wernsaar
3e33afef2e
Merge pull request #592 from wernsaar/develop
added benchmark scripts
10 years ago
Werner Saar
8614057ea9
added benchmark scripts for numpy, octave and R
10 years ago
Werner Saar
7f375f9e8f
updated geev benchmark
10 years ago
wernsaar
69c5169e7d
Merge pull request #589 from wernsaar/develop
small modification of gemm.c
10 years ago
Werner Saar
e19948baa1
small modification of gemm.c
10 years ago
wernsaar
a2eaf234fc
Merge pull request #587 from wernsaar/develop
added gesv benchmark
10 years ago
Werner Saar
6a13a94e71
added gesv benchmark
10 years ago
wernsaar
eff43d3289
Merge pull request #585 from wernsaar/develop
bugfix for benchmark Makefile on MAC
10 years ago
Werner Saar
9c4817d07b
bugfix for Makefile on mac
10 years ago
wernsaar
319f3a0451
Merge pull request #584 from wernsaar/develop
bugfixes, to build benchmarks with mingw on Windows OS
10 years ago
Werner Saar
02c7766f68
bugfixes, to build benchmarks with mingw on Windows OS
10 years ago
wernsaar
f38cb67ca8
Merge pull request #581 from wernsaar/develop
bugfix for arm locking
10 years ago
Werner Saar
eea2e30b74
bugfix for arm locking
10 years ago
Werner Saar
19b8fd2aed
smp lock bugfix
10 years ago
wernsaar
0cc5212741
Merge pull request #580 from wernsaar/develop
added blas level1 swap benchmark
10 years ago
Werner Saar
c47c8e8cf5
added blas level1 swap benchmark
10 years ago
Zhang Xianyi
a11555c715
Support Android NDK armeabi-v7a-hard ABI. (-mfloat-abi=hard)
e.g.
make HOSTCC=gcc CC=arm-linux-androideabi-gcc NO_LAPACK=1 TARGET=ARMV7
In Android NDK, it uses armeabi-v7a-hard ABI.
TARGET_CFLAGS += -mhard-float -D_NDK_MATH_NO_SOFTFP=1
TARGET_LDFLAGS += -Wl,--no-warn-mismatch -lm_hard
For more information, please check hard-float example at
android_ndk/tests/device/hard-float/jni/.
10 years ago
wernsaar
897d03518e
Merge pull request #578 from wernsaar/develop
added blas level1 copy benchmark
10 years ago
Werner Saar
23fbc5728e
added blas level1 copy benchmark
10 years ago
Zhang Xianyi
6d40fa587f
Fix f_check bug.
10 years ago
wernsaar
22dcd79959
Merge pull request #577 from wernsaar/develop
Bugfix for armv6 memory barrier
10 years ago
Werner Saar
ea4df0aad3
Ref #574 : Bugfix for armv6 memory barrier
10 years ago
Zhang Xianyi
e127fb8fd8
1) Refs #575 . Remove g77 from compiler list.
2) If OpenBLAS cannot find Fortran compiler, it will only build BLAS
(without LAPACK).
10 years ago
wernsaar
7fb718a7d8
Merge pull request #572 from wernsaar/develop
added optimized cscal and zscal functions for steamroller
10 years ago
Werner Saar
24f58c8bb1
added optimized cscal and zscal kernels for steamroller
10 years ago
Werner Saar
95b1faf667
added optimized cscal and zscal kernels for steamroller and piledriver
10 years ago
Werner Saar
2d9e406050
added optimized cscal kernel for sandybridge
10 years ago
Werner Saar
59083e3ce1
added optimized cscal kernel for bulldozer
10 years ago
wernsaar
685be40339
Merge pull request #571 from wernsaar/develop
added optimized cscal and zscal functions
10 years ago
Werner Saar
31c9e399e9
added optimized cscal kernel for haswell
10 years ago
Werner Saar
7de6bb9889
added optimized zscal kernel for bulldozer
10 years ago
Werner Saar
d63034303b
added optimized zscal kernel for haswell
10 years ago
Zhang Xianyi
51ff17d46e
Add AMD Excavator target.
10 years ago
wernsaar
905534942a
Merge pull request #568 from wernsaar/develop
added optimized dscal kernel
10 years ago
Werner Saar
18e90ee2e3
bugfix: added static to functions
10 years ago
Werner Saar
e00cccc41e
added optimized dscal kernel for piledriver
10 years ago
Werner Saar
73f09bf64f
optimized dscal kernel for increment != 1
10 years ago
Werner Saar
02e772c7e4
added optimized dscal kernel for haswell
10 years ago
Werner Saar
7aee913991
added optimized dscal kernel for sandybridge
10 years ago
Werner Saar
e50a933037
added optimized dscal kernel for bulldozer
10 years ago
Zhang Xianyi
5f9011d6ef
Merge pull request #566 from powderluv/develop
Fix build with ALLOC_SHM=0 (Android NDK)
10 years ago
powderluv
ebb9eba987
Fix build with ALLOC_SHM=0 (Android NDK)
Refactor such that you can build with ALLOC_SHM=0. HughTLB
implicity depends on ALLOC_SHM=1. This patch allows
building for Android NDK r10d.
10 years ago
Zhang Xianyi
8e5a1083bb
Refs #532 . Improve gemv paralel with small m and large n case.
Splite the matrix and reduction.
10 years ago
Zhang Xianyi
6743beb748
Refs #565 . Fix the bug of generate FEXTRALIB.
10 years ago
Zhang Xianyi
bcabf72c08
Refs #565 . Merge branch 'andreasnoack-anj/bench' into develop
10 years ago
Andreas Noack
cda29f183b
Add vecLib benchmarks
10 years ago
wernsaar
e52d36450a
Merge pull request #564 from wernsaar/develop
Use only 1 thread in trsm if m or n < 2*GEMM_MULTITHREAD_THRESHOLD
10 years ago
Werner Saar
f8f2e261fe
use only 1 thread if m or n < 2*GEMM_MULTITHREAD_THRESHOLD
10 years ago
Werner Saar
be3c843700
added loops to trsm.c
10 years ago
wernsaar
e6f57db846
Merge pull request #563 from wernsaar/develop
Bugfix for gemm3m tests
10 years ago