04 2003 -0500 cannam@127: cannam@127: eliminate warning cannam@127: cannam@127: commit 87d217e8cd045402dbb4d9a4bc7ac81481edbcf9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 5 13:12:56 2003 -0500 cannam@127: cannam@127: SIMD_CFLAGS only for simd code cannam@127: cannam@127: commit 8346b6688d8e88aa91864685b77de030e8cb2549 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Mar 5 11:06:41 2003 -0500 cannam@127: cannam@127: Minor changes. cannam@127: cannam@127: commit 181d6c8fbdca0f24c1feb199c9a29edcf2187977 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Mar 5 02:13:34 2003 -0500 cannam@127: cannam@127: cross-compiling with MinGW can't detect f77 mangling, so add an option to use what seems to be the most common styles cannam@127: cannam@127: commit 17f9e2aabc5526c6614d7055960c5e7f5fda3720 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 20:00:31 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit b0715eb2e0f6662e3b3b41adf70799a31c2ab630 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 20:00:13 2003 -0500 cannam@127: cannam@127: we only use our-malloc-16 on machines where size_t == uintptr_t, so don't bother doing the right thing with the benchmark cannam@127: cannam@127: commit 72d331d4dbb9bf0bed0796e05eaf970a17c2975a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 19:46:09 2003 -0500 cannam@127: cannam@127: support WITH_OUR_MALLOC16 cannam@127: cannam@127: commit d2ee17676db2b01e1d57b6f6fcebe4c9c8987fff cannam@127: Author: fftw cannam@127: Date: Tue Mar 4 18:50:53 2003 -0500 cannam@127: cannam@127: automatically add -msse etcetera for --enable-sse etcetera cannam@127: cannam@127: commit 0a7cb6363f8effac8a34176c7b31d1dfbe4e71d0 cannam@127: Author: fftw cannam@127: Date: Tue Mar 4 18:24:26 2003 -0500 cannam@127: cannam@127: got rid of const warning cannam@127: cannam@127: commit f27a29dff516ba8bf8bd22a3affe1e881a045389 cannam@127: Author: fftw cannam@127: Date: Tue Mar 4 18:22:48 2003 -0500 cannam@127: cannam@127: missing header cannam@127: cannam@127: commit 58b8d88bdb16fde7d1400c93b1d976af4a29acaf cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 15:55:47 2003 -0500 cannam@127: cannam@127: fixes cannam@127: cannam@127: commit a636d3b26c9ca10c0225bb058035e2f99ae41383 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 15:53:26 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 530bdb066779445d91537bb42fafd03d98d24bd1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 02:22:14 2003 -0500 cannam@127: cannam@127: started guru reference cannam@127: cannam@127: commit c44336102065022482f5d8a4eda068247672c05c cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 01:44:09 2003 -0500 cannam@127: cannam@127: use same FFTW_IODIM between precisions cannam@127: cannam@127: commit da6302aba33f0dc74c9da6d7cd4824a6c431c948 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 00:25:57 2003 -0500 cannam@127: cannam@127: renamed section cannam@127: cannam@127: commit bf45437f266c9ce170d54e87466ba34f41b1937d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 00:21:49 2003 -0500 cannam@127: cannam@127: no need for "advanced" in subheadings cannam@127: cannam@127: commit 5fb9bd9fe4b93abeb0aa4b00e1ca6e9057da2fbd cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 00:20:05 2003 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 0127b618539bcb2ddf8634d4bb09c10673ba26a5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Mar 4 00:17:23 2003 -0500 cannam@127: cannam@127: finished advanced interface cannam@127: cannam@127: commit 76aa5434ffee4220caa0b1935d813723d43d55eb cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 23:26:12 2003 -0500 cannam@127: cannam@127: more advance interface docs cannam@127: cannam@127: commit 05a9b164357317a362a1f4e0acb2067faa66910e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 23:12:09 2003 -0500 cannam@127: cannam@127: fail for win32 cannam@127: cannam@127: commit c49ad63f2fb49af4c81ea1fde51303013e637d7b cannam@127: Author: fftw cannam@127: Date: Mon Mar 3 17:18:48 2003 -0500 cannam@127: cannam@127: shortened help string cannam@127: cannam@127: commit 52ebcb06b186e8f796fdc71ae30d3ac7e9e35017 cannam@127: Author: fftw cannam@127: Date: Mon Mar 3 17:16:17 2003 -0500 cannam@127: cannam@127: fixed cross-refs cannam@127: cannam@127: commit 331a793c80e1bb04018aad92d07791ff432d792e cannam@127: Author: fftw cannam@127: Date: Mon Mar 3 17:07:27 2003 -0500 cannam@127: cannam@127: FFTW_POSSIBLY_UNALIGNED -> simpler FFTW_UNALIGNED in API, added bench option cannam@127: cannam@127: commit 3ba1c479988c55e2f9244fac654f491c5b1c4b78 cannam@127: Author: fftw cannam@127: Date: Mon Mar 3 16:58:07 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 62a1622e28fcc9408467bccee64c50f977243b7f cannam@127: Author: fftw cannam@127: Date: Mon Mar 3 16:52:58 2003 -0500 cannam@127: cannam@127: noted assumption cannam@127: cannam@127: commit b6a1f1234fe0834ad8c7a313fc15c710bffafdc6 cannam@127: Author: fftw cannam@127: Date: Mon Mar 3 16:50:33 2003 -0500 cannam@127: cannam@127: provide our own malloc16 routine because of Windows lossage cannam@127: cannam@127: commit 22de7295407d77062d3611d326295950f90d4907 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 13:28:12 2003 -0500 cannam@127: cannam@127: capitalization cannam@127: cannam@127: commit 5756c9b659e1dda142a21c8c4c8fed00015bf29d cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 13:26:32 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 0f92b4f922681df3c6ea4a35bafb8c32907a028d cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 12:55:57 2003 -0500 cannam@127: cannam@127: vertical skip looks better than indenting for setting off short paragraphs cannam@127: cannam@127: commit 767a89f2268461313cb0a3666be311640bb288af cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Mar 3 06:34:09 2003 -0500 cannam@127: cannam@127: Removed franz-mode. Automake was distributing franz files cannam@127: whether franz mode was enabled or not. cannam@127: cannam@127: commit d40ea4ed0a561aa7f85008bb970d07b33010a0eb cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 01:44:00 2003 -0500 cannam@127: cannam@127: made output boundary conditions more prominent; they are important, cannam@127: because they make the different transform types inequivalent in cannam@127: parity cannam@127: cannam@127: commit 909ed5b34a848e505c9a62fcb5b07d346183a43d cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 01:17:28 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit feb1fc01699f139143e536e1d0f961b904bba74e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 01:17:07 2003 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 8d2e91da57095741496a5ae8b809cee8bd01bdb9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 01:10:28 2003 -0500 cannam@127: cannam@127: started advanced reference cannam@127: cannam@127: commit 34cc962abf3c75c27328c21fb2c9b053426870f8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Mar 3 00:52:02 2003 -0500 cannam@127: cannam@127: r2r reference cannam@127: cannam@127: commit 6a32d0463a93a19f01e9b13bdc2e0d73857c7eaa cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 23:51:21 2003 -0500 cannam@127: cannam@127: workaround for info formatting bug cannam@127: cannam@127: commit a76009f2fb554d2af97e39f4857b70d26a263bf5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 23:47:19 2003 -0500 cannam@127: cannam@127: noted lack of fftw_malloc in Fortran cannam@127: cannam@127: commit 53555b1acdefbc4b092702bcd7defa71dd523ee7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 23:42:52 2003 -0500 cannam@127: cannam@127: parallelism cannam@127: cannam@127: commit a84b5314b96882b7495c7d0fbdd91a73f678683f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 23:39:54 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit f4b30c1aeb9cadcb0ef3586a40e2a41a6087304f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 23:33:02 2003 -0500 cannam@127: cannam@127: r2c/c2r reference cannam@127: cannam@127: commit 9afb0869850070a47c3b45df511efdaef0c19292 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 22:44:10 2003 -0500 cannam@127: cannam@127: table of contents was being included twice cannam@127: cannam@127: commit 9433ef02af21f2e3ee1c5a5e6034a2e5a02663af cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 22:42:29 2003 -0500 cannam@127: cannam@127: minor changes cannam@127: cannam@127: commit 34aaf0acd96dc522e8b71c3844077a7d28149690 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 21:54:13 2003 -0500 cannam@127: cannam@127: started reference section cannam@127: cannam@127: commit 10afdab4c99f7d367227f61d6ea87e43113379ef cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 19:10:02 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit da1655a272a6bd0bf3db360605818d3684e01919 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 19:03:23 2003 -0500 cannam@127: cannam@127: started ref. section cannam@127: cannam@127: commit a3cc56c2b538f79864f787f9480a7da21017624a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 18:50:58 2003 -0500 cannam@127: cannam@127: fftw_flops takes const plan cannam@127: cannam@127: commit e6c9dd42b944a416f6cca057b2277acb2a00d370 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:54:14 2003 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 9d97e6245d45d65061499080021f2e0c877803b6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:52:41 2003 -0500 cannam@127: cannam@127: added "Wisdom of Fortran?" section cannam@127: cannam@127: commit 85f80c144fc9da705ddc7da87d0e437a4125d1db cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:50:37 2003 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 0f4d81b32a7ddf1e011dcc66a7ca3a6f01602aa9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:49:57 2003 -0500 cannam@127: cannam@127: wording cannam@127: cannam@127: commit 6c6dd67d7f64ce4ab293456c0b4fce7397b4204f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:46:13 2003 -0500 cannam@127: cannam@127: added comments cannam@127: cannam@127: commit d9ecf01ce4b7d0bb1c81de9097941541d96f68d0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:44:01 2003 -0500 cannam@127: cannam@127: added example file cannam@127: cannam@127: commit 37b6da9ec0958f78193e343ff5adbb7221039698 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:37:32 2003 -0500 cannam@127: cannam@127: don't print out READ WISDOM unless we have cannam@127: cannam@127: commit c476c76dbda1de2cfcfed5db46f8eb6a59ca5eda cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 15:36:28 2003 -0500 cannam@127: cannam@127: EOF is not a space cannam@127: cannam@127: commit 789f94ba726188b22495dffa33536923784cc893 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 2 09:14:37 2003 -0500 cannam@127: cannam@127: Turn on inline by default cannam@127: cannam@127: commit f76cd82b2e8d570d38aafcd3bc479871a6bfef71 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Mar 2 07:11:56 2003 -0500 cannam@127: cannam@127: Optionally inline loop in notw codelets cannam@127: cannam@127: commit 4ee60a97aba5df7daa9a1f0f20fc8a18b4caeef9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 01:37:41 2003 -0500 cannam@127: cannam@127: updated nodes cannam@127: cannam@127: commit a760bacb99bcb4d1b37deac1a0d03048564f06ae cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 01:37:19 2003 -0500 cannam@127: cannam@127: wrote most of Fortran chapter cannam@127: cannam@127: commit 84b26fd1d2d412fc5dae194fa4f49ea8c5ad803b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 00:58:37 2003 -0500 cannam@127: cannam@127: citation cannam@127: cannam@127: commit 90c66908b4f24f05f5a77a85d890ef77a5946747 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Mar 2 00:57:22 2003 -0500 cannam@127: cannam@127: added parallel FFTW chapter cannam@127: cannam@127: commit 1a89e4fc8d30e58c46d409543e5641d74d82012b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 20:42:23 2003 -0500 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 125c6e2e61c2977a10fe882134b6daa518d211b6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 20:34:38 2003 -0500 cannam@127: cannam@127: added inlining to TODO cannam@127: cannam@127: commit 86f19bdcd118e4f74034a5acf2a9f46ae0dd563b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 19:36:26 2003 -0500 cannam@127: cannam@127: added K cannam@127: cannam@127: commit c471cfe8ed04c68bd3ba96de578160018676966f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 19:15:18 2003 -0500 cannam@127: cannam@127: use K for constants cannam@127: cannam@127: commit c9132f12b56356608c7430b1aa8674c57982cf6f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 19:14:54 2003 -0500 cannam@127: cannam@127: fixed cross-ref cannam@127: cannam@127: commit 2c552e93b7ac76c6ed2cb15d84fb724e71d90901 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 19:14:16 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit ffd88e528368512ad6260f9829d093be01b0b8e0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 18:50:43 2003 -0500 cannam@127: cannam@127: cleanup cannam@127: cannam@127: commit eb500b0aee97bc247fadc5f14053addd510f8911 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 18:46:38 2003 -0500 cannam@127: cannam@127: "words of wisdom" by itself is a little too obscure cannam@127: cannam@127: commit c110b9bcf9dc0e3ac3bd0a9dc0aa04a3003808ab cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 18:43:21 2003 -0500 cannam@127: cannam@127: re-added multi-dimensional array stuff cannam@127: cannam@127: commit 3c1809be37bedc7b19bb0ad1645d2d0c55fb24af cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 18:15:22 2003 -0500 cannam@127: cannam@127: added alignment section cannam@127: cannam@127: commit 5ea9d154e8d9b180445e82c228f66dc620435630 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Mar 1 16:34:21 2003 -0500 cannam@127: cannam@127: shrunk code cannam@127: cannam@127: commit 969e6184c37360147d4377765e4209f740bbbc63 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 20:22:00 2003 -0500 cannam@127: cannam@127: slight compression cannam@127: cannam@127: commit 61f49745af277cf662c0b684d812bb937991da02 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 19:01:20 2003 -0500 cannam@127: cannam@127: style cannam@127: cannam@127: commit 7a450c9741b7d712c4b0647c8348b6f5c16c5b5b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 18:46:53 2003 -0500 cannam@127: cannam@127: noted not in API cannam@127: cannam@127: commit dcb2c790e6afe7674f917a64a27a5d757de04d54 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 18:43:14 2003 -0500 cannam@127: cannam@127: more updates cannam@127: cannam@127: commit 9c734e0be5f7e454d53ea076c85b07a1563d12d0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 18:38:42 2003 -0500 cannam@127: cannam@127: slight updates cannam@127: cannam@127: commit 3e0a26ba8c35cc39e451dddb4ff538a9b6897853 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 18:28:58 2003 -0500 cannam@127: cannam@127: great const-ification of apply/solve and print cannam@127: cannam@127: commit 7531ed4ba4a1cd9a4e9caf11c225f930a72efc73 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 17:51:15 2003 -0500 cannam@127: cannam@127: make fftw_execute take a const plan, to remind the user that it is re-entrant (or should be)... cannam@127: cannam@127: commit 4688736baa020b3ea5f442e36b70d793b431c5c5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 17:29:40 2003 -0500 cannam@127: cannam@127: weakening cannam@127: cannam@127: commit 0318454412dbe1cd837ddb068bd343ca6e112011 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 17:28:48 2003 -0500 cannam@127: cannam@127: note cannam@127: cannam@127: commit 91b816d6c3f80bdb7e0d0116306ae7ffd2c455a8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 17:27:10 2003 -0500 cannam@127: cannam@127: footnote about why DHT is provided cannam@127: cannam@127: commit ade0a1b900ff7aad1f0b34334d0aeef444f9c6f1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 15:07:03 2003 -0500 cannam@127: cannam@127: index cannam@127: cannam@127: commit 02af64c2431e9ebe1f95750c16596bb16b0130e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 15:05:48 2003 -0500 cannam@127: cannam@127: added DHT tutorial cannam@127: cannam@127: commit 4c0a2b93c6fabdbd47e06a9f6ba76008bcb560a1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 14:36:45 2003 -0500 cannam@127: cannam@127: fixed O(n log n) cannam@127: cannam@127: commit fd7ecdadbf64ae5027bac415310c4a98a276db60 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 14:12:15 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit ee8d32cc161fa77c6d9566dfb000a80af883f835 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 14:06:22 2003 -0500 cannam@127: cannam@127: slight improvements cannam@127: cannam@127: commit 0b2ef4ccfd465403919403e5151753a4280f683e cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 00:55:50 2003 -0500 cannam@127: cannam@127: addition cannam@127: cannam@127: commit 22bd399df29e7380522c5bac340a3f04a466fd79 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 28 00:54:09 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit 1b357d49f4d4ee22c59374391be91ddb42813a2d cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 23:49:37 2003 -0500 cannam@127: cannam@127: fix cannam@127: cannam@127: commit 1c30eacc33d5c9d5daf303cfbbc5fa74e6a5bfa4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 23:43:56 2003 -0500 cannam@127: cannam@127: slight changes cannam@127: cannam@127: commit 053b9356142e3b05c1ee11800f497813e5c9f119 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 23:27:48 2003 -0500 cannam@127: cannam@127: added R{E,O}DFTab tutorial cannam@127: cannam@127: commit a793a4024b69b7e4ec4bbbeedb00508845c0cab2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 17:24:20 2003 -0500 cannam@127: cannam@127: fixes cannam@127: cannam@127: commit de5b2994a11c8c2b3d1948f43525864b0ac5d265 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 17:20:42 2003 -0500 cannam@127: cannam@127: fixes cannam@127: cannam@127: commit 027014da3b7f99190c9c1edbe0f6d0c0d15e043a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 17:11:54 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit 6359d6080ac4a827218faee02ba1bfe5a5a676bf cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 17:07:45 2003 -0500 cannam@127: cannam@127: documented r2hc/hc2r cannam@127: cannam@127: commit a44e1bc64be97cffdf71bf77dcb526786daa8efe cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 16:19:16 2003 -0500 cannam@127: cannam@127: minor changes cannam@127: cannam@127: commit 7186d1f0701c1507ce6b57f943f0d069c69e09d1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 13:54:06 2003 -0500 cannam@127: cannam@127: timed planner and unifying radix-2 butterfly loops are not critical for release cannam@127: cannam@127: commit e22ae82e9d2c007712ae8e8523a2ba4844265b26 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 13:51:20 2003 -0500 cannam@127: cannam@127: reodft/verify.c no longer exists cannam@127: cannam@127: commit d562aee6ca0c9e2c375d31a2f283ef5188b8819a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 13:44:19 2003 -0500 cannam@127: cannam@127: optimization: REDFT00 of size 2 is same as R2HC cannam@127: cannam@127: commit 35bca2a3e6f2b887fe4517dfed61eb4cc614f9ff cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 12:35:33 2003 -0500 cannam@127: cannam@127: R{E,O}DFT01 of size-1 is identity cannam@127: cannam@127: commit 3e86434a19f94bd85e576be96fb26b0db8456b7e cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 12:15:10 2003 -0500 cannam@127: cannam@127: minor simplification cannam@127: cannam@127: commit 23aeb956f45a31061c6f0bee5c78119e332e9d20 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 02:46:31 2003 -0500 cannam@127: cannam@127: fixed add count cannam@127: cannam@127: commit 629bf73abe3666100c7a3cdb795cdf85f1c3467c cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 02:25:04 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 46350e9b4b06fc596f73c2e8297276e38871fbcc cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 02:22:03 2003 -0500 cannam@127: cannam@127: another optimization cannam@127: cannam@127: commit 16310c985bad6d32fa0da6362c37fd375822d813 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 01:43:00 2003 -0500 cannam@127: cannam@127: added op counts cannam@127: cannam@127: commit 870808939ac67893ae3193d1eaf47d6722399743 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 01:29:32 2003 -0500 cannam@127: cannam@127: cleanup cannam@127: cannam@127: commit e13936e36480509c10d5f8da4806a17a1f2c9d34 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 01:17:23 2003 -0500 cannam@127: cannam@127: typo in comment cannam@127: cannam@127: commit 32c3d158f7f210901f1c16a8c8cbdfff05024993 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 01:13:49 2003 -0500 cannam@127: cannam@127: fixed comment cannam@127: cannam@127: commit 6e65b622f4e11f6b75ce19b92715054e01726a87 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 01:12:05 2003 -0500 cannam@127: cannam@127: use E instead of R cannam@127: cannam@127: commit b2dbcc1af3dac45c5dc937090de39d8c50f79f04 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 01:05:39 2003 -0500 cannam@127: cannam@127: more unrolling to eliminate if statements in loops, for speedups of 25-40% cannam@127: cannam@127: commit efdfcd1ab423b3b5f4c226859c38fe82ef8d5ee3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Feb 27 00:27:00 2003 -0500 cannam@127: cannam@127: some loop splitting to touch each element of output buf only once and eliminate some conditionals...speeds up by 30-40% cannam@127: cannam@127: commit ac2585fa04303d0a9733f25529a4de770165a96a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 26 17:48:26 2003 -0500 cannam@127: cannam@127: comma cannam@127: cannam@127: commit 12f6863d7ba56d03a828d47d95226914f7624343 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 26 17:46:17 2003 -0500 cannam@127: cannam@127: pointer to odd case cannam@127: cannam@127: commit b305de27048e5d88018afd557b9853fcfd938e7e cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 26 17:40:54 2003 -0500 cannam@127: cannam@127: precision -> accuracy (c.f. Kahan) cannam@127: cannam@127: commit 8cce3f1c36041dfd0f3099ccd2b4d07af10ba0ae cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 26 17:36:13 2003 -0500 cannam@127: cannam@127: added time limit for wisdom generation cannam@127: cannam@127: commit 57f9db2fb5d1498630bc04fa9ce59c0362383dc2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 26 13:24:36 2003 -0500 cannam@127: cannam@127: caps cannam@127: cannam@127: commit 194e3fe2a23b43433042f38567d615508f0219f0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 25 20:56:01 2003 -0500 cannam@127: cannam@127: another note cannam@127: cannam@127: commit 74d5d37f8b5f57257ac2996c1b78cd6e178009b2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 25 20:54:57 2003 -0500 cannam@127: cannam@127: note cannam@127: cannam@127: commit 4c454a521c659245d7d5328a0428abe8e0e65ca0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 25 20:42:08 2003 -0500 cannam@127: cannam@127: added new, more accurate (hopefully) reodft11 algorithms; added --disable-debug-malloc; added --impulse-accuracy-rounds=rounds flags to libbench2 for impulse-response accuracy tests cannam@127: cannam@127: commit 56c91af19d265df468a1c332950285ccc35cadf2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 23 14:07:48 2003 -0500 cannam@127: cannam@127: fftw_wisdom.1 is in $builddir, not $srcdir cannam@127: cannam@127: commit afb274d60def917682dcfb6752788ae69feb0e89 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 17 03:42:19 2003 -0500 cannam@127: cannam@127: pde cannam@127: cannam@127: commit 57844d17a4a5e42a9b3a6e264d4b9ef96a48b7d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 17 03:40:19 2003 -0500 cannam@127: cannam@127: consistent number cannam@127: cannam@127: commit da10f4a095936c4a272edf95561177e0ba1e0976 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 17 03:39:02 2003 -0500 cannam@127: cannam@127: started r2r doc cannam@127: cannam@127: commit 9339401bc1db11ab2b3ea8332adf2b7f8d2bd39d cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 17 02:31:51 2003 -0500 cannam@127: cannam@127: rfftwnd cannam@127: cannam@127: commit e9481965be99453d16fce50a2cec8a7189d50e5a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 15 17:02:07 2003 -0500 cannam@127: cannam@127: continued cannam@127: cannam@127: commit 387c70c9f598cc84949f9b36c3a7ec3aee478107 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 15 15:16:26 2003 -0500 cannam@127: cannam@127: started r2c/c2r docs cannam@127: cannam@127: commit 0df57f98fa114607c9ea5a9e17e8aa4fa92bd0c1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 15 01:12:52 2003 -0500 cannam@127: cannam@127: added r{e,o}dft11 accuracy test cannam@127: cannam@127: commit e24081ffd7a170743a930c91ec251fb1fa590072 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 15 00:42:48 2003 -0500 cannam@127: cannam@127: added more r2r accuracy checks cannam@127: cannam@127: commit da37c854fdf95a2cfc3cf2c6ef698ab1ed9e8a70 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Feb 14 19:19:54 2003 -0500 cannam@127: cannam@127: $< is a GNUism cannam@127: cannam@127: commit 01c0739002308b926e8ed648f93c2b46ef885404 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 21:02:16 2003 -0500 cannam@127: cannam@127: r2r test cases are in cannam@127: cannam@127: commit e0d1053729fe6e63cfc19bf040c14593ced050c5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 21:01:28 2003 -0500 cannam@127: cannam@127: added vector radix to TODO cannam@127: cannam@127: commit eabfd75e1f96eb039ac8ba4f612ad92a5de3f3f2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 17:21:33 2003 -0500 cannam@127: cannam@127: fixed cross-ref cannam@127: cannam@127: commit fe1a1f526ac5401ffbb69ddc61b07af2f9c08cfc cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 17:19:56 2003 -0500 cannam@127: cannam@127: shorter synopsis cannam@127: cannam@127: commit 73464a04bcc91f1244cca8812515833da6cad60c cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 12:53:19 2003 -0500 cannam@127: cannam@127: obsolete cannam@127: cannam@127: commit f235c4cdb767ed752563b5a12b609f4a606ae89d cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 12:52:53 2003 -0500 cannam@127: cannam@127: removed old dotens cannam@127: cannam@127: commit 1b45907552bf8c3c7e91e77b9256f904a7dc46db cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 12:52:16 2003 -0500 cannam@127: cannam@127: removed old verify files cannam@127: cannam@127: commit d2baa62fc65ce7b8c09581f2feaacd90466c07e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Feb 12 12:37:17 2003 -0500 cannam@127: cannam@127: disable threads support by default cannam@127: cannam@127: commit 6fc7d66c60a2e9bfbac7bba821b5329c9fde4b0b cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Feb 12 11:03:28 2003 -0500 cannam@127: cannam@127: Removed old test program cannam@127: cannam@127: commit 948df3a1949a1a5d9d8924a1c51c49d015477b73 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 22:30:55 2003 -0500 cannam@127: cannam@127: joke cannam@127: cannam@127: commit 06377bf381dbb2e1a05674678924168ee9235d46 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 22:27:44 2003 -0500 cannam@127: cannam@127: add --help and --version, to be GNU-lly correct cannam@127: cannam@127: commit c59c2fb43df57981f39141efe881ade700dffb3f cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 22:27:18 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit ecc46199c7967a7164deaa4f6be2ad734eb6c986 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 22:17:35 2003 -0500 cannam@127: cannam@127: better help cannam@127: cannam@127: commit e73d1cfefcfdffa9a318c184463973e309e1f421 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 21:47:35 2003 -0500 cannam@127: cannam@127: comma cannam@127: cannam@127: commit dc27e6924a7f6e054e0d542d855d4f62c9545ce2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 