platform/upstream/openblas.git
9 years agoupdated haswell optimized sgmv_n kernel
wernsaar [Tue, 5 Aug 2014 06:04:47 +0000 (08:04 +0200)]
updated haswell optimized sgmv_n kernel

9 years agomodified sgemv_n for haswell
wernsaar [Mon, 4 Aug 2014 14:22:11 +0000 (16:22 +0200)]
modified sgemv_n for haswell

9 years agoadded a better optimized sgemv_n kernel for bulldozer and piledriver
wernsaar [Mon, 4 Aug 2014 12:29:01 +0000 (14:29 +0200)]
added a better optimized sgemv_n kernel for bulldozer and piledriver

10 years agoadded experimental support for big numa machines
wernsaar [Sat, 2 Aug 2014 11:40:16 +0000 (13:40 +0200)]
added experimental support for big numa machines

10 years agobugfix for linux affinity code
wernsaar [Fri, 1 Aug 2014 21:10:08 +0000 (23:10 +0200)]
bugfix for linux affinity code

10 years agoadded benchmarks for lapack potrf, potrs and potri functions
wernsaar [Fri, 1 Aug 2014 19:08:37 +0000 (21:08 +0200)]
added benchmarks for lapack potrf, potrs and potri functions

10 years agoadded benchmark program for lapack ?getri functions
wernsaar [Fri, 1 Aug 2014 06:55:20 +0000 (08:55 +0200)]
added benchmark program for lapack ?getri functions

10 years agobugfix for zgeev
wernsaar [Thu, 31 Jul 2014 10:35:38 +0000 (12:35 +0200)]
bugfix for zgeev

10 years agoadded lapack geev benchmark
wernsaar [Thu, 31 Jul 2014 08:35:25 +0000 (10:35 +0200)]
added lapack geev benchmark

10 years agobugfix in cholesky.c
wernsaar [Wed, 30 Jul 2014 12:00:19 +0000 (14:00 +0200)]
bugfix in cholesky.c

10 years agoextented plot-filter.sh for linpack and cholesky benchmarks
wernsaar [Wed, 30 Jul 2014 11:03:42 +0000 (13:03 +0200)]
extented plot-filter.sh for linpack and cholesky benchmarks

10 years agooptimization of sandybridge cgemm-kernel
wernsaar [Tue, 29 Jul 2014 17:07:21 +0000 (19:07 +0200)]
optimization of sandybridge cgemm-kernel

10 years agoMerge pull request #421 from wernsaar/develop
Zhang Xianyi [Tue, 29 Jul 2014 07:50:00 +0000 (15:50 +0800)]
Merge pull request #421 from wernsaar/develop

optimized sgemm- and cgemm-kernel for haswell

10 years agooptimized cgemm kernel for haswell
wernsaar [Tue, 29 Jul 2014 06:53:09 +0000 (08:53 +0200)]
optimized cgemm kernel for haswell

10 years agooptimized sgemm kernel for haswell
wernsaar [Mon, 28 Jul 2014 09:50:01 +0000 (11:50 +0200)]
optimized sgemm kernel for haswell

10 years agoMerge pull request #420 from wernsaar/develop
Zhang Xianyi [Sun, 27 Jul 2014 15:30:14 +0000 (23:30 +0800)]
Merge pull request #420 from wernsaar/develop

Optimizations for HASWELL

10 years agooptimized zgemm kernel for haswell
wernsaar [Sun, 27 Jul 2014 09:51:42 +0000 (11:51 +0200)]
optimized zgemm kernel for haswell

10 years agooptimizations for trsm
wernsaar [Fri, 25 Jul 2014 09:59:17 +0000 (11:59 +0200)]
optimizations for trsm

10 years agooptimizations for trmm
wernsaar [Fri, 25 Jul 2014 08:00:23 +0000 (10:00 +0200)]
optimizations for trmm

10 years agoallow to set custom value for ?GEMM_DEFAULT_UNROLL_MN, optimizations for syrk
wernsaar [Thu, 24 Jul 2014 16:43:31 +0000 (18:43 +0200)]
allow to set custom value for ?GEMM_DEFAULT_UNROLL_MN, optimizations for syrk

10 years agoadded a sample plot-filter scripts and a header file for gnuplot
wernsaar [Mon, 21 Jul 2014 12:50:24 +0000 (14:50 +0200)]
added a sample plot-filter scripts and a header file for gnuplot

10 years agoadded symv benchmark
wernsaar [Mon, 21 Jul 2014 05:50:54 +0000 (07:50 +0200)]
added symv benchmark

10 years agoadd benchmark for ger
wernsaar [Mon, 21 Jul 2014 04:25:42 +0000 (06:25 +0200)]
add benchmark for ger

10 years agoMerge pull request #419 from wernsaar/develop
Zhang Xianyi [Sun, 20 Jul 2014 15:35:17 +0000 (23:35 +0800)]
Merge pull request #419 from wernsaar/develop

added optimized sgemv kernels for Sandy Bridge, Haswell, Bullldozer, and Piledriver.

