a9b51b844
Merge pull request #1798 from martin-frbg/cmake-avx512 by
2018-10-08 21:15:17 +0200
eba394c71
(refs/pull/1798/head)
Add -march=skylake-avx512 when required by
2018-10-08 19:18:12 +0200
6b0c7c6d0
optimize thread lock on mips64 by
2018-10-08 16:06:43 +0800
582c58972
(refs/pull/1793/head)
dgemm/skylakex: replace discrete mul/add with fma by
2018-10-06 23:13:26 +0000
adbf6afa2
Add vector optimizations for ncopy as well for dgemm/skylakex by
2018-10-06 21:18:12 +0000
32bec8afb
add a skylakex optimized dgemm beta function by
2018-10-06 16:36:26 +0000
6e2c49455
Merge pull request #1791 from dev-zero/develop by
2018-10-06 16:29:29 +0200
20c5d668f
dgemm/avx512 simplify and speed up the 4x4 kernel by
2018-10-06 14:12:32 +0000
6d43c51cc
undo slow dgemm/skylake microoptimization by
2018-10-06 14:00:37 +0000
d74dc39b0
Add optimized *copy versions for skylakex by
2018-10-06 13:47:20 +0000
41951da6d
(refs/pull/1792/head)
Merge pull request #6 from xianyi/develop by
2018-10-06 14:36:36 +0200
474f7e958
Add SYMBOLPREFIX and -SUFFIX options and improve help output by
2018-10-06 14:28:04 +0200
79ea839b6
(refs/pull/1791/head)
fix parallel build issues with APFS/HFS+/ext2/3 in netlib-lapack by
2018-10-06 14:10:02 +0200
f7f97c614
Merge pull request #1789 from brada4/develop by
2018-10-05 20:42:37 +0200
6f22e1cfb
Merge pull request #1788 from fenrus75/avx512-8x16 by
2018-10-05 20:40:38 +0200
66b43affb
(refs/pull/1788/head)
Add a 24x8 kernel to the skylakex dgemm implementation by
2018-10-05 13:22:21 +0000
1938819c2
skylake dgemm: Add a 16x8 kernel by
2018-10-05 11:49:43 +0000
bda3dbe2e
(refs/pull/1789/head)
update travis alpine chroot with avx512 intrinsics headers by
2018-10-05 15:47:55 +0300
c3e0f0eb3
update travis alpine chroot with avx512 intrinsics headers by
2018-10-05 15:41:52 +0300
a980953bd
Merge pull request #1785 from brada4/develop by
2018-10-05 08:25:38 +0200
78c99d523
Merge pull request #1784 from fenrus75/dgemm-avx512 by
2018-10-05 08:03:27 +0200
b7496c363
(refs/pull/1784/head)
Function name needs to be CNAME, set from outside to allow suffixing for dynamic_arch by
2018-10-04 19:14:59 +0200
95f4e8757
Merge pull request #1787 from jeromerobert/develop by
2018-10-04 18:41:47 +0200
b095f2fad
(refs/pull/1787/head)
Fix unknown type name __WAIT_STATUS on RHEL5 by
2018-10-04 12:27:44 +0200
02ef20a1e
Merge pull request #1786 from martin-frbg/immintrin by
2018-10-04 09:07:09 +0200
4c3643ed7
(refs/pull/1786/head)
Check availability of immintrin.h in the AVX512 compatibility test by
2018-10-04 07:36:49 +0200
591cca7cb
Check availability of immintrin.h in the AVX512 compatibility test by
2018-10-04 07:35:30 +0200
3439158de
(refs/pull/1785/head)
address #1782 2nd loop by
2018-10-03 21:20:50 +0200
45fe8cb0c
Create a AVX512 enabled version of DGEMM by
2018-10-03 14:45:25 +0000
f27916d11
use blas lock on LOONGSON3A by
2018-09-30 16:55:13 +0800
bc024fe00
(refs/pull/1760/merge)
Merge 80f7c5daf5 into 544b069e85 by
2018-09-29 20:34:39 +0000
544b069e8
Merge pull request #1780 from martin-frbg/issue1774-2 by
2018-09-29 09:27:47 +0200
9b2a7ad40
(refs/pull/1780/head)
Convert fldmia/fstmia instructions to UAL syntax for clang7 by
