Commit Graph

  • cfc4acc221
    typo Martin Kroeker 2018-12-16 16:19:51 +0100
  • 545c2b1bbb
    Add -mavx2 on Haswell only if the compiler supports it Martin Kroeker 2018-12-16 13:09:19 +0100
  • 69d206440a Make the skylakex/haswell sgemm code compile and run even with compilers without avx2 support Arjan van de Ven 2018-12-16 00:19:41 +0000
  • 3843e3e017
    use -maxv2 on haswell Martin Kroeker 2018-12-15 23:30:31 +0100
  • fbcb14a74b
    should be core-avx2 Martin Kroeker 2018-12-15 20:18:59 +0100
  • 2a3190dc76
    fix elseifeq and use older option core2-avx for compatibility Martin Kroeker 2018-12-15 20:17:44 +0100
  • 1ebe5c0f49
    Add -march=haswell to HASWELL part of DYNAMIC_ARCH build Martin Kroeker 2018-12-15 19:35:35 +0100
  • 0586899a10 Use sgemm_ncopy_4_skylakex.c also for Haswell Arjan van de Ven 2018-12-15 13:43:07 +0000
  • 00dc09ad19 Use the skylake sgemm beta code also for haswell Arjan van de Ven 2018-12-15 13:18:59 +0000
  • 78d877b54b
    Merge pull request #1914 from fenrus75/smallmatrix Martin Kroeker 2018-12-13 19:08:14 +0100
  • cdc668d82b Add a "sgemm direct" mode for small matrixes Arjan van de Ven 2018-12-12 16:45:57 +0000
  • 87718807f0
    Merge pull request #1910 from martin-frbg/issue1909 Martin Kroeker 2018-12-12 14:56:25 +0100
  • 51aec8e96b
    make sure the added march=skylake-avx512 does not cause problems on Windows Martin Kroeker 2018-12-11 22:47:32 +0100
  • 91d6a56931 reduce scope of avx512 cflags Andrew 2018-12-11 22:33:36 +0100
  • 863162a1c8 init Andrew 2018-12-11 22:32:23 +0100
  • 06f7d78d70
    Add -march=skylake-avx512 to SkylakeX part of DYNAMIC_ARCH builds Martin Kroeker 2018-12-11 21:10:38 +0100
  • 38cc638591
    Avoid adding blanket march=skylake-avx512 to dynamic_arch builds Martin Kroeker 2018-12-11 21:09:26 +0100
  • 0bf6d74e5f
    Fix typo in previous commit for arm dynamic arch Martin Kroeker 2018-12-07 19:37:33 +0100
  • 133c278ee5
    Add DYNAMIC_CORE list for ARM64 Martin Kroeker 2018-12-07 17:42:23 +0100
  • 2b355592e3
    Make sure to use the arm version of dynamic.c in ARM64 DYNAMIC_ARCH Martin Kroeker 2018-12-07 16:25:55 +0100
  • 5f99b61aa7 try to maintain cache line alignment for odd BLAS L1 work splits Andrew 2018-12-07 00:15:27 +0100
  • 5839734284 init Andrew 2018-12-07 00:13:45 +0100
  • ff3eb1d474
    Merge pull request #1904 from martin-frbg/issue1870 Martin Kroeker 2018-12-06 23:01:23 +0100
  • 0b09516678
    Fix missing parameter in popen call Martin Kroeker 2018-12-06 18:33:05 +0100
  • 7639f2e1f0
    Rewrite the conditional for OSX to fix cmake parsing on others Martin Kroeker 2018-12-06 14:04:27 +0100
  • 2fc712469d
    Avoid creating spurious non-suffixed c/zgemm_kernels Martin Kroeker 2018-12-06 13:56:06 +0100
  • 6ba30e270d
    Fix typo that broke CNRM2 on ARMV8 since 0.3.0 Martin Kroeker 2018-12-06 13:42:25 +0100
  • bf23518e36
    Merge pull request #1903 from rengolin/armv8 Martin Kroeker 2018-12-05 22:10:53 +0100
  • 31a490ea88 Fix two mistakes on Arm64 builds Renato Golin 2018-12-05 18:51:38 +0000
  • 701ea88347
    Use p2align instead of align for OSX compatibility Martin Kroeker 2018-12-03 13:06:43 +0100
  • 721c56c224
    Merge pull request #1899 from brada4/fbsd12 Martin Kroeker 2018-12-03 12:50:27 +0100
  • c5f8aeff2d
    Merge branch 'develop' into fbsd12 Martin Kroeker 2018-12-03 12:50:14 +0100
  • 8278cbe7f8