21:46:12 2003 -0500 cannam@127: cannam@127: formatting cannam@127: cannam@127: commit 06c5acf858b96e548a3d5664252103486c5dbb5e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 21:45:23 2003 -0500 cannam@127: cannam@127: man pages for tools cannam@127: cannam@127: commit d643ece55b08510928523882ac2213361d1eaf43 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 19:07:12 2003 -0500 cannam@127: cannam@127: added -V cannam@127: cannam@127: commit ad12cdca62eb5030d1388f12f7278fd1a3eb8a3a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 18:42:17 2003 -0500 cannam@127: cannam@127: added install-wisdom target cannam@127: cannam@127: commit 83162f468afd0941a99c408ae84e6c35ce43dbb3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 18:23:02 2003 -0500 cannam@127: cannam@127: another note cannam@127: cannam@127: commit 52735853d05221978df609981a95f9d89ec03c0a cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 11 17:32:56 2003 -0500 cannam@127: cannam@127: started r2r accuracy tests (only three kinds covered so far) cannam@127: cannam@127: commit 6fb598e12ddd2e595289c0d399cd7c283425540b cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 10 22:04:18 2003 -0500 cannam@127: cannam@127: silence warning cannam@127: cannam@127: commit b94eaa910fb2a707a185e743514f009a77663600 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 10 20:55:20 2003 -0500 cannam@127: cannam@127: gcc bug is now avoided. cannam@127: cannam@127: commit d142433a2935361da613eef685c306e1f86ef8cb cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 10 20:37:54 2003 -0500 cannam@127: cannam@127: Accuracy test cannam@127: cannam@127: commit 3e6c6925a0daf524ddff6ef711ebe2dbf07ebda2 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Feb 10 07:59:57 2003 -0500 cannam@127: cannam@127: There is no point in precomputing strides for the long-double code, as cannam@127: multiplication by sizeof(long double) cannot be folded into the cannam@127: addressing mode. This change also fixes the gcc-2.95 bug that causes cannam@127: miscompilation of certain codelets. cannam@127: cannam@127: commit 1cdf3be30717cb411fcb7272628ab72dc31ea3d0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 10 02:54:35 2003 -0500 cannam@127: cannam@127: added random r2r tests cannam@127: cannam@127: commit 13fd49dc504be79d65f5c3b254b08572689fcd71 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 10 02:44:58 2003 -0500 cannam@127: cannam@127: whoops, bugfix: missing stride for ro10 cannam@127: cannam@127: commit f0926d171845f84e02584361b0a6a9b6c4d68e71 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Feb 10 02:21:50 2003 -0500 cannam@127: cannam@127: formatting cannam@127: cannam@127: commit 2ec7cca77de0ed39b104a090158f4f3994f18343 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 23:24:52 2003 -0500 cannam@127: cannam@127: flop counts for reodft cannam@127: cannam@127: commit 1ec87d09b3698d5c2093d8436ea885225d67191a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 23:22:15 2003 -0500 cannam@127: cannam@127: declare aligned_main cannam@127: cannam@127: commit 9c3374ad54ec97ed408760b77234ea4980fcd311 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 20:56:06 2003 -0500 cannam@127: cannam@127: corrected rader op counts cannam@127: cannam@127: commit 6803f88282e3117c77721aff1a96515236b27fb9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 20:25:32 2003 -0500 cannam@127: cannam@127: punctuation cannam@127: cannam@127: commit e8cbdde425f97261b79551ea78f87322a4983bf3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 20:25:17 2003 -0500 cannam@127: cannam@127: noted need for better estimator cannam@127: cannam@127: commit 156eefce1a365107071ac016b4c818354a98e60b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 19:58:59 2003 -0500 cannam@127: cannam@127: noted F77 api fix for g77 mangling incompatibility cannam@127: cannam@127: commit e160cbe881f0f509fa09e6eedd76141b439c3ad9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 19:30:55 2003 -0500 cannam@127: cannam@127: build f77 header file of constants from fftw3.h cannam@127: cannam@127: commit 370b6e68c535ab81d29047d5fd3a9a48f7e3ebec cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 19:04:53 2003 -0500 cannam@127: cannam@127: updates cannam@127: cannam@127: commit f2c761d6d435ea22fc390b1e388dc0d01a747bd4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 19:03:34 2003 -0500 cannam@127: cannam@127: threads f77 api cannam@127: cannam@127: commit b84617e3c6d025d4f13cfa3056ddbdbd5227b961 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 18:54:00 2003 -0500 cannam@127: cannam@127: finished f77 serial api cannam@127: cannam@127: commit 86446f99fc266c435826ab0f0ca77b48117dd21f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 18:32:26 2003 -0500 cannam@127: cannam@127: added flops, slight cleanups cannam@127: cannam@127: commit b02c6ea6492b370ac0dde405bc4d899b3b4d4ab7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 9 18:11:48 2003 -0500 cannam@127: cannam@127: Oops, forgot #include cannam@127: cannam@127: commit 218af736c45f2ac117c4fe70c79029a7bb26ae33 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 9 18:08:26 2003 -0500 cannam@127: cannam@127: Removed duplication of stack-alignment code cannam@127: cannam@127: commit 5b5fc6186df8fa5214ae22ebaf84922aab584d90 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 15:48:15 2003 -0500 cannam@127: cannam@127: allow - to read problems from stdin cannam@127: cannam@127: commit c8e7f4b0b4ed904a7dc8b474f220d17bd061809e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 15:22:23 2003 -0500 cannam@127: cannam@127: added fftw-wisdom tool cannam@127: cannam@127: commit 216bb0693d91019be789666644d90c1f9afde7a5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 15:06:38 2003 -0500 cannam@127: cannam@127: elim. warning cannam@127: cannam@127: commit 8ff159c3583032eb2b661bb50b34d77344f1898e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 14:24:19 2003 -0500 cannam@127: cannam@127: destroy_input should not contaminate flags of other problems cannam@127: cannam@127: commit c1e578a3c33cee071a10e2f8f49a5dd29f4749ae cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 13:06:11 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 8b09de262bcd31d3ef04cff36791c389f75b733b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 13:01:45 2003 -0500 cannam@127: cannam@127: removed overzealous inplace check, which caused problems for rdft2 cannam@127: cannam@127: commit bfb7a5cab2f68265d33dea80716baec602a7c5ef cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 9 08:14:03 2003 -0500 cannam@127: cannam@127: Consistent syntax for RNK_MINFTY tensors cannam@127: cannam@127: commit 0f87db2efc6a2d72c3bb8584c195ee3682e09870 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Feb 9 07:31:13 2003 -0500 cannam@127: cannam@127: lisply-correct tensor print. We no longer need to parse tensors. cannam@127: cannam@127: commit 14826af57fa8cd1490d3d4d8111e64336ad638a4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 03:35:56 2003 -0500 cannam@127: cannam@127: removed completed items cannam@127: cannam@127: commit ba72775e69c1b4ff00b77a37bd0c80312bcc072f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 03:27:56 2003 -0500 cannam@127: cannam@127: slight renaming cannam@127: cannam@127: commit a96011aa06fe98812ad45afba51a6f1c3ceeab31 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 03:15:28 2003 -0500 cannam@127: cannam@127: multi-dimensional r2r verifier cannam@127: cannam@127: commit d2c2e3058ab81d087848fdd251e8bb6e92416710 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 02:40:22 2003 -0500 cannam@127: cannam@127: comments cannam@127: cannam@127: commit 5553af4969fd029313dc53f63201fa9c40acd051 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 02:38:26 2003 -0500 cannam@127: cannam@127: slight simplification cannam@127: cannam@127: commit 8aa7d693d055305129c2518385e9816529c9a334 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 9 02:36:25 2003 -0500 cannam@127: cannam@127: added 1d r2r verifier (triple ugh) cannam@127: cannam@127: commit ef489a80e6559cf2828da23340df129302681dd5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 8 22:23:00 2003 -0500 cannam@127: cannam@127: added vector transforms to random tests cannam@127: cannam@127: commit 826567b4d872cb6920840a850e0a584e0cc015e9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 8 20:59:07 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit b4d28e3488a63128b0ad1500d2e8b5777eadc8e4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 8 19:52:58 2003 -0500 cannam@127: cannam@127: fixed interaction between dwims for sz/vecsz with rdft2 transforms cannam@127: cannam@127: commit 0c8c54737beaa3db6a119769716f40416ddc7718 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 8 19:35:56 2003 -0500 cannam@127: cannam@127: added destroy_input flag/check cannam@127: cannam@127: commit 0e205231678541426c1a0bcd61b0442e7e24ad4a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 8 19:11:58 2003 -0500 cannam@127: cannam@127: added rdft2 verifier cannam@127: cannam@127: commit a20a05830b52221eda2f16ab7da1dc80e0e5a050 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Feb 8 13:31:14 2003 -0500 cannam@127: cannam@127: an additional check for in-place case cannam@127: cannam@127: commit 6096b268ec7fb9e5c1ad5d41aff355e8f674fd22 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 7 17:36:56 2003 -0500 cannam@127: cannam@127: slight fix: hc2r constraints are mostly determined by sub-plan cannam@127: cannam@127: commit 668b0af47a07011aaa3202ee70d3588aeca0ddd9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Feb 7 16:28:55 2003 -0500 cannam@127: cannam@127: make radix2-dft inapplicable to in-place/split case (r == rio, iio >= rio + n/2+1 != r + 1) cannam@127: cannam@127: commit 533f0a1824842664dfe63287e03800c2426b8ba5 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Feb 4 06:36:29 2003 -0500 cannam@127: cannam@127: Allow plnr->hook to be 0 cannam@127: cannam@127: commit 5c89a91a0efc3714980409d12055f91a2bd33693 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 4 03:25:36 2003 -0500 cannam@127: cannam@127: moved dft stuff into verify-dft cannam@127: cannam@127: commit 941da36d0fa1562ef98fd796b05c0a8a94c4ff94 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 4 03:25:00 2003 -0500 cannam@127: cannam@127: cruft cannam@127: cannam@127: commit 727cc86ece827ba21a236149c66ef6c7e7890d6b cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Feb 4 03:18:28 2003 -0500 cannam@127: cannam@127: further unify libbench2 and paranoid verifiers cannam@127: cannam@127: commit 6fb68912913cd9ab647b0206a713470e1bad462b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Feb 2 01:45:37 2003 -0500 cannam@127: cannam@127: typo in comment cannam@127: cannam@127: commit 3d1a5701f571ec275672faf3da2d7ea6f1e34b93 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 1 09:30:03 2003 -0500 cannam@127: cannam@127: Fixed p==2 case cannam@127: cannam@127: commit 723093b36b481e0f742822129f33998ba5acff14 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Feb 1 09:23:43 2003 -0500 cannam@127: cannam@127: Incorporated new find_generator by Greg Dionne. cannam@127: cannam@127: commit 93a75fda2dad56fbf69030eabdb09af0987e5316 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 31 20:46:24 2003 -0500 cannam@127: cannam@127: Removed nonportable call to gettext() cannam@127: cannam@127: commit 0d937fc4f0800cdad67d7a6a496c30c67c70b0ae cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 29 19:03:43 2003 -0500 cannam@127: cannam@127: uintptr_t is in in openbsd cannam@127: cannam@127: commit 9ffa4f6b400e1818a4c50a1385d916d501ff16b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 29 15:41:56 2003 -0500 cannam@127: cannam@127: Huge speedups in wisdom I/O. cannam@127: cannam@127: commit 426e786cc0662f3926cd79d3d76b0825a65ff445 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 19:36:51 2003 -0500 cannam@127: cannam@127: Added appropriate warning against likely future bug. cannam@127: cannam@127: commit b254ecc51abc22f1642e0bae9d6d22fbb2efb771 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 19:00:24 2003 -0500 cannam@127: cannam@127: Don't attempt to remove bogus wisdom entries. cannam@127: cannam@127: commit c19570082c79ce6d86613248e700ee17bb3582b8 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 18:16:24 2003 -0500 cannam@127: cannam@127: Fixed a couple of very very very nasty bugs---pointers became cannam@127: invalid after the hash table was relocated. cannam@127: cannam@127: commit 123972fa083c9fb07f18c3ee3a902a79606f5987 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 07:34:10 2003 -0500 cannam@127: cannam@127: Read wisdom at can_do() time, otherwise wisdom is destroyed. cannam@127: cannam@127: commit d1e805e6353a689a61b6aec66a28d568723717fc cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 06:54:38 2003 -0500 cannam@127: cannam@127: More conservative inheritance of blessings cannam@127: cannam@127: commit e718fe3fa7a7c4194011493e0bd86b78b222c0b9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 06:50:20 2003 -0500 cannam@127: cannam@127: Print the same info as it is hashed cannam@127: cannam@127: commit 84199fe5035171395b24754b6f4428513b850e84 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 28 06:49:48 2003 -0500 cannam@127: cannam@127: Print name of executable when FAILURE cannam@127: cannam@127: commit 3919d8a49fb4779e470deefd35cc3c7fc09c20ce cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 27 06:59:40 2003 -0500 cannam@127: cannam@127: New NO_SEARCH planner flag, which avoids searching altogether. cannam@127: A wisdom entry must lead to a NO_SEARCH-grade plan, or else the cannam@127: wisdom entry is bogus. cannam@127: cannam@127: commit 9534126e49e082098917ef5500133d8ef8a7289a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 20:45:21 2003 -0500 cannam@127: cannam@127: Use cosl()/sinl() when appropriate cannam@127: cannam@127: commit 5cc66fc2964feb54cff148e70280c083715d371f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 16:29:18 2003 -0500 cannam@127: cannam@127: Use null pointers when estimating. The estimator should never cannam@127: time anything. cannam@127: cannam@127: commit 8a54d02af36535be471d8326bf4e061165295320 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 15:19:01 2003 -0500 cannam@127: cannam@127: note cannam@127: cannam@127: commit 127681d03bd37c45649032138e7c976ec3395c99 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 15:16:22 2003 -0500 cannam@127: cannam@127: support multiple mangling schemes with g77 cannam@127: cannam@127: commit 757b13e27cfe6317d5c871796c129ec5b693e89b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 12:58:57 2003 -0500 cannam@127: cannam@127: fixed verbose, made random tests only use selected rank, use rank <= 4, fixed final flush_problems call cannam@127: cannam@127: commit c379edca317112097e76dacd0dfb69c83c319023 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 12:42:49 2003 -0500 cannam@127: cannam@127: fixed typo (count instead of maxcount) cannam@127: cannam@127: commit 6c0c2a4aac442f27536a584f1e619c69f6aa7ca6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 12:12:07 2003 -0500 cannam@127: cannam@127: hypot is no longer used cannam@127: cannam@127: commit 8466e0fb929081a67acbe832ddd155f33ee13734 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 12:07:43 2003 -0500 cannam@127: cannam@127: check for _alloca (MSVC) cannam@127: cannam@127: commit 34321edf6b705ea8f04c0ac903baf9a2d0239cd2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 26 11:56:53 2003 -0500 cannam@127: cannam@127: slight fix in assert cannam@127: cannam@127: commit c099d12f16d6131750ccde572b7651661b84881b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 11:55:39 2003 -0500 cannam@127: cannam@127: Allocate problem in all cases--- can_do may need correct pointers. cannam@127: cannam@127: commit 93ba509b275e8a2b798b237dae50927c04da5b74 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 11:51:27 2003 -0500 cannam@127: cannam@127: Nastier checks cannam@127: cannam@127: commit 91419140c877e227d804c4cbb18cb89b350527b3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 11:51:16 2003 -0500 cannam@127: cannam@127: X(use_plan) is a relic. cannam@127: cannam@127: commit 9cc664aacbc213b2cdbca13e686ca9f15f4d89f4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 09:23:16 2003 -0500 cannam@127: cannam@127: Print full pathname of the bench executable, so that I don't get cannam@127: confused when running multiple tests for different configurations. cannam@127: cannam@127: commit a755e0b1e768f7624d20ba4d564d9b658fc8aa45 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 26 07:35:46 2003 -0500 cannam@127: cannam@127: Split done() into done() and cleanup(), in order to test cannam@127: multiple problems with the same planner from the command line. cannam@127: cannam@127: commit c9a2310aa41b815190cd73c801d28f6b68635734 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 20:44:49 2003 -0500 cannam@127: cannam@127: Improved readability cannam@127: cannam@127: commit 6a7d0ba4578fa4f1989e521e80cd1504dddb5ff9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 25 19:17:26 2003 -0500 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 7e5332d67aa4dd505518874798560834170c2d1c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 25 19:16:53 2003 -0500 cannam@127: cannam@127: added macos9 mpallocatealigned function cannam@127: cannam@127: commit f3bba67e15e3ff2cc63e615a97ef4161af9fbb6d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 25 18:59:55 2003 -0500 cannam@127: cannam@127: sometimes __APPLE__ is defined instead of __MACOSX__ cannam@127: cannam@127: commit eb44a626174b2c4b5a3b91799f929f087ab89b90 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 25 18:54:39 2003 -0500 cannam@127: cannam@127: macos x malloc is already 16-byte aligned cannam@127: cannam@127: commit e1f4dfe3d7d517b655cbf1d9f34910cf5b91f16f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 13:38:32 2003 -0500 cannam@127: cannam@127: Include because uintptr_t is defined there cannam@127: on solaris. cannam@127: cannam@127: commit 2e0d88fe660fa8d5dd70ac8b4d7ce327b8e3143a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 13:22:59 2003 -0500 cannam@127: cannam@127: Oops---forgot getopt_long cannam@127: cannam@127: commit 8ca5ca5adffa7f09e53fd6876720807c9e09b526 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 13:17:29 2003 -0500 cannam@127: cannam@127: Include default includes when checking for uintptr_t. cannam@127: (Otherwise solaris breaks.) cannam@127: cannam@127: commit b2e7887137a70e836841860650f673a32d8fd0e0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 12:39:52 2003 -0500 cannam@127: cannam@127: distribute check.pl cannam@127: cannam@127: commit f523570817e6d4e02d1229eb4fae65aa54b39c90 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 12:38:34 2003 -0500 cannam@127: cannam@127: Check split format, too. cannam@127: cannam@127: commit 4cf6b31bc1d606a85ebe86b81538440c32ba16d4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 25 11:48:19 2003 -0500 cannam@127: cannam@127: New tests, added make check cannam@127: cannam@127: commit cc595c7702af171d1850e32593ad093a1884fa98 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 23 08:34:24 2003 -0500 cannam@127: cannam@127: More tests cannam@127: cannam@127: commit 132d24bf7371a5738a8703d6700452432c1ff8d6 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 21 20:32:12 2003 -0500 cannam@127: cannam@127: Deal with rnk(sz)=-infinity cannam@127: cannam@127: commit dbf5eba2cfe458f7fa1853b8b73ac880f50268d3 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 21 10:07:16 2003 -0500 cannam@127: cannam@127: Crazy idea cannam@127: cannam@127: commit 538d043b618e83f8c9dd443618e30fad09412560 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 21 07:14:22 2003 -0500 cannam@127: cannam@127: Test program, still barely worthy of the name. cannam@127: cannam@127: commit bd13e47fca93beafd8c87bd039e4c7f6f9843cc0 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 20 08:29:21 2003 -0500 cannam@127: cannam@127: Stylistic changes cannam@127: cannam@127: commit 7a7f938bfa0596d8a971476e304a584e80c9af3e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 20 07:03:38 2003 -0500 cannam@127: cannam@127: Implemented flops api cannam@127: cannam@127: commit 06f9de2ae48e8508332300af57ce4a892d5d7327 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 19 14:27:21 2003 -0500 cannam@127: cannam@127: cleanup cannam@127: cannam@127: commit 0004f3c1d04b2dbf2cd0c329464a761f513b17e8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 19 14:14:49 2003 -0500 cannam@127: cannam@127: 'v' syntax now defaults to an 'internal' (stride 1) vector, which is a more interesting case and corresponds more closely to the intuitive notion of a 'vector' transform, while '*' does the old 'external' (stride n) vector cannam@127: cannam@127: commit 3ae6aeb8ad69f728e24a22eaff8cb1c2d769dbfd cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 19 13:55:35 2003 -0500 cannam@127: cannam@127: removed '/' overloading cannam@127: cannam@127: commit 16e5b7c653597353fa972d5da6226e3d1c21f09c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 19 13:52:09 2003 -0500 cannam@127: cannam@127: get rid of '*' and ',' synonyms for 'x' in problem parser; there's no need to clutter the namespace with syntax we never use cannam@127: cannam@127: commit fe570b1a3ef49b842a35c74088e1893023c924a3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 19 07:28:27 2003 -0500 cannam@127: cannam@127: Signed/unsigned fixes. cannam@127: cannam@127: commit 39087e0b7d51d64ce70403c94042723a27ebd90e cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 19 07:09:54 2003 -0500 cannam@127: cannam@127: Test split arrays. cannam@127: cannam@127: commit f5c448ba8c68ad9343a147b9ee0edddd48101248 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 23:46:57 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit ce827c93bd4b66fca1e4c6925c9638fc061f2a9c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 21:53:18 2003 -0500 cannam@127: cannam@127: caps cannam@127: cannam@127: commit 8e0bc243bfa4a19e901e09af2175220823a29fda cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 21:52:51 2003 -0500 cannam@127: cannam@127: brackets cannam@127: cannam@127: commit 205193db72e85418b6db84064c2d0c417d3622ae cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 21:52:32 2003 -0500 cannam@127: cannam@127: quote cannam@127: cannam@127: commit 53b6dc0784f2f573114f99a64e3c3a3f5c25d144 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:53:11 2003 -0500 cannam@127: cannam@127: referencing cannam@127: cannam@127: commit dc903b262a3cdbfacda95f8cacf08a79b26a3725 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:33:28 2003 -0500 cannam@127: cannam@127: fix cannam@127: cannam@127: commit 34867e8b93f1464aeb74afe7a57e6db29a6bf6ef cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:31:41 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit de2f4e199030747045d6b15f10f81015e6fa77c9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 20:31:22 2003 -0500 cannam@127: cannam@127: Print errors when --verify. cannam@127: cannam@127: commit a241dce3b13972ae124686d2a73d6845172dca10 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:30:27 2003 -0500 cannam@127: cannam@127: improved description, noted that FFTW_ESTIMATE does not destroy arrays cannam@127: cannam@127: commit de9ad7da59f6f405cb9698340a708c51879074fa cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:23:12 2003 -0500 cannam@127: cannam@127: FFTW_DEFAULTS isn't really needed cannam@127: cannam@127: commit 742ec9578cb87f7e8640c998b6455f0c1347cbad cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:21:09 2003 -0500 cannam@127: cannam@127: added FFTW_MEASURE synonym for FFTW_DEFAULTS cannam@127: cannam@127: commit 7e4c0117633ecc6c774e5747fb88e5d9b901ade1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:18:29 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit 1f5d8e6883c07b8b55b3ccdd76728dba0db83b51 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 20:16:08 2003 -0500 cannam@127: cannam@127: Clearer name cannam@127: cannam@127: commit 72f6ff219f76d5836c974d7739c9deb1fdaae1b1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 20:13:14 2003 -0500 cannam@127: cannam@127: Completed dft api test cannam@127: cannam@127: commit d98d355d8025c3244f40cb21d3c13fd49b95bb31 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:07:33 2003 -0500 cannam@127: cannam@127: index cannam@127: cannam@127: commit 5abf9be2a1e971d3911958c2f2b0f830c1e94507 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:05:50 2003 -0500 cannam@127: cannam@127: fix cannam@127: cannam@127: commit 6ce8d648e9020903839bb75540e2c7f31c350a77 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:04:11 2003 -0500 cannam@127: cannam@127: parallel structure cannam@127: cannam@127: commit bf5e342ad8e4ec2778b7cb07e9fe04c3b88fef1c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:03:18 2003 -0500 cannam@127: cannam@127: fix cannam@127: cannam@127: commit 27f73ffc519eca88af9d51fccafa9d0a9eaec3d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 20:00:24 2003 -0500 cannam@127: cannam@127: joke cannam@127: cannam@127: commit fc0561411a690340303ab579fe66b5b919e94706 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 19:59:28 2003 -0500 cannam@127: cannam@127: recommendation to read tutorial in-order cannam@127: cannam@127: commit 23f008eece1d5af1cc9aff5c21f16c4b78626a43 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 19:54:55 2003 -0500 cannam@127: cannam@127: expanded outline cannam@127: cannam@127: commit dd05ed963f2b3b6248a90c8a28ec92ed0748447f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 19:35:52 2003 -0500 cannam@127: cannam@127: clarification cannam@127: cannam@127: commit 45f4203a263004153eb30c5e2b6d5fbf7d363ebe cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 19:17:27 2003 -0500 cannam@127: cannam@127: draft complex-dft tutorial cannam@127: cannam@127: commit a1cf23e6204a958c2adb5fa5ad6908ed9ae8d5aa cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 17:27:15 2003 -0500 cannam@127: cannam@127: Paranoid mode is back. Fixed dwim to do what I mean. cannam@127: cannam@127: commit 02a981bcb005fd082e832f912a0d6970469af2db cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 18 17:13:51 2003 -0500 cannam@127: cannam@127: started tut. cannam@127: cannam@127: commit db27392f9335988028063634e188cd6e0329b2db cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 16:13:15 2003 -0500 cannam@127: cannam@127: Great renaming, so that we can include both bench-user.h and cannam@127: ifftw.h to implement the paranoid-mode hook. cannam@127: cannam@127: commit 272ce9998c6c2ba1440c85c89adf525029c3713c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 15:41:18 2003 -0500 cannam@127: cannam@127: Trying to tweak the verifier so that I can use it in cannam@127: bench.c for paranoid mode cannam@127: cannam@127: commit 9406410c744c3d040dcf53cab0033e6289315e0d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 10:24:05 2003 -0500 cannam@127: cannam@127: Added stride_factor for complex arrays. cannam@127: cannam@127: commit be5440925e131346debad7cb5c52ec9ccca20838 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 10:02:11 2003 -0500 cannam@127: cannam@127: can_do now calls the planner. cannam@127: cannam@127: commit 1c2aa801bd04200c319430596f26e33c57ade5b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 09:59:24 2003 -0500 cannam@127: cannam@127: Call guru api in bench.c cannam@127: cannam@127: commit 5ccc685036846da380536544c08668012a62953a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 08:17:23 2003 -0500 cannam@127: cannam@127: Fixed prototype. cannam@127: cannam@127: commit 6a0efba859963432de8d7ddef8a68615fba215df cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 08:14:48 2003 -0500 cannam@127: cannam@127: Attempt to make the signed/unsigned use of flags consistent. cannam@127: cannam@127: commit 6c6caca90a9df0f2f76cae61abf4d5b4108e5a16 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 08:03:07 2003 -0500 cannam@127: cannam@127: Implemented useropt. cannam@127: cannam@127: commit 7165449ca5470fe7104141090f15d804f8fa3d58 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 08:02:05 2003 -0500 cannam@127: cannam@127: The first map_flags pass must be transitive, i.e., always use the cannam@127: latest flags value as opposed to the original value. (I think.) cannam@127: cannam@127: commit b5ff8655a95e88173c98942113dedb0b8f293154 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 18 07:20:19 2003 -0500 cannam@127: cannam@127: Started working on verifier cannam@127: cannam@127: commit 6c1864f54390a4ba6483dd6f6af716030275af40 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 17 14:53:28 2003 -0500 cannam@127: cannam@127: added X(threads_cleanup) cannam@127: cannam@127: commit 53ccbeeb98ce85aeee2dfdc73a79518dd428cabd cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 17 10:35:56 2003 -0500 cannam@127: cannam@127: Use C style for upper and lower array bounds. Free tensors properly. cannam@127: cannam@127: commit 8f979d12529a8cdcbc19773db64b203d396667f3 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 17 08:50:42 2003 -0500 cannam@127: cannam@127: Fixed ambiguous syntax cannam@127: cannam@127: commit b4a79fbfcd90a04148e114c3dc9ffeec57475b91 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 17 08:20:57 2003 -0500 cannam@127: cannam@127: Parse minus sign, bugfixes cannam@127: cannam@127: commit 74b7faa7902bc94ba6a2cb2229b29a0ae7fc1ae6 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 17 08:11:56 2003 -0500 cannam@127: cannam@127: Skeleton libbench2 implemented (probably still buggy) cannam@127: cannam@127: commit e589fb07c231478fcaac2ff1747634bf9f06ea8f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jan 17 04:23:37 2003 -0500 cannam@127: cannam@127: Formatting cannam@127: cannam@127: commit ca9524db0c137f154e83a76d36cf935f00674f2e cannam@127: Author: fftw cannam@127: Date: Fri Jan 17 03:15:24 2003 -0500 cannam@127: cannam@127: slight updates cannam@127: cannam@127: commit 9cf580eecfb6efdc94025f0016482c3b39e42d44 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 17 01:44:44 2003 -0500 cannam@127: cannam@127: eliminated obsolete uimin/uimax cannam@127: cannam@127: commit 43e7097cd5f50fec4d5cba68968062d735c70118 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 17 01:40:10 2003 -0500 cannam@127: cannam@127: threads needs to have its own library, lest all programs linking to libfftw3.so need -lpthread cannam@127: cannam@127: commit dfbd69e73262bfd32e4238660b05e9e66f2d4639 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 19:53:46 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 2270fad47a873f34165771451625eb46b32f8934 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 19:53:30 2003 -0500 cannam@127: cannam@127: better name cannam@127: cannam@127: commit 90d92f5270d46d9e8f4775937e55433d425a5706 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 19:52:36 2003 -0500 cannam@127: cannam@127: added more functions cannam@127: cannam@127: commit 3f06842ca4733e7ecabf350ae1e679d52ed7698a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 16:57:06 2003 -0500 cannam@127: cannam@127: if 'long' is big enough, use it for mulmod in preference to 'long long' cannam@127: cannam@127: commit 66e1948c825bca967b2ad7e6746242e8b23f2b00 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 14:53:41 2003 -0500 cannam@127: cannam@127: use uintptr_t for pointer alignment arithmetic cannam@127: cannam@127: commit 7eb1f83c40d65241a97769cbd182b979f54b3694 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 16 07:58:28 2003 -0500 cannam@127: cannam@127: More signed/unsigned cleanup cannam@127: cannam@127: commit 45b331a5c2824f7d0d08df9385910c66db337edf cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 16 07:57:40 2003 -0500 cannam@127: cannam@127: null function pointers are technically nonportable cannam@127: cannam@127: commit 67822e08115a08b056287208aa8db4cf8679eeb5 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 16 07:17:45 2003 -0500 cannam@127: cannam@127: Free short_options cannam@127: cannam@127: commit 477c8d3241c4d9943d025ae59f9305a0b149231b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 16 05:48:30 2003 -0500 cannam@127: cannam@127: Oops, forgot STACK_FREE cannam@127: cannam@127: commit 0e20238b7462741468c08db5854a75106766b2ef cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 16 05:40:39 2003 -0500 cannam@127: cannam@127: Do not require memalign() unless HAVE_SIMD cannam@127: cannam@127: commit e1ab6010079824a6d2eba12510455609646681fc cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 01:03:31 2003 -0500 cannam@127: cannam@127: MS VC++ _aligned_malloc cannam@127: cannam@127: commit b60bc7e076569eb05d30aea259d6d6347e6a2da0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 00:44:45 2003 -0500 cannam@127: cannam@127: added api fftw_malloc/free cannam@127: cannam@127: commit fce03e8f76d32e4642d3e3abe4ace0d60e5e14f5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 16 00:43:48 2003 -0500 cannam@127: cannam@127: silence warning cannam@127: cannam@127: commit 641795cb961dfc1336f70563c2c7ad1ed3192395 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 22:39:04 2003 -0500 cannam@127: cannam@127: send error output to stderr cannam@127: cannam@127: commit f6710096b2309498d0d21582380e4edf3f3cc75c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 15 13:20:35 2003 -0500 cannam@127: cannam@127: Pure paranoia. cannam@127: cannam@127: commit 91f5030882cd7a147a68a99634aa5e2b962998cf cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 15 06:51:34 2003 -0500 cannam@127: cannam@127: Fixed formatting that was messed up by the conversion uint->int. cannam@127: Ensure that iodims etc are kosher. cannam@127: cannam@127: commit e013a32092d6ec5aa0e9f2d9ae6c26d4b8659c6f cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 01:32:18 2003 -0500 cannam@127: cannam@127: added version stamp cannam@127: cannam@127: commit 50b479b4aac66242696e7fd98f58455325526959 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 01:28:20 2003 -0500 cannam@127: cannam@127: added warning cannam@127: cannam@127: commit a1084fccb8215cfd46c69f6b5eeb7ff22f358d82 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 01:23:25 2003 -0500 cannam@127: cannam@127: add fftw-wisdom-to-conf to BUILT_SOURCES cannam@127: cannam@127: commit 4b8e34f3dba941ca2f59b9705ee49a9f29951906 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 01:09:29 2003 -0500 cannam@127: cannam@127: added const cannam@127: cannam@127: commit e3063ad93de5985a0cea8fcc35052dfdd31d3f24 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 01:04:10 2003 -0500 cannam@127: cannam@127: added wisdom-to-conf cannam@127: cannam@127: commit f1bc153c63191407f4af84ca6641b4153481abca cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 00:23:36 2003 -0500 cannam@127: cannam@127: include type prefix in wisdom preamble cannam@127: cannam@127: commit 564b63e0eb961ab85824847dd4171323d185f2d3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 15 00:02:31 2003 -0500 cannam@127: cannam@127: updates cannam@127: cannam@127: commit eed0a2c1a6165c360b7f87ff1aa77341dc112be5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 23:59:26 2003 -0500 cannam@127: cannam@127: check the_plan before printing cannam@127: cannam@127: commit b90c45ecd325b1cbb5821b7d22b7d1003a01e11b cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 14 21:10:25 2003 -0500 cannam@127: cannam@127: Eliminated those unsigned values that would break LP64 machines. cannam@127: cannam@127: commit 2cfc97931df736f5090ba7eec7fa6d13686c6899 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 15:14:29 2003 -0500 cannam@127: cannam@127: comments cannam@127: cannam@127: commit 3b9adee3905d5c9686dd26e6af706297c57d3e6e cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 14 08:00:08 2003 -0500 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit b8ef56b0756c8db296926946f027105168ac91c9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 14 07:59:14 2003 -0500 cannam@127: cannam@127: int/uint confusion cannam@127: cannam@127: commit 4063890615e1ebdd337cd0b6b79e3d8c191f7ac7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 02:25:33 2003 -0500 cannam@127: cannam@127: updated introduction and some organization cannam@127: cannam@127: commit 23ce88399655bc3c3f102fb81927f18f964381d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 01:34:46 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit b165736884413d29ac6ea2d63b7784ebf40c8400 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 01:33:04 2003 -0500 cannam@127: cannam@127: newline cannam@127: cannam@127: commit 02a1859f44c60c5452b4d9fb3e89ecdac0d57873 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 00:23:04 2003 -0500 cannam@127: cannam@127: added win32 timer cannam@127: cannam@127: commit d0e64f8319671968827241d6923c1dcc613734ec cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 00:12:21 2003 -0500 cannam@127: cannam@127: sync with kernel/alloc.c cannam@127: cannam@127: commit 1e179069c40aafd83bbaedf588ced907c60d8f7d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 14 00:03:20 2003 -0500 cannam@127: cannam@127: handle missing F77_FUNC_ cannam@127: cannam@127: commit d1e7472bbe33eaf99e4464fea7629ea9dc2549d8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 17:42:50 2003 -0500 cannam@127: cannam@127: used fint instead of int to make Fortran integer type easier to change cannam@127: cannam@127: commit 2a5dd8f944a6ed354d8245abf6cc67de05ca7457 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 17:38:56 2003 -0500 cannam@127: cannam@127: slight abbreviation cannam@127: cannam@127: commit 1371e68a5061a7de34681052e5c7f31139752046 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 17:35:20 2003 -0500 cannam@127: cannam@127: the great lengthening, part I: int -> long in api; mv mktensor-rowmajor to api cannam@127: cannam@127: commit 1011711ec3ed3d1252ee9ea5134e8e18a9925081 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 15:23:22 2003 -0500 cannam@127: cannam@127: long types cannam@127: cannam@127: commit 2f236bb6f4b8d4b68a2799c59eed45c3fa5d9bef cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jan 13 04:20:37 2003 -0500 cannam@127: cannam@127: Renamed fftw_malloc -> MALLOC, X(free) -> X(ifree), X(free0) -> cannam@127: X(ifree0), non_fftw_malloc -> NATIVE_MALLOC cannam@127: cannam@127: commit ab8d02fc9ecab18a2639a0167616e782995592eb cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 02:37:22 2003 -0500 cannam@127: cannam@127: added beginning of Fortran interface cannam@127: cannam@127: commit f2c44ba05c22d8cab1e72c5393ba64e97fb4eb57 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 01:05:29 2003 -0500 cannam@127: cannam@127: add fortran mangling check cannam@127: cannam@127: commit 59c96c1e23c8b4b4830a0f1aa70d8715a57db138 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 00:33:28 2003 -0500 cannam@127: cannam@127: added guru r2r interface cannam@127: cannam@127: commit d1b297f4a235356f816342e21c1f69617d836a4f cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 00:23:26 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 07839004aaa3e10e2493cf14fcaf6c603703ecf6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jan 13 00:16:20 2003 -0500 cannam@127: cannam@127: added r2r planner cannam@127: cannam@127: commit f0e64dbc84dcd207fcf9ab13bf270707e878b9cb cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 22:58:18 2003 -0500 cannam@127: cannam@127: more long-double checks cannam@127: cannam@127: commit fc870a86543c6166b8f3c689278c3e6c429c8fb5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 20:01:51 2003 -0500 cannam@127: cannam@127: slight regrouping cannam@127: cannam@127: commit 889820ff1b5fe4f8ccf4c0f321cbcfc6066facc6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 19:58:46 2003 -0500 cannam@127: cannam@127: added joke cannam@127: cannam@127: commit 709357f9b9ad15dbc409491672174b0369364de5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 19:53:58 2003 -0500 cannam@127: cannam@127: simplified rdft2 padding cannam@127: cannam@127: commit b724cc6adafccd6b09b69ea2433567634d2b18fc cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 19:02:09 2003 -0500 cannam@127: cannam@127: added comment cannam@127: cannam@127: commit f7e00499811c30295febdd6d70699c0bee9c2260 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 18:54:49 2003 -0500 cannam@127: cannam@127: use latest api cannam@127: cannam@127: commit 17dfd8a88bbd556d885e59de2c75f6ed10666ee3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 18:49:58 2003 -0500 cannam@127: cannam@127: nembed should only be in advanced (many) interface, not basic interface...only a handful of people over the years have ever requested that functionality. cannam@127: cannam@127: commit de10a37b79b7222dff049d7b17a2f52c4d8818f7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 18:41:57 2003 -0500 cannam@127: cannam@127: impatient is default; generalize mapping functions using xor trick cannam@127: cannam@127: commit d759ad32aae7ec487f4bacbacd50c36e9b9252de cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 14:39:42 2003 -0500 cannam@127: cannam@127: use NULL nembed to signal padding cannam@127: cannam@127: commit c52303f271b00a8388b368b13e26e492e34e8ac1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 14:23:00 2003 -0500 cannam@127: cannam@127: accept NULL nembed cannam@127: cannam@127: commit 130e62b15d813a605dbe261661f3d4f73eefd869 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 13:57:13 2003 -0500 cannam@127: cannam@127: added execute-dft-r2c/c2r cannam@127: cannam@127: commit 1fd627fbdef40e0158d1880e765131d7316614eb cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 13:43:20 2003 -0500 cannam@127: cannam@127: don't need dft.h cannam@127: cannam@127: commit 116ca5713809a7a18bea146e4e1d2c13679f0570 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 13:22:14 2003 -0500 cannam@127: cannam@127: tensors are compressed in the problem, duh cannam@127: cannam@127: commit 2ede363d3dc04c22c6d801931c613f8acb365f20 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 12:45:26 2003 -0500 cannam@127: cannam@127: noted that posix_memalign bug is now fixed, thanks to bug report by yours truly cannam@127: cannam@127: commit aa78a752a45559c3cd10009619c38714715b4bd8 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 12 12:44:43 2003 -0500 cannam@127: cannam@127: Bug: n[3] instead of n[2]. Bug was propagated by copy-and-paste. cannam@127: Grrr... cannam@127: cannam@127: commit da61449b6d55793e890eaf9246c2cef570656949 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 12 12:41:43 2003 -0500 cannam@127: cannam@127: Express plan_dft() in terms of plan_many_dft() cannam@127: cannam@127: commit f50b2491505035d8da53cdc2f807f777ab7f2fa2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 12 12:19:53 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 91650cec6fb479345ace984c86a3d0bf8dd45fa3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jan 12 06:00:46 2003 -0500 cannam@127: cannam@127: Manual skeleton. cannam@127: cannam@127: commit 58983b0fcb5dde376eef5290f5afeda420bb3516 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 23:46:34 2003 -0500 cannam@127: cannam@127: added r2c/c2r guru api cannam@127: cannam@127: commit 7ab4791d61cf0a563110b7c4458f092ad3209452 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 23:42:10 2003 -0500 cannam@127: cannam@127: FFTW_DESTROY_INPUT is default for c2r transforms cannam@127: cannam@127: commit 702d37e3aa20ac9e2007f9415a6c09875dc58eec cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 23:36:26 2003 -0500 cannam@127: cannam@127: added more of r2c/c2r api cannam@127: cannam@127: commit 877b1c30a8de302c16d17618928ea9eeafa1d840 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 21:09:41 2003 -0500 cannam@127: cannam@127: r2c doesn't have adjustible sign cannam@127: cannam@127: commit d7e17c10e9b94495bf5b8d91ee938bddb15a778f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 21:07:55 2003 -0500 cannam@127: cannam@127: note that copyright year is out of date cannam@127: cannam@127: commit 414ef3efe34a68c1d1886ebc7bf3696c9888312d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 21:04:23 2003 -0500 cannam@127: cannam@127: updated api for r2c cannam@127: cannam@127: commit f55aa9fa016782becff68f499151eb9b1142f48c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 21:00:07 2003 -0500 cannam@127: cannam@127: removed annoying nophys == niphys case cannam@127: cannam@127: commit 5ac383f909dedb9038b26d2534d1c50f831bb622 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 20:58:13 2003 -0500 cannam@127: cannam@127: added basic r2c/c2r planner cannam@127: cannam@127: commit 67a72b6fe3950808458f9db07fd17fdb5a23f174 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 19:34:14 2003 -0500 cannam@127: cannam@127: dist should be in terms of complex values cannam@127: cannam@127: commit 38330465de7aa72398c41e080a538abe90f11525 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 19:14:24 2003 -0500 cannam@127: cannam@127: added plan-with-nthreads cannam@127: cannam@127: commit 250cd26e00612f247ec647a8b1cd12757c6bc2dd cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 19:12:51 2003 -0500 cannam@127: cannam@127: added function to set nthr cannam@127: cannam@127: commit 2a3a928928d0a1a720099f63d46f9c9335e60d07 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 18:04:57 2003 -0500 cannam@127: cannam@127: slight cleanup cannam@127: cannam@127: commit c70d4ae24f18ed8573746982ded357d7a66e45d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 17:57:29 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit f6eead982eac7fba05e3e1b211e92218fa75ac0e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 17:55:39 2003 -0500 cannam@127: cannam@127: maxlen is maximum string length, not including null termination cannam@127: cannam@127: commit 7d5ced1616625f49a8064e213c195cb0dfdfd015 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 17:50:49 2003 -0500 cannam@127: cannam@127: imprt reverts hashtable on failure cannam@127: cannam@127: commit 8b8397f47fca460f6f4d799ed4d3523dbb9febe4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 16:43:54 2003 -0500 cannam@127: cannam@127: slight move cannam@127: cannam@127: commit ef10382faf88c76dbed1b15712ea4385f5c53d60 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 16:34:56 2003 -0500 cannam@127: cannam@127: stdio.h should be inlcuded outside of extern "C" cannam@127: cannam@127: commit a12b4db5cb652f45955ba8597ac6aad5494d10f2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 16:26:35 2003 -0500 cannam@127: cannam@127: added guru planner API cannam@127: cannam@127: commit 6612a3a2112dc6db386ca1a7c9e80e0cdff8060a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 15:54:57 2003 -0500 cannam@127: cannam@127: added FFTW_FORWARD/BACKWARD cannam@127: cannam@127: commit e9182c7a6103c19c79d2d457aadfd3cbafc8e7f8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 15:52:17 2003 -0500 cannam@127: cannam@127: added plan_many_dft cannam@127: cannam@127: commit 93581dbc842eb787a6a1f514d9ae4a3af66da1ae cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 15:44:37 2003 -0500 cannam@127: cannam@127: indenting cannam@127: cannam@127: commit 4b42a448907aaef4bce3be179fe2676f89dc7580 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 14:49:08 2003 -0500 cannam@127: cannam@127: Final \n cannam@127: cannam@127: commit 3174ca24fb957b047983215e5651a9f5db6a6687 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 14:47:31 2003 -0500 cannam@127: cannam@127: Do not compile if not defined(FFTW_DEBUG), in order to avoid cannam@127: unused code in the shared library. cannam@127: cannam@127: commit e011c0ebee3524df3ebfe4c485e34247e5167ffd cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 14:45:56 2003 -0500 cannam@127: cannam@127: Implemented print_plan() cannam@127: cannam@127: commit 83d6f1227a7413bf1cee8e8fda10b15569e6391f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 13:12:01 2003 -0500 cannam@127: cannam@127: changed the OOP-like plan_destroy to the more-grammatical destroy_plan cannam@127: cannam@127: commit 7f9077eb88fc4deb6d8c1457988ad518ab450a92 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 12:58:04 2003 -0500 cannam@127: cannam@127: added guru execute_dft cannam@127: cannam@127: commit eb4083006537a4a3dc5ee3d202d1bc9c07909a3b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 12:38:40 2003 -0500 cannam@127: cannam@127: allow for malloc errors in wisdom string, since non-fftw-malloc cannam@127: cannam@127: commit a84ffa432d7e480e83d87090ed763a8ba8deefc1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 12:16:05 2003 -0500 cannam@127: cannam@127: cleanup should reset plnr to zero so that fftw can be restarted cannam@127: cannam@127: commit 5c64b4a2e584e795861b00a8c2683f1ab740c5f4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 12:13:18 2003 -0500 cannam@127: cannam@127: NO_UGLY is an internal planner flag cannam@127: cannam@127: commit f9e7b4ae52caaf33854eeab2f49cc98c0ed76431 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 11:23:13 2003 -0500 cannam@127: cannam@127: Written 1d api in terms of generic n-d api. The code is less compact cannam@127: but easier to test cannam@127: cannam@127: commit 96c701ecb75dbd0236023c61f59a0cde3f0f330d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 11:07:25 2003 -0500 cannam@127: cannam@127: Added wisdom to header file, made scanners/printer static. stdio.h cannam@127: no longer needed in fftw.h, removed. Probably the printer_file cannam@127: should be reintroduced in a separate file if we ever want to cannam@127: print plans... cannam@127: cannam@127: commit ea3e4d45a1aec83b6e5534b2f0fbdd8271601ebd cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 09:49:30 2003 -0500 cannam@127: cannam@127: Implemented more APIs cannam@127: cannam@127: commit 41044feee9e55eb6ff29128fca2ad7458087146b cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 09:21:53 2003 -0500 cannam@127: cannam@127: Added cleanup() to API cannam@127: cannam@127: commit 125c89f921354d7d4e18aa61700b2d2ce8704e5f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jan 11 09:17:34 2003 -0500 cannam@127: cannam@127: Started new bench.c. I had to rename plan_destroy -> cannam@127: plan_destroy_internal to avoid conflicts with API cannam@127: cannam@127: commit f315b29db425d56e3daffc4a5710b8e6542a0c91 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 02:45:39 2003 -0500 cannam@127: cannam@127: fix types cannam@127: cannam@127: commit 9419d5287867213b7dec8bbb7e594a3f77157be2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 02:13:25 2003 -0500 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 8a271133e0891ed171ae642860ef03dff81e4bce cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 02:10:50 2003 -0500 cannam@127: cannam@127: added wisdom api cannam@127: cannam@127: commit ce93efb2d481b23ccb261df25cb2021dd38b5668 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 01:01:17 2003 -0500 cannam@127: cannam@127: grammar cannam@127: cannam@127: commit faefac80f41e5203c91c356619f97c3c1cdf8b13 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 00:54:54 2003 -0500 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit c01969dd0125889865e81c33fff6f5a0055f71b5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 00:52:04 2003 -0500 cannam@127: cannam@127: implemented api/mapflags cannam@127: cannam@127: commit fb2e4c252410f4b6d3e26ce97ca17083ca45a773 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 11 00:48:27 2003 -0500 cannam@127: cannam@127: IMPATIENT is an api issue cannam@127: cannam@127: commit 21879988984f816cbd2ff5250d85b004a7dac217 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 10 01:57:41 2003 -0500 cannam@127: cannam@127: removed un-needed headers cannam@127: cannam@127: commit d0d8c732879727d7af40c9301c58d661c4c62343 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jan 10 01:56:59 2003 -0500 cannam@127: cannam@127: mkplanner initializes nthr to 1 already cannam@127: cannam@127: commit e0b0e74169c6b2ee44abc01b035e356b5ab1aaa8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 18:53:09 2003 -0500 cannam@127: cannam@127: boilerplate cannam@127: cannam@127: commit f90417638448166e44f56b0f7bcc61f0263c40b7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 18:16:39 2003 -0500 cannam@127: cannam@127: fold vecloop into r{e,o}dft apply function to share buffer, etcetera cannam@127: cannam@127: commit aa1101d19e86b64d4753f8bf562df5db7ea5de73 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 18:10:19 2003 -0500 cannam@127: cannam@127: whoops, bugfix in impulse test for vecn > 1 cannam@127: cannam@127: commit d70526c96f5f1959cf5fa3df3e15ff71a8e66487 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 14:23:51 2003 -0500 cannam@127: cannam@127: bugfix, grr cannam@127: cannam@127: commit 6c4923f6c3d24b14b644dea8a85adaddfb165ef5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 14:21:16 2003 -0500 cannam@127: cannam@127: fixed signed-ness enum problem cannam@127: cannam@127: commit 656713c6b00d9f53d81820fd4675fe568a204088 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 9 14:12:42 2003 -0500 cannam@127: cannam@127: Explicit cast cannam@127: cannam@127: commit 06f32cbb5be575b9880ff2b1a0e4031fa9be68d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 9 13:41:51 2003 -0500 cannam@127: cannam@127: Added configure_planner(). mkplan() behaves properly when plan is null. cannam@127: cannam@127: commit 0c9627b61142790be11d642e3348808cbfa7cd5e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 9 06:48:53 2003 -0500 cannam@127: cannam@127: More API work cannam@127: cannam@127: commit e21443ac067af4615dc8513d68d880f78801b983 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jan 9 05:40:34 2003 -0500 cannam@127: cannam@127: First skeleton of API infrastructure cannam@127: cannam@127: commit 8c1212b04af0632d827194223919a73133593c54 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 03:19:35 2003 -0500 cannam@127: cannam@127: unsigned strikes again cannam@127: cannam@127: commit d9142b307e261d5d50a1b2086eef9012e2c36602 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 01:51:45 2003 -0500 cannam@127: cannam@127: put rdft2_inplace_strides and rdft2_tensor_max_index in their own files for tighter linking cannam@127: cannam@127: commit 7c048dc37ce30e18367fc3e84ec7759a2c2f0b7e cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 01:43:13 2003 -0500 cannam@127: cannam@127: added rdft2_tensor_max_index...incorrect use of tensor_max_index was preventing proper loop ordering for rnk > 2 rdft2 cannam@127: cannam@127: commit 561ca9cb4f10d1710cea9126fc0fa63366814127 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 00:44:45 2003 -0500 cannam@127: cannam@127: arbitrary spltrnk in rdft2 rank-geq2 cannam@127: cannam@127: commit f17e0e00c61ac65b9353fc879a5e33f185bc8f36 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 00:40:17 2003 -0500 cannam@127: cannam@127: don't mention wisdom when non-verbose cannam@127: cannam@127: commit be48b68a4776f2add565a8ff0b0b0c4b8095518e cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jan 9 00:02:35 2003 -0500 cannam@127: cannam@127: bug fix: printing %T should pass tensor *, not tensor ** cannam@127: cannam@127: commit 13e8d5776b965f625f836ffb3ed0541c5ec1c3b4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 23:40:48 2003 -0500 cannam@127: cannam@127: correct(?) normalization for rodft00 ... all of the even/odd transforms should be normalized according to the expanded'' DFT of ~twice the length cannam@127: cannam@127: commit 08581922580b63f9723d7bd0da7e341d49b0225a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 23:18:23 2003 -0500 cannam@127: cannam@127: fixed tests for n=1 cannam@127: cannam@127: commit cd3f97118a39ef25691a86f62df1a53abfe3f15b cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 22:10:08 2003 -0500 cannam@127: cannam@127: fixed bug in vector tests for rdft(2) cannam@127: cannam@127: commit 81b7636d1db4f7c33fa315720b91077f0c189f22 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 20:12:00 2003 -0500 cannam@127: cannam@127: fixed handling when first rnk-1 dimensions compress to nothing (ugh) cannam@127: cannam@127: commit a1150e27e85473748f0705407bb3858272d25bca cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 20:02:35 2003 -0500 cannam@127: cannam@127: fixed incorrect/missing rdft2 rank-0 handling cannam@127: cannam@127: commit 77ab86cd9b8adef254ae54ee9f5f1355efb82b6f cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 19:49:05 2003 -0500 cannam@127: cannam@127: bug fix: for rnk > 1, must compress rnk-1 dims separately (ugh) cannam@127: cannam@127: commit cba19ba921fb1d660ea71804cd40ba3d14fac750 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 17:39:14 2003 -0500 cannam@127: cannam@127: added trailing newline cannam@127: cannam@127: commit 925276da406dd1908a70b57c584cab6719dfb44b cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 17:38:02 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 3740fe7538b1f9e0c2776a305f1c46dde0e12082 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 16:53:16 2003 -0500 cannam@127: cannam@127: got rid of compiler warning cannam@127: cannam@127: commit 2dfda812b6ddde932b9dd627cfbc2677ec4caeb0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 16:49:48 2003 -0500 cannam@127: cannam@127: whoops, test r2hc and not rodft00 by default cannam@127: cannam@127: commit 12f2eb610a61a32de3a2d961676f005a3c7bc0c9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jan 8 16:46:24 2003 -0500 cannam@127: cannam@127: got rid of real_n...use physical n everywhere in rdft; fixed rdft sz compression; fixed rodft00 verify bug cannam@127: cannam@127: commit 17233aac9a159de06ecf2dec334205094e3e0a03 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 8 07:20:47 2003 -0500 cannam@127: cannam@127: icc-6.0 bug workaround cannam@127: cannam@127: commit 8490d0c5c383dcfe1910afc3e006557fb7c9aa76 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 8 04:21:40 2003 -0500 cannam@127: cannam@127: Reclaimed the fftw_real identifier, because I need it for the API cannam@127: cannam@127: commit 82c0ab6a22809a05739960cb8c06c9d14d5e7968 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 8 04:14:55 2003 -0500 cannam@127: cannam@127: Use recommended AC_OUTPUT syntax cannam@127: cannam@127: commit 38010c2e123c85caeb3c0827f769f304b8f77c87 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jan 8 04:00:22 2003 -0500 cannam@127: cannam@127: Removed FFTW(foo) as a synonym for X(foo). This is an API issue. cannam@127: cannam@127: commit 863cf56f79b7eac7b70f307d24f431d71bfbdd52 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 7 17:45:52 2003 -0500 cannam@127: cannam@127: get rid of warning cannam@127: cannam@127: commit fbc87e15fead24d239286af63e298620ac46b30b cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 7 16:22:39 2003 -0500 cannam@127: cannam@127: Renamed conflicting files */codelet.h into dft/codelet-dft.h and cannam@127: rdft/codelet-rdft.h cannam@127: cannam@127: commit 683c665e1da6396f9b2c2dc8ecc749b90e666907 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 7 16:21:16 2003 -0500 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 13ef7881b1b28e1772271d54f92e7f7d96059c25 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 7 15:47:24 2003 -0500 cannam@127: cannam@127: Silence warnings cannam@127: cannam@127: commit f35b6c4c226aa4cba7f2a0b30b5493cc94517e13 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 7 15:00:14 2003 -0500 cannam@127: cannam@127: fftw2 used spltrnk=1 cannam@127: cannam@127: commit 9a9b9463c83f021eeefa0743fd50b9e11c008103 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 7 14:32:06 2003 -0500 cannam@127: cannam@127: Silence warning cannam@127: cannam@127: commit 86d050e48df435dd7091a75e4ee9647cc31d65e0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jan 7 12:13:50 2003 -0500 cannam@127: cannam@127: noted deficiency cannam@127: cannam@127: commit 97269b487afae721bc3efc07d4510284d184500e cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 7 07:18:51 2003 -0500 cannam@127: cannam@127: Strengthened conditions for a problem to be POSSIBLY_UNALIGNED cannam@127: cannam@127: commit d135e51da8af9610080ca861eec8a12f04e33617 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jan 7 05:09:42 2003 -0500 cannam@127: cannam@127: Strengthened conditions for a plan to be POSSIBLY_UNALIGNED cannam@127: cannam@127: commit 41d4363cc830c074d8e602a4046fcfb361714aa3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 5 02:43:45 2003 -0500 cannam@127: cannam@127: added copyright todo cannam@127: cannam@127: commit 81f531aeaa2fb148c7f8b5519a792c7e226060dd cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 5 02:37:31 2003 -0500 cannam@127: cannam@127: modified comment cannam@127: cannam@127: commit e17581aca74e377a94b5506199a6f3c0d95dd218 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 5 02:34:36 2003 -0500 cannam@127: cannam@127: fixed comment cannam@127: cannam@127: commit f33e50cd3dedd8472c0b37116e337749dd80efa8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jan 5 02:31:56 2003 -0500 cannam@127: cannam@127: implemented rdft2 verify cannam@127: cannam@127: commit 352eadf383e28c25c7132ace3c4179e561c54aa8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jan 4 16:20:42 2003 -0500 cannam@127: cannam@127: fix --enable-single cannam@127: cannam@127: commit 3cd824b965de4c51a977683e83bfaa1f2d8b37ab cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 23 12:59:12 2002 -0400 cannam@127: cannam@127: slight fixes cannam@127: cannam@127: commit 64f0f3180cf46058053d0a452152f3fb7e4d5363 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Oct 23 12:42:39 2002 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit b6cffe0e74206ccd7ae7726181a361bea4d94986 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Oct 1 09:32:56 2002 -0400 cannam@127: cannam@127: Experimental stuff cannam@127: cannam@127: commit b92e96518b5b9ac3275a6f7194d5e1ec49b36e7d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 28 13:03:53 2002 -0400 cannam@127: cannam@127: Experimental Franz mode cannam@127: cannam@127: commit fd2ac8fb21fc75eccec5c5352069388b52ab00ea cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 26 15:14:38 2002 -0400 cannam@127: cannam@127: const-correct cannam@127: cannam@127: commit ec5733489ef85cbe78e5253358fdb320be5b2642 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 26 15:06:38 2002 -0400 cannam@127: cannam@127: Reuse dimcmp routine for other purposes cannam@127: cannam@127: commit 6fa12bfc6f9ac208da72478981473011a292f57d cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 25 07:37:38 2002 -0400 cannam@127: cannam@127: Use tornk1 correctly. cannam@127: cannam@127: commit 97b84fbe4c90ade6b9cad6ac2efba9b6fb305412 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 25 07:36:38 2002 -0400 cannam@127: cannam@127: Hmm... I thought I had fixed this before... cannam@127: cannam@127: commit 69de6d4b5d66e405c267001886d8a7ae9e84224b cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 24 21:27:49 2002 -0400 cannam@127: cannam@127: Collect more common idioms cannam@127: cannam@127: commit 1bbba9625dca12e70a6e26402ba1a2262b7ca984 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 24 21:15:57 2002 -0400 cannam@127: cannam@127: Still collecting common idioms... cannam@127: cannam@127: commit 01a7139392f8170c8563510d0c489bfd91687520 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 24 21:13:00 2002 -0400 cannam@127: cannam@127: More garbage collection. cannam@127: cannam@127: commit 45bb1a6c49ce1569ebc75896da0ed42b0b03ee59 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 24 21:08:19 2002 -0400 cannam@127: cannam@127: More compact code cannam@127: cannam@127: commit eed5c4ed8045a26be389b99e1492aedc5017f448 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 24 20:54:43 2002 -0400 cannam@127: cannam@127: Collect common pattern if (foo) free(foo) ==> free0(foo) cannam@127: cannam@127: commit e7d2657d2d3bb77eb2403856e102678d865de742 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 24 20:08:44 2002 -0400 cannam@127: cannam@127: Collect some common code in */buffered*.c cannam@127: cannam@127: commit 61cd95889228f7a100d853c42e461780fd01dd92 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 24 19:39:22 2002 -0400 cannam@127: cannam@127: use STRUCT_HACK #define to determing rdft kind[] allocation cannam@127: cannam@127: commit 337af322b345f45b275182f7bc8f5949794ea140 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 24 17:21:09 2002 -0400 cannam@127: cannam@127: report total pcost of measured/estimated plans...epcost is especially useful to estimate the effects of various impatience flags on planning time for large transforms cannam@127: cannam@127: commit 5cbf8b44eabe724a226d58fbeac341b7f3c13e49 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 23 18:49:10 2002 -0400 cannam@127: cannam@127: Prevent unwanted inlining cannam@127: cannam@127: commit 7342f004be53b759052eaf9a01a9a574dc64631f cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 23 18:37:59 2002 -0400 cannam@127: cannam@127: Space compaction cannam@127: cannam@127: commit d8299eef074631210e64b01453a7602dad45d6b8 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 23 11:49:32 2002 -0400 cannam@127: cannam@127: Still reducing size cannam@127: cannam@127: commit 5df9269dc8d95153c138fd44e41effd6ed1f58e2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 16:03:30 2002 -0400 cannam@127: cannam@127: Saved another 5KB by redesigning opcnt protocol. (gasp!) cannam@127: cannam@127: commit 074344d84ab955d0ad7efdc9b58f8414952a0372 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 15:00:59 2002 -0400 cannam@127: cannam@127: More code compression cannam@127: cannam@127: commit 7e2e90935398c3d3d50cc2bbcab66d4b188bf757 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 13:27:46 2002 -0400 cannam@127: cannam@127: Smaller code size. cannam@127: cannam@127: commit 1da75a085efd3348694dafb0905fb59e2c6cee27 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 12:50:36 2002 -0400 cannam@127: cannam@127: Started unification of rader cannam@127: cannam@127: commit e0cb464fbac3602192afd97211885e814674d246 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 12:35:30 2002 -0400 cannam@127: cannam@127: Typo cannam@127: cannam@127: commit 363f9b3b1a6bbf78e371c46a74645b55281ec0ca cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 12:25:20 2002 -0400 cannam@127: cannam@127: Changed protocol for destroy_plan so as to save space. cannam@127: cannam@127: commit daf930d4450cc9caa5d528b631f964bfbf16a208 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 11:08:57 2002 -0400 cannam@127: cannam@127: Introduced convenient function X(mkplan_d) cannam@127: cannam@127: commit e74d86afcd19e77f275c86c916449ae2b82310be cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 10:21:36 2002 -0400 cannam@127: cannam@127: Split tensor/md5 into separate files to allow independent linking cannam@127: and/or prevent undesidred inlining cannam@127: cannam@127: commit 249329f66447c68d67536d4a868ac589b264a9ff cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 22 09:49:09 2002 -0400 cannam@127: cannam@127: Treat all tensors as dynamically allocated objects. They were cannam@127: dynamically allocated in part anyway, so there is no point in cannam@127: complicating the object code with the clumsy calling conventions cannam@127: for by-value structs. cannam@127: cannam@127: commit 53cf5c7cab96e0657153327e660e787279e77c4f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 21 18:24:55 2002 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit e36da5f9b63af8a62dab370b005e2472e5edc33f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 18:10:07 2002 -0400 cannam@127: cannam@127: Avoid generating NaN when n = 0. cannam@127: cannam@127: commit a49b921ea278fcb353b2be6338d04daf3b0a72dc cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 18:04:05 2002 -0400 cannam@127: cannam@127: Saved more. cannam@127: cannam@127: commit 2008afba6889d6f2b9d3f00dcbcf0bc9edd8c7c2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 17:47:36 2002 -0400 cannam@127: cannam@127: Save 1200 bytes of object code. Do not pass structs by value whenever cannam@127: practical, because the calling protocol generates clumsy code. cannam@127: cannam@127: commit 8dbaef7c3531ccca29ae4f52528ed11c5089700d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 12:10:21 2002 -0400 cannam@127: cannam@127: Do not allocate buffers for rader omegas. Let the planner do it cannam@127: if necessary. cannam@127: cannam@127: commit 0cd3107a7fe058042f7e23b73658bacf82d08805 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 12:03:46 2002 -0400 cannam@127: cannam@127: Check rank *before* reading kind[0], which may be undefined if rnk < 1 cannam@127: cannam@127: commit ffab113d0748937a80de8e046d5d971a7cfde97f cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 11:48:50 2002 -0400 cannam@127: cannam@127: Second step towards rader unification. cannam@127: cannam@127: commit 054daf75a708d4b060c35b13a48ee8e8b1732cc1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 11:37:06 2002 -0400 cannam@127: cannam@127: First step towards unification of Rader code cannam@127: cannam@127: commit fc97f7d9567238bd1930e63614352160ff2bc202 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 21 07:58:11 2002 -0400 cannam@127: cannam@127: Fix ugliness condition for cooley-tukey. cannam@127: cannam@127: commit 28fe4962b2e634dc302c3fba3853b87788b411ad cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 20 16:53:45 2002 -0400 cannam@127: cannam@127: Removed RADER_MIN_GOOD and associated machinery cannam@127: cannam@127: commit dc40093700e7a00e3808b606108137c7ce5cb592 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 20 14:49:12 2002 -0400 cannam@127: cannam@127: Proper cast cannam@127: cannam@127: commit 2eec2b720ae866f16db023e3815f27875f572a56 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 20 14:45:54 2002 -0400 cannam@127: cannam@127: Typo cannam@127: cannam@127: commit d55f46a0acf7e75a5c216964aa0016166254876f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 20 14:38:13 2002 -0400 cannam@127: cannam@127: Implemented NO_LARGE_GENERIC cannam@127: cannam@127: commit 535ecb44b8e4450306cf760afb294431e5595ae6 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 19 07:48:25 2002 -0400 cannam@127: cannam@127: Consistent macroization of NO_DHT_R2HC cannam@127: cannam@127: commit fe02be9d79515c92b53d929977c270b46a8b7fdd cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 18 21:47:17 2002 -0400 cannam@127: cannam@127: NO_DHT_R2HC is a planner flag, otherwise the EXHAUSTIVE planner loops. cannam@127: cannam@127: commit 20e70850bb3d2cd4590c9bfce7777b8a2f9a80fc cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 18 20:47:31 2002 -0400 cannam@127: cannam@127: Resurrected NO_EXHAUSTIVE cannam@127: cannam@127: commit 4e477d8e68603cc899c8d0104fc6897817fd74d9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Sep 18 19:31:57 2002 -0400 cannam@127: cannam@127: au revoir, score() cannam@127: cannam@127: commit c3f01031fa05a9088d18e643a9b3476fa6a6437d cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Sep 18 19:31:05 2002 -0400 cannam@127: cannam@127: eliminated unused cannam@127: cannam@127: commit 25e32538394211412f3aac06baa6677ae148ea03 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Sep 18 18:28:44 2002 -0400 cannam@127: cannam@127: capitalize and parenthesize SUBSUMES cannam@127: cannam@127: commit 7115ad27ce3a4390e6c81800126315f757abbdb5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Sep 18 18:26:58 2002 -0400 cannam@127: cannam@127: comment cannam@127: cannam@127: commit 3ec48dd0fd8e9cc88fd85a1b7b74f9ec5ef1789d cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 18 18:03:18 2002 -0400 cannam@127: cannam@127: Use flags from wisdom if wisdom is applicable. cannam@127: cannam@127: commit e16b332f900b1872044fe195f7e40ae15e5ed5e6 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 18 17:16:17 2002 -0400 cannam@127: cannam@127: Removed score() machinery cannam@127: cannam@127: commit bc4041b9adab2d69de986123e38bee24f480eb3a cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 18 14:12:21 2002 -0400 cannam@127: cannam@127: Revised planner hack cannam@127: cannam@127: commit 4f3717ebf2eca24ac5e8017eaf8856bf5270020c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Sep 18 10:14:41 2002 -0400 cannam@127: cannam@127: Fix warning cannam@127: cannam@127: commit b627b00f2f1adf8c8839b27618ac1765064c0b78 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 17:54:07 2002 -0400 cannam@127: cannam@127: Type qualifiers. cannam@127: cannam@127: commit 48fc716d339ceb08432ab2a6704e79de578ad5fc cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 16:17:55 2002 -0400 cannam@127: cannam@127: ESTIMATE is no longer subsumed by everything else. cannam@127: cannam@127: commit 1c6447f56fe864b172d70f8940dc9de74a15a499 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 10:55:15 2002 -0400 cannam@127: cannam@127: NO_BUFFERING is a planner flag, not a problem flag cannam@127: cannam@127: commit 458afba08480c2115e585e898153352ca125ff39 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 09:36:16 2002 -0400 cannam@127: cannam@127: Maintain flags in canonical form. cannam@127: cannam@127: commit 1a01c050c68d0ffa9380d7bd780194509c75a31e cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 09:09:57 2002 -0400 cannam@127: cannam@127: In dramatic break with tradition, SUBSUME is now a partial order. I cannam@127: swear. cannam@127: cannam@127: commit 7c1f9aafa59ca2c68f98f1b2f88ca8b029506e09 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 07:29:00 2002 -0400 cannam@127: cannam@127: Added comment cannam@127: cannam@127: commit 8b8f6515fa6ddcc9ac579c80062a9b9aa55917c1 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 17 07:27:17 2002 -0400 cannam@127: cannam@127: Inverted ESTIMATE flag, renamed USE_SCORE for consistency with the cannam@127: convention that 0 subsumes 1. cannam@127: cannam@127: commit 1fd38e50ba038d57947daa7c999bab9da4a33836 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 17 02:50:15 2002 -0400 cannam@127: cannam@127: NO_INDIRECT -> NO_INDIRECT_OP (out-of-place only) cannam@127: cannam@127: commit ff2617c02989df82c99fd064ec298e107afe627e cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Sep 17 00:40:04 2002 -0400 cannam@127: cannam@127: hpux needs -D_REENTRANT (thanks to Clinton Roy for the bug report) cannam@127: cannam@127: commit ef127fa967046516cf5658be72eb70c93b817120 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 23:54:34 2002 -0400 cannam@127: cannam@127: Oops. cannam@127: cannam@127: commit 70546cbffe33ede1657b54f626e133039ba26528 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 23:44:47 2002 -0400 cannam@127: cannam@127: Yet another attempt at getting the planner right. cannam@127: cannam@127: commit ac2a09b8c81db49fcc9c770b94723577beee286c cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 21:56:14 2002 -0400 cannam@127: cannam@127: Better coding. cannam@127: cannam@127: commit a0a3d5520d53b44194f63fe2873207b57a07d544 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 21:51:06 2002 -0400 cannam@127: cannam@127: NO_UGLY is no longer a flag, but a separate planner field that does not cannam@127: interfere with wisdom. cannam@127: cannam@127: commit 69253431765ca3b9cfce2c4a56c846512f6c9968 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 19:04:41 2002 -0400 cannam@127: cannam@127: Did not compile without FFTW_DEBUG cannam@127: cannam@127: commit 43a0347e3daa0fd854eec1d7ded5c6f45ce727e4 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 18:37:06 2002 -0400 cannam@127: cannam@127: Changed scoring mechanism. cannam@127: cannam@127: commit e4f00711d6784b6f4196859738a039ae1f7b9edd cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 17:13:45 2002 -0400 cannam@127: cannam@127: Count infeasible plans cannam@127: cannam@127: commit 18299388527442d1dfc7b7a5748da0d987c019bc cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 16 16:36:12 2002 -0400 cannam@127: cannam@127: curse subsumed plans before export cannam@127: cannam@127: commit fb22a4fc3b6fa45b0f46f605ed1c94eba5960359 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 15:40:46 2002 -0400 cannam@127: cannam@127: removed ESTIMATE_BIT vs. ESTIMATE... ESTIMATE | IMPATIENT is a UI issue cannam@127: cannam@127: commit 548808e1fefe66c9b882d332d70488986e3b073d cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 15:31:39 2002 -0400 cannam@127: cannam@127: cleanup cannam@127: cannam@127: commit 14a42333af6152472262413b8d1a97207a1aff59 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 15:28:47 2002 -0400 cannam@127: cannam@127: use CONSERVE_MEMORY flag to prevent buffered for large sizes cannam@127: cannam@127: commit 688cb6fee87d6ba5ed14e9e9899ba46c96eaddbb cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 15:16:16 2002 -0400 cannam@127: cannam@127: moved NO_DHT_R2HC back into planner flags: there's no reason we would want this flag to block plan reuse cannam@127: cannam@127: commit b06ee447ad1ee0f95af06c2d91092db1475f44a5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 14:59:14 2002 -0400 cannam@127: cannam@127: whoops, commas cannam@127: cannam@127: commit 396a6523178fa8aa79f3b716e6a14577bb83c337 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 14:58:26 2002 -0400 cannam@127: cannam@127: problem_flags == checked in applicable, planner_flags == checked in score cannam@127: cannam@127: commit b7ef5ad344bcd298e14a30b30bd2d6f2b3c7442f cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 14:53:16 2002 -0400 cannam@127: cannam@127: ESTIMATE should not *include* all impatience flags, even if it subsumes them; some impatience flags, like NO_INDIRECT, might make a problem unsolvable cannam@127: cannam@127: commit 81a60e6002c427a15cbb298654f954c09954c9a4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 16 00:56:29 2002 -0400 cannam@127: cannam@127: quotatio marks cannam@127: cannam@127: commit 0833118f7818c740e7387c607c320e79e088c6be cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 23:55:44 2002 -0400 cannam@127: cannam@127: delete blank line cannam@127: cannam@127: commit 4cbe17440ce2d074a4c0a0d3245d25c63dd469fb cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 23:51:14 2002 -0400 cannam@127: cannam@127: substitution cannam@127: cannam@127: commit 3963051622d435d96083c0d753dcd8f503bac2f5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 23:49:50 2002 -0400 cannam@127: cannam@127: note that we are not GNUlly correct cannam@127: cannam@127: commit ec9b8c84419f5dd8cd533eca7b07391696019046 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 23:41:01 2002 -0400 cannam@127: cannam@127: indenting cannam@127: cannam@127: commit bb5f5581a5d05566bf679da7ed67a2e59e68781e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 23:37:46 2002 -0400 cannam@127: cannam@127: more jokes cannam@127: cannam@127: commit 0db38cc3a3f7215cdd2e9c308fa9d88c0422024e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 23:20:14 2002 -0400 cannam@127: cannam@127: NONTHREADED_ICKYP includes nthr > 1 check cannam@127: cannam@127: commit a1900e4f7fdc5bc663fe60ec30d99f342ac06d34 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 22:56:44 2002 -0400 cannam@127: cannam@127: use md5sig cannam@127: cannam@127: commit 7f2631f48f2874827ca50e2c9ee4d59ddf861ba3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 22:55:41 2002 -0400 cannam@127: cannam@127: md5sig typedef cannam@127: cannam@127: commit c83d9aa4a856e2b34011a4285df02dc43937d982 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 22:35:13 2002 -0400 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 849fd22c4bdead7cab04a20c1b63966946b2355a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Sep 15 22:30:26 2002 -0400 cannam@127: cannam@127: partially-ordered impatience cannam@127: cannam@127: commit f811a39af185c82590b34fd1439901b8cbf32d03 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 14 19:47:56 2002 -0400 cannam@127: cannam@127: Removed all that planner inheritance crap. cannam@127: cannam@127: commit 74cf5ca97fc18b5d64c869c64575c0095f8b81c2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Sep 14 16:35:28 2002 -0400 cannam@127: cannam@127: string.h is used for more than strlen cannam@127: cannam@127: commit 7f974585ea055241b1339d303ffdb472305c7d75 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 14 12:19:13 2002 -0400 cannam@127: cannam@127: Reduced hashtable size by 1/6 (on 32-bit machines) at the expense cannam@127: of messier planner. cannam@127: cannam@127: commit 8b1efa0ba0e2a490fc04c66900ad41248a55c86c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 14 08:31:29 2002 -0400 cannam@127: cannam@127: Only print wisdom if verbose > 3 cannam@127: cannam@127: commit abd7a17545150645bb864c140559ba794257a897 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Sep 14 07:56:56 2002 -0400 cannam@127: cannam@127: Changed syntax of temporaries to avoid shadowing library functions cannam@127: (which is harmless but I hate the warning) cannam@127: cannam@127: commit a120b53fd271fad4f9b879ff6247840764061813 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Sep 13 23:07:39 2002 -0400 cannam@127: cannam@127: only add warnings in debug/maintainer mode, and add a few more warning flags; eliminate more warnings; add support for posix_memalign (broken in glibc, grrr) cannam@127: cannam@127: commit 7832eabf884004c42c3e4089fe637e205f47732e cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 21:57:50 2002 -0400 cannam@127: cannam@127: Explicit cast cannam@127: cannam@127: commit d5127e37db1b557049933fe9aff91d9c3b0a1dc0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 21:54:50 2002 -0400 cannam@127: cannam@127: Use double-hashing. This allows a slightly higher load factor cannam@127: at the expense of a messier computation of the hashtable size. cannam@127: cannam@127: commit e689e22a6e1b981c379989a760186035fa18939c cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Sep 13 17:53:13 2002 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit 1d2a159b2078ef8c6063fad80d7358fd30bd4f9d cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 15:36:07 2002 -0400 cannam@127: cannam@127: Slight change in hash table growth functions. cannam@127: cannam@127: commit 9e1d9f0454bc70a807bcdb0f9ff25ed18a7c9903 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 14:58:22 2002 -0400 cannam@127: cannam@127: More statistics. cannam@127: cannam@127: commit e1049bcf9446871a4ed34cebaae6b5b542dbca53 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 10:13:02 2002 -0400 cannam@127: cannam@127: Clearer logic. cannam@127: cannam@127: commit 4dda68614273939d7843da329ab6c8b4da2e7bb6 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 10:11:10 2002 -0400 cannam@127: cannam@127: Oops. cannam@127: cannam@127: commit 223f36a95fc7bd42e3b2d4bac8ad506cc2e4c8d9 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 09:31:46 2002 -0400 cannam@127: cannam@127: Cleaned up cannam@127: cannam@127: commit c78314bd598dfb1c7e54f18fc2d050240910de4c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 09:16:07 2002 -0400 cannam@127: cannam@127: Deal properly with infeasible problems. cannam@127: cannam@127: commit 875f159755baff40c2dd02f462c6bc36c49fbc59 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Sep 13 07:15:06 2002 -0400 cannam@127: cannam@127: Redundantly initialize hash table to prevent valgrind warnings. cannam@127: cannam@127: commit b7047a11d0b3e41e53e1890f73135be6f2ce3b2c cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 19:00:22 2002 -0400 cannam@127: cannam@127: Removed relics from past. cannam@127: cannam@127: commit 4cae827eaf456e69f3155183afd52e4c0216c980 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 18:53:44 2002 -0400 cannam@127: cannam@127: md5hash a problem only once. cannam@127: cannam@127: commit a1ef1699bff263e0141cd43801c7a4ff3431389e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 16:33:49 2002 -0400 cannam@127: cannam@127: Renamed k7 codelets cannam@127: cannam@127: commit c4367d998eeed60d4618b8a5b54162d500b84271 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Sep 12 16:32:03 2002 -0400 cannam@127: cannam@127: FORBID_DHT_R2HC -> DHT_R2HC_VERBOTEN for consistency cannam@127: cannam@127: commit 39a9858e45ca542695b9419c09ec6b61b09a6004 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Sep 12 16:28:43 2002 -0400 cannam@127: cannam@127: removed obsolete macro cannam@127: cannam@127: commit bd1d1de9ba94e111921e911a49de82fe4ff2d16e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 16:20:39 2002 -0400 cannam@127: cannam@127: Split flags in SIMD code. cannam@127: cannam@127: commit b9fbfffc3ee26e56c4c16448ced8db523670de55 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 16:18:51 2002 -0400 cannam@127: cannam@127: Forgot to fix threads cannam@127: cannam@127: commit 1d3447ab63c27d4bd97beb41882ca34addd0df18 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 16:10:05 2002 -0400 cannam@127: cannam@127: Split flags into planner_flags and problem_flags cannam@127: cannam@127: commit 075ff4047c6c5a98bd268a3bae692df6a9ec1d84 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Sep 12 15:46:56 2002 -0400 cannam@127: cannam@127: tetrameter cannam@127: cannam@127: commit ff7f0235dd4460a5e3c332725151037107514954 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 15:11:21 2002 -0400 cannam@127: cannam@127: Overwrite less impatient solutions properly. cannam@127: cannam@127: commit b470f419700398ec14357990abeb69aa6eb4d857 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 11:29:16 2002 -0400 cannam@127: cannam@127: Oops. cannam@127: cannam@127: commit 1ae2a65b2895b51b43f316fa11fcc932ed127ae2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 10:58:56 2002 -0400 cannam@127: cannam@127: Keep less impatient solution in case of conflict. Paranoid cannam@127: cast to uint in certain places. cannam@127: cannam@127: commit 640b1f4df72028daa4293c304e53af0da7f31c36 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 10:02:51 2002 -0400 cannam@127: cannam@127: Complete reimplementation of planner hash table. cannam@127: cannam@127: commit 40f47f4111154bd1e17e44f87908228ede39af18 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Sep 12 07:58:45 2002 -0400 cannam@127: cannam@127: planner->cnt was not properly decremented. cannam@127: cannam@127: commit fda67f21284b158043d7ba171a81a933a3891e1a cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Sep 11 17:52:39 2002 -0400 cannam@127: cannam@127: typo cannam@127: cannam@127: commit ebe84b30659823364a95bfd646512b387bef4629 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 9 17:10:45 2002 -0400 cannam@127: cannam@127: Simplified cannam@127: cannam@127: commit 230458a658da6fa62fac4ccd66918f38442df00b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 9 17:03:32 2002 -0400 cannam@127: cannam@127: Always overwrite old wisdom with new, in case the old is cannam@127: corrupt/conclicting. cannam@127: cannam@127: commit 2d91c8d00acc06eb228d7ba5492e8f6ec7ee24d9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 9 16:56:03 2002 -0400 cannam@127: cannam@127: added quote/joke cannam@127: cannam@127: commit 0173e3dc140cec9b061b4cf0cc0a626e41105e1f cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 9 15:04:47 2002 -0400 cannam@127: cannam@127: Completed wisdom import cannam@127: cannam@127: commit dee4de2b0b7a9c60d9501d67e5ab17c5e828e474 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 9 10:14:22 2002 -0400 cannam@127: cannam@127: Slight cleanup of md5 interface. cannam@127: cannam@127: commit b990a36ddb7c8ec04d248fcbbbc4a135827cf7b2 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 22:32:43 2002 -0400 cannam@127: cannam@127: More consistent protocol between planner and inferior. cannam@127: cannam@127: commit fbf287fea9e51a6eb2a62030a115aea58ef2f630 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 21:08:30 2002 -0400 cannam@127: cannam@127: I can't think of any situation where saving infeasible problems would cannam@127: be desirable. Removed relevant code. cannam@127: cannam@127: commit c194f7f7a34d6909408bcd55e543f4cbf7a60a2f cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 20:57:03 2002 -0400 cannam@127: cannam@127: Encoder registrar's names in wisdom. Remove export_conf, since cannam@127: a separate program can now generate it. cannam@127: cannam@127: commit e9a30d633c905ac3eba878af7839ad73f153dd06 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 15:11:06 2002 -0400 cannam@127: cannam@127: Fixed typo cannam@127: cannam@127: commit 51b8ddee6eda85459d68909df089b7e251fb65ca cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 14:52:45 2002 -0400 cannam@127: cannam@127: Fixed broken trochaic meter. cannam@127: cannam@127: commit eb531c7d3242141b4603cca8f270d88bab0f48e3 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 09:49:50 2002 -0400 cannam@127: cannam@127: Initialize planner->score. It is correct to leave it uninitialized, cannam@127: but I don't want people to send reports about purify complaining. cannam@127: cannam@127: commit 6a000fc379ab96b4ea9310e76f5391af7d6131a8 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Sep 3 09:03:46 2002 -0400 cannam@127: cannam@127: More latin silliness cannam@127: cannam@127: commit 3b9fecd5d0365958954cc149251fed5b9ce07ddc cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 2 17:57:32 2002 -0400 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 28a40bce8ba8e91b240d4f6e7ddcf55b68f05e77 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 2 17:33:49 2002 -0400 cannam@127: cannam@127: added clock() getseconds timer cannam@127: cannam@127: commit c004f7f51d23ac8f1c6220ff4a18e83d2e4fe7cf cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 16:16:58 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 3f227ec57b346fe8688fbf2e08dfcf6cc3c3c955 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 15:58:19 2002 -0400 cannam@127: cannam@127: Experimental INDIRECT_VERBOTEN flag (not used) cannam@127: cannam@127: commit 802f348a8a8c0bd97e15ed827e8092ab358abbe2 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 15:36:21 2002 -0400 cannam@127: cannam@127: Do not allow buffering in children of indirect solvers. cannam@127: cannam@127: commit f081fc5e202141350664e3d9adbe947d5331ab9f cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 15:02:11 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 08826857c28146b6366770565a9971eb4bdd9505 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 14:32:28 2002 -0400 cannam@127: cannam@127: Hash sizeof(R) as part of wisdom. cannam@127: cannam@127: commit ff803ad22f9a5a980be0ab10aebd18ab210557de cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Sep 2 13:47:57 2002 -0400 cannam@127: cannam@127: added --enable-float synonym for --enable-single (since with have --enable-long-double) cannam@127: cannam@127: commit 99672a129c0032eeb3c99424d16f16547e1fff5d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 13:46:08 2002 -0400 cannam@127: cannam@127: zerotens is now in its own file, so it does not cause dft to be linked cannam@127: in if only rdft is used. cannam@127: cannam@127: commit fe35f517845e4692b46077037dc40b155c9fa500 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 11:56:37 2002 -0400 cannam@127: cannam@127: Removed unused var. cannam@127: cannam@127: commit 2a4e8a9a1121f4adb852256865ae2b52743d6f40 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 11:55:33 2002 -0400 cannam@127: cannam@127: Split insert() in preparation for wisdom import cannam@127: cannam@127: commit fc1cf1dfa9a72fe8968426c4e694b7336926a03d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 11:46:57 2002 -0400 cannam@127: cannam@127: Moved debugging infrastructure to test directory so that it is not cannam@127: linked into the shared library. cannam@127: cannam@127: commit 07a825b37e850981bc9fa18460538346cc7dd137 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Sep 2 11:04:54 2002 -0400 cannam@127: cannam@127: Reactivated wisdom export cannam@127: cannam@127: commit dfcc8fd2b155015d11c95762e4384ae139f922c6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 1 21:30:58 2002 -0400 cannam@127: cannam@127: Dump errors to stderr, not stdout. cannam@127: cannam@127: commit 44e1a88d3527239c405ca268888b3695c902c807 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 1 21:26:38 2002 -0400 cannam@127: cannam@127: Removed traverse.c. cannam@127: traverse.c is no longer need for plan blessing. I figured out cannam@127: a way to avoid using it in planner-score.c, so the file is cannam@127: now redundant. cannam@127: cannam@127: commit 1c9ef6ccf7c373d274215c83bbede0fea30682c7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 1 19:51:50 2002 -0400 cannam@127: cannam@127: Removed code made obsolete by new MD5 scheme: problem equality cannam@127: tests, scanners, and associated list of problem kinds. cannam@127: cannam@127: commit 726b571dc2949d3369fc09237b41f99ef8ce058b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Sep 1 19:22:54 2002 -0400 cannam@127: cannam@127: Started md5 implementation cannam@127: cannam@127: commit d89348364f45c8f635356d8ae1047f6cb8508158 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 31 14:00:04 2002 -0400 cannam@127: cannam@127: Keep track of hit rate cannam@127: cannam@127: commit 971b014a7698a2b54a6dec16b486b7953a1f41e0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 31 12:44:04 2002 -0400 cannam@127: cannam@127: Only dump when verbose > 4 cannam@127: cannam@127: commit 018df5704e37547209aeaecadc50811defcf5f2b cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 31 09:55:57 2002 -0400 cannam@127: cannam@127: Debugging infrastructure cannam@127: cannam@127: commit 837cecb7a386caabb14f3b41518083f18b364fc7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 31 09:21:48 2002 -0400 cannam@127: cannam@127: Use debug infrastructure to dump planner. cannam@127: cannam@127: commit dfaf407162a50f7a77ef3496dbaf754b27551029 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 30 21:29:10 2002 -0400 cannam@127: cannam@127: Do not store plans in planner, plus general planner cleanup. cannam@127: cannam@127: commit e74dd299a89b62e7d0e51be3293163325b5044ea cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 18:07:52 2002 -0400 cannam@127: cannam@127: renamed IN_DHT_R2HC to the more general FORBID_DHT_R2HC cannam@127: cannam@127: commit 904ff75e31a716b0c9d97f7e4efd425bd4309c06 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 18:07:21 2002 -0400 cannam@127: cannam@127: eliminated unused var cannam@127: cannam@127: commit 60f3382238de7ce933a192a923f06657900e4853 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 30 12:09:48 2002 -0400 cannam@127: cannam@127: Score planner was not working correctly when using wisdom. Fixed. cannam@127: cannam@127: commit 0522d2fcf8b00ec9f1f3cdbd38cfa03d764e1154 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 30 08:20:48 2002 -0400 cannam@127: cannam@127: Use hash table in debug malloc cannam@127: cannam@127: commit e451f6d74620eefd71304a543d0eb48ab9ea953d cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 02:45:15 2002 -0400 cannam@127: cannam@127: listed some good stuff cannam@127: cannam@127: commit 4d5aeb7a8917367942f1e798bd19d5c3be2feaf1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 02:17:30 2002 -0400 cannam@127: cannam@127: timed planner cannam@127: cannam@127: commit 3bb76589f0d1efc68b0a82eeecbdf30748a991fd cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 02:07:00 2002 -0400 cannam@127: cannam@127: fma? cannam@127: cannam@127: commit e231f879031ab34c3a8fa8a8da0fbf642cd88a0b cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 02:05:55 2002 -0400 cannam@127: cannam@127: update cannam@127: cannam@127: commit 1f29de7940d8c29acc6b608deec341cd982cc706 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 01:31:47 2002 -0400 cannam@127: cannam@127: rader-dht -> dht-rader cannam@127: cannam@127: commit 888439dcea7af2d64300c776238afa7c5c18a372 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 30 01:21:37 2002 -0400 cannam@127: cannam@127: add DHT solver, and break up rader-dht and r2hc-hc2r cannam@127: cannam@127: commit df668dba33c6181cb6feb58dba6f649c89d73fe6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 23:20:35 2002 -0400 cannam@127: cannam@127: another option cannam@127: cannam@127: commit 213d66b5bcf73e5525c97bb6dc9461808600d483 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 22:55:29 2002 -0400 cannam@127: cannam@127: generalized indirect solvers for fftw2-like buffering and more cannam@127: cannam@127: commit 795353001d9c3db1d30d5def55b8671cb4eb67b0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 18:08:16 2002 -0400 cannam@127: cannam@127: tensor_max_index and tensor_min_stride are now both unsigned cannam@127: cannam@127: commit d59c4e92144b248504c9c01cae0ea5d3f4e0aa60 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 17:58:35 2002 -0400 cannam@127: cannam@127: added iabs.c, and tensor_min_stride returns min absolute value cannam@127: cannam@127: commit 9247f8665076f514844f3fd4a0478f9988313251 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 17:31:39 2002 -0400 cannam@127: cannam@127: bug fix in cldrest hc2c/c2hc copy loops cannam@127: cannam@127: commit c8d575230e1d18331b0a1d8fe22d6c8fcfd70ce4 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 29 13:45:08 2002 -0400 cannam@127: cannam@127: Added things to do. cannam@127: cannam@127: commit c8b62313d7952baa412a1c18427473c010451303 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 13:10:04 2002 -0400 cannam@127: cannam@127: added automake prereq cannam@127: cannam@127: commit 3a6be2745802ae618bbb0521fe5e54e17e1eecb2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 29 08:36:36 2002 -0400 cannam@127: cannam@127: Use indexed addressing cannam@127: cannam@127: commit b27567a442018af7ec7cf782f117bfc9f5476e08 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 29 08:20:55 2002 -0400 cannam@127: cannam@127: Ooops cannam@127: cannam@127: commit dc19f8daf30d11be571d8856cb10e8d8a9046b6c cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 29 07:45:37 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit ee4bb3eb29224fe9ef1e5c3c8416256b04a22599 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 02:32:13 2002 -0400 cannam@127: cannam@127: updates to win32 threads code (ick) cannam@127: cannam@127: commit 6cfa2e60b0f2831342f21a72eb3732bf80366250 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 29 01:44:33 2002 -0400 cannam@127: cannam@127: added threaded version cannam@127: cannam@127: commit 8cc323cbe02b7c270bb664e2c8acedce4ddf48b5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Aug 28 19:47:21 2002 -0400 cannam@127: cannam@127: fix make dist cannam@127: cannam@127: commit 43fa7922084aae1bdb1b69aff7109aed56e1f025 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Aug 28 15:09:03 2002 -0400 cannam@127: cannam@127: whoops, bugfix for inverse cannam@127: cannam@127: commit b3136883e3f02c5bbb68338e01d134aa2b2a25eb cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 28 14:50:34 2002 -0400 cannam@127: cannam@127: Use C9x convention for naming (fftwf etc.). Removed installable header cannam@127: files since they will be part of the API. cannam@127: cannam@127: commit ed3b5e17a932fe32d1a9397642c6e043eb3dbc40 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Aug 27 23:34:00 2002 -0400 cannam@127: cannam@127: allow _1 variants to accept rnk 0 (sz 1) problems cannam@127: cannam@127: commit 004227a749ea2b1990047e29b15826437335239f cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Aug 27 15:56:09 2002 -0400 cannam@127: cannam@127: updated cannam@127: cannam@127: commit ca46171fc03dc63c4a93462abded63a039acdd82 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 26 20:14:56 2002 -0400 cannam@127: cannam@127: Loop unroll is useless cannam@127: cannam@127: commit 6528250479b8700bc9082f776d5f3c340080175a cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 26 20:00:41 2002 -0400 cannam@127: cannam@127: Use indexed addressing cannam@127: cannam@127: commit 8f4d60a4abc3e10d5e4e2739119422babe433a1d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 26 19:46:46 2002 -0400 cannam@127: cannam@127: Use indexed addressing in transpose routines. (Seems to be cannam@127: slightly better on athlon.) cannam@127: cannam@127: commit aac3c6a8800ddbc174774e9eeeb32f054c8af6a6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 26 12:59:44 2002 -0400 cannam@127: cannam@127: added comment about stability cannam@127: cannam@127: commit bdaced931410f8e984ac5c3a833e842d6ffa8965 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 26 07:43:53 2002 -0400 cannam@127: cannam@127: Approximate opcount cannam@127: cannam@127: commit 46c5151b696b0d6f0ff98f952d8a13283d95877b cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 26 06:38:49 2002 -0400 cannam@127: cannam@127: Finished rdft2 via dft/rdft cannam@127: cannam@127: commit c9122c8dbfcac3ae13893442c0ad348e410b646b cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 26 00:15:59 2002 -0400 cannam@127: cannam@127: some updates cannam@127: cannam@127: commit b049bb9502bd3c00a3a8f1ff3cfd3c5596e9e7c0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 26 00:05:53 2002 -0400 cannam@127: cannam@127: rdft kind is now per-dimension, added rdft/rank-geq2 cannam@127: cannam@127: commit 32db021f7eabd57af68c88e1e6266589a828df35 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 25 22:45:38 2002 -0400 cannam@127: cannam@127: added note cannam@127: cannam@127: commit e174f61162d8e897e181f1ae8f01b5c8ba4122d8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 25 22:28:12 2002 -0400 cannam@127: cannam@127: must zero real sz cannam@127: cannam@127: commit 7881bf396852115443bd3ce1dbdce177d8d64b6e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 25 22:06:52 2002 -0400 cannam@127: cannam@127: unified pickdim funcs cannam@127: cannam@127: commit 9b588fdbd50ab5d47cb936aae0569b6f9b54fc4c cannam@127: Author: fftw cannam@127: Date: Sun Aug 25 14:10:55 2002 -0400 cannam@127: cannam@127: silence warnings cannam@127: cannam@127: commit f58dff38cc5c75f8c508e971a1ffd286be572f87 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 25 13:16:49 2002 -0400 cannam@127: cannam@127: I had to add another planner flag to record whether pointers could cannam@127: become unaligned because of vrank-geq1 solvers (these solvers only cannam@127: plan the first element of a vector problem, but the second element cannam@127: may have a different alignment). This addition is ugly, but I don't cannam@127: see any way around it. cannam@127: cannam@127: commit 3633f42453ac103289d0c471630892680f1b0625 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 25 10:18:25 2002 -0400 cannam@127: cannam@127: Added thoughts cannam@127: cannam@127: commit fb9c1acef1a3499a8629190b172a1ec0430260b4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 25 10:08:59 2002 -0400 cannam@127: cannam@127: Implemented rdft2 via vector rdft + radix2 step cannam@127: cannam@127: commit 8bc1aed075f15afc6de9d82adc44d6ab8b5e50f6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 24 17:43:54 2002 -0400 cannam@127: cannam@127: Stylistic changes cannam@127: cannam@127: commit 188add2600049e4313ba1e77c1976b887544ae90 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 24 11:19:30 2002 -0400 cannam@127: cannam@127: Simplified mktwiddle interface cannam@127: cannam@127: commit 1c91434a8fea606141e28014376c2d2c9937f1b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 24 11:05:08 2002 -0400 cannam@127: cannam@127: Unification of certain vector computations. rdft2-dft is now a cannam@127: vector transform. cannam@127: cannam@127: commit f9311503a90a428a78350116e1bf47c6ffefddcc cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 23 20:21:25 2002 -0400 cannam@127: cannam@127: Intel compiler seems to be still buggy cannam@127: cannam@127: commit de1bb9192bc2740a6fbe70bbac497a1ac34450f0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 23 16:07:12 2002 -0400 cannam@127: cannam@127: Streamlined twiddle protocol cannam@127: cannam@127: commit e2a28ed96b7b40db5f3fe6b72852acf550cdaca4 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 23 13:22:17 2002 -0400 cannam@127: cannam@127: Implemented rdft2 via dft (forward only for now) cannam@127: cannam@127: commit 17d57ef85db79c55dcd5c77260618e798833e1d0 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 22 11:29:29 2002 -0400 cannam@127: cannam@127: More cleanup of verify cannam@127: cannam@127: commit 0eb03788a7c1c13953638a19182235738511b77a cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 22 11:16:03 2002 -0400 cannam@127: cannam@127: Changed error criterion because old one was too strict cannam@127: cannam@127: commit e97f092fbcd955e08b36522352e0e9b94cfd473e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 22 11:15:17 2002 -0400 cannam@127: cannam@127: Disable shared cannam@127: cannam@127: commit f611df2e683eea128367d652cf12e2c4a2b3dfb1 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 22 09:19:12 2002 -0400 cannam@127: cannam@127: Added thoughts cannam@127: cannam@127: commit 6ad63b2082094c83de1fe5a01400423c249564a1 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 22 09:17:28 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 42381c1bcded2b7d1854300ff5de0addfca36575 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 22 09:11:34 2002 -0400 cannam@127: cannam@127: Do not use inline. Minor changes. cannam@127: cannam@127: commit b9b2448db23e3d5d0d6d10b4bf3fe1858a847a05 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Aug 21 16:23:26 2002 -0400 cannam@127: cannam@127: more commented flags cannam@127: cannam@127: commit 1c316981c3ed81a4550d6ba9bfa733745532bcf2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Aug 20 19:44:43 2002 -0400 cannam@127: cannam@127: added DCT-IV and DST-IV cannam@127: cannam@127: commit c15e995b01901e957e42c369b5341b6dbeaf3ac2 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 20 16:01:36 2002 -0400 cannam@127: cannam@127: Slight improvement in twiddle scheme cannam@127: cannam@127: commit 8ba1ef4db7a3866fae35bd1825a5a5c35ae5673f cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Aug 20 15:31:54 2002 -0400 cannam@127: cannam@127: name fix cannam@127: cannam@127: commit 5b56bb057b02f41413a9d5436a3d9d1c50a0e7f5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Aug 20 15:16:48 2002 -0400 cannam@127: cannam@127: removed extraneous variable cannam@127: cannam@127: commit d8b1080be08abcfa55a88f38d7bc677d99a9c2d9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 20 11:46:29 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit b500a0d285f5b95a9b88952b8830aa8423be9332 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 20 08:37:45 2002 -0400 cannam@127: cannam@127: Still playing around cannam@127: cannam@127: commit 7a44ac35618394bd3715c928e9dc0b3a7a149f0e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Aug 19 19:56:29 2002 -0400 cannam@127: cannam@127: Playing around with addition chain cannam@127: cannam@127: commit ac8dfff733ce38c8b013523ff4e9fc9888456989 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 19 19:48:56 2002 -0400 cannam@127: cannam@127: comments cannam@127: cannam@127: commit 175b3b2cec3441b1a0e34d03343e9f581f0e030e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 19 19:45:35 2002 -0400 cannam@127: cannam@127: comment fixes cannam@127: cannam@127: commit 0b6386c1ef38de51383f1306c82e7966b0db0d52 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 19 19:40:18 2002 -0400 cannam@127: cannam@127: added reodft stuff cannam@127: cannam@127: commit 8d4aef3c5738367c010a1bfd3004c94f73281950 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 18 19:44:14 2002 -0400 cannam@127: cannam@127: Sync with nbenchfft cannam@127: cannam@127: commit 6ec9197550f61b20dad2a8e238bd00da3bf2cf23 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 18 16:02:37 2002 -0400 cannam@127: cannam@127: Economy of thought cannam@127: cannam@127: commit 70610d2a45fcebc6b9c3c61e5dd6caa96d292b4f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 17 15:52:05 2002 -0400 cannam@127: cannam@127: distribute addchain.c cannam@127: cannam@127: commit 2fea59351e2abaeec9bb1ea094d06097282bdf7e cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 17 14:09:11 2002 -0400 cannam@127: cannam@127: Nothing serious cannam@127: cannam@127: commit c5ef4a2ddba0963a8c9a388edf050c5ee2fbbb00 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 17 10:47:59 2002 -0400 cannam@127: cannam@127: New twiddle policy (disabled for now) cannam@127: cannam@127: commit bf62c3f3bb4be6257869db7d46f69b694c7a2688 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 16 23:44:28 2002 -0400 cannam@127: cannam@127: bug fix for hc2r (must use inverse dft) cannam@127: cannam@127: commit e7434c44d6c3f77e761da5e7a8e850f48c6fb872 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 20:27:10 2002 -0400 cannam@127: cannam@127: New log3 twiddle policy cannam@127: cannam@127: commit d0f1857c45c12d35cbd9fded016c3b7ceac70aa7 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 18:10:33 2002 -0400 cannam@127: cannam@127: More verify cleanup cannam@127: cannam@127: commit c8f750da8aab093581b6be29ff8d781906ca771b cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 16:31:19 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit eae86b4ff97b783a93ebd6f7b0a6352cea48359c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 15:22:36 2002 -0400 cannam@127: cannam@127: Economy of thought (and code) cannam@127: cannam@127: commit d7bdf2e9b5a011b19bb16a9d12d5d763a3196c2a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 14:05:45 2002 -0400 cannam@127: cannam@127: Added comment cannam@127: cannam@127: commit 4bec01a2c6ea089d18a81b8d7d3ce649cbe80fe8 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 12:57:43 2002 -0400 cannam@127: cannam@127: Cleaner rounding algorithm cannam@127: cannam@127: commit d91cc0e5f23a61e226b2a575c23f35c79ec3ca06 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 11:27:43 2002 -0400 cannam@127: cannam@127: Can get away with shorter length in bluestein (I think). cannam@127: cannam@127: commit 56113aa7d008511b8387a1d1652e03d9fd8844e7 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 11:08:09 2002 -0400 cannam@127: cannam@127: Portability improvements cannam@127: cannam@127: commit b58468b7ecd1e0ff7b9a2b1236d64e357627d8a0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 08:06:31 2002 -0400 cannam@127: cannam@127: Optionally average accuracy test over many rounds cannam@127: cannam@127: commit 21b850aeaafa046e663e6bc5a42a9538c9571180 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 07:50:24 2002 -0400 cannam@127: cannam@127: More accurate formula for trig tables cannam@127: cannam@127: commit ce0241125c235817e2132e938e8c9dcd3166773f cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 16 06:42:02 2002 -0400 cannam@127: cannam@127: Implemented accuracy test for all integers