3094fc6c8
(refs/pull/2876/head)
Lazyly reinit threads after a fork in OMP mode by
2020-10-01 15:41:42 +0200
3c05f54df
(refs/pull/2874/head)
Avoid out of bounds access on invalid memory free by
2020-10-01 10:48:45 +0200
dee7c4993
Fix TABs and trailing space by
2020-10-01 10:43:16 +0200
d3c0d6811
Merge pull request #2873 from martin-frbg/issue2871 by
2020-10-01 06:38:22 +0200
9637cd1fd
Merge pull request #2865 from thisch/backticks by
2020-10-01 06:38:06 +0200
236772657
Remove redundant status message by
2020-09-30 23:28:49 +0200
5464eb13e
(refs/pull/2873/head)
Change ifdef linux to __linux for C11 compatibility by
2020-09-30 22:59:41 +0200
e1574cbc8
Change ifdef linux to __linux for C11 compatibility by
2020-09-30 22:50:21 +0200
0b2bb5696
Change ifdef linux to __linux for C11 compatibility by
2020-09-30 22:47:25 +0200
a7d5d0078
Change ifdef linux to __linux for C11 compatibility by
2020-09-30 22:46:25 +0200
be40440ec
Change ifdef linux to __linux for C11 compatibility by
2020-09-30 22:45:18 +0200
2bf70c8e3
Change ifdef linux to __linux for C11 compatibility by
2020-09-30 22:43:25 +0200
60e6c68e3
Adapt ARM architect by
2020-09-29 16:36:14 +0800
64629cb5c
Merge pull request #91 from xianyi/develop by
2020-09-28 22:48:53 +0200
1b1a757f5
Optimize the performance of dot by using universal intrinsics in X86/ARM by
2020-09-28 20:36:53 +0800
0d98ce202
Merge pull request #2866 from RajalakshmiSR/p10_dcopy by
2020-09-28 07:22:54 +0200
2df4235e0
(refs/pull/2866/head)
Optimize dcopy/zcopy for POWER10 by
2020-09-27 21:42:32 -0500
fe8cd5ae7
(refs/pull/2865/head)
Consolidate usage of backticks for build options by
2020-09-28 00:42:17 +0200
ba31c8f5f
Merge pull request #2853 from Qiyu8/usimd-daxpy by
2020-09-27 23:19:59 +0200
e961d4d60
Merge pull request #2864 from martin-frbg/lapack445 by
2020-09-27 23:11:17 +0200
7ed25e9e1
(refs/pull/2864/head)
FIx underflow/rounding errors in LAPACK (S,D)LANV2 by
2020-09-27 22:59:20 +0200
7b169379e
Merge pull request #2863 from martin-frbg/readmefixes by
2020-09-27 22:50:25 +0200
7f539fb85
(refs/pull/2863/head)
Update cpu list, outline cmake build, clarify scope of set_num_threads extension by
2020-09-27 22:48:41 +0200
caf7a1229
Merge pull request #90 from xianyi/develop by
2020-09-27 22:35:45 +0200
72b5b7364
Merge pull request #2850 from xiaojiayuan111/develop by
2020-09-27 12:12:35 +0200
881c15179
(refs/pull/2853/head)
remove default support for FMA4 on zen architect by
2020-09-27 09:35:50 +0800
896bbd55e
Add support for building only selected variable types by
2020-09-26 23:25:55 +0200
c5a32288c
Work around sgemm_r/dgemm_r not being properly defined with BUILD_COMPLEX/BUILD_COMPLEX16 by
2020-09-26 23:24:37 +0200
dfaafd3b5
Merge pull request #2854 from martin-frbg/travis-graviton by
2020-09-23 21:59:18 +0200
f2e9a24e1
(refs/pull/2854/head)
Add AWS Graviton2 build by
2020-09-23 19:02:20 +0200
98153875e
Adapt tests to having only a subset of types in the library by
2020-09-22 23:28:57 +0200
0eaae30e8
Adapt tests to having only a subset of types in the build by
2020-09-22 23:28:03 +0200
dfbc62ef7
Support building only a subset of types by