    Merge pull request #1894 from pkubaj/patch-2 Martin Kroeker 2018-12-03 12:48:53 +0100
  • ea6d1b96bd
    Update Makefile.system Martin Kroeker 2018-12-03 08:59:10 +0100
  • 360374be62
    Update with the changes from 0.3.4 Martin Kroeker 2018-12-02 23:44:13 +0100
  • f5acaad8f0
    Increment version to 0.3.5.dev Martin Kroeker 2018-12-02 23:43:15 +0100
  • 93fa6b7b76
    Increment version to 0.3.5.dev Martin Kroeker 2018-12-02 23:42:33 +0100
  • c0827a7164
    Update with changes from 0.3.4 v0.3.4 Martin Kroeker 2018-12-02 23:41:17 +0100
  • 86cff4effc
    Merge pull request #1900 from xianyi/develop Martin Kroeker 2018-12-02 23:40:21 +0100
  • b028960aba
    Merge branch 'release-0.3.0' into develop Martin Kroeker 2018-12-02 23:38:49 +0100
  • 3c9e3faedb
    fixup BSD naming of powerpc arch Martin Kroeker 2018-12-02 23:24:53 +0100
  • 44c81fd135 oops Andrew 2018-12-02 20:27:53 +0100
  • 26b3710485 Add architecture mappings for FreeBSD12 Andrew 2018-12-02 12:07:41 +0100
  • 84e614d0fd init Andrew 2018-12-02 12:05:15 +0100
  • dceff5542c
    Handle Android environments that identify as Linux (#1898) Martin Kroeker 2018-12-01 20:56:11 +0100
  • 191014f940
    Update system_check.cmake Martin Kroeker 2018-12-01 18:53:53 +0100
  • c1a060645b
    Handle Android environments that identify as Linux Martin Kroeker 2018-12-01 12:46:34 +0100
  • 6c7b691083
    Really revert xDOT changes from 1832 Martin Kroeker 2018-11-30 21:32:01 +0100
  • 5f4c550c27
    Merge pull request #1892 from martin-frbg/mipsdot Martin Kroeker 2018-11-30 21:28:21 +0100
  • 731b2722ba Fix build on POWER, remove DragonFly, add NetBSD pkubaj 2018-11-30 16:04:07 +0000
  • f85ce54d4a
    Use correct Makefile on powerpc64 pkubaj 2018-11-30 16:05:49 +0000
  • bbeb8a0b23
    Fix build on POWER, remove DragonFly, add NetBSD pkubaj 2018-11-30 16:04:07 +0000
  • 2601cd58ab remove surplus locking code , only enabled w x86, disabled or never enabled on all others Andrew 2018-10-11 23:29:34 +0300
  • 95a5542e3c
    Revert DOT kernel changes from #1834 Martin Kroeker 2018-11-30 11:16:24 +0100
  • 7a2e1bc804
    Use generic kernel for DSDOT/SDSDOT Martin Kroeker 2018-11-30 10:57:09 +0100
  • 35653e38b3
    Merge pull request #1834 from fengrl/develop Martin Kroeker 2018-11-30 10:48:46 +0100
  • 71e25ae42f
    Merge pull request #1890 from martin-frbg/issue1889 Martin Kroeker 2018-11-29 15:47:35 +0100
  • 97d7298973
    call it OpenBLAS not just version Martin Kroeker 2018-11-29 11:52:08 +0100
  • de0d0ed52f
    Improve formatting of config output Martin Kroeker 2018-11-29 11:28:19 +0100
  • 081ceb3e02
    Propagate version number for openblas_get_config Martin Kroeker 2018-11-29 00:12:04 +0100
  • a29ec458c2
    propagate verison number for openblas_config_version Martin Kroeker 2018-11-29 00:10:49 +0100
  • 816775e309
    Add version information to openblas_get_config output Martin Kroeker 2018-11-29 00:06:44 +0100
  • c067c1fb7b _AXPY calibration for >2 cores Andrew 2018-11-26 17:52:22 +0100
  • c6dff2803c init Andrew 2018-11-26 17:49:25 +0100
  • b6363f4539
    Merge pull request #1885 from brada4/freebsd Martin Kroeker 2018-11-25 22:20:13 +0100
  • 19c4bdd8b3 Add return value so that freebsd system clang does not err out Andrew 2018-11-25 21:35:01 +0100
  • f049a4c84f init Andrew 2018-11-25 21:34:09 +0100
  • f72fdf525c