10 years agoadded sgemv_n microkernel for haswell
wernsaar [Sun, 20 Jul 2014 12:52:25 +0000 (14:52 +0200)]
added sgemv_n microkernel for haswell

10 years agooptimized sgemv_n_microk_sandy.c
wernsaar [Sun, 20 Jul 2014 12:08:04 +0000 (14:08 +0200)]
optimized sgemv_n_microk_sandy.c

10 years agoadded sgemv_n microkernel for sandybridge
wernsaar [Sun, 20 Jul 2014 11:17:47 +0000 (13:17 +0200)]
added sgemv_n microkernel for sandybridge

10 years agoadded sgemv_t microkernel for haswell
wernsaar [Sun, 20 Jul 2014 09:30:32 +0000 (11:30 +0200)]
added sgemv_t microkernel for haswell

10 years agobugfix in sgemv_t_microk_sandy.c
wernsaar [Sun, 20 Jul 2014 08:48:41 +0000 (10:48 +0200)]
bugfix in sgemv_t_microk_sandy.c

10 years agoadded sgemv_t microkernel for sandybridge
wernsaar [Sun, 20 Jul 2014 08:21:08 +0000 (10:21 +0200)]
added sgemv_t microkernel for sandybridge

10 years agoadded optimized sgemv_t for bulldozer and piledriver
wernsaar [Sat, 19 Jul 2014 13:48:07 +0000 (15:48 +0200)]
added optimized sgemv_t for bulldozer and piledriver

10 years agodon't use this sgemv_n on Windows
wernsaar [Sat, 19 Jul 2014 05:15:34 +0000 (07:15 +0200)]
don't use this sgemv_n on Windows

10 years agoMerge pull request #414 from staticfloat/sf/symlinkfix
Zhang Xianyi [Fri, 18 Jul 2014 15:13:18 +0000 (23:13 +0800)]
Merge pull request #414 from staticfloat/sf/symlinkfix

Don't create an absolute symlink when installing on Darwin

10 years agoperformance optimizations for sgemv_n
wernsaar [Fri, 18 Jul 2014 09:25:21 +0000 (11:25 +0200)]
performance optimizations for sgemv_n

10 years agoadded blocked sgemv_n and microkernel for bulldozer and piledriver
wernsaar [Thu, 17 Jul 2014 21:15:07 +0000 (23:15 +0200)]
added blocked sgemv_n and microkernel for bulldozer and piledriver

10 years agoRefs #415. Fixed the x86/i386 compiling bug with DYNAMIC_ARCH=1.
Zhang Xianyi [Thu, 17 Jul 2014 07:02:01 +0000 (15:02 +0800)]
Refs #415. Fixed the x86/i386 compiling bug with DYNAMIC_ARCH=1.

10 years agoDon't create an absolute symlink when installing on Darwin
Elliot Saba [Wed, 16 Jul 2014 19:31:27 +0000 (15:31 -0400)]
Don't create an absolute symlink when installing on Darwin

10 years agochanged string GFORTRAN to lowercase
wernsaar [Wed, 16 Jul 2014 15:08:43 +0000 (17:08 +0200)]
changed string GFORTRAN to lowercase

10 years agoOpenBLAS 0.2.10 version.
Zhang Xianyi [Wed, 16 Jul 2014 10:04:18 +0000 (18:04 +0800)]
OpenBLAS 0.2.10 version.

