Martin Kroeker
680f744abf
Merge pull request #108 from xianyi/develop
rebase
5 years ago
Martin Kroeker
6f9460f0f6
Merge pull request #2937 from martin-frbg/pwr-buffersz
Increase and unify BUFFERSIZE on POWER;fix gcc inline warning
5 years ago
Martin Kroeker
6c970fa998
Merge pull request #2938 from martin-frbg/2934-3
Fix twisted spelling that broke the gfortran version test again
5 years ago
Martin Kroeker
b23cb05231
Fix twisted spelling that broke the gfortran version test again
5 years ago
Martin Kroeker
1d4c96fa0c
Increase BUFFERSIZE further
5 years ago
Martin Kroeker
34c3c407ef
label always_inline function as inline to silence a gcc warning
5 years ago
Martin Kroeker
3f84a9ca15
Merge pull request #2936 from martin-frbg/issue2934-2
Fix compiler version check for -mavx2 support (DYNAMIC_ARCH case)
5 years ago
Martin Kroeker
7e265c50bf
Merge pull request #2935 from martin-frbg/lapack458
Fix macro used in argument conversion (LAPACK PR 458)
5 years ago
Martin Kroeker
ee90f30384
Increase BUFFERSIZE for POWER8-10 and use same value for POWER6
to fix overflow warning for PWR8 ZGEMM and PWR9 C/ZGEMM and avoid size mismatches in DYNAMIC_ARCH
5 years ago
Martin Kroeker
2e48d560ba
Fix compiler version check
5 years ago
Martin Kroeker
ab7f466467
Merge pull request #106 from xianyi/develop
rebase
5 years ago
Martin Kroeker
f95031204e
Fix macro used in argument conversion (LAPACK PR 458)
5 years ago
Martin Kroeker
909068facf
Merge pull request #2932 from RajalakshmiSR/copyp10
Optimize scopy/ccopy for POWER10
5 years ago
Martin Kroeker
5b7438fdde
Merge pull request #2934 from thrasibule/improve_version_check
actually check that version is greater than 4.7
5 years ago
Guillaume Horel
47696b43e9
actually check that version is greater than 4.7
5 years ago
Rajalakshmi Srinivasaraghavan
ad745c0bae
Optimize scopy/ccopy for POWER10
This patch makes use of new POWER10 vector pair instructions for
loads and stores. Also reorganized all variants of copy functions
to make use of same kernel.
5 years ago
Martin Kroeker
17c46bf06a
Merge pull request #2930 from ismail/fix-no-return
Fix build with -Werror=return-type
5 years ago
Martin Kroeker
28242096cd
Merge pull request #2928 from martin-frbg/issue2917
Enable -mavx2 for flang as well where supported
5 years ago
İsmail Dönmez
4a1d00f589
Fix build with -Werror=return-type
dgemm_tcopy_16_skylakex.c CNAME function should return an int, add a
return 0 similar to other files.
5 years ago
Martin Kroeker
00813363be
Enable -mavx2 for flang as well
5 years ago
Martin Kroeker
336e35469a
Merge pull request #105 from xianyi/develop
rebase
5 years ago
Martin Kroeker
29668458f7
Merge pull request #2925 from martin-frbg/issue2911-2
Add binutils version check as prerequisite for POWER10 in DYNAMIC_ARCH build
5 years ago
Martin Kroeker
ee83e29046
Merge pull request #2926 from bartoldeman/vzeroupper-clobber-all
x86_64: clobber all xmm registers after vzeroupper
5 years ago
Martin Kroeker
1a0f57c8f0
Fix missing backquotes
5 years ago
Bart Oldeman
b073d759d0
x86_64: clobber all xmm registers after vzeroupper
As observed using GCC 10 using -march=native -ftree-vectorize
on Knights Landing, it is now smart enough to find clobbers inside
non-inlined static functions.
In particular, sgemv counted on a kernel to preserve the whole
%ymm2 register (since it was not in the clobber list), but the top
part was destroyed by vzeroupper. This caused many tests to fail.
This patch makes sure all xmm (and ymm/zmm by extension) registers
are listed as clobbered to avoid this happening, as most kernels
already did correctly in fact.
5 years ago
Martin Kroeker
eddc65c7b7
Add POWER10 support flag (unconditionally for now)
5 years ago
Martin Kroeker
bb8c3f6861
Add ld/binutils version check for POWER10 support
5 years ago
Martin Kroeker
ff65952e46
Move HAVE_P10_SUPPORT to the build system
to be able to include a binutils version check
5 years ago
Martin Kroeker
6208c9899e
Merge pull request #104 from xianyi/develop
rebase
5 years ago
Martin Kroeker
8e20ab21c8
Merge pull request #2924 from martin-frbg/issue2920
Put back all symbols accidentally dropped in the reorganization of gensymbol
5 years ago
Martin Kroeker
dc6e44c3f8
Merge pull request #2916 from martin-frbg/issue2911
Clean up duplicate definitions in POWER8 kernels and fix power10 option passing
5 years ago
Martin Kroeker
4ad33c46b0
Add back symbols that got dropped when splitting by type
5 years ago
Martin Kroeker
fe2a922ada
Add POWER10 compiler options to CCOMMON_OPT rather than COMMON_OPT
5 years ago
Martin Kroeker
9cac379655
Merge pull request #103 from xianyi/develop
rebase
5 years ago
Martin Kroeker
a61c086408
Fix spurious trailing whitespace in comment
5 years ago
Martin Kroeker
5b9ebe4f8a
Merge pull request #2919 from isuruf/export
Fix exporting some lapack and cblas symbols
5 years ago
Martin Kroeker
7eddaf0d6f
Remove -mmma again (reduntant with cpu=power10) and add override statements
5 years ago
Isuru Fernando
14b1d33933
Fix exporting some lapack and cblas
5 years ago
Martin Kroeker
77669b019d
Merge pull request #2915 from bartoldeman/no-empty_sgemm_direct_skylakex
sgemm_direct_skylakex: fix 75eeb26 regression.
5 years ago
Martin Kroeker
5e8ddc9001
Merge pull request #2913 from martin-frbg/issue2910
Support cross-compiling for Apple Vortex
5 years ago
Bart Oldeman
03e781b766
sgemm_direct_skylakex: fix 75eeb26 regression.
The
`#if defined(SKYLAKEX) || defined (COOPERLAKE)`
from that commit was before #include "common.h" so caused the
compiled function to be empty, returning garbage results for
qualifying sgemm's on those architectures.
Closes #2914
5 years ago
Martin Kroeker
f1a4071d8c
Clean up STACKSIZE redefinition
5 years ago
Martin Kroeker
97cf10062f
Clean up STACKSIZE redefinition
5 years ago
Martin Kroeker
17e288e18d
Clean up STACKSIZE redefinition
5 years ago
Martin Kroeker
c1422f3e46
Clean up STACKSIZE redefinition
5 years ago
Martin Kroeker
d85b24e103
Clean up STACKSIZE redefinition
5 years ago
Martin Kroeker
7d6c85f9da
Add compiler option -mmma for POWER10
5 years ago
Martin Kroeker
2e7ee7c716
Fix naming of L2 cache size item reported for Vortex
5 years ago
Martin Kroeker
efd47b0104
Merge pull request #2909 from isuruf/patch-1
Need a space when redirecting to file
5 years ago
Martin Kroeker
f5902ab0a1
Support cross-compiling for Apple Vortex
5 years ago