2018-09-28 23:05:15 +0200
10ce70701
Merge pull request #1778 from fengrl/develop by
2018-09-26 11:14:58 +0200
6fc85a635
(refs/pull/1778/head)
test_axpy work error on LOONGSON3A platform #1777 by
2018-09-26 15:14:04 +0800
831c66138
Merge pull request #1775 from martin-frbg/issue1774 by
2018-09-25 18:58:39 +0200
7e5df34e6
(refs/pull/1775/head)
Convert fldmia/fstmia instructions to UAL syntax for clang7 by
2018-09-25 09:41:58 +0200
4f45040b8
Merge pull request #1773 from martin-frbg/issue1767 by
2018-09-23 23:25:15 +0200
e3bdfe28c
(refs/pull/1768/merge)
Merge 3ba26743ee into 56e7c68810 by
2018-09-22 16:56:53 +0000
28aa94bf4
(refs/pull/1773/head)
Include thread numbers in failure message from blas_thread_init by
2018-09-22 14:00:15 +0200
56e7c6881
Merge pull request #1771 from staticfloat/sf/ldflags by
2018-09-22 13:11:39 +0200
cf6df9464
Document the stub status of the QUAD_PRECiSION code (#1772) by
2018-09-22 12:31:37 +0200
b2b77f478
(refs/pull/1772/head)
Typo fix by
2018-09-21 22:40:04 +0200
e73c52de8
Document the stub status of the QUAD_PRECiSION code inherited from GotoBLAS2 by
2018-09-21 12:42:31 +0200
6f77af2ee
(refs/pull/1771/head)
Add `$(LDFLAGS)` to `$(CC)` and `$(FC)` invocations within `exports/Makefile` by
2018-09-21 09:19:51 +0000
3ba26743e
(refs/pull/1768/head)
Update trmv_U.c by
2018-09-21 08:23:05 +0200
a2fb50edb
Try harder at disabling gemv_n unrolling by
2018-09-20 22:50:25 +0200
4d183e556
Merge pull request #1765 from martin-frbg/issue1761 by
2018-09-19 22:02:21 +0200
34d55fd16
Merge pull request #1764 from yurivict/64-suffix by
2018-09-19 18:16:38 +0200
b99157021
Merge pull request #1762 from martin-frbg/issue1710-2 by
2018-09-19 18:16:21 +0200
288aeea8a
(refs/pull/1765/head)
Fix default settings - USE_TLS and USE_SIMPLE_THREADED_LEVEL3 should both be off by
2018-09-19 18:08:31 +0200
1ad1e7906
Catch inadvertent USE_TLS=0 declaration by
2018-09-19 18:03:43 +0200
b40262650
Do not use the new TLS code for non-threaded builds even if USE_TLS is set by
2018-09-16 12:43:36 +0200
ec0cac166
Merge pull request #4 from xianyi/develop by
2018-09-16 12:36:49 +0200
2349e1514
(refs/pull/1764/head)
Allow to install the 'interfare64' version concurrently with the regular version by
2018-09-15 19:59:17 -0700
f3c262156
(refs/pull/1762/head)
Add an explicit cast to silence a warning by
2018-09-13 14:24:29 +0200
30f5a69ab
Add explicit cast to silence a warning by
2018-09-13 14:23:31 +0200
80f7c5daf
(refs/pull/1760/head)
had to clean build tree before test build myself 2 by
2018-09-11 21:51:44 +0200
74c4b9165
had to clean build tree before test build myself by
2018-09-11 21:44:55 +0200
c88cd1555
remove unused 'levelness' parameter from num_cpu_avail() by
2018-09-11 21:23:58 +0200
fd081a91e
Merge pull request #1759 from martin-frbg/lapack283 by
2018-09-11 13:52:09 +0200
094f8c3b5
(refs/pull/1759/head)
remove unused variable ldb_t by
2018-09-11 10:53:47 +0200
5cf090f51
remove unused variable ldb_t by
2018-09-11 10:52:30 +0200
58363542e
remove unused variable ldb_t by
2018-09-11 10:51:17 +0200
3abc22a5b
Merge pull request #1757 from brada4/develop by
2018-09-09 22:55:15 +0200
1e531701b
(refs/pull/1757/head)
