Commit Graph

5682 Commits

Author SHA1 Message Date
Martin Kroeker
b67a92c19f Merge pull request #3211 from martin-frbg/lapack471
Handle norm NaN value in xGESDD (Reference LAPACK PR471)
2021-05-02 18:44:29 +02:00
Martin Kroeker
d77d9bc920 Handle norm NaN value (Reference LAPACK PR471) 2021-05-02 11:24:50 +02:00
Martin Kroeker
37d3e2bd94 Merge pull request #3210 from martin-frbg/lapack502
Fix possible division by zero in LAPACK xTGSJA (Reference-LAPACK PR502)
2021-05-02 09:02:11 +02:00
Martin Kroeker
de8656769c Fix possible division by zero in xTGSJA (Reference-LAPACK PR502) 2021-05-01 21:31:13 +02:00
Martin Kroeker
d43e07198d Merge pull request #3208 from martin-frbg/lapack534
Apply MKL team fixes to the LAPACKE interfaces (Reference-LAPACK PR 534)
2021-05-01 20:18:29 +02:00
Martin Kroeker
da16764c7a Merge pull request #3209 from martin-frbg/issue3160
Add casts to prevent overflow of intermediate results
2021-05-01 20:08:24 +02:00
Martin Kroeker
98ebc8ac59 Add casts to prevent overflow of intermediate result 2021-05-01 14:48:19 +02:00
Martin Kroeker
904b221f03 Add cast to prevent overflow of intermediate result 2021-05-01 14:47:22 +02:00
Martin Kroeker
5cc35abc3d Apply MKL team fixes to the LAPACKE interfaces (Reference-LAPACK PR 534)
Removed spurious checks for INFO in xLACPY,xLASET after routines not returning any,and redundant requirements for ldvt in xGESVD_WORK
2021-05-01 13:22:10 +02:00
Martin Kroeker
254774f5a6 Add const qualifiers 2021-05-01 13:10:16 +02:00
Martin Kroeker
ae9cdee753 Merge pull request #3207 from hjl-tools/hjl/cet/develop
x86: Enable Intel CET
2021-05-01 12:42:54 +02:00
H.J. Lu
53ee0b76bb x86: Enable Intel CET
When Intel CET is enabled, we need to include <cet.h> in assembly codes
to mark Intel CET support and place _CET_ENDBR at the function entry.
2021-04-30 19:45:39 -07:00
Martin Kroeker
dc6b04c375 Merge pull request #3206 from martin-frbg/lapack480535
Import packing improvements to LAPACK xLAQR from Reference-LAPACK (PR 480+535)
2021-04-30 21:42:44 +02:00
Martin Kroeker
87d2e314db Import packing improvements in LAPACK xLAQR from Reference-LAPACK PR 480+535 2021-04-30 13:50:55 +02:00
Martin Kroeker
3a30c12019 Merge pull request #25 from xianyi/develop
rebase
2021-04-30 13:47:17 +02:00
Martin Kroeker
c9a82f54d1 Merge pull request #3204 from martin-frbg/lapack506
Correct INFO value returned by SLASQ2/DLASQ2 (Reference-LAPACK 506)
2021-04-30 13:25:48 +02:00
Martin Kroeker
444cb78be5 correct INFO value (Reference-LAPACK 506) 2021-04-30 09:26:54 +02:00
Martin Kroeker
171c20e3b6 Merge pull request #3202 from martin-frbg/issue3201
Fix division by zero in the non-x86 codepath of C/ZROTG
2021-04-29 18:58:27 +02:00
Martin Kroeker
c5fb91f1bc Fix division by zero in the non-x86 codepath 2021-04-29 09:47:18 +02:00
Martin Kroeker
9a36a283d3 Merge pull request #3199 from martin-frbg/lapack537
Add LAPACKE fixes from Reference-LAPACK PR 537
2021-04-29 05:39:50 +02:00
Martin Kroeker
7e35d25ea0 Merge pull request #3198 from martin-frbg/lapack539
Apply fixes from Reference-LAPACK PR468 and 539 for array declarations in ?ORGBR/?UNGBR
2021-04-29 05:39:35 +02:00
Martin Kroeker
3704f5e5b0 Add missing break statements in the ?lascl functions 2021-04-28 20:56:55 +02:00
Martin Kroeker
6b76066632 Add const qualifiers 2021-04-28 20:55:37 +02:00
Martin Kroeker
2b01132515 Clean up misdeclaration of the dummy stand-in for A in ?ORGBR/?UNGBR workspace queries (Reference-LAPACK PR 468 and 530) 2021-04-28 19:20:08 +02:00
Martin Kroeker
8e95a1e18d Merge pull request #3195 from martin-frbg/lapack536
Apply lapack-testing fix from Reference-LAPACK PR536
2021-04-28 18:17:25 +02:00
Martin Kroeker
13a29d13fd Apply lapack-testing fix from Reference-LAPACK PR536
fixes changing back from a single OMP thread for error exit testing to the originally requested number of threads for computational tests
2021-04-27 15:48:22 +02:00
Martin Kroeker
a6c2cb8417 Merge pull request #3193 from martin-frbg/lapack538
Apply lapack-testing fixes from Reference-LAPACK PR538
2021-04-27 15:40:51 +02:00
Martin Kroeker
d511a7bb4f Merge pull request #3191 from martin-frbg/issue3188
Delay creation of the (soft)link until after the library has been built
2021-04-27 13:35:16 +02:00
Martin Kroeker
3526ff2507 Apply fixes from Reference-LAPACK PR538 2021-04-27 12:52:49 +02:00
Martin Kroeker
adcfe7b789 Merge pull request #3190 from martin-frbg/issue3128-2
Replace spurious AVX512 requirement in the Haswell drot microkernel with an AVX2/FMA3 guard
2021-04-27 06:36:28 +02:00
Martin Kroeker
0608bc5d82 delay creation of the softlink until after the library has been created 2021-04-26 22:32:23 +02:00
Martin Kroeker
3d511f0e66 replace spurious avx512 requirement with fma check 2021-04-26 21:55:30 +02:00
Martin Kroeker
0b8a436af9 Add mixed clang/ifort build on OSX to Azure CI (#3185)
* Add mixed clang/ifort build on OSX to the Azure CI config based on https://github.com/oneapi-src/oneapi-ci
(and remove debugging tools from the clang+gfortran job)