10 years agoRefs #401. Added NO_AVX2 flag for old binutils (e.g. RHEL6)
Zhang Xianyi [Wed, 16 Jul 2014 00:37:55 +0000 (08:37 +0800)]
Refs #401. Added NO_AVX2 flag for old binutils (e.g. RHEL6)

10 years agoadjust number of threads for small size in cgemv and zgemv
wernsaar [Tue, 15 Jul 2014 14:27:02 +0000 (16:27 +0200)]
adjust number of threads for small size in cgemv and zgemv

10 years agoadjust number of threads for sgemv and dgemv
wernsaar [Tue, 15 Jul 2014 14:04:46 +0000 (16:04 +0200)]
adjust number of threads for sgemv and dgemv

10 years agoadjusted number of threads for small size
wernsaar [Tue, 15 Jul 2014 12:41:35 +0000 (14:41 +0200)]
adjusted number of threads for small size

10 years agoadded benchmark for gemv
wernsaar [Tue, 15 Jul 2014 11:35:36 +0000 (13:35 +0200)]
added benchmark for gemv

10 years agoMerge pull request #413 from wernsaar/develop
Zhang Xianyi [Mon, 14 Jul 2014 14:39:22 +0000 (22:39 +0800)]
Merge pull request #413 from wernsaar/develop

additional benchmarks

10 years agoadded additional test value to dstest.in
wernsaar [Sun, 13 Jul 2014 16:29:19 +0000 (18:29 +0200)]
added additional test value to dstest.in

10 years agoadded additional test value
wernsaar [Sun, 13 Jul 2014 16:26:38 +0000 (18:26 +0200)]
added additional test value

10 years agosegment violation in x86_64 sgemv kernels
wernsaar [Sun, 13 Jul 2014 08:49:43 +0000 (10:49 +0200)]
segment violation in x86_64 sgemv kernels

10 years agosegment violation in sgemv kernels
wernsaar [Sun, 13 Jul 2014 08:46:14 +0000 (10:46 +0200)]
segment violation in sgemv kernels

10 years agomodified pathes to atlas, mkl and acml
wernsaar [Sat, 12 Jul 2014 14:20:29 +0000 (16:20 +0200)]
modified pathes to atlas, mkl and acml

10 years agoadded conf option for number of loops
wernsaar [Sat, 12 Jul 2014 09:54:39 +0000 (11:54 +0200)]
added conf option for number of loops

10 years agoadded her2k benchmark
wernsaar [Fri, 11 Jul 2014 14:31:05 +0000 (16:31 +0200)]
added her2k benchmark

10 years agoadded herk benchmark
wernsaar [Fri, 11 Jul 2014 14:16:48 +0000 (16:16 +0200)]
added herk benchmark

10 years agoadd hemm benchmark
wernsaar [Fri, 11 Jul 2014 13:26:34 +0000 (15:26 +0200)]
add hemm benchmark

10 years agoadded syr2k benchmark
wernsaar [Fri, 11 Jul 2014 12:48:25 +0000 (14:48 +0200)]
added syr2k benchmark

10 years agoadded syrk benchmark
wernsaar [Fri, 11 Jul 2014 12:21:25 +0000 (14:21 +0200)]
added syrk benchmark

10 years agoadded trsm benchmark
wernsaar [Fri, 11 Jul 2014 11:51:08 +0000 (13:51 +0200)]
added trsm benchmark

10 years agoadded trmm benchmark
wernsaar [Fri, 11 Jul 2014 11:20:42 +0000 (13:20 +0200)]
added trmm benchmark

10 years agoadded benchmark for symm
wernsaar [Fri, 11 Jul 2014 10:47:48 +0000 (12:47 +0200)]
added benchmark for symm

10 years agoadded gemm benchmark and modified Makefile for benchmark
wernsaar [Fri, 11 Jul 2014 09:09:47 +0000 (11:09 +0200)]
added gemm benchmark and modified Makefile for benchmark

10 years agoMerge pull request #411 from wernsaar/develop
Zhang Xianyi [Thu, 10 Jul 2014 14:38:15 +0000 (22:38 +0800)]
Merge pull request #411 from wernsaar/develop

Lapack-test on x86 32bit now runs without errors.

10 years agoRef #410: disabled optimized potri functions ( single threading bug)
wernsaar [Thu, 10 Jul 2014 11:42:32 +0000 (13:42 +0200)]
Ref #410: disabled optimized potri functions ( single threading bug)

10 years agoLapack-test Windows 32bit now error free
wernsaar [Thu, 10 Jul 2014 09:01:47 +0000 (11:01 +0200)]
Lapack-test Windows 32bit now error free

10 years agoLapack-test: cleanup of x86 32bit KERNEL file
wernsaar [Wed, 9 Jul 2014 14:08:19 +0000 (16:08 +0200)]
Lapack-test: cleanup of x86 32bit KERNEL file

