summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2017-07-25Merge pull request #1247 from martin-frbg/revert-cgroups-prsandbox/tyzeroy/develZhang Xianyi2-81/+5
Revert #1246, "honor cgroup/cpuset limits" for now
2017-07-25Revert #1246, "honor cgroup/cpuset limits" for nowMartin Kroeker2-81/+5
Unsafe usage of the __GLIBC_PREREQ macro lead to build breakage on non-glibc systems
2017-07-24Merge branch 'develop'v0.2.20upstream/0.2.20Zhang Xianyi4371-75017/+305046
0.2.20 version
2017-07-24Update doc for 0.2.20 version.Zhang Xianyi3-2/+43
2017-07-24Merge pull request #1239 from martin-frbg/cgroupsZhang Xianyi2-5/+81
Honor cgroup/cpuset limits when enumerating cpus
2017-07-24Merge pull request #1244 from martin-frbg/micmuc_cimatcopyZhang Xianyi1-7/+6
Fix complex imatcopy for Trans cases with non-square matrix
2017-07-21Use in-place transform shortcut only if matrix is squareMartin Kroeker1-6/+4
2017-07-20Add files via uploadMartin Kroeker1-6/+7
2017-07-15Exchange rows and cols in final omatcopy with BlasTransMartin Kroeker1-1/+1
This is MicMuc's patch from #899
2017-07-15More fixes for silly miseditsMartin Kroeker1-2/+4
2017-07-15Fixup braces lost in previous editMartin Kroeker1-8/+8
2017-07-15Merge branch 'develop' into cgroupsMartin Kroeker169-843/+22840
2017-07-13Disable ReLAPACK by default (#1238)Martin Kroeker3-2/+5
* Disable ReLAPACK by default; mention it in final build message if included * Add files via upload * Add files via upload * Add files via upload
2017-07-13Merge pull request #1214 from martin-frbg/relapackZhang Xianyi103-9/+21584
Initial import of ReLAPACK
2017-07-13Merge pull request #1234 from brada4/developZhang Xianyi6-6/+6
Fix write past fixed size buffer
2017-07-12Merge pull request #1235 from xianyi/revert-1233-cpuid-fixMartin Kroeker1-14/+12
Revert "Fix unintentional fall-through cases in get_cacheinfo"
2017-07-12Revert "Fix unintentional fall-through cases in get_cacheinfo"Martin Kroeker1-14/+12
2017-07-12Fix write past fixed size bufferAndrew6-6/+6
2017-07-11Add files via uploadMartin Kroeker0-0/+0
2017-07-11Add files via uploadMartin Kroeker1-5/+23
2017-07-11Honor cgroup/cpuset constraints when enumerating cpusMartin Kroeker2-26/+98
2017-07-11Merge pull request #1233 from martin-frbg/cpuid-fixMartin Kroeker1-12/+14
Fix unintentional fall-through cases in get_cacheinfo
2017-07-11Fix unintentional fall-through cases in get_cacheinfoMartin Kroeker1-12/+14
These appear to be unintended side effects of PR #1091, probably causing #1232
2017-07-10Merge pull request #1226 from ashwinyes/develop_arm_clang_ual_fixZhang Xianyi15-148/+148
arm: Fix clang compilation for ARMv7
2017-07-10Merge pull request #1221 from ashwinyes/develop_arm_softfpZhang Xianyi39-311/+683
arm: add support for softfp in arm vfp assembly files
2017-07-10Merge branch 'develop' into develop_arm_softfpZhang Xianyi21-370/+432
2017-07-09Merge pull request #1230 from martin-frbg/rhel5Martin Kroeker1-2/+19
Add sched_getcpu implementation for pre-2.6 glibc
2017-07-09Do not add -lpthread on Android builds (#1229)Martin Kroeker1-1/+1
* Do not add -lpthread on Android builds * Do not add -lpthread on Android cmake builds
2017-07-09Add sched_getcpu implementation for pre-2.6 glibcMartin Kroeker1-2/+19
Fixes #1210, compilation on RHEL5 with affinity enabled
2017-07-07Merge pull request #1225 from martin-frbg/stolen_from_wernsaar_forkZhang Xianyi1-2/+2
fixed syrk_thread.c taken from wernsaar
2017-07-07arm: Fix clang compilation for ARMv7Ashwin Sekhar T K15-148/+148
clang is not recognizing some pre-UAL VFP mnemonics like fnmacs, fnmacd, fnmuls and fnmuld. Replaced them with equivalent UAL mnemonics which are vmls.f32, vmls.f64, vnmul.f32 and vnmul.f64 respectively.
2017-07-06fixed syrk_thread.c taken from wernsaarMartin Kroeker1-2/+2
Stride calculation fix copied from https://github.com/wernsaar/OpenBLAS/commit/88900e1
2017-07-06Handle different object extensions in MakefileMartin Kroeker1-11/+45
The optimized LAPACK functions from interface use OS-dependent suffixes .o/.obj for the object files, while netlib LAPACK uses .o throughout. ReLAPACK object names have to match in order for function replacement in the growing library file to work.
2017-07-05Link -lm or -lm_hard for Android ARMv7.Zhang Xianyi1-0/+8
2017-07-03Merge pull request #1218 from m-brow/power9Zhang Xianyi9-328/+328
Optimise loads on Power9 LE
2017-07-03Merge pull request #1212 from neilsh-msft/developZhang Xianyi7-35/+73
Add Microsoft Windows 10 UWP build support
2017-07-02Update MakefileMartin Kroeker1-1/+1
2017-07-02Add files via uploadMartin Kroeker1-2/+3
2017-07-02arm: Remove unnecessary files/codeAshwin Sekhar T K5-1443/+8
Since softfp code has been added to all required vfp kernels, the code for auto detection of abi is no longer required. The option to force softfp ABI on make command line by giving ARM_SOFTFP_ABI=1 is retained. But there is no need to give this option anymore. Also the newly added C versions of 4x4/4x2 gemm/trmm kernels are removed. These are longer required. Moreover these kernels has bugs.
2017-07-02arm: add softfp support in zgemm/ztrmm vfp kernelsAshwin Sekhar T K6-18/+66
2017-07-02arm: add softfp support in cgemm/ctrmm vfp kernelsAshwin Sekhar T K6-4/+63
2017-07-02arm: add softfp support in dgemm/dtrmm vfp kernelsAshwin Sekhar T K6-14/+55
2017-07-02arm: add softfp support in sgemm/strmm vfp kernelsAshwin Sekhar T K6-22/+52
2017-07-02generic: Bug fixes in generic 4x2 and 4x4 gemm kernelsAshwin Sekhar T K2-67/+67
2017-07-02arm: add softfp support in vfp gemv kernelsAshwin Sekhar T K10-86/+252
2017-07-01Merge pull request #1220 from ashwinyes/develop_aarch64_20170701_t99_optionsMartin Kroeker1-2/+2
arm64: Change mtune/mcpu options for THUNDERX2T99 target
2017-07-01arm64: Change mtune/mcpu options for THUNDERX2T99 targetAshwin Sekhar T K1-2/+2
2017-07-01arm: add softfp support in kernel/arm/swap_vfp.SAshwin Sekhar T K2-5/+42
2017-07-01arm: add softfp support in kernel/arm/nrm2_vfp*.SAshwin Sekhar T K4-12/+24
2017-06-30arm: add softfp support in kernel/arm/*dot_vfp.SAshwin Sekhar T K5-24/+66