cannam@127: cannam@127: commit db374e203e4d37c399e6b3d877da8cdf192ec649 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 18:54:44 2002 -0400 cannam@127: cannam@127: inv, neg: make static cannam@127: cannam@127: commit 79f1c53641c0cec5612621c1f72726a81d56144e cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 17:25:37 2002 -0400 cannam@127: cannam@127: Verify was not complete for real transforms cannam@127: cannam@127: commit c60e8fcdedd600b93f30f098ca49f794375e8377 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 16:30:03 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 688a0ef88f8ef6a8d060ad2b04ce51b4d94870b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 16:29:16 2002 -0400 cannam@127: cannam@127: Fixed hb codelets cannam@127: cannam@127: commit 8a4d71183c0b5cd62b6f9f53f41cfd68a8b602fc cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 14:10:45 2002 -0400 cannam@127: cannam@127: Changed twiddle policy cannam@127: cannam@127: commit 9905db7579db957d5ebc9f472847910d24b65e10 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 15 13:32:24 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 59c9d170edb4001d3f37b64cfb2d8d48e9cb9b7d cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 11:01:04 2002 -0400 cannam@127: cannam@127: No point in libbench being a shared library cannam@127: cannam@127: commit b75824c63085764eb1fbf97b83961eb0411bd969 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 15 09:48:37 2002 -0400 cannam@127: cannam@127: Moved accuracy test to libbench cannam@127: cannam@127: commit ebac0dde6d77f268c45cfc7ba17230c547e60800 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 19:48:23 2002 -0400 cannam@127: cannam@127: Modified accuracy test cannam@127: cannam@127: commit d5e2c4a63b8f9b195e7812f817cefc61617accf9 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 08:34:26 2002 -0400 cannam@127: cannam@127: Fixes for long double cannam@127: cannam@127: commit 41c23eb9e18add9786e959871bcba7d1ffc44bb6 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 08:17:57 2002 -0400 cannam@127: cannam@127: Normalize input cannam@127: cannam@127: commit d83e36740eaf0cd2d8ffecb25d38ffdcd6412bcb cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 07:26:41 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 0d312034a4b8a8ece11903c0b81aa4ce57151783 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 07:25:34 2002 -0400 cannam@127: cannam@127: Also compute relative error cannam@127: cannam@127: commit 10c281df8a8195c0eb497cd3b73955807c64f06e cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 07:08:20 2002 -0400 cannam@127: cannam@127: Loop over N cannam@127: cannam@127: commit 588a70753f53ef9fe7801fd3c0cd1b1f2c5da7fc cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 14 06:54:50 2002 -0400 cannam@127: cannam@127: simple-minded accuracy test cannam@127: cannam@127: commit b25380fae97bb3af2b38f592f5393b10be1ff430 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Aug 14 03:26:06 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit d32e62c62857ed17fdf4a9fa7ebb12007c8e32bc cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 13 11:42:41 2002 -0400 cannam@127: cannam@127: fma() stuff is too nonportable, removed cannam@127: cannam@127: commit 583c58e086a592a198619d6c36fcb6137b4ea068 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 12 14:07:44 2002 -0400 cannam@127: cannam@127: slight fix cannam@127: cannam@127: commit 2be67d85e7e8c8b4db5644bdeca6cfa1a0211959 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 12 14:07:18 2002 -0400 cannam@127: cannam@127: use table for rdft_kind_str cannam@127: cannam@127: commit 331ca343e551313e04bc1c88ae8c8cd3e1dfb4fd cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 12 13:43:08 2002 -0400 cannam@127: cannam@127: slight fixes cannam@127: cannam@127: commit 8bf7bf1145ef67937cf020d64f0e9913aef84d58 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 12 13:31:37 2002 -0400 cannam@127: cannam@127: multidimensional rdft2 cannam@127: cannam@127: commit 4457a7cf6bf60cff0e842bfd127e22e7b3de55e5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 10 19:33:23 2002 -0400 cannam@127: cannam@127: use tensor_copy_inplace cannam@127: cannam@127: commit 5e370a1a072a67b940639f311e296a97150acf1c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 10 19:32:03 2002 -0400 cannam@127: cannam@127: bugfix, use tensor_copy_inplace cannam@127: cannam@127: commit 92f280c99e002d9ee78e42967ee81bb4fcf84d2b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 10 19:30:39 2002 -0400 cannam@127: cannam@127: use tensor_copy_inplace cannam@127: cannam@127: commit 55ee1b50c140e81f41abc05975f01393c8bd4cbd cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 10 19:28:07 2002 -0400 cannam@127: cannam@127: added tensor_copy_inplace cannam@127: cannam@127: commit ce8083b65d5ae7952d40c253896ae0e6759e73e8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 10 19:25:50 2002 -0400 cannam@127: cannam@127: fixed trig-function table type cannam@127: cannam@127: commit 9b354635204711389328f487a058a54604d58e0a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 10 14:41:04 2002 -0400 cannam@127: cannam@127: Improved trig scheme cannam@127: cannam@127: commit 466d2a03411d082ab673c73582a08842f12f6846 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 21:05:01 2002 -0400 cannam@127: cannam@127: Allow for testing using long double instead of pari cannam@127: cannam@127: commit 14b243d1d509236a5b19e8783570989cdfda6333 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 20:49:32 2002 -0400 cannam@127: cannam@127: Yet another trig scheme. cannam@127: cannam@127: commit 361e112752a93e14cab74d86d92fccb88686fed1 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 20:38:07 2002 -0400 cannam@127: cannam@127: Yet another scheme cannam@127: cannam@127: commit b3ca7c941515736b0ebd97c7d1195cd736d2b8d8 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 20:31:16 2002 -0400 cannam@127: cannam@127: Careful with overflow cannam@127: cannam@127: commit c1af0a91c6bbcd3482427d1be4a812a0c061d879 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 20:16:23 2002 -0400 cannam@127: cannam@127: Avoid overflow cannam@127: cannam@127: commit f06cb59c469661f10f65f220b91d79e8d98097f7 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 19:26:57 2002 -0400 cannam@127: cannam@127: New(er) trig routines cannam@127: cannam@127: commit ba6e2f6487663745c402856288f95441c6191fe8 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 19:25:44 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 267f53395f5e34f83a0664c9405e4d7b17583695 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 18:49:04 2002 -0400 cannam@127: cannam@127: New file cannam@127: cannam@127: commit cc25b36b4369f7fd773b614e416185078bc3e20c cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 13:04:00 2002 -0400 cannam@127: cannam@127: Commented about likely gcc bug cannam@127: cannam@127: commit 745572695256ffc140f9b3bd828b561f56bea1a5 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 9 13:01:49 2002 -0400 cannam@127: cannam@127: Improved accuracy of twiddle factors cannam@127: cannam@127: commit b90ec91c045668caabc583c27da9400331fc34cc cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 8 06:36:23 2002 -0400 cannam@127: cannam@127: Wrong comment cannam@127: cannam@127: commit 01653dbd957c931c5e562c6cdf727c26a4570680 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 7 17:14:09 2002 -0400 cannam@127: cannam@127: Experimental 3dnow port using gcc, to compare it with Stefan's stuff. cannam@127: cannam@127: commit 9716316af3a8a84ac9888e8b184fad1f8b34279d cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 7 12:58:10 2002 -0400 cannam@127: cannam@127: End of AREF experiment cannam@127: cannam@127: commit 03365b937b905ad6dd6dad3ec0044f010f2cec51 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 7 07:47:19 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 00d1519ee07579c41da9738b4bd0d9e130c252df cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Aug 7 07:46:38 2002 -0400 cannam@127: cannam@127: Pathetic attempt to reduce size of configure script cannam@127: cannam@127: commit 882c809b6257b73377a20a807a20a61f5cc5a655 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 6 20:38:11 2002 -0400 cannam@127: cannam@127: Changed array syntax for experiments. cannam@127: cannam@127: commit 06bf9f0b7d08eb0a66a07b4b517fede0514a4a2c cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 6 19:58:20 2002 -0400 cannam@127: cannam@127: Fix warning cannam@127: cannam@127: commit dd2b973d27111516233a46e5d44734f2d1cea503 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 6 13:35:28 2002 -0400 cannam@127: cannam@127: Move nonportable stuff in one place. cannam@127: cannam@127: commit 3a3a36d48074544b746b464bd194f93a371615b9 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 6 10:32:53 2002 -0400 cannam@127: cannam@127: Economy of thought: I didn't like having two algorithms for removing cannam@127: solutions, both correct. At least now we have the same algorithm cannam@127: copied twice. cannam@127: cannam@127: commit e0cf8fd96853061b2160a99ed871b621a69bacbe cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Aug 6 09:12:21 2002 -0400 cannam@127: cannam@127: Added things to do cannam@127: cannam@127: commit f96ded332986cff7099c0dd6bf2cff07d3e59217 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 5 19:54:31 2002 -0400 cannam@127: cannam@127: improved interaction of planner with patience flags cannam@127: cannam@127: commit f37ad7a0a0a7009a6c29c02ff53b06440f12e846 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Aug 5 14:17:58 2002 -0400 cannam@127: cannam@127: set up for real-even/odd DFTs, where n is not the size of the data cannam@127: cannam@127: commit 1a2ea854fa6156b907c817752dc47a1c07ef5c2e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 4 23:57:51 2002 -0400 cannam@127: cannam@127: DESTROY_INPUT flag cannam@127: cannam@127: commit 18483232ce3afae0412e565222de6c48891700d7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 4 22:50:19 2002 -0400 cannam@127: cannam@127: CLASSIC -> IMPATIENT cannam@127: cannam@127: commit 0fee1c8d39ed87aaab3387028cc3ff4422261a41 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Aug 4 19:05:43 2002 -0400 cannam@127: cannam@127: Require make maintainer-clean to remove the generator, as opposed cannam@127: to make clean. In this way we can type make clean without regenerating cannam@127: all codelets. cannam@127: cannam@127: commit b633708685610bf42bb69bbe71f31f0fd849aff5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 4 17:34:04 2002 -0400 cannam@127: cannam@127: ESTIMATE plans are not blessed cannam@127: cannam@127: commit 17f106f814fd30121f7fcc2de65cc78f77a6448d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 4 17:24:37 2002 -0400 cannam@127: cannam@127: use flags in wisdom cannam@127: cannam@127: commit 342928973eaf98429367ce537b088761c391505c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 4 17:03:45 2002 -0400 cannam@127: cannam@127: score now takes plnr, not flags, as arg cannam@127: cannam@127: commit 5ef96008dcfb0e7428716122ea8ea56d0637898a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Aug 4 16:37:46 2002 -0400 cannam@127: cannam@127: align initial stack in alignment check, which should now pass for gcc 3.1.1 cannam@127: cannam@127: commit ce14480bda337274a988627272fbe696bcaf5589 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 3 20:04:57 2002 -0400 cannam@127: cannam@127: Detect ultrasparc (sort of) cannam@127: cannam@127: commit 946e964b908a9fcd9b98345a5f525049b8143cce cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 3 19:38:17 2002 -0400 cannam@127: cannam@127: added solvtab_rdft_r2r placeholder cannam@127: cannam@127: commit db8c63ea924d244e0c207d514dd425bfab39f2b6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 3 19:34:49 2002 -0400 cannam@127: cannam@127: Damn solaris cannam@127: cannam@127: commit 6f4f2a31d28db1040f796b703d9b6c9fd7b4052d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 3 17:55:44 2002 -0400 cannam@127: cannam@127: use E extended precision in solvers cannam@127: cannam@127: commit eb1a98695f9827716943ddc0ca00475c2d61d9c2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 3 17:53:29 2002 -0400 cannam@127: cannam@127: an alternative notation for D{C,S}T: DXTio, where i/o are {0,1} cannam@127: according to whether the input/output are shifted, respectively. cannam@127: Alternatively, io is the binary representation of the usual cannam@127: DXT-{I,II,III,IV} nomenclature, minus 1. cannam@127: cannam@127: commit 24b13985e810f08cbef3c5dac739433c5ac0161a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Aug 3 17:49:11 2002 -0400 cannam@127: cannam@127: use E extended precision in solvers cannam@127: cannam@127: commit 46b2fc024b187b4356bf6a7977d508a4c4ba22c1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 3 15:39:49 2002 -0400 cannam@127: cannam@127: More portability fixes, compiler bugs workarounds, etc. cannam@127: cannam@127: commit ca88f96aed7b0399f4d2199342c5287639e51d3b cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 3 15:09:56 2002 -0400 cannam@127: cannam@127: More portability work cannam@127: cannam@127: commit 3cfd742c2225f91d295d75af9e6ddc46cd4c39f4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 3 14:33:40 2002 -0400 cannam@127: cannam@127: Improved portability, removed gnu make dependencies cannam@127: cannam@127: commit ac8aa3edbc9864af3b3e3e8d753cc2388b80732c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Aug 3 13:48:53 2002 -0400 cannam@127: cannam@127: Remember to thank XXX cannam@127: cannam@127: commit 807dc0e147fedfa044a4ae2a03dbff426e155136 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 2 17:38:18 2002 -0400 cannam@127: cannam@127: Multiplication on altivec requires FMA with -0.0 to be IEEE754 compliant. cannam@127: cannam@127: commit dfa0ebdb72edd084c82b682b62fffdbd8f9a7611 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 2 15:26:37 2002 -0400 cannam@127: cannam@127: Allow for extended precision in codelets cannam@127: cannam@127: commit 2eee7899ea3308e919dbeafffeee423dd0c810b5 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Aug 2 08:52:04 2002 -0400 cannam@127: cannam@127: Shortened names cannam@127: cannam@127: commit 239f0f6f2197b4761abad5f8ac2f1da6736a5ccd cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Aug 2 03:49:09 2002 -0400 cannam@127: cannam@127: added infrastructure for future r2r transforms cannam@127: cannam@127: commit 4f64527883bd151d5f597abec9870dc9e6d0c8b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 1 21:29:14 2002 -0400 cannam@127: cannam@127: Version info cannam@127: cannam@127: commit 1f6a7039b9fe3a439d6aa9fa83d179fb864ab920 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 1 21:06:22 2002 -0400 cannam@127: cannam@127: Listened to one customer and added radix-12. Added radix-15 for cannam@127: consistency (whatever that is) cannam@127: cannam@127: commit ece6187a35d44322c45b0fc946187615d8d3bebd cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 19:50:53 2002 -0400 cannam@127: cannam@127: whoops again, fixed the wrong line cannam@127: cannam@127: commit 53c48f4c8eb4f39a1bcea9b47a2cf78c669e2dd2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 19:50:16 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit afb281f39223c26fe968873928fd8ca0c69c1fe7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 16:01:15 2002 -0400 cannam@127: cannam@127: use new AC_INIT and add VERSION to wisdom cannam@127: cannam@127: commit 1d4b7a029734d0948b44713fb94429ffd4ce40d4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 14:56:45 2002 -0400 cannam@127: cannam@127: mygetR -> getR cannam@127: cannam@127: commit 010ffe455949d901be083a52aeb485e933d4c252 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 14:56:02 2002 -0400 cannam@127: cannam@127: scanner cleanups: just return 0/1, simplify integer reads cannam@127: cannam@127: commit 052184d84276b884548c95a76e89d5f2ccd124d2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 1 08:04:01 2002 -0400 cannam@127: cannam@127: Reverted back to casting pointer to ulong cannam@127: cannam@127: commit c61b1e4aa77a06a1565d2f816bc2b6a22c82f6d8 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Aug 1 08:03:46 2002 -0400 cannam@127: cannam@127: Cast to unsigned long, not long cannam@127: cannam@127: commit bc2a8794eec9dbdd2eaa2d10070974ab1cdcc3bf cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 03:14:50 2002 -0400 cannam@127: cannam@127: additional comment cannam@127: cannam@127: commit 72bc55e7f202b4d772bc8a50263870f1434becb0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 03:12:37 2002 -0400 cannam@127: cannam@127: added comment cannam@127: cannam@127: commit 980a9e749d1361de03ea2256209ee0216942a6aa cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Aug 1 03:03:18 2002 -0400 cannam@127: cannam@127: added wisdom import cannam@127: cannam@127: commit b9bcf9486c742271f7c9fa64f41791666cf16cb6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 31 23:12:05 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 183a8a7311c571981db4ef087608b599de96b062 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 31 22:06:46 2002 -0400 cannam@127: cannam@127: use %u for alignment_of cannam@127: cannam@127: commit f9cc3f2e326569214e7ac246b5dacabe10f9f4aa cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 31 21:47:15 2002 -0400 cannam@127: cannam@127: ptrdiff_t form cannam@127: cannam@127: commit 26346129bd45ff91529e18e5770220025ae5cc8c cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 31 21:33:35 2002 -0400 cannam@127: cannam@127: Cast to avoid warning from C++ compiler cannam@127: cannam@127: commit dc8c0c64365fd7f14a579a730f50107f4c01839e cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 31 18:57:04 2002 -0400 cannam@127: cannam@127: Make problem equality depend on alignments. cannam@127: cannam@127: commit 185babf3691983eb1fc109f4d2864ea80070319f cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 31 15:45:31 2002 -0400 cannam@127: cannam@127: Shorter names cannam@127: cannam@127: commit d0a23f2a7ca0ef90c893e1bc9fe38562bf4b97c4 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 31 14:38:00 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit db553c5b6c9be77013e5e6862aecb074abd05daf cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 31 14:37:19 2002 -0400 cannam@127: cannam@127: Fix warning cannam@127: cannam@127: commit 20ce4a31106f745c8765cafa87b94df7b152ba01 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 31 07:52:53 2002 -0400 cannam@127: cannam@127: Removed silly abstraction barrier. Also, cons() terminology was cannam@127: no longer appropriate. cannam@127: cannam@127: commit 6e519e71ee2bff45a45acc9860e6688b5a2ac0ca cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 22:35:24 2002 -0400 cannam@127: cannam@127: removed register_registrar and solvtab_exec_reverse hacks cannam@127: cannam@127: commit 3bb2201fd6c0b2a0e2e6e1cb07849fc640c23fe4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 19:54:41 2002 -0400 cannam@127: cannam@127: register_registrar doesn't search whole solver list (maybe we should change register_solver instead) cannam@127: cannam@127: commit acf987d04a520c14c0d452f2036338e4d89e91a0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 19:36:37 2002 -0400 cannam@127: cannam@127: credit cannam@127: cannam@127: commit 1ae9a399e262ce07b3733a11fcb23ea08541bd45 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 19:34:16 2002 -0400 cannam@127: cannam@127: added HP/UX ia64 support, courtesy of Teresa L. Johnson cannam@127: cannam@127: commit 76ce2ea38a0a18376e316ee3348e8ffd069aebe1 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 30 13:28:33 2002 -0400 cannam@127: cannam@127: Fixed alignment checks cannam@127: cannam@127: commit 7356d1bc11f552e41d0de8df8fc9e0ef4f83b1a0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 01:20:11 2002 -0400 cannam@127: cannam@127: ugh, wisdom id fixes in exprt_conf cannam@127: cannam@127: commit 110cfd3d5abb89da042f3953d99179c04fcb6839 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 00:41:15 2002 -0400 cannam@127: cannam@127: exprt_registrars -> exprt_conf, added missing SOLVTAB_END cannam@127: cannam@127: commit 331b32dd8322273182a47c852416afaac4f6007b cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 30 00:36:26 2002 -0400 cannam@127: cannam@127: exprt_registrars should output self-contained configuration cannam@127: cannam@127: commit ddd63d9b49d333a58f352f4f561a6ff1fbe17a5a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 29 23:52:07 2002 -0400 cannam@127: cannam@127: added exprt_registrars cannam@127: cannam@127: commit 691ba278639460f94cfd6ff45e14e10007d4f62c cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 29 23:42:27 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit ebcd431d564b1f5f86f3bb274ed123971d449415 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 21:24:51 2002 -0400 cannam@127: cannam@127: More stringent requirements on strides for SIMD codelets cannam@127: cannam@127: commit 4fa11627e55b15059ce9b91dce1383c29040f2bb cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 29 21:05:49 2002 -0400 cannam@127: cannam@127: remove warning cannam@127: cannam@127: commit 30f4b2f2ca2fd97ae591c98d812ec38546a1cd8b cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 29 20:51:19 2002 -0400 cannam@127: cannam@127: use %td for ptrdiff_t and %T for tensors cannam@127: cannam@127: commit 33c7a10abb7b7b1e3250654481f536b6e33de824 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 16:17:11 2002 -0400 cannam@127: cannam@127: Fix for SIMD cannam@127: cannam@127: commit 1688dda0ec01678ac3d2e16af154c4898a56b568 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 15:40:53 2002 -0400 cannam@127: cannam@127: Missing lfftw_mkstride and lfftw_stride_destroy cannam@127: cannam@127: commit 2e84b7c68c4270593cc2a1c152520b6f55e3c0c9 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 14:34:46 2002 -0400 cannam@127: cannam@127: Implement LDA/STA cannam@127: cannam@127: commit 385b21d8dc7b1c465acbb83b5414caefa80960d2 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 14:19:21 2002 -0400 cannam@127: cannam@127: More SIMD work cannam@127: cannam@127: commit 00e43e5facae3c33e901ca12dd57cf5905c8508d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 13:16:12 2002 -0400 cannam@127: cannam@127: Cleanup cannam@127: cannam@127: commit 6fb8177180cf59f95bc37163f4e8d4c68b1657e8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 29 13:02:38 2002 -0400 cannam@127: cannam@127: update cannam@127: cannam@127: commit 8354486a52f87afe52440aa3316acec7c768ac75 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 12:45:33 2002 -0400 cannam@127: cannam@127: Also check strides in SIMD codelets cannam@127: cannam@127: commit 7b48f56b4e18bd9799c46214829e7b15531e5244 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 29 11:26:08 2002 -0400 cannam@127: cannam@127: Minor changes, mostly for consistency with the big-endian processor cannam@127: cannam@127: commit ec8f6e4c58d50603587e0bba533ec2086e6174b8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 29 00:50:06 2002 -0400 cannam@127: cannam@127: added comment cannam@127: cannam@127: commit a7cc792884a9acb1d81a346cdab1d9b07e9b2bdf cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 28 21:19:35 2002 -0400 cannam@127: cannam@127: added code for icc's _mm_malloc (memalign replacement) cannam@127: cannam@127: commit d1398d4a205bae587e99b40049dd6a301c1f58f9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 28 17:33:07 2002 -0400 cannam@127: cannam@127: slight fixes cannam@127: cannam@127: commit 5f21f0a04a242174ff85c63925c88e15e6ff101e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 28 16:28:43 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit d5256b19914cddf9b241ebce04f10042f4837e9b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 16:13:19 2002 -0400 cannam@127: cannam@127: Use vec_xor to change sign cannam@127: cannam@127: commit ec0a29c8d03cbed27c09a96fcb3f022bfc9f647f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 28 16:10:59 2002 -0400 cannam@127: cannam@127: added rdft2 cannam@127: cannam@127: commit 516c9c1117a0811dba416bfa1ba20a5c93e91532 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 15:45:54 2002 -0400 cannam@127: cannam@127: Optimized cannam@127: cannam@127: commit 8619a2039d6723004a1fef760203e5d6f33f9469 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 15:11:14 2002 -0400 cannam@127: cannam@127: Changed ALIGNMENT cannam@127: cannam@127: commit 37c7c1fd79835b212e94e565fb3abe0352243919 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 15:09:40 2002 -0400 cannam@127: cannam@127: alignment := 8 cannam@127: cannam@127: commit 27b891f615494d0f1996ed1acf0665eb386704e6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 14:57:22 2002 -0400 cannam@127: cannam@127: Avoid warning cannam@127: cannam@127: commit fd53f4d5bdfc8c8daf0126fab34270b09cf49aa2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 14:53:03 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 8b749b4406276bb4b4d6e3b4c78486d0fea38fff cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 14:50:09 2002 -0400 cannam@127: cannam@127: New altivec experiment cannam@127: cannam@127: commit 87bd001083f039c6728a3a19d03b7e14eac11666 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 13:48:20 2002 -0400 cannam@127: cannam@127: Nothing cannam@127: cannam@127: commit 3a5876fd4582a075560988801d7c958b0ca75a74 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 13:47:50 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 551ad6c0e199fcd5fce5defd470ce7d975dcacb4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 13:44:28 2002 -0400 cannam@127: cannam@127: Nothing cannam@127: cannam@127: commit a13f42aa3c37a3065a08a59220529d5292683ac9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 10:38:10 2002 -0400 cannam@127: cannam@127: Constants are now in separate file. cannam@127: cannam@127: commit 94226e68396c790ce6bfbbf8db0c299fed32e2f9 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 07:58:37 2002 -0400 cannam@127: cannam@127: More precise comment cannam@127: cannam@127: commit 4009a4d5579eb5520346c956632ef0c2df5273d2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 28 07:56:40 2002 -0400 cannam@127: cannam@127: gcc-3.1 bug workaround cannam@127: cannam@127: commit 092830f99bf3fd15390980b4e441d4c7d1a9826c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 28 01:39:54 2002 -0400 cannam@127: cannam@127: slight optimization, and exported zerotens functions cannam@127: cannam@127: commit e3797dbb5984f5f1272b452c7005c775badb6fb2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 28 00:54:59 2002 -0400 cannam@127: cannam@127: should be a plan_dft, not a plan_rdft cannam@127: cannam@127: commit ab69981af2f5c80981e7f1432560cbdaae08770c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 27 21:36:46 2002 -0400 cannam@127: cannam@127: Optimizations. Make it work with vanilla non-Apple gcc. cannam@127: cannam@127: commit 9a7ad02a4bfebfa91a4afe01756023a3a74f5d8b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 27 19:20:09 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 4aac8a4d98395b964b16b1251d8d52410fc232f8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 27 18:54:01 2002 -0400 cannam@127: cannam@127: added hc2r (dif) cannam@127: cannam@127: commit abe907208a2a3e7ff558b3f12bb0b254768d670a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 27 18:31:43 2002 -0400 cannam@127: cannam@127: add hc2r (dif) case cannam@127: cannam@127: commit b933474c3373bdca65dd9cce3b16272c2b197ee8 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 27 15:09:40 2002 -0400 cannam@127: cannam@127: Altivec port cannam@127: cannam@127: commit 0884acf4e8fc2cd9ec4144877e5a0879bbf779e6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 