10 years agoMerge pull request #409 from wernsaar/develop
Zhang Xianyi [Wed, 9 Jul 2014 13:11:00 +0000 (21:11 +0800)]
Merge pull request #409 from wernsaar/develop

some fixes for Lapack and ARM platform

10 years agobugfixes for lapack on ARM Platform
wernsaar [Wed, 9 Jul 2014 10:21:39 +0000 (12:21 +0200)]
bugfixes for lapack on ARM Platform

10 years agoOpenBLAS 0.2.10 rc2 version.
Zhang Xianyi [Wed, 9 Jul 2014 00:47:36 +0000 (08:47 +0800)]
OpenBLAS 0.2.10 rc2 version.

10 years agoadded cross compiler examples for 32bit and 64bit ARM
wernsaar [Tue, 8 Jul 2014 10:55:18 +0000 (12:55 +0200)]
added cross compiler examples for 32bit and 64bit ARM

10 years agoRefs #406. Fixed utest building bug.
Zhang Xianyi [Tue, 8 Jul 2014 09:26:49 +0000 (17:26 +0800)]
Refs #406. Fixed utest building bug.

10 years agoLapack bug114: replaced cgesvd.f and zgesvd.f
wernsaar [Tue, 8 Jul 2014 08:21:10 +0000 (10:21 +0200)]
Lapack bug114: replaced cgesvd.f and zgesvd.f

10 years agoLapack bug117: replaced zstemr.f
wernsaar [Tue, 8 Jul 2014 08:08:34 +0000 (10:08 +0200)]
Lapack bug117: replaced zstemr.f

10 years agoLapack bug118: replaced clanhf.f and zlanhf.f
wernsaar [Tue, 8 Jul 2014 07:57:40 +0000 (09:57 +0200)]
Lapack bug118: replaced clanhf.f and zlanhf.f

10 years agoFixed #407. Support outputing the CPU corename on runtime.
Zhang Xianyi [Tue, 8 Jul 2014 04:48:08 +0000 (12:48 +0800)]
Fixed #407. Support outputing the CPU corename on runtime.
The user can use char * openblas_get_config() or char * openblas_get_corename().

10 years agoMerge pull request #404 from wernsaar/develop
Zhang Xianyi [Sun, 6 Jul 2014 16:39:33 +0000 (00:39 +0800)]
Merge pull request #404 from wernsaar/develop

A lot of fixes for v0.2.10-rc2

10 years agoremoved reference to daxpy_bulldozer kernel (Windows bug in lapack-test)
wernsaar [Sun, 6 Jul 2014 14:39:32 +0000 (16:39 +0200)]
removed reference to daxpy_bulldozer kernel (Windows bug in lapack-test)

10 years agobugfix for fortran compiler
wernsaar [Sun, 6 Jul 2014 11:33:42 +0000 (13:33 +0200)]
bugfix for fortran compiler

10 years agoadded definitions for PILEDRIVER and HASWELL
wernsaar [Sun, 6 Jul 2014 10:08:27 +0000 (12:08 +0200)]
added definitions for PILEDRIVER and HASWELL

10 years agobugfix for CORE2
wernsaar [Sun, 6 Jul 2014 09:47:28 +0000 (11:47 +0200)]
bugfix for CORE2

10 years agofallback to zgemm_kernel_4x2_sse.S
wernsaar [Sun, 6 Jul 2014 09:05:28 +0000 (11:05 +0200)]
fallback to zgemm_kernel_4x2_sse.S

10 years agoadded missing definition for DUNNINGTON
wernsaar [Sun, 6 Jul 2014 08:17:07 +0000 (10:17 +0200)]
added missing definition for DUNNINGTON

10 years agoremoved reference to zgemm_kernel_4x2_sse3.S (bug in lapack-test)
wernsaar [Sat, 5 Jul 2014 14:13:17 +0000 (16:13 +0200)]
removed reference to zgemm_kernel_4x2_sse3.S (bug in lapack-test)

10 years agoenabled compiling of *3M functions
wernsaar [Wed, 2 Jul 2014 12:11:53 +0000 (14:11 +0200)]
enabled compiling of *3M functions

10 years agofixed my bug in ger.c
wernsaar [Wed, 2 Jul 2014 08:39:33 +0000 (10:39 +0200)]
fixed my bug in ger.c

10 years agodisabled *3M functions for x86_64 platforms
wernsaar [Tue, 1 Jul 2014 14:18:05 +0000 (16:18 +0200)]
disabled *3M functions for x86_64 platforms

