b37516add
Add BGEMM parameters by
2025-07-10 14:59:01 +0200
406b0c597
deploy: d030f81380 by
2025-07-10 08:53:16 +0000
d030f8138
Merge pull request #5369 from martin-frbg/lapack1144 by
2025-07-10 10:46:15 +0200
b746f0eda
(refs/pull/5369/head)
Allocate IWORK to hold at least the one element for workspace queries by
2025-07-10 08:58:16 +0200
b8f66ba0e
Merge pull request #5367 from Mousius/bgemm-init by
2025-07-10 00:57:41 +0200
cdebb4fd4
Merge pull request #5365 from martin-frbg/issue5324 by
2025-07-09 22:50:54 +0200
ff614575c
(refs/pull/5365/head)
Fix arm64 HAVE_SME setting for DYNAMIC_ARCH builds by
2025-07-09 14:44:25 +0200
1f00593af
deploy: 0e11537cab by
2025-07-09 07:35:28 +0000
0e11537ca
Merge pull request #5357 from Mousius/bgemm-init by
2025-07-09 09:34:58 +0200
8cd4be8d4
(refs/pull/5367/head)
Temporarily disable test_bgemm by
2025-07-09 08:27:18 +0100
66d9185eb
(refs/pull/5357/head)
Fix CMake support by
2025-07-08 22:47:20 +0000
87c802bd6
Add files via upload by
2025-07-08 22:22:38 +0200
98aefb70b
Merge pull request #5292 from isharif168/optimized_gemv_n_1x3 by
2025-07-08 21:05:43 +0200
fd3740681
(refs/pull/5292/head)
Merge branch 'develop' into optimized_gemv_n_1x3 by
2025-07-08 21:05:30 +0200
fabdf0726
fix options rewriting by
2025-07-08 19:18:11 +0200
48394384e
Use correct constants for per-target BGEMM/SBGEMM by
2025-07-07 11:09:26 +0000
73bf0b941
Add bgemm to gensymbol by
2025-07-07 10:40:33 +0000
f95e7b0e3
Add infrastructure for BGEMM by
2025-07-03 17:47:08 +0000
15d6e5851
Merge pull request #5364 from martin-frbg/blashalf by
2025-07-08 17:14:50 +0200
04bb5acd7
(refs/pull/5364/head)
change BLAS_HALF to BLAS_BFLOAT16 (another missed rename) by
2025-07-08 14:40:22 +0200
3d3188707
Merge pull request #5362 from Mousius/fix-bf16 by
2025-07-08 14:35:50 +0200
436fc0836
deploy: d2ea9bbb6d by
2025-07-08 10:00:49 +0000
0ddf8ebd4
Merge pull request #5354 from pratiklp00/p11 by
2025-07-08 11:52:18 +0200
d2ea9bbb6
Merge pull request #5363 from guoyuanplct/develop by
2025-07-08 11:47:18 +0200
4ff549a45
(refs/pull/5363/head)
Update CONTRIBUTORS.md by
2025-07-08 17:16:51 +0800
309c48e32
Update CONTRIBUTORS.md by
2025-07-08 17:13:27 +0800
ac45a7ea7
Create CNAME by
2025-07-07 15:23:31 +0200
552e1c7a7
(refs/pull/5362/head)
Correct compiler flags for NEOVERSEV1 target by
2025-07-07 11:26:36 +0000
46b9b7a08
Also enable BFLOAT16 for make cirun by
2025-07-07 10:41:12 +0000
eaaa628af
Enable BUILD_BFLOAT16 in cirun by
2025-07-07 10:20:17 +0000
7a97c4ca9
Rename HALF -> BFLOAT16 in some more places by
2025-07-07 10:03:26 +0000
76168bb63
deploy: ee6560c89f by
2025-07-07 05:50:37 +0000
ee6560c89
Merge pull request #5360 from sertonix/cpuid-arm by
2025-07-07 07:41:56 +0200
8d11e4630
(refs/pull/5360/head)
Fix cpuid.S on arm by
2025-07-06 23:48:10 +0200
03a4afcf1
Merge pull request #5359 from martin-frbg/gitign_isnan by
2025-07-05 22:26:55 +0200
901de8f33
(refs/pull/5359/head)
remove lapacke_mangling.h and add la_xisnan.mod by