27 15:06:21 2002 -0400 cannam@127: cannam@127: Fixed signed/unsigned bug. cannam@127: cannam@127: commit 11508c3160c5d3a404a58eb143139d9088a213e5 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 25 20:11:26 2002 -0400 cannam@127: cannam@127: Make rank0 unapplicable to in-place problems. cannam@127: cannam@127: commit 81a49b1e405be525a9ee5476ddfa16e8c70ef702 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 25 17:10:52 2002 -0400 cannam@127: cannam@127: only works for r odd cannam@127: cannam@127: commit 2b54747fb0e87bbd03b3c7b04ed1cb752a470796 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 25 15:30:06 2002 -0400 cannam@127: cannam@127: Reinserted much better timing-avoidance heuristic cannam@127: cannam@127: commit 171716115f0f318397186964ecc341ac9268fd84 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 25 15:21:13 2002 -0400 cannam@127: cannam@127: Score is now a property of the plan, not of the solver. cannam@127: Revised representation of closures. cannam@127: cannam@127: commit 67c69e319a7ca8ac6c81a45a1d0f6dde9efc2e12 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jul 25 06:36:51 2002 -0400 cannam@127: cannam@127: Cosmetic changes. Added hc2r_128.c cannam@127: cannam@127: commit 0a22b8dd9629f62d1a682af581c17d6dc71e244a cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 25 01:37:53 2002 -0400 cannam@127: cannam@127: added hc2r cannam@127: cannam@127: commit 22bad3aea85c62120134db4652c6ac990c8607e2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 25 00:51:45 2002 -0400 cannam@127: cannam@127: added hc2hc-difbuf cannam@127: cannam@127: commit aac8e9d03008ccbe1c244717e404e283c03eabe1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 25 00:25:06 2002 -0400 cannam@127: cannam@127: added rdft-dif cannam@127: cannam@127: commit 39d632acade375e06e60dc11cd0b693ed29bbf07 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 25 00:22:36 2002 -0400 cannam@127: cannam@127: whoops, hc2r must be conjugated to have right sign cannam@127: cannam@127: commit ebc9e7b4083f1d545cc47032a7bffbcc5d5a26ce cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 23:27:45 2002 -0400 cannam@127: cannam@127: slight change cannam@127: cannam@127: commit 6c5a0b11d3a86a418e02108a90472ff19d97bae0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 23:24:24 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 28adebe469b82ee53e436f33389b459d8707a603 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 22:46:39 2002 -0400 cannam@127: cannam@127: support hc2r codelets cannam@127: cannam@127: commit af7b3ec85871349e26698fb5edf95c6a1e96bbbf cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 22:01:53 2002 -0400 cannam@127: cannam@127: use vector plan for r/i instead of two separate plans cannam@127: cannam@127: commit b31e3e7d86ef1ab3aa58145768cc801979ba5cd6 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 20:36:34 2002 -0400 cannam@127: cannam@127: hack to allow rader/generic to work in-place for small prime sizes, instead of always using buffered cannam@127: cannam@127: commit cddf15b3b7c1d3baec98982550f18344c3361216 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 18:04:41 2002 -0400 cannam@127: cannam@127: added rdft-generic cannam@127: cannam@127: commit 76637f738e056d7e4fcba907ffd4ab52db457fed cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 17:27:34 2002 -0400 cannam@127: cannam@127: fixed add count cannam@127: cannam@127: commit 7c1f6a8f3b35a5034daacc521a10c06424144047 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 14:52:26 2002 -0400 cannam@127: cannam@127: again cannam@127: cannam@127: commit ab910c9e4a7fc66e0a19e1b9557669e896ac465b cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 14:51:58 2002 -0400 cannam@127: cannam@127: slight fix cannam@127: cannam@127: commit 2169c91de93a2c096765218e2b25e32e6f2d47f0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 14:51:07 2002 -0400 cannam@127: cannam@127: fixed comment cannam@127: cannam@127: commit b6ed79694396f04555b0009027b94355c81a4019 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 14:41:24 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 10fabba80f177e1ee4bfca04ac09836c798998ef cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 14:38:15 2002 -0400 cannam@127: cannam@127: added rader-hc2hc cannam@127: cannam@127: commit 3015fea221f119cf88e68c12087c0ca8fbb508a9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Wed Jul 24 00:07:59 2002 -0400 cannam@127: cannam@127: whoops, initialize W cannam@127: cannam@127: commit d48486c4715a0db6bb2653a34d868f5f52732f66 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 23:03:09 2002 -0400 cannam@127: cannam@127: strides should not be unsigned cannam@127: cannam@127: commit 7d6e7cacd21c97ef1622d681de2543e71ac2171d cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 23:02:08 2002 -0400 cannam@127: cannam@127: more stride sign fixes cannam@127: cannam@127: commit b967fadc107addb8cec4effc1f0e7ae7d6ce1f86 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 23:01:04 2002 -0400 cannam@127: cannam@127: strides should not be unsigned! cannam@127: cannam@127: commit 0ad85517c669d39fcf0ac6f77e73ed8c2fa80e89 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 14:55:25 2002 -0400 cannam@127: cannam@127: added comment cannam@127: cannam@127: commit 5d278e1ac3640bc39cd6b7e19aaa5563cd319de4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 14:52:04 2002 -0400 cannam@127: cannam@127: another fix to op count cannam@127: cannam@127: commit 9260aed8161a66eb5de14e68c932d11bc113cd56 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 14:51:01 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 3f42b7510d2c0f2b1e7bc34342041f8123667897 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 14:49:43 2002 -0400 cannam@127: cannam@127: slight fix to op counts cannam@127: cannam@127: commit 1288dec288612070c531c98067255cf3de3d90b1 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 14:09:19 2002 -0400 cannam@127: cannam@127: added dft-r2hc cannam@127: cannam@127: commit ad4bf834d8b55b38d2766779e5d00c4f61e30dbe cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 02:50:12 2002 -0400 cannam@127: cannam@127: better comment and var. name cannam@127: cannam@127: commit f1ab8ef1b9cf77432f6bb627a5c3ec2f586ebcd9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 02:39:11 2002 -0400 cannam@127: cannam@127: fixed tests for hc2r, and added r2hc-hc2r cannam@127: cannam@127: commit d3b91945fd199f6bb99711479972b7074c00b352 cannam@127: Author: Steven G. Johnson cannam@127: Date: Tue Jul 23 00:45:23 2002 -0400 cannam@127: cannam@127: added rader-dht cannam@127: cannam@127: commit 57036068d38970156c0bcf5d4edd72cdb20a09fd cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 22 21:05:12 2002 -0400 cannam@127: cannam@127: Added r2hc_128, what the hell. cannam@127: cannam@127: commit d82c1c99be202e2cc55851a4bd406b4682cb0b4a cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 22 20:48:59 2002 -0400 cannam@127: cannam@127: Added codelets that compute twiddle factors cannam@127: cannam@127: commit f98ad798168562c0da5714110eb0c37972178728 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 19:57:16 2002 -0400 cannam@127: cannam@127: added rdft-buffered cannam@127: cannam@127: commit 50b0158abe60a9e62698baf54e31623bf29a34f9 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 19:43:39 2002 -0400 cannam@127: cannam@127: added hc2hc-ditbuf cannam@127: cannam@127: commit 4b7abfd7514cb4d98a0c87746c25fcafe6d263b3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 14:29:04 2002 -0400 cannam@127: cannam@127: use STACK_MALLOC (alloca), since generic radix is always small cannam@127: cannam@127: commit d083d389c40a363c4b90a6ca2efd202c52b81713 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 14:22:43 2002 -0400 cannam@127: cannam@127: small cleanup cannam@127: cannam@127: commit 851d792b2de11df3620f32093e02632f78aeef6e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 22 07:42:13 2002 -0400 cannam@127: cannam@127: What the hell was I thinking? cannam@127: cannam@127: commit 7237f72026e6aad4325427a52b0fb683ec4b2e0d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 22 07:37:12 2002 -0400 cannam@127: cannam@127: Reduced code size by using table instead of switch statement. cannam@127: cannam@127: commit f253821d2c79215c87e18cf134e218e02d0235ed cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 22 07:27:06 2002 -0400 cannam@127: cannam@127: Changed hash function to avoid collisions with DFT. cannam@127: cannam@127: commit 602ef947b9122139e2b55fca3e007ca6bcda4bbf cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 01:37:06 2002 -0400 cannam@127: cannam@127: added missing file, whoops cannam@127: cannam@127: commit 6b3144d456eb3c0caee53880e7fe60ddbd2c48d5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 01:24:17 2002 -0400 cannam@127: cannam@127: whoops, generate enough twiddles for odd m cannam@127: cannam@127: commit 4738a6cbbc5206c3fdc7b0bf7cdc481609439497 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 01:10:21 2002 -0400 cannam@127: cannam@127: don't try to verify R2HCII or HC2RIII plans cannam@127: cannam@127: commit ec9e9517ca4ac29008a9b1b8f79f4543ef4ae90a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 01:05:00 2002 -0400 cannam@127: cannam@127: recursive case now works, I think cannam@127: cannam@127: commit 7ebf4c56ae4cc7861840cb8ee5d8a482c5e3f64a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 01:04:40 2002 -0400 cannam@127: cannam@127: add extra impulse test for debugging cannam@127: cannam@127: commit 7dacfd5778747c8ae3b803ddf37d7921eeab713f cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 01:02:38 2002 -0400 cannam@127: cannam@127: whoops, multiply ios offset by stride (and rename to ioffset) cannam@127: cannam@127: commit ca3c5bf3c6de8946f1caf40e779487110e5b59ce cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 22 00:22:02 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit e40381e0407f8570c076968fb7c7138ffebe9ae2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 23:58:14 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 030d0f7f52cbc000070d885b815768bfadb86649 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 23:43:03 2002 -0400 cannam@127: cannam@127: added hc2hc-dit cannam@127: cannam@127: commit c1c28b632b9cc24c542610dbdb3bf424eb740810 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 23:15:12 2002 -0400 cannam@127: cannam@127: twiddles can be shared with smaller m's cannam@127: cannam@127: commit cbb0b11b1a8aa023f1d81dba688278012451de8e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 22:34:28 2002 -0400 cannam@127: cannam@127: preparing for recursive rdft... cannam@127: cannam@127: commit 8f48e0e3caf86690c7328cd128cc981364c9026f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 19:31:22 2002 -0400 cannam@127: cannam@127: slight fix, to match libbench/verify.c cannam@127: cannam@127: commit d9aec187c49dbc272df30d040d4acfc160220b07 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 18:43:12 2002 -0400 cannam@127: cannam@127: r2hcII has imag parts offset by n-1, not n. We can also allocate fewer strides. cannam@127: cannam@127: commit 00e3acce93c910450482c37155200244adfc51b4 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 18:27:09 2002 -0400 cannam@127: cannam@127: delete unused var cannam@127: cannam@127: commit 469254570eda6c6851c8c76ac2ce631c7e160d40 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 02:06:53 2002 -0400 cannam@127: cannam@127: added some rdft solvers cannam@127: cannam@127: commit c267ad079a4ef7cb7a9fdbe4556c89121137be02 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 01:52:54 2002 -0400 cannam@127: cannam@127: pass identifier in FFTW() through another macro so that the mangled name cannam@127: can itself be a preprocessor symbol cannam@127: cannam@127: commit cf660c4cf10b80d7ec37cd99825c9663738d77e7 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 01:05:21 2002 -0400 cannam@127: cannam@127: fix in comment cannam@127: cannam@127: commit 14081a9d216ccc757b9ce46631d956f5135628ea cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 00:47:03 2002 -0400 cannam@127: cannam@127: bench tests rdft plans cannam@127: cannam@127: commit 61ed41f792df937cc80b9fa0c643871ab7449968 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 00:22:14 2002 -0400 cannam@127: cannam@127: make rdft.h and dft.h compatible cannam@127: cannam@127: commit d314a5c84e70803b32075884ee96c0099c695d9a cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 21 00:12:19 2002 -0400 cannam@127: cannam@127: first-draft rdft verify cannam@127: cannam@127: commit 01403979be858145b4f1f61f03c9f1f98c59587d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 20 22:09:15 2002 -0400 cannam@127: cannam@127: got rid of annoying warnings cannam@127: cannam@127: commit 710e4e4687092d0d823f7fe442c0bc981d99a598 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 20 22:07:37 2002 -0400 cannam@127: cannam@127: added stub codelet registration for linking purposes cannam@127: cannam@127: commit 9b9775415f67f53910d95e1ac963f1bed389ff9b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 20 21:46:03 2002 -0400 cannam@127: cannam@127: basic rdft stuff cannam@127: cannam@127: commit 9c7a553bedb1f7f2fce816ae284d4867ffc0924d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 20 21:06:50 2002 -0400 cannam@127: cannam@127: rdft codelets now compile cannam@127: cannam@127: commit fb7686cdfd1674f03c35ed523fcc2d11db157ecb cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 20 18:40:31 2002 -0400 cannam@127: cannam@127: Oops, was generating rdfts instead of hdfts cannam@127: cannam@127: commit f82dcb63a673b994a4677ed9f8d065766c79c31c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 20 18:25:47 2002 -0400 cannam@127: cannam@127: Added hc2r codelets cannam@127: cannam@127: commit d8127083d80d0f0b9de30e6e3c9ae1b90f92a90d cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 20 17:54:39 2002 -0400 cannam@127: cannam@127: return W in hc2hc codelets cannam@127: cannam@127: commit 354e28470103a92db21d621263a687a6bf437595 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jul 20 17:51:06 2002 -0400 cannam@127: cannam@127: Some work on rdft codelets cannam@127: cannam@127: commit 8a7b5a3242b8bd823c8d70e1b04e9492d6d65d43 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 16 13:55:50 2002 -0400 cannam@127: cannam@127: fix const cannam@127: cannam@127: commit 6a3576889b8683e9ee15d2f95fb76f6fef645667 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 16 07:00:10 2002 -0400 cannam@127: cannam@127: Separate CFLAGS in codelets. Fix const in certain places. cannam@127: cannam@127: commit 7870f6dff402e655def77265bc4ac0225608f677 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 21:10:42 2002 -0400 cannam@127: cannam@127: note buffering problem cannam@127: cannam@127: commit 0bd6af885007a6f0561577c521d4999c903f27df cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 15 20:27:51 2002 -0400 cannam@127: cannam@127: Removed unpredictable branch from inner loop cannam@127: cannam@127: commit 8a40f059239add905fa4c0abd6c20a40644559fa cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 19:35:04 2002 -0400 cannam@127: cannam@127: update cannam@127: cannam@127: commit 45bb25aa64ce46c8821c9717770a28a5ab60e492 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 19:31:39 2002 -0400 cannam@127: cannam@127: optimization cannam@127: cannam@127: commit 526958106e6a43bfc1d4a7cab335fc3df41a7d9f cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 19:28:30 2002 -0400 cannam@127: cannam@127: added generic dit cannam@127: cannam@127: commit aebc10cb69074f33b3370de5aff7bc20b684800b cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 17:03:53 2002 -0400 cannam@127: cannam@127: whoops, mksolver should be static cannam@127: cannam@127: commit e2b6303fa6575e6796c2834f222b77d221e1a77e cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 15 16:46:36 2002 -0400 cannam@127: cannam@127: First implementation of gen_hc2hc, probably still buggy. cannam@127: cannam@127: commit 0105f03531806f86cc1c0e89c18b91947b15cb10 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 16:40:23 2002 -0400 cannam@127: cannam@127: don't count loading of twiddle factors in ops.other, since it isn't cannam@127: counted for the codelets cannam@127: cannam@127: commit 47f3220441f5da7ee844e0abd36f41f32b4bc17e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 15:13:19 2002 -0400 cannam@127: cannam@127: plan_destroy puts plan to sleep before deallocating it, to eliminate duplicate free calls in solvers cannam@127: cannam@127: commit 90347b282680ec2b611ededef7ab7272beb2857a cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jul 15 15:07:41 2002 -0400 cannam@127: cannam@127: fftw2-like vector recursion flag cannam@127: cannam@127: commit df45d5d1588019d57e80aee326c92e5ffb49715f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 14 21:01:44 2002 -0400 cannam@127: cannam@127: More jokes cannam@127: cannam@127: commit 5efd22c7d5499d7bad84216e526ec11551fb81f0 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 14 20:36:01 2002 -0400 cannam@127: cannam@127: Bless plan for testing purposes cannam@127: cannam@127: commit dd45761d063a5473473f44d5800a55b35794c8a6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 14 20:35:49 2002 -0400 cannam@127: cannam@127: Canonical linked-list deletion (hope it is right) cannam@127: cannam@127: commit 04cbcbfe2eb171da6ee678d000f1cf7aa2435f5d cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 18:26:19 2002 -0400 cannam@127: cannam@127: use estimating planner for cld_omega cannam@127: cannam@127: commit c2e125a60dc8101c25c1f08debd9a4b1661b1658 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 18:10:56 2002 -0400 cannam@127: cannam@127: better internal naming cannam@127: cannam@127: commit fac5147b9b14fe2997cde8bbd5a39c956f577eaf cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 18:10:01 2002 -0400 cannam@127: cannam@127: printing should really be fixed now, grrr cannam@127: cannam@127: commit 8dbd07648153ea12bd52c12aad39b58adc479140 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 17:57:12 2002 -0400 cannam@127: cannam@127: print all distinct child plans cannam@127: cannam@127: commit af0c968546d0c39197e3c7925e21bf1897f1b5ab cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 17:49:21 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit af601a5405861e68cedd4314f70b677b6c36e640 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 17:45:54 2002 -0400 cannam@127: cannam@127: whoops, destroy should delete twiddle/omega from list cannam@127: cannam@127: commit a20712e3a4b5d2364f092fd222b540cbc8c2df44 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 17:33:02 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit a43e6c2aeb20b20987891fa7461cc6a2898d1785 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 17:12:14 2002 -0400 cannam@127: cannam@127: added plan_bless and FORGET_ACCURSED cannam@127: cannam@127: commit 8da186b0e85df747bbd0a91db772c869e9b35e3c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 16:15:43 2002 -0400 cannam@127: cannam@127: further cleanup cannam@127: cannam@127: commit 255479f4ad3175722fe32fd4a2b7cefa858b3b35 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 16:14:15 2002 -0400 cannam@127: cannam@127: slight cleanup cannam@127: cannam@127: commit d5346f1dfc7154d9a4fbade8fba1dcce90c7bec5 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 16:09:17 2002 -0400 cannam@127: cannam@127: added traverse_plan via print (ugh) cannam@127: cannam@127: commit 1edcc5b2fb3cf8741aec7b32042453803a1eb145 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 15:08:29 2002 -0400 cannam@127: cannam@127: added TW_FULL, and additional n parameter for twiddles cannam@127: cannam@127: commit c300c4c9e119ee5e657fe52fa48ce8251321f8a2 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 15:03:51 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit 8eb08032b56dac1d0b4200c2a1e17e6a33674395 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jul 14 13:49:20 2002 -0400 cannam@127: cannam@127: save flags before invoking solver mkplan cannam@127: cannam@127: commit 68d1b66d54458eb685bee1a95bd4433421a68f58 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 14 09:28:37 2002 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit 8f335f455b54a35089073c690ecd93c7380c1b95 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 22:17:29 2002 -0400 cannam@127: cannam@127: added support for UNICOS _rtc() real-time-clock intrinsic function cannam@127: cannam@127: commit e0550777d2519623392bd1678c39e7491fd3d38c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 22:06:35 2002 -0400 cannam@127: cannam@127: fixed typo: HAVE_TIME_H should include time.h, not sys/time.h cannam@127: cannam@127: commit fcff09d063384ac24b87c16cfed1c246de45623f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 21:46:02 2002 -0400 cannam@127: cannam@127: support AIX read_real_time timer cannam@127: cannam@127: commit ca89737634be3d5de4851c1f3fbc19d900cf22b0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 17:02:51 2002 -0400 cannam@127: cannam@127: use && instead of the (sigh) unportable -a cannam@127: cannam@127: commit 769cf9267df8b75c3e2849a171e995136bacd4af cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 16:38:18 2002 -0400 cannam@127: cannam@127: use AC_HELP_STRING cannam@127: cannam@127: commit 6600ee1ae97f1919117e4c3877092249443c545b cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 16:05:43 2002 -0400 cannam@127: cannam@127: support long-double precision cannam@127: cannam@127: commit d7aff35e1553f8735b56597fd524c5b52d8e475f cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 15:48:10 2002 -0400 cannam@127: cannam@127: whoops whoops cannam@127: cannam@127: commit 9a20964a145ceef9018cf8bf0977be7ba63ecb6e cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 15:47:39 2002 -0400 cannam@127: cannam@127: whoops cannam@127: cannam@127: commit d040c7ef353abc5accf76a9953a26feb2d11fc0c cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 14:13:42 2002 -0400 cannam@127: cannam@127: buffered solver strides have been fixed cannam@127: cannam@127: commit 6bcbee663a0b8b8b23b70a180e1ca12ee1141724 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 13:48:13 2002 -0400 cannam@127: cannam@127: convention cannam@127: cannam@127: commit 4d3d49e4b447b49a45b803fea4ff4d23a31288a0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 12:50:06 2002 -0400 cannam@127: cannam@127: share twiddle arrays in Rader cannam@127: cannam@127: commit 91dbf0b319de38c0b67df70aa4c39ccac0b523da cannam@127: Author: Steven G. Johnson cannam@127: Date: Sat Jul 13 12:48:10 2002 -0400 cannam@127: cannam@127: call done() after verify cannam@127: cannam@127: commit b5b07111cda5f2b5b1130829d05b698575f4a5f8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 15:42:04 2002 -0400 cannam@127: cannam@127: output planner time with -v cannam@127: cannam@127: commit b71bd73584d1e960018fbda1b8f078fa4e2ea542 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 15:40:14 2002 -0400 cannam@127: cannam@127: support double outputs cannam@127: cannam@127: commit e808db8fbfe2b7d4afbabe558d5a757379c49dd8 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 15:09:19 2002 -0400 cannam@127: cannam@127: removed extraneous parens cannam@127: cannam@127: commit b1ca74610947a0007932fb0eb65c794868f15977 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 15:08:13 2002 -0400 cannam@127: cannam@127: increase maxbufsz to 64k; makes a big difference for large 2d transforms cannam@127: cannam@127: commit 92dfa732c944f37774e1b4b9b889ba64a7621ccb cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 12 05:59:26 2002 -0400 cannam@127: cannam@127: Fix cannam@127: cannam@127: commit fdae83d7830d6df98ac417066e0c5ea8bc254d35 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 01:22:38 2002 -0400 cannam@127: cannam@127: fix comment cannam@127: cannam@127: commit 8b316634377ad2f829c26d6d107005638c6ab8ba cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 00:59:29 2002 -0400 cannam@127: cannam@127: fix in comment cannam@127: cannam@127: commit efb8ce2f3a2e518f02245f8eb12425a30bb420c0 cannam@127: Author: Steven G. Johnson cannam@127: Date: Fri Jul 12 00:13:13 2002 -0400 cannam@127: cannam@127: updated cannam@127: cannam@127: commit 21a94bd1783b4cde2728d54932cdf1ecf2485a21 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 11 23:39:27 2002 -0400 cannam@127: cannam@127: buffered malloc's buffers cannam@127: cannam@127: commit 2cec064ce9f3fd0ccd891651557a5739409c19c3 cannam@127: Author: Steven G. Johnson cannam@127: Date: Thu Jul 11 23:30:26 2002 -0400 cannam@127: cannam@127: share more code between apply and apply_dit in Rader cannam@127: cannam@127: commit afd5fe37e6c3060145530115b6c2f2d676ddfe37 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 8 12:30:34 2002 -0400 cannam@127: cannam@127: Polished cannam@127: cannam@127: commit 71ae7be079984537f7770d1b70280b77ad23c722 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 8 09:47:11 2002 -0400 cannam@127: cannam@127: [empty commit message] cannam@127: cannam@127: commit a5760afe5aed6838383937fad0a3243528ce15fc cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 8 09:42:08 2002 -0400 cannam@127: cannam@127: SIMD/FMA stuff cannam@127: cannam@127: commit ec76a60088a86df970b3cbf4005506ade4570040 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 8 07:43:51 2002 -0400 cannam@127: cannam@127: Avoid code duplication cannam@127: cannam@127: commit ffce0587abc26960f1bffb08b99f61280176d25b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 7 20:56:15 2002 -0400 cannam@127: cannam@127: Fixes for FMA+SIMD cannam@127: cannam@127: commit cd1278e203d5014ee55026d00eef0c5cd87091a7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jul 7 20:32:01 2002 -0400 cannam@127: cannam@127: Major changes in SIMD fftw cannam@127: cannam@127: commit 47b31e4f895a8385d29297182fd4ab3cbe756486 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 5 17:32:09 2002 -0400 cannam@127: cannam@127: Use unpck instructions instead of shuffles cannam@127: cannam@127: commit 8390c4b680fa05c264d6308d70aeb8b69e97b44a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 5 15:49:14 2002 -0400 cannam@127: cannam@127: Minor tweaks cannam@127: cannam@127: commit 9939d14df8ec43f0f3724eccce6907f723ae7bcd cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 5 15:02:54 2002 -0400 cannam@127: cannam@127: Use score planner cannam@127: cannam@127: commit db780c34509c8cc70385f63815249dbb982371ab cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jul 5 14:49:59 2002 -0400 cannam@127: cannam@127: Added verifier cannam@127: cannam@127: commit a4c35fbcefcca25a0e31431dcdabb0d44a2bfb98 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jul 3 20:32:28 2002 -0400 cannam@127: cannam@127: More simd codelets cannam@127: cannam@127: commit 0b41b3e8c38d89cca050b1b0df6110056a55463e cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 2 16:18:09 2002 -0400 cannam@127: cannam@127: Oops cannam@127: cannam@127: commit 4aa14927597947c2f2c0c38064e92ce29850f3eb cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 2 16:13:24 2002 -0400 cannam@127: cannam@127: Fixed classic mode cannam@127: cannam@127: commit b32c4fa8b6fddb6a4af23a7d2794adb53733fc2d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 2 15:38:36 2002 -0400 cannam@127: cannam@127: Use LDK for constants so that we can play games. cannam@127: cannam@127: commit 38505faa2d20e4d958d80dce05620dbf20420822 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 2 13:15:58 2002 -0400 cannam@127: cannam@127: Improved support for fixed strides cannam@127: cannam@127: commit 2c99260bbd5e86361b09120817f07543384fc5e0 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 2 