fix small typo by
2018-09-09 16:52:25 +0200
5d42b6ea0
Merge pull request #1756 from martin-frbg/issue1754 by
2018-09-07 11:02:18 +0200
ba4f43332
Merge pull request #1749 from martin-frbg/issue1531 by
2018-09-07 11:02:01 +0200
d93cf1126
Merge pull request #1753 from dloghin/risc-v by
2018-09-07 11:01:23 +0200
4cf7315a5
(refs/pull/1749/head)
Adjust ARMV8 SGEMM unrolling when using the C fallback kernel_2x2 for IOS by
2018-09-06 21:41:54 +0200
b57af9379
(refs/pull/1756/head)
just make CBLAS_LAYOUT an alias of the existing CBLAS_ORDER by
2018-09-06 16:54:31 +0200
8aeab0601
Follow netlib renaming/aliasing CBLAS_ORDER to CBLAS_LAYOUT by
2018-09-06 16:39:52 +0200
a1bdc308b
(refs/pull/1753/head)
override ARCH (archiver) in lapack-netlib/make.inc by
2018-09-06 13:13:36 +0800
0b7ccb9e3
Revert "replace ARCH with AR in lapack-netlib" by
2018-09-06 13:08:30 +0800
8450c13fb
(refs/pull/1752/head)
Introduce TARGET fallbacks in CMake by
2018-09-06 11:05:31 +1000
bb0d01e7b
(refs/pull/1751/merge)
Merge db17ce896f into cbc46163bd by
2018-09-05 04:54:54 +0000
db17ce896
(refs/pull/1751/head)
replace ARCH with AR in lapack-netlib by
2018-09-05 12:49:37 +0800
1cb7b9015
Conditional compilation of assembly files that IOS does not like by
2018-09-04 11:06:51 +0200
a4bd41e9f
Fix paths to C kernels for nrm2 by
2018-09-04 10:51:19 +0200
9e2bb0c64
Update with the changes from 0.3.3 by
2018-08-31 00:21:13 +0200
dbfd7524c
Update version to 0.3.4.dev by
2018-08-31 00:19:21 +0200
2982ce505
Update version to 0.3.4.dev by
2018-08-31 00:18:37 +0200
fd8d1868a
(tag: v0.3.3)
Updates for 0.3.3 by
2018-08-31 00:07:48 +0200
f0563f14b
Version 0.3.3 by
2018-08-30 23:43:57 +0200
3197f8676
Version 0.3.3 by
2018-08-30 23:43:14 +0200
422a8fa95
Merge pull request #1747 from xianyi/develop by
2018-08-30 23:42:19 +0200
5bac15adb
(refs/pull/1747/head)
Merge pull request #1746 from martin-frbg/issue1674 by
2018-08-30 17:48:07 +0200
e17f969fa
(refs/pull/1746/head)
Assume cross-compilation if host and target os differ by
2018-08-30 13:28:46 +0200
e11126b26
Merge pull request #1745 from martin-frbg/issue1743 by
2018-08-29 07:43:58 +0200
74608e470
Merge pull request #1744 from martin-frbg/lapack272 by
2018-08-28 22:58:58 +0200
f3fd44a73
(refs/pull/1745/head)
Set USE_TRMM for all ZARCH variants to fix TRMM faults with zarch-generic by
2018-08-28 21:34:07 +0200
9e917b16d
(refs/pull/1744/head)
Fix missing replacements of ILAENV by ILAENV_2STAGE (lapack PR 272) by
2018-08-28 21:11:54 +0200
8440a4cb1
Merge pull request #1742 from martin-frbg/interim033 by
2018-08-28 08:02:15 +0200
b55690a65
(refs/pull/1742/head)
typo fix by
2018-08-26 11:31:07 +0200
b902a4098
Rewrite glibc version check by
2018-08-26 11:18:02 +0200
5991d1a6c
Update memory.c by
2018-08-25 22:12:40 +0200
b1b743f43
Merge branch 'develop' into interim033 by
2018-08-25 19:45:19 +0200
2caa2210b
Add USE_TLS option to choose between old and new implementation of memory.c by
2018-08-25 19:37:11 +0200
2a589c4b2
Add USE_TLS option to switch between old and new memory.c by
2018-08-25 19:36:12 +0200
fd42ca462
Combo of default pre-0.3.1 memory.c and band-aided version of PR1739 by
2018-08-25 19:35:16 +0200