2020-09-22 23:25:59 +0200
b475b4bd0
Support building only a subset of types by
2020-09-22 23:25:04 +0200
357bff06b
Add BUILD_vartype defines by
2020-09-22 23:24:22 +0200
988a6f429
Add BUILD_vartype defines by
2020-09-22 23:23:33 +0200
e5e2fbd59
Support building only selected types by
2020-09-22 23:21:30 +0200
3287848c8
Support building only seleced types by
2020-09-22 23:20:51 +0200
26611af8e
fix grouping of sources used for more than one type by
2020-09-22 23:20:05 +0200
b886bd672
add defines for building a subset of types by
2020-09-22 23:18:55 +0200
61fae5929
Merge pull request #88 from xianyi/develop by
2020-09-22 23:15:33 +0200
33d22f99f
Merge pull request #2851 from martin-frbg/travis-xcode12 by
2020-09-22 21:44:55 +0200
5ba01dd1a
(refs/pull/2851/head)
Add an OSX build with xcode12 by
2020-09-22 17:26:19 +0200
14f7dad3b
performance improved by
2020-09-22 16:52:15 +0800
06cf73a23
(refs/pull/2850/head)
fix a bug of trmm by
2020-09-22 16:47:10 +0800
ebe64a3c0
(refs/pull/2849/head)
fix a bug of trmm by
2020-09-22 15:39:59 +0800
325b539c2
Optimize the performance of daxpy by using universal intrinsics by
2020-09-22 10:38:35 +0800
0f112077e
Merge pull request #2847 from mhillenibm/fixup_cscal by
2020-09-21 22:22:43 +0200
22aa81f3e
(refs/pull/2847/head)
s390x: fix cscal and zscal implementations by
2020-09-14 18:36:31 +0200
77ea73f5e
s390x: for clang use fp-contract=on instead of fast by
2020-09-16 15:55:38 +0200
f91057cba
s390x: move common vector definitions and utils into header by
2020-09-15 10:54:37 +0200
992d7ca63
Merge pull request #2845 from martin-frbg/lapack443 by
2020-09-18 23:18:41 +0200
7e4d5c237
(refs/pull/2845/head)
Fix workspace query in xGELQ (Reference-LAPACK PR443) by
2020-09-18 09:19:46 +0200
8d12027a7
Merge pull request #86 from xianyi/develop by
2020-09-18 09:17:49 +0200
b1e0bccee
Merge pull request #2844 from RajalakshmiSR/daxpy_p10 by
2020-09-17 23:46:32 +0200
be43d2cb9
(refs/pull/2844/head)
Optimize daxpy/zaxpy for POWER10 by
2020-09-17 12:56:28 -0500
2855e6000
Merge pull request #2841 from martin-frbg/cpp_gemvtest by
2020-09-17 17:29:56 +0200
144a03446
Merge pull request #2843 from mhillenibm/fixup_merge_dynamic_zarch by
2020-09-17 17:28:43 +0200
75d440caa
(refs/pull/2843/head)
s390x/DYNAMIC_ARCH: fixup broken merge and reapply simplification by
2020-09-17 16:45:07 +0200
6abca76c4
(refs/pull/2841/head)
Add option for running only the less demanding GEMV version of the thread safety tests by
2020-09-17 13:49:24 +0200
84c00c3c6
Support running just the GEMV version of the thread safety test by
2020-09-17 13:46:41 +0200
8c5c991bd
Add cpp_thread_test options by
2020-09-17 13:45:40 +0200
2e3b15d68
Add CMakeLists.txt by
2020-09-17 13:43:55 +0200
eaf7f825b
Merge pull request #85 from xianyi/develop by
2020-09-17 13:42:47 +0200
4c10a1673
Merge pull request #2840 from martin-frbg/fixup2833 by
2020-09-16 18:55:50 +0200
c4aeeeb9f
(refs/pull/2840/head)
Activate all BUILD_ options if none was specified by
2020-09-15 23:15:34 +0200
3843bd188
Merge pull request #84 from xianyi/develop by
2020-09-15 23:13:30 +0200
ddec244a5
Merge pull request #2838 from austinpagan/gordon_trmm by