10:30:58 2002 -0400 cannam@127: cannam@127: Changed accounting of flops cannam@127: cannam@127: commit ae3999cb3d0ec0d5140c9dead499de0cf4318e5e cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jul 2 08:51:38 2002 -0400 cannam@127: cannam@127: Wrong code in non-fma mode cannam@127: cannam@127: commit 55015bd78bfbcbffb833554b7df558efd1a826cc cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 1 23:17:06 2002 -0400 cannam@127: cannam@127: sse2 stuff cannam@127: cannam@127: commit 021b59968903662e9727f7833c19c613f1b192b7 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 1 14:05:56 2002 -0400 cannam@127: cannam@127: Identify CPUs for special codelets cannam@127: cannam@127: commit f304f0151ac1ad333b9450a6d78c8cd8f8724d1a cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 1 09:26:42 2002 -0400 cannam@127: cannam@127: Change split problem syntax cannam@127: cannam@127: commit 48a37449fcbd51779bb3dde3f3b8c2d02bbc323d cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jul 1 09:11:39 2002 -0400 cannam@127: cannam@127: Removed -fma flag cannam@127: cannam@127: commit 38d63d20ab1bd900f5f89914ca7b7f90191d866f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 30 20:08:26 2002 -0400 cannam@127: cannam@127: Work around gcc bug cannam@127: cannam@127: commit 6b96cea114710b35caa1d65a669f92a46e20f27f cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 30 18:34:06 2002 -0400 cannam@127: cannam@127: New simd stuff cannam@127: cannam@127: commit 5dd26fcef247d6776f4b5b587b15a1b317a97431 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 30 17:00:09 2002 -0400 cannam@127: cannam@127: Added altivec support cannam@127: cannam@127: commit a81146d09b6762b1c82b9f998720454245ee5e25 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 30 14:47:47 2002 -0400 cannam@127: cannam@127: Forgot file cannam@127: cannam@127: commit 734f270fddd3a03066eb4323706090f99c82bf9c cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 30 14:37:55 2002 -0400 cannam@127: cannam@127: Progress towards simd implementation cannam@127: cannam@127: commit 05dc59aa43296d7fb6b79edf5decf65e9a100401 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 25 20:23:29 2002 -0400 cannam@127: cannam@127: Add 128- codelet cannam@127: cannam@127: commit a94cde7e798f7da510f20b25d1a4bd4aea4dc58b cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 20:47:28 2002 -0400 cannam@127: cannam@127: More simd changes. Ensure proper stack alignment in k7 codelets. cannam@127: cannam@127: commit c4ef1c53884a0a7fee4587ef0b1f4317ae83a644 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 13:01:33 2002 -0400 cannam@127: cannam@127: Fixed prototypes cannam@127: cannam@127: commit 5383095f6dd63b1db544695c4b0f7f244344d698 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 12:53:26 2002 -0400 cannam@127: cannam@127: Sparc cycle counter requires v9 cannam@127: cannam@127: commit a25e9ee31d663307683038fbf703f0ed60a502e3 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 11:45:48 2002 -0400 cannam@127: cannam@127: Minor fixes cannam@127: cannam@127: commit 01922e916a7428cc0f7f188518720710b101e9e6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 11:11:46 2002 -0400 cannam@127: cannam@127: Fixed ev67 detection cannam@127: cannam@127: commit ff1e337f46f9f1bbdcc32756bcdaa0e86566df2e cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 10:19:11 2002 -0400 cannam@127: cannam@127: Print flops cannam@127: cannam@127: commit 3f2d94c9ccabcf93d7f8b1230c5b2f97ac6cc4d2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 22 09:36:47 2002 -0400 cannam@127: cannam@127: Nothing really cannam@127: cannam@127: commit 8843a01025a26b7bb5a3f26b94c46243332cca25 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 21 22:19:20 2002 -0400 cannam@127: cannam@127: More simd work cannam@127: cannam@127: commit 3a71fc73cd6c17ba8df8d7aad29fa1bf5ad71a96 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 20 21:22:41 2002 -0400 cannam@127: cannam@127: More simd work cannam@127: cannam@127: commit fcbb846544687bd8c14e9d004746230e8b6a49e2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 20 18:51:33 2002 -0400 cannam@127: cannam@127: More simd work cannam@127: cannam@127: commit 4e25f887266e572b5418dcd0bad8db69265ba3b3 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 20 15:04:37 2002 -0400 cannam@127: cannam@127: Moving towards incorporation of simd stuff cannam@127: cannam@127: commit 8b98bf67f5e3c976f17988acc8c2e063d3848408 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 19 18:47:55 2002 -0400 cannam@127: cannam@127: Reorganized k7 stuff into own directory cannam@127: cannam@127: commit 30f0669ed255025d8e4bad6adb9737022a7a6b34 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 19 13:21:13 2002 -0400 cannam@127: cannam@127: Minor experimental stuff cannam@127: cannam@127: commit 51b0055df9ca6f4c9ea05ccbeb4714bd61d8469b cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 19 11:20:29 2002 -0400 cannam@127: cannam@127: Cosmetic changes cannam@127: cannam@127: commit 3309bd7a80d6dfb0da60d6c2a185d5213505b036 cannam@127: Author: fftw cannam@127: Date: Wed Jun 19 01:43:31 2002 -0400 cannam@127: cannam@127: allocate buffers on the fly cannam@127: cannam@127: commit d290c98ff9fdac1646b56bf9fd80913ecf4b42ed cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 17:48:41 2002 -0400 cannam@127: cannam@127: Added ct-ditbuf-k7.c . Major changes required in generator. cannam@127: cannam@127: commit c672bc817602bb03a8e6d7b637e78e84dc2f1f1d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 11:55:57 2002 -0400 cannam@127: cannam@127: Nothing, really cannam@127: cannam@127: commit 327d908d5d396a22fa85216b24203c6b709f9379 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 11:19:59 2002 -0400 cannam@127: cannam@127: !SINGLE ==> !K7_MODE (for some reason the contrapositive sounds wrong) cannam@127: cannam@127: commit fc34a6a92882caf5e8346225e9c25a2a1b6fcb2c cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 11:07:13 2002 -0400 cannam@127: cannam@127: Buffer is now symmetric wrt forward/backward transform cannam@127: cannam@127: commit 2dfa3580ad3c4d0960a3b18aba5464c6d9cc91c1 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 10:33:58 2002 -0400 cannam@127: cannam@127: Fixed applicable() in indirect.c cannam@127: cannam@127: commit 40e9e2373f160b3760c87aa707e20aa91a30479d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 08:41:18 2002 -0400 cannam@127: cannam@127: Fixed attempt to free() uninitialized pointer. cannam@127: cannam@127: commit 854771dbd1a262127ab6ef87b31e6ec3645dbce4 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 18 08:13:55 2002 -0400 cannam@127: cannam@127: Added reference counts for awake() cannam@127: cannam@127: commit b65907ea5ba6b84347976ae59688acbc4a982b82 cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jun 17 20:49:05 2002 -0400 cannam@127: cannam@127: updated comment cannam@127: cannam@127: commit 11cbdda3f3c4e688e59eb28d43946c6053ce461e cannam@127: Author: Steven G. Johnson cannam@127: Date: Mon Jun 17 20:46:45 2002 -0400 cannam@127: cannam@127: slight update cannam@127: cannam@127: commit df79d1312e84ddc47b65056c2b59fb6eec20915b cannam@127: Author: fftw cannam@127: Date: Mon Jun 17 02:30:16 2002 -0400 cannam@127: cannam@127: moved prime-number stuff into primes.c, so it can be shared with generic codelet and with rfftw rader cannam@127: cannam@127: commit 92c3a4be474dca7d3e4f2076f2d5c61e040e6739 cannam@127: Author: fftw cannam@127: Date: Mon Jun 17 02:01:58 2002 -0400 cannam@127: cannam@127: added comment cannam@127: cannam@127: commit b580b3303f83589a5b87ad88f104ed41b8c641a3 cannam@127: Author: fftw cannam@127: Date: Mon Jun 17 01:39:55 2002 -0400 cannam@127: cannam@127: added rader-dit cannam@127: cannam@127: commit 51f015c4a6df06fa475e6e291d7e0f8a0293a891 cannam@127: Author: fftw cannam@127: Date: Sun Jun 16 23:50:16 2002 -0400 cannam@127: cannam@127: added initial Rader (no DIT yet) cannam@127: cannam@127: commit d45d1bbd591b1c30885baa7b1025ae016b23e937 cannam@127: Author: fftw cannam@127: Date: Sun Jun 16 22:29:51 2002 -0400 cannam@127: cannam@127: don't warn about long long cannam@127: cannam@127: commit 16ad72af138d08605df2fadf239e142ace2680eb cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 21:30:42 2002 -0400 cannam@127: cannam@127: Added k7 DIF codelets cannam@127: cannam@127: commit 3728b053dde52469a7deea4654ca0352c40bd6d4 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 19:13:31 2002 -0400 cannam@127: cannam@127: Added stuff to do cannam@127: cannam@127: commit 6eadf663f6d05339378e742de11484c6d46283a5 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 19:05:58 2002 -0400 cannam@127: cannam@127: Handle dual case R = I + 1 cannam@127: cannam@127: commit f6e99e6f7959f1e91fb66ff344ce372c75567e9b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 18:54:31 2002 -0400 cannam@127: cannam@127: Removed useless flag cannam@127: cannam@127: commit d38b5396624973e2ddc9531587b799eaa13f9daf cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 18:30:32 2002 -0400 cannam@127: cannam@127: Removed useless file cannam@127: cannam@127: commit 376194067d9decb0e20df12443a0ac64a16f67a2 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 18:30:18 2002 -0400 cannam@127: cannam@127: More k7 work. Switched to runtime CLASSIC mode. cannam@127: cannam@127: commit 331b9479423ed2ad02e75cb8a759241178460ede cannam@127: Author: Steven G. Johnson cannam@127: Date: Sun Jun 16 17:15:18 2002 -0400 cannam@127: cannam@127: spelling cannam@127: cannam@127: commit c278f9c639c8c850773474634acb51104857ad21 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 15:51:44 2002 -0400 cannam@127: cannam@127: Do not compile if not K7_MODE cannam@127: cannam@127: commit 375f66850303af4e6eb9c7263fe5c5c769876065 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 15:35:02 2002 -0400 cannam@127: cannam@127: Do not require K7 definitions to compile cannam@127: cannam@127: commit 684b95447b420d52337cd499589d31dce430ca4a cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 16 08:05:17 2002 -0400 cannam@127: cannam@127: More k7 stuff cannam@127: cannam@127: commit 535c1c74ca2099099e3edbab3aa7acafa02ea7d6 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 15 18:30:43 2002 -0400 cannam@127: cannam@127: Try to be compatible with automake-1.6 cannam@127: cannam@127: commit ba06164744e743ea9ea8dabb9359dd0e3b830c9c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 15 18:23:40 2002 -0400 cannam@127: cannam@127: More merging of Stefan's generator with main genfft branch cannam@127: cannam@127: commit fbe6e3f25188363b31caadf2a598be16eeb2a6a7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 15 13:51:39 2002 -0400 cannam@127: cannam@127: Slowly merging genfft-k7 with main genfft branch cannam@127: cannam@127: commit 8a567ee1d079e0cb9ba06e66a3f06d9be0eb34b0 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 21:33:02 2002 -0400 cannam@127: cannam@127: Fixed, really cannam@127: cannam@127: commit 6bd3b52c770b109a025d9982098066f91a84c937 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 21:27:12 2002 -0400 cannam@127: cannam@127: Oops... cannam@127: cannam@127: commit 5b6a71f941de2e28005b7821317ec7032c0b9743 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 21:25:34 2002 -0400 cannam@127: cannam@127: Work properly when $(ALL_CODELETS) = "" cannam@127: cannam@127: commit 07399bfca3c51c6e59dc5f7f9a0f1e7600ab2f41 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 21:11:16 2002 -0400 cannam@127: cannam@127: Fixed k7 build machinery cannam@127: cannam@127: commit 015e0a9b55cfef97ad18e06af166a22d02240d09 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 17:42:35 2002 -0400 cannam@127: cannam@127: More work on k7 stuff cannam@127: cannam@127: commit cf8b11788ed23e9e651cbadf1950302c67102f62 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 15:54:29 2002 -0400 cannam@127: cannam@127: More work on k7 stuff cannam@127: cannam@127: commit 8bd8bb064c77d442eb9da44432a3030b2b48f88a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 14:18:15 2002 -0400 cannam@127: cannam@127: Changed my mind again cannam@127: cannam@127: commit 1cfe47c1940dab749e1c63c15bfde4d0eed29495 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 11:53:09 2002 -0400 cannam@127: cannam@127: Removed some useless stuff. cannam@127: cannam@127: commit 315edea8537b5df64d2eb456cd20f12e16c50544 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 11:01:39 2002 -0400 cannam@127: cannam@127: Hmm... cannam@127: cannam@127: commit fb33fef8ea963d61445cfdef5dcf576cbd616cd8 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 10:28:12 2002 -0400 cannam@127: cannam@127: More work in preparation for k7 stuff cannam@127: cannam@127: commit 1efd1ce4b0d507eab8ff1b618a75bb66303b690d cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 07:25:28 2002 -0400 cannam@127: cannam@127: Still preparing to include k7 stuff cannam@127: cannam@127: commit 4c4195fba4ee1d73c35a4d8695a714d788a542cb cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 07:06:02 2002 -0400 cannam@127: cannam@127: Create .depend cannam@127: cannam@127: commit 4a55307470a6737b48d2cfb10be690ce60b7ac25 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 14 06:56:15 2002 -0400 cannam@127: cannam@127: Imported Stefan's K7 generator cannam@127: cannam@127: commit 61a7a73d101594961d9b63fb34fb2340635aaf1b cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 15:30:41 2002 -0400 cannam@127: cannam@127: Generator for real->halfcomplex and halfcomplex->real codelets cannam@127: cannam@127: commit d43342dc8c0aa4ac1802eff495ca96ac37c9fb22 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 11:54:02 2002 -0400 cannam@127: cannam@127: Improved hash functions, printers cannam@127: cannam@127: commit b26a2a40553deea06a47abfd3ba1f45d54b28cc2 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 11:17:31 2002 -0400 cannam@127: cannam@127: Only regenerate codlist.c in maintainer mode cannam@127: cannam@127: commit 39f18e53eaf87dc26e7001bccb2839fca11d1bf8 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 11:04:24 2002 -0400 cannam@127: cannam@127: Planner can export solution list cannam@127: cannam@127: commit 94b38d1890179198beac49ead7263d4c0a345a59 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 08:59:53 2002 -0400 cannam@127: cannam@127: Fixed for intel compiler cannam@127: cannam@127: commit bd4fbd5827358275240d991af98b0c477a4158cf cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 08:48:51 2002 -0400 cannam@127: cannam@127: Revised strategy for constants in codelets cannam@127: cannam@127: commit 9d1d76e5a32774ac2c7d9c43d2594032aeb20933 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 13 06:21:31 2002 -0400 cannam@127: cannam@127: Enable score planner in classic mode, naive planner in pro mode. cannam@127: cannam@127: commit 81ae7fad037e7325c65da97a3c842c220ea68a2d cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 12 19:18:18 2002 -0400 cannam@127: cannam@127: Report classic/pro cannam@127: cannam@127: commit 2c6576feda81b17c61be5678e97ca1a4db5f7935 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 12 19:07:48 2002 -0400 cannam@127: cannam@127: Fixed behavior of buffered solver for large buffers. cannam@127: cannam@127: commit e241b59af22b29dd33ada31e9dcfc93eadfc594f cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 12 18:57:19 2002 -0400 cannam@127: cannam@127: Make assumption COST(vector) = length * COST(scalar) in classic mode. cannam@127: cannam@127: commit 2ccbe15f93c5f1bd3bdf6997f28ef1f10a2a5c76 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 12 18:19:48 2002 -0400 cannam@127: cannam@127: Revised planner implementation in preparation for wisdom. cannam@127: cannam@127: commit 40ff868c177cd7afc80a4b35f2771aadd59ec6a4 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 12 08:27:36 2002 -0400 cannam@127: cannam@127: Manually hoist loop invariants. cannam@127: cannam@127: commit fca6f800f5e3a40bf20f80e4a2b3da5fe64da13e cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 12 07:47:41 2002 -0400 cannam@127: cannam@127: Revised loop to compile better with gcc -O cannam@127: cannam@127: commit c3e9fb16552f161f337915be9bba97d50aeca6a6 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 17:24:09 2002 -0400 cannam@127: cannam@127: Changed tensor syntax cannam@127: cannam@127: commit c2fb4345e13f698ea829e790e9e50bf5f218d740 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 16:39:45 2002 -0400 cannam@127: cannam@127: Added stuff to do. cannam@127: cannam@127: commit 79b41ae4b2cf1c9d44e386a0fad68531b6fc78a7 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 16:28:14 2002 -0400 cannam@127: cannam@127: Report classic/pro in version number cannam@127: cannam@127: commit e80c221c61c5f10652272a7b86dc61d461ace6d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 14:22:49 2002 -0400 cannam@127: cannam@127: Renamed versions into classic/pro cannam@127: cannam@127: commit 49f5d6fa06030e9ee51346622a68b20e37eff60b cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 14:06:06 2002 -0400 cannam@127: cannam@127: Revised planners, estimator cannam@127: cannam@127: commit cd4556586e22a54132fe9b13920474a2b404fe21 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 11:45:41 2002 -0400 cannam@127: cannam@127: I don't know what I am doing. cannam@127: cannam@127: commit 2feca6ebca91d4c762240f32f2381be534adb289 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 10:35:52 2002 -0400 cannam@127: cannam@127: Massive revision of estimator cannam@127: cannam@127: commit 9147712cee87c8b5aff5dc3deebcadf1c159ea90 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 11 07:32:20 2002 -0400 cannam@127: cannam@127: Many changes cannam@127: cannam@127: commit a68d4a6962649cd7996d2320934764d7490bf685 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 21:35:29 2002 -0400 cannam@127: cannam@127: Keep it simple, stupid. cannam@127: cannam@127: commit 6561b587e1de9ea7d45bd354ad1907512733d3e2 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 19:24:28 2002 -0400 cannam@127: cannam@127: Fixed when #undef PRECOMPUTE_ARRAY_INDICES cannam@127: cannam@127: commit dc412dcd137b131ea524fedf1e7012ad405068bc cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 17:58:13 2002 -0400 cannam@127: cannam@127: Minor changes cannam@127: cannam@127: commit 9a633a7a248735746b8e8e4ef7675a2c0b2a8330 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 16:30:37 2002 -0400 cannam@127: cannam@127: Added ct-ditbuf.c, many changes everywhere cannam@127: cannam@127: commit c8406b59941a26cd2145a45edbfa10d136572b94 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 10:55:40 2002 -0400 cannam@127: cannam@127: More name mangling cannam@127: cannam@127: commit 0889cbac352d22d2214500301375f43bf584a6f6 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 10:08:27 2002 -0400 cannam@127: cannam@127: Fixed build system for single/double precision cannam@127: cannam@127: commit c27e0f15eeb5336b8eebc47c215d2ba092e68aa9 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 09:04:21 2002 -0400 cannam@127: cannam@127: Massive renaming to support both single and double precision. cannam@127: (Must recompile everything twice). cannam@127: cannam@127: commit a0b9a19548d7a6278163b36fa1d3088a3ae24cd7 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 10 06:49:55 2002 -0400 cannam@127: cannam@127: Preliminary crude support for vector transforms in benchmark library. cannam@127: cannam@127: commit ac333b5f92707c9dd867b283d23a53bc64ef69dc cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 21:11:51 2002 -0400 cannam@127: cannam@127: Wrong cast cannam@127: cannam@127: commit 669c861be6b66386cd4e8bf2763a8beacebe9bf7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 16:48:54 2002 -0400 cannam@127: cannam@127: Added things to do. cannam@127: cannam@127: commit 9bfab3a2538b271e95b879402f2269d3cb796144 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 16:07:12 2002 -0400 cannam@127: cannam@127: twlen0: make static cannam@127: cannam@127: commit 89e7b8c3a0488bd401d699312ce92ac4fc7ad872 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 15:34:54 2002 -0400 cannam@127: cannam@127: Nothing cannam@127: cannam@127: commit d8b5a396d1befb9c8709b66e032d5babd304779b cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 15:30:13 2002 -0400 cannam@127: cannam@127: Forgot break in switch statement. cannam@127: cannam@127: commit 36c2890f3950c1994c94a98ddf84e218064211d1 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 15:27:24 2002 -0400 cannam@127: cannam@127: Fix for c++ compatibility cannam@127: cannam@127: commit 71882a5547eeb7ca329df8033e7b2eddb759bf76 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 15:16:43 2002 -0400 cannam@127: cannam@127: Added printer, changed everything cannam@127: cannam@127: commit fea5c4c9f8ba921c22c6d7388e5babe9c73b8e26 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 11:37:07 2002 -0400 cannam@127: cannam@127: Removed redundant nop solver cannam@127: cannam@127: commit 382dec87c6ce525bdbf78111112feabc152eb8f7 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 11:06:31 2002 -0400 cannam@127: cannam@127: More things to do cannam@127: cannam@127: commit e471ce8b9ea5726e0a4d2b63770ca8e242327e71 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 11:01:41 2002 -0400 cannam@127: cannam@127: Introduced idea of rank -infinity and associated NOP plans cannam@127: cannam@127: commit 408125e1afdc419d7a914c6475dc663199d1e372 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 08:36:27 2002 -0400 cannam@127: cannam@127: Fixed comment cannam@127: cannam@127: commit eb28a96b485229b8d6883369b0cab54fe3df5b3d cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 08:20:13 2002 -0400 cannam@127: cannam@127: Removed useless assertions. cannam@127: cannam@127: commit a05a666ec036474e44f87cecc89a76f13249c4cd cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 08:19:26 2002 -0400 cannam@127: cannam@127: Don't malloc(0). cannam@127: cannam@127: commit 53cf44cd85d0b4fe9fd486f24c591d66fd4cb9ff cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 08:08:13 2002 -0400 cannam@127: cannam@127: Fixed signed/unsigned puns cannam@127: cannam@127: commit e2aae81e2ec289bb799c02ff72172ffe7f34d076 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 9 07:52:22 2002 -0400 cannam@127: cannam@127: Added buffered.c cannam@127: cannam@127: commit bb2eb63fb1a1ddf2a8c28a918c2108f9b128053a cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 16:57:54 2002 -0400 cannam@127: cannam@127: Fixed printout cannam@127: cannam@127: commit 5e196db13b6e3ea78d999f2a2958c7e65c4e5e68 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 16:42:52 2002 -0400 cannam@127: cannam@127: Fixed comment cannam@127: cannam@127: commit 00c4dc36263864bb6edf511df147475ca8049e35 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 16:40:58 2002 -0400 cannam@127: cannam@127: Added vrank3-transpose, renamed vrank0-transpose -> vrank2-transpose cannam@127: cannam@127: commit 70736b265541c54ab69946ad312e21300b0b2f47 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 15:51:46 2002 -0400 cannam@127: cannam@127: Added vrank0-transpose cannam@127: cannam@127: commit 21fa46aab36c0306bf64d254b8718c8eaf96118c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 15:11:09 2002 -0400 cannam@127: cannam@127: Added planner-score.c cannam@127: cannam@127: commit 3700ef27b625aa5dd4e42e19969b9d5d0e546b39 cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 11:10:44 2002 -0400 cannam@127: cannam@127: Added indirect.c cannam@127: cannam@127: commit 92cdbe47f89adcf0d5ae5a1535e171fa58b94f0c cannam@127: Author: Matteo Frigo cannam@127: Date: Sat Jun 8 09:34:58 2002 -0400 cannam@127: cannam@127: dif, ditf solvers cannam@127: cannam@127: commit 9382caabae5da944eb93a194ed963efa5b0dcd2a cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 7 18:07:53 2002 -0400 cannam@127: cannam@127: Implemented rank_geq2. Revised build system cannam@127: cannam@127: commit fd9d18f8d6fbe910aa48673bfcfa7b3e9a0ab7e4 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 7 07:12:25 2002 -0400 cannam@127: cannam@127: Fixed printout cannam@127: cannam@127: commit 3f7d553cb7059f0f68b15004eb6a62a2a396c092 cannam@127: Author: Matteo Frigo cannam@127: Date: Fri Jun 7 07:07:46 2002 -0400 cannam@127: cannam@127: Added rank0. Revised codelet organization. cannam@127: cannam@127: commit 9b91cf2747d0b49d1220a4e1fc494fe72d2a6f64 cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 6 18:03:17 2002 -0400 cannam@127: cannam@127: Added memoization cannam@127: cannam@127: commit 61d0f601e04b8e1f9ff8fe350c4ef61211be20de cannam@127: Author: Matteo Frigo cannam@127: Date: Thu Jun 6 08:07:33 2002 -0400 cannam@127: cannam@127: Added vecloop cannam@127: cannam@127: commit 5995d0f346270a179f5156150ce1997613636f8f cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 5 19:02:56 2002 -0400 cannam@127: cannam@127: First DIT solver/plan cannam@127: cannam@127: commit 70714a1ca5a962fc4acf3c91b5b1dbb2b5518310 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 5 16:03:44 2002 -0400 cannam@127: cannam@127: More work on ct cannam@127: cannam@127: commit 479c74ce55bbc9e78c22e4ed8506db9c494871c5 cannam@127: Author: Matteo Frigo cannam@127: Date: Wed Jun 5 11:28:09 2002 -0400 cannam@127: cannam@127: Only use cycle counters cannam@127: cannam@127: commit f397307a4a2ecb675736dbfaee0e3cb117ecafc4 cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 4 20:22:23 2002 -0400 cannam@127: cannam@127: Signed/unsigned fixup cannam@127: cannam@127: commit 26f9cec858cb382ad71e3d025ef4258dffce6e0d cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 4 20:03:56 2002 -0400 cannam@127: cannam@127: New file twiddle.c cannam@127: cannam@127: commit e31060a0f867b05a6cde5ade185fa6edc487c43b cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 4 17:49:39 2002 -0400 cannam@127: cannam@127: Made tensor ranks and vector lengths unsigned. Hopefully fixed cannam@127: all places where it matters. cannam@127: cannam@127: commit 33459884b1d822c57fec05becfccd6cf4a5c1efd cannam@127: Author: Matteo Frigo cannam@127: Date: Tue Jun 4 16:28:58 2002 -0400 cannam@127: cannam@127: System is in working state now (but very incomplete) cannam@127: cannam@127: commit 53e8499d5241b1194dcf9d39bb617b45ba00beed cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 3 18:10:12 2002 -0400 cannam@127: cannam@127: Started implementing planners cannam@127: cannam@127: commit 41ec7720369ea65ef877dacee5b43c73d36e9dc0 cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 3 11:44:18 2002 -0400 cannam@127: cannam@127: Imported libbench from the new benchfft. We will use libbench cannam@127: for benchmarking and testing. cannam@127: cannam@127: commit 936272ab0ecd1ecb585f889ef7d8c996a9e9920c cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 3 09:18:46 2002 -0400 cannam@127: cannam@127: Removed useless rand.c cannam@127: cannam@127: commit c0d9815658256675268110505f21e80b6a850e2c cannam@127: Author: Matteo Frigo cannam@127: Date: Mon Jun 3 08:09:05 2002 -0400 cannam@127: cannam@127: Added timer cannam@127: cannam@127: commit 1135ab2472586a6ddfd23571aa9724686470238c cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 2 21:03:51 2002 -0400 cannam@127: cannam@127: Split codelets into standard and inplace cannam@127: cannam@127: commit 8394a2d62bb726c0af14d1ca1bf721d538bbb6ce cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 2 19:49:03 2002 -0400 cannam@127: cannam@127: Many many changes cannam@127: cannam@127: commit 069ab6b1a318b8d31556dafc6b07953becc026ab cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 2 15:00:11 2002 -0400 cannam@127: cannam@127: Fixed anachronism cannam@127: cannam@127: commit 9a85d188479075e3f6d1de0d33c35008578e5b97 cannam@127: Author: Matteo Frigo cannam@127: Date: Sun Jun 2 14:42:32 2002 -0400 cannam@127: cannam@127: Initial revision