2025-07-05 20:35:16 +0200
93eb2587b
deploy: ce6991780a by
2025-07-05 17:20:03 +0000
ce6991780
Merge pull request #5356 from ilina-linaro/ilina-woa by
2025-07-05 19:07:45 +0200
3c811feea
deploy: df013c5e28 by
2025-07-04 21:39:03 +0000
df013c5e2
Merge pull request #5358 from iha-taisei/dot_unroll by
2025-07-04 23:38:32 +0200
b5bf50a03
Add GEMMTR tests by
2025-07-04 16:16:59 +0200
f7ad906b4
(refs/pull/5358/head)
Performance improvements of [SD]DOT with loop-unrolling on A64FX by
2025-07-04 22:57:44 +0900
fcd502c6d
Merge branch 'OpenMathLib:develop' into gemmt_tests by
2025-07-03 16:36:50 +0200
7f360001f
(refs/pull/5356/head)
Update README.md to include Windows on Arm64 by
2025-07-03 07:15:20 -0600
698de96ee
deploy: 36c2589d3a by
2025-07-02 07:14:33 +0000
36c2589d3
Merge pull request #5355 from tetsuzo-usui/add_parallel_laed3 by
2025-07-02 09:14:03 +0200
14107e37d
(refs/pull/5355/head)
Add parallel laed3 by
2025-07-01 22:12:27 +0900
556cab3b4
deploy: a06bcf836b by
2025-07-01 12:07:21 +0000
a06bcf836
Merge pull request #5353 from nakagawa-fj/feature/gemm_divide_rate_for_A64FX by
2025-07-01 14:06:53 +0200
5253c8f16
(refs/pull/5353/head)
Multi-thread Performance Improvement of GEMM with DIVIDE_RATE=1 for A64FX. by
2025-06-30 21:35:16 +0900
8f0a1a3f8
Merge pull request #5303 from martin-frbg/issue5289 by
2025-06-29 22:47:56 +0200
2c0dd2468
Merge pull request #5350 from martin-frbg/issue5341 by
2025-06-29 21:10:18 +0200
7ae24d0b8
Merge pull request #5351 from martin-frbg/lapack1140 by
2025-06-29 19:20:17 +0200
5aeca597f
(refs/pull/5351/head)
Fix documentation error and ordering bug (Reference-LAPACK PR 1140) by
2025-06-29 17:42:15 +0200
b833ba191
deploy: dcb289539b by
2025-06-29 15:40:09 +0000
dcb289539
Merge pull request #5344 from MaartenBaert/fix-dlasd7 by
2025-06-29 17:39:41 +0200
9bcffbd65
(refs/pull/5350/head)
Declare the server_lock mutex volatile in addition to static by
2025-06-29 15:42:43 +0200
334cd242d
Merge pull request #5348 from hideaki-motoki/issue5343_prefered_size_for_a64fx by
2025-06-27 14:57:37 +0200
bba75d5e4
(refs/pull/5348/head)
GEMM_PREFERED_SIZE parameter has been changed for A64FX. by
2025-06-27 19:37:36 +0900
5374ca377
deploy: 4062c10370 by
2025-06-27 07:45:36 +0000
4062c1037
Merge pull request #5345 from OpenMathLib/revert-5251-issue5250 by
2025-06-27 09:45:10 +0200
bc1600ba7
deploy: b78d1dc0ae by
2025-06-26 18:08:12 +0000
b78d1dc0a
Merge pull request #5342 from martin-frbg/cmake_ampere by
2025-06-26 18:46:33 +0200
83a01d29c
(refs/pull/5345/head, revert-5251-issue5250)
Revert "Fix out-of-bounds accesses in ?/SCAL/?GEEV triggered by preceding errrors/invalid inputs" by
2025-06-26 17:47:20 +0200
3df31756c
Add WoA to the list of platforms for which binaries are provided by
2025-06-26 12:28:24 +0200
2207d80f7
Update for 0.3.30 by
2025-06-26 12:26:10 +0200
560fa88c9
(refs/pull/5342/head)
Add cross-build parameters for Ampere One by
2025-06-26 10:57:30 +0200
55bb5ef86