10 years agoadded optimized sdot- and dsdot-kernel, written in C
wernsaar [Mon, 30 Jun 2014 12:46:38 +0000 (14:46 +0200)]
added optimized sdot- and dsdot-kernel, written in C

10 years agodisabled SMP for sbmv and zsbmv again
wernsaar [Sun, 29 Jun 2014 19:18:38 +0000 (21:18 +0200)]
disabled SMP for sbmv and zsbmv again

10 years agoenabled SMP for sbmv and zsbmv, but only for 64bit binaries
wernsaar [Sun, 29 Jun 2014 18:35:56 +0000 (20:35 +0200)]
enabled SMP for sbmv and zsbmv, but only for 64bit binaries

10 years agoenabled smp for ger.c and zger.c, but only for 64bit binaries
wernsaar [Sun, 29 Jun 2014 14:43:04 +0000 (16:43 +0200)]
enabled smp for ger.c and zger.c, but only for 64bit binaries

10 years agomodification, to run blas-test on Windows
wernsaar [Sun, 29 Jun 2014 08:15:29 +0000 (10:15 +0200)]
modification, to run blas-test on Windows

10 years agoOpenBLAS 0.2.10 rc1 version.
Zhang Xianyi [Sun, 29 Jun 2014 02:45:50 +0000 (10:45 +0800)]
OpenBLAS 0.2.10 rc1 version.

10 years agoMerge branch 'wernsaar-develop' into develop
Zhang Xianyi [Sun, 29 Jun 2014 02:40:54 +0000 (10:40 +0800)]
Merge branch 'wernsaar-develop' into develop

10 years agoFixed #395. Enable optimized cgemm for Sandybridge. Added optimized sdot kernel.
Zhang Xianyi [Sun, 29 Jun 2014 02:34:51 +0000 (10:34 +0800)]
Fixed #395. Enable optimized cgemm for Sandybridge. Added optimized sdot kernel.
Fixed c/zgemm, zgemv computational error of haswell, piledriver, bullldozer, and
barcelona on Windows.

Merge branch 'develop' of https://github.com/wernsaar/OpenBLAS into wernsaar-develop

Conflicts:
kernel/Makefile.L1
kernel/x86_64/KERNEL
param.h

10 years agofixed zgemv bug for older AMD Processors
wernsaar [Sat, 28 Jun 2014 17:04:49 +0000 (19:04 +0200)]
fixed zgemv bug for older AMD Processors

10 years agoMerge branch 'TimothyGu-develop' into develop
Zhang Xianyi [Sat, 28 Jun 2014 12:52:07 +0000 (20:52 +0800)]
Merge branch 'TimothyGu-develop' into develop
Fixed #398. Remove all trailing whitespace except lapack-netlib.

10 years agoMerge branch 'develop' of https://github.com/TimothyGu/OpenBLAS into TimothyGu-develop
Zhang Xianyi [Sat, 28 Jun 2014 12:51:31 +0000 (20:51 +0800)]
Merge branch 'develop' of https://github.com/TimothyGu/OpenBLAS into TimothyGu-develop

Conflicts:
driver/others/memory.c

10 years agoMerge pull request #399 from TimothyGu/upstr
Zhang Xianyi [Sat, 28 Jun 2014 12:40:23 +0000 (20:40 +0800)]
Merge pull request #399 from TimothyGu/upstr

Build import libs as .dll.a instead of .lib

10 years agoMerge pull request #397 from vtjnash/develop
Zhang Xianyi [Sat, 28 Jun 2014 12:38:48 +0000 (20:38 +0800)]
Merge pull request #397 from vtjnash/develop

fix #394

10 years agobugfix for barcelona zgemv-kernel
wernsaar [Sat, 28 Jun 2014 10:36:11 +0000 (12:36 +0200)]
bugfix for barcelona zgemv-kernel

10 years agobugfix for bulldozer cgemm-, zgemm- and zgemv-kernel
wernsaar [Sat, 28 Jun 2014 10:16:20 +0000 (12:16 +0200)]
bugfix for bulldozer cgemm-, zgemm- and zgemv-kernel

10 years agobugfix for piledriver cgemm-, zgemm- and zgemv-kernel
wernsaar [Sat, 28 Jun 2014 09:46:58 +0000 (11:46 +0200)]
bugfix for piledriver cgemm-, zgemm- and zgemv-kernel