2020-09-15 21:17:48 +0200
dfeca4609
(refs/pull/2838/head)
Adding performance patch for trmm, just like #2836 by
2020-09-15 08:59:50 -0500
f8950f40a
Merge pull request #2836 from austinpagan/gordon_trsm by
2020-09-15 11:26:37 +0200
274d6e015
(refs/pull/2836/head)
Fixing a performance bug in trsm_[LR].c. by
2020-09-14 13:10:48 -0500
91c84e1c0
Merge pull request #2796 from Guobing-Chen/BF16_dot_coversion_apis by
2020-09-14 15:00:19 +0200
1ee1e7b49
Merge pull request #2833 from martin-frbg/issue2830 by
2020-09-14 07:24:23 +0200
ba644378d
(refs/pull/2833/head)
Copy BUILD_ options available to the compiler flags by
2020-09-14 00:03:33 +0200
9e11c2d62
Add BUILD_SINGLE etc by
2020-09-13 23:55:11 +0200
4d250d0cd
Rearrange ifdefs by
2020-09-13 23:29:01 +0200
de139337b
Remove spurious tests for complex ASUM and NRM2 by
2020-09-13 22:20:41 +0200
ec2948f14
Make tests conditional on BUILD_DOUBLE by
2020-09-13 22:17:46 +0200
ce8939863
Make tests for individual variable types conditional on the respective BUILD_ option by
2020-09-13 21:52:18 +0200
593ce9e23
Make building individual tests depend on BUILD_SINGLE etc defines by
2020-09-13 21:50:12 +0200
74e358bcd
Remove spurious complex16 tests by
2020-09-13 21:49:01 +0200
26792d209
Copy BUILD_* directives to the compiler options to allow ifdef in tests by
2020-09-13 21:47:55 +0200
6b52c7e17
Merge pull request #2832 from martin-frbg/issue2831 by
2020-09-13 21:20:30 +0200
746ad3bd1
(refs/pull/2832/head)
Fix vendor match for GCC gfortran by
2020-09-13 18:40:59 +0200
55d4d470e
Merge pull request #83 from xianyi/develop by
2020-09-13 18:30:11 +0200
a27089473
Merge pull request #2829 from mhillenibm/clang_s390x by
2020-09-08 23:36:41 +0200
047b8d7af
(refs/pull/2829/head)
Add an s390 build with clang to the Travis configuration by
2020-09-08 19:30:37 +0200
f7731a358
Update CONTRIBUTERS.md - clang build fixes for IBM z by
2020-09-08 15:15:15 +0200
a55fe06f2
s390x/DYNAMIC_ARCH: define a HW_CAP flag to support slightly older glibc versions by
2020-09-07 17:13:03 +0200
4f34bcfb5
s390x/DYNAMIC_ARCH: pass supported arch levels from Makefile to run-time code by
2020-09-07 17:04:03 +0200
0629d8ebd
s390x/DYNAMIC_ARCH: generalize detecting supported archs for clang by
2020-09-04 16:32:45 +0200
15da2f9ac
Merge pull request #2828 from martin-frbg/lapack438 by
2020-09-08 10:25:19 +0200
7d9c77f42
(refs/pull/2828/head)
Correct dimension argument to xLASET by
2020-09-07 22:03:46 +0200
c8f029a51
Merge pull request #82 from xianyi/develop by
2020-09-07 21:59:13 +0200
e72430fe4
Merge pull request #2803 from xiegengxin/AVX2-asum by
2020-09-06 18:32:15 +0200
6e0f6c5f0
Merge pull request #2824 from martin-frbg/asumbench by
2020-09-06 10:05:47 +0200
6f8fad87c
(refs/pull/2824/head)
Use POSIX2001 clock.gettime for higher resolution by
2020-09-05 19:44:01 +0200
ed0f2d3dd
Merge pull request #2816 from martin-frbg/silicon by
2020-09-05 19:17:59 +0200
43a31b778
Merge pull request #2823 from martin-frbg/fix2778 by
2020-09-05 17:29:38 +0200
8a2a137a9
(refs/pull/2823/head)
Correct argument to SLASET (Improves fix from PR2778) by
2020-09-05 13:06:31 +0200