Add compiler options for Ampere One by
2025-06-26 10:50:44 +0200
b37889e52
(refs/pull/5344/head)
Merge branch 'OpenMathLib:develop' into fix-dlasd7 by
2025-06-26 09:29:07 +0200
1dde4a13c
(refs/pull/5354/head)
p11 changes by
2025-06-26 00:03:38 -0500
d6fc32110
deploy: 11ce79a4f0 by
2025-06-25 14:45:12 +0000
11ce79a4f
Merge pull request #5329 from foxtran/fix/docs by
2025-06-25 16:44:44 +0200
0904a42fa
Fix documentation error and ordering bug in DLASD7 by
2025-06-25 15:47:48 +0200
4db92bf9e
deploy: d24195e9a1 by
2025-06-25 09:25:14 +0000
d24195e9a
Merge pull request #5295 from Pengzhou0810/develop by
2025-06-25 11:09:46 +0200
fe783000d
(refs/pull/5338/head)
Update install.md by
2025-06-25 10:50:34 +0200
134b21ae6
(refs/pull/5295/head)
Fix some hyperthreading errors. by
2025-05-26 10:57:25 +0800
d96daa220
Merge pull request #5290 from Srangrang/develop by
2025-06-24 23:10:15 +0200
fdc1c3234
Merge pull request #5336 from martin-frbg/issue5332 by
2025-06-24 21:58:58 +0200
5aa483e16
(refs/pull/5336/head)
Use response files on old PPC/Intel Macs in single-target builds too by
2025-06-24 17:37:34 +0200
12591caa9
Merge pull request #5334 from azuresky01/develop by
2025-06-24 16:09:25 +0200
ebc4ab8f9
deploy: 8b08df5c5a by
2025-06-24 11:09:01 +0000
ee26caffb
Merge pull request #5309 from davidz-ampere/dev-ampereone by
2025-06-24 12:27:08 +0200
8b08df5c5
Merge pull request #5335 from martin-frbg/issue5330 by
2025-06-24 12:25:46 +0200
3bba35b8f
(refs/pull/5335/head)
Remove non-portable option from objcopy calls by
2025-06-24 09:01:47 +0200
8953ba9c2
(refs/pull/5334/head)
Fix INTERFACE64 builds on Loongarch64 with LLVM by
2025-06-24 14:27:15 +0800
ed457343d
(refs/pull/5326/head)
CMake: Make sure to find OpenMP dependency before usage. by
2025-06-21 05:46:53 +0100
aa90ab414
(refs/pull/5309/head)
Add support for Ampere AmpereOne processors by
2025-06-24 00:12:34 -0400
dcf5acec2
(refs/pull/4027/merge)
Merge 82827762c0 into b4945057b7 by
2025-06-23 13:27:40 +0100
46b0dfef8
(refs/pull/5329/head)
Use links to issues by
2025-06-21 11:35:02 +0200
83efceb3c
Keep dgemm_snb_1thread.png in repo by
2025-06-21 11:24:42 +0200
b4945057b
Merge pull request #5319 from imciner2/im/armtypes by
2025-06-20 06:02:22 -0700
f4ff8c32a
deploy: b3904aeed7 by
2025-06-20 12:55:50 +0000
b3904aeed
Merge pull request #5323 from imciner2/im/ofast by
2025-06-20 05:55:21 -0700
721c80644
(refs/pull/5323/head)
Switch power to use O3 instead of Ofast by
2025-06-20 09:23:05 +0100
badef1d32
(refs/pull/5319/head)
Update sbgemm_tcopy_4_neoversev1 kernel to use standard C types by
2025-06-19 14:26:16 +0100
4e6da5ed3
Update version to 0.3.30.dev by
2025-06-19 11:57:35 +0200
8dff37827
Update version to 0.3.30.dev by
2025-06-19 11:56:55 +0200
c055c36b4
Merge pull request #5317 from OpenMathLib/release-0.3.0 by
2025-06-19 02:56:01 -0700
993fad6ae
(tag: v0.3.30, refs/pull/5317/head, release-0.3.0)
Update version to 0.3.30 by
2025-06-19 11:45:39 +0200