    Merge pull request #1875 from martin-frbg/issue1851 Martin Kroeker 2018-11-25 20:53:46 +0100
  • 5393759a98
    Merge pull request #1869 from martin-frbg/axpy0 Martin Kroeker 2018-11-25 20:52:49 +0100
  • 5cf18e2875
    Merge pull request #1878 from kiwifb/PGI_f_check Martin Kroeker 2018-11-25 20:51:50 +0100
  • 910050985a
    Merge pull request #1876 from rengolin/armv8-cleanup Martin Kroeker 2018-11-25 20:51:24 +0100
  • da5894a2ed Limit _AXPY threading due to observed regression with excess threads Andrew 2018-11-23 16:08:15 +0100
  • 6c044d5967 init Andrew 2018-11-23 16:03:42 +0100
  • 0184713e1a Correct link flags for PGI compiler. François Bissey 2018-11-21 14:24:56 +1300
  • 459cc83803 initialize z/cdot result variable at declaration time to avoid confusing static checker Andrew 2018-11-20 20:59:35 +0100
  • ae895f8648 initialize result at declaration instead of quick-exit only Andrew 2018-11-20 20:58:08 +0100
  • c1fa2c801c init Andrew 2018-11-20 20:56:02 +0100
  • 45c3c459e1
    Merge pull request #1868 from martin-frbg/aix_cpuid Martin Kroeker 2018-11-20 17:25:57 +0100
  • 113cb00b95
    fix missing parenthesis Martin Kroeker 2018-11-19 21:01:36 +0100
  • 5192651706
    Add CriticalSection handling instead of mutexes for Windows Martin Kroeker 2018-11-19 17:58:22 +0100
  • 310ea55f29 Simplifying ARMv8 build parameters Renato Golin 2018-11-16 15:45:12 +0000
  • 2e6fae2aad
    Serialize accesses to parallelized level3 functions from multiple callers Martin Kroeker 2018-11-19 14:02:50 +0100
  • 368d14f8c8
    Fix harmless typo Martin Kroeker 2018-11-16 14:58:28 +0100
  • 42bc2a9202
    Fix copy-paste errors (POWER8/9 and extraneous return) Martin Kroeker 2018-11-16 12:10:44 +0100
  • 43bb386b10 fix dot problem on 64bit mips fengruilin 2018-11-15 11:11:59 +0800
  • c171b8ad13
    Handle special case INCX=0,INCY=0 in the axpy interface Martin Kroeker 2018-11-13 13:57:18 +0100
  • 2f04cf22ac
    Detect POWER9 as POWER8 on AIX and Linux Martin Kroeker 2018-11-13 08:16:14 +0100
  • 807f6e6922
    Use prtconf to determine CPU type on AIX Martin Kroeker 2018-11-12 18:52:29 +0100
  • ecbeb802a0
    Merge pull request #1865 from martin-frbg/issue1844 Martin Kroeker 2018-11-12 17:30:44 +0100
  • 2c5725cc39
    Merge pull request #1864 from aytekinar/patch-1 Martin Kroeker 2018-11-12 14:30:28 +0100
  • e3666931d8
    Update .travis.yml Arda Aytekin 2018-11-09 00:25:30 +0100
  • ae02a57261
    Merge pull request #1866 from martin-frbg/issue1859 Martin Kroeker 2018-11-10 19:23:31 +0100
  • a6a52a73f7
    Fix argument in SLASET call to zero S Martin Kroeker 2018-11-10 17:16:53 +0100
  • 0427277cef
    Allow optimization for small m, large n only if it can be made threadsafe Martin Kroeker 2018-11-10 15:45:54 +0100
  • 4f43668eec
    Merge pull request #2 from xianyi/develop Martin Kroeker 2018-11-10 15:37:25 +0100
  • b0c15bacc1
    Merge pull request #1863 from martin-frbg/aix_install3 Martin Kroeker 2018-11-09 13:12:06 +0100
  • cfb0f5b0f8
    Set LIBSONAME suffix to .a for AIX Martin Kroeker 2018-11-08 22:39:10 +0100
  • 667fed579d
    Merge pull request #1856 from rengolin/armv8-a57 Martin Kroeker 2018-11-07 21:01:29 +0100
  • 96d2f2c9b2
    Merge pull request #1831 from brada4/hemv Martin Kroeker 2018-11-07 08:49:21 +0100
  • 653e657a58
    Merge pull request #1857 from brada4/fc-1847 Martin Kroeker 2018-11-07 08:48:31 +0100