* Remove extraneous libgfortran dependency of ifort builds

* remove FEXTRALIB from link line of shared library as ifort keeps track of dependencies (and they are different for a .dylib than what f_check got for an executable)
2021-04-22 02:11:20 +02:00
Martin Kroeker
352efdd13a Merge pull request #24 from xianyi/develop
rebase
2021-04-20 21:30:28 +02:00
Martin Kroeker
4855af02a3 Merge pull request #3184 from martin-frbg/ctestfix
Fix obscure ctest crashes on OSX and add OSX builds to Azure CI
2021-04-20 07:31:07 +02:00
Martin Kroeker
94a5a1f0f1 Add OSX build variations to Azure CI 2021-04-19 22:27:08 +02:00
Martin Kroeker
751d127d7c Include cblas_test.h to achieve int/long size change with INTERFACE64 2021-04-19 22:26:34 +02:00
Martin Kroeker
fc101b67e5 Merge pull request #23 from xianyi/develop
rebase
2021-04-19 22:24:12 +02:00
Martin Kroeker
b0239a05fd Merge pull request #3183 from martin-frbg/2715-x
Restore __volatile__ keyword in ARM64 DYNAMIC_ARCH detection mechanism
2021-04-16 14:52:12 +02:00
Martin Kroeker
623d580b4c Restore __volatile__ keyword 2021-04-16 10:27:32 +02:00
Martin Kroeker
974acb39ff Merge pull request #3181 from RajalakshmiSR/dgemmp10vp
POWER10: Improve dgemm performance
2021-04-14 22:43:02 +02:00
Rajalakshmi Srinivasaraghavan
2379abaa5e POWER10: Improve dgemm performance
This patch uses vector pair pointer for input load operation
which helps to generate power10 lxvp instructions.
2021-04-13 22:30:06 -05:00
Martin Kroeker
3caf781d7c Merge pull request #3179 from RajalakshmiSR/zgemvp10
POWER10: Optimized zgemv
2021-04-11 10:01:09 +02:00
Rajalakshmi Srinivasaraghavan
55bb9f639a POWER10: Optimized zgemv
This patch makes use of Matrix-Multiply Assist (MMA)
feature introduced in POWER ISA v3.1 for zgemv_n and zgemv_t.
2021-04-10 19:00:24 -05:00
Martin Kroeker
0dba04bb58 Merge pull request #3178 from martin-frbg/fix2864
Fix unwanted fallback to implicit typing in slanv2/dlanv2
2021-04-09 13:38:05 +02:00
Martin Kroeker
e96f5e3c65 Fix implicit typing of new variable TWO 2021-04-09 10:04:15 +02:00
Martin Kroeker
558724e99f Fix implicit typing of new variable TWO 2021-04-09 10:03:31 +02:00
Martin Kroeker
067c96a873 Merge pull request #3177 from martin-frbg/issue3176
Use "old" compute(24) function with clang due to register limitations
2021-04-07 08:22:42 +02:00
Martin Kroeker
4b380c0b40 Merge pull request #3175 from LYP951018/develop
Pass NO_AVX512 macro def when `DYNAMIC_ARCH` is enabled
2021-04-07 08:22:28 +02:00
Martin Kroeker
2dfb24730d Use "old" compute(24) function with clang due to register limitations 2021-04-06 19:58:32 +02:00