platform/upstream/openblas.git
4 years agoImplementaion of dasum, sasum with AVX2 & AVX512 intrinsic
Gengxin Xie [Fri, 21 Aug 2020 06:44:36 +0000 (14:44 +0800)]
Implementaion of dasum, sasum with AVX2 & AVX512 intrinsic

4 years ago[WIP] Refactor the driver code for direct SGEMM (#2782)
Martin Kroeker [Wed, 19 Aug 2020 12:51:09 +0000 (14:51 +0200)]
[WIP] Refactor the driver code for direct SGEMM (#2782)

Move "direct SGEMM" functionality out of the SkylakeX SGEMM kernel and make it available
(on x86_64 targets only for now) in DYNAMIC_ARCH builds
* Add  sgemm_direct targets in the kernel Makefile.L3 and CMakeLists.txt
* Add direct_sgemm functions to the gotoblas struct in common_param.h
* Move sgemm_direct_performant helper to separate file
* Update gemm.c  to macros for sgemm_direct to support dynamic_arch naming via common_s,h
* (Conditionally) add sgemm_direct functions in setparam-ref.c

4 years agoMerge pull request #2785 from albertziegenhagel/always-generate-pkg-config
Martin Kroeker [Wed, 19 Aug 2020 12:42:58 +0000 (14:42 +0200)]
Merge pull request #2785 from albertziegenhagel/always-generate-pkg-config

Do not require pkg-config to generate the *.pc file

4 years agoDo not require pkg-config to generate the *.pc file
Albert Ziegenhagel [Tue, 18 Aug 2020 06:48:48 +0000 (08:48 +0200)]
Do not require pkg-config to generate the *.pc file

Generating the pkg-config file does not actually depend on pkg-config being available.

4 years agoMerge pull request #2784 from martin-frbg/issue2783
Martin Kroeker [Mon, 17 Aug 2020 17:06:13 +0000 (19:06 +0200)]
Merge pull request #2784 from martin-frbg/issue2783

Add fallback typedef for bfloat16 to openblas_config.h template

4 years agoAdd typedef for bfloat16 if needed
Martin Kroeker [Mon, 17 Aug 2020 13:32:14 +0000 (15:32 +0200)]
Add typedef for bfloat16 if needed

4 years agoMerge pull request #77 from xianyi/develop
Martin Kroeker [Mon, 17 Aug 2020 13:28:15 +0000 (15:28 +0200)]
Merge pull request #77 from xianyi/develop

rebase

4 years agorevert
Martin Kroeker [Mon, 17 Aug 2020 13:20:41 +0000 (15:20 +0200)]
revert

4 years agorevert
Martin Kroeker [Mon, 17 Aug 2020 13:20:16 +0000 (15:20 +0200)]
revert

4 years agorevert
Martin Kroeker [Mon, 17 Aug 2020 13:19:40 +0000 (15:19 +0200)]
revert

4 years agoUpdate .drone.yml
Martin Kroeker [Sat, 15 Aug 2020 13:46:18 +0000 (15:46 +0200)]
Update .drone.yml

4 years agoUpdate Makefile
Martin Kroeker [Sat, 15 Aug 2020 12:46:26 +0000 (14:46 +0200)]
Update Makefile

4 years agoAdd simple MT sgemm precision test and INTERFACE64 build
Martin Kroeker [Sat, 15 Aug 2020 11:38:05 +0000 (13:38 +0200)]
Add simple MT sgemm precision test and INTERFACE64 build

4 years agoAdd simple sgemm preicsion test
Martin Kroeker [Sat, 15 Aug 2020 11:33:52 +0000 (13:33 +0200)]
Add simple sgemm preicsion test

4 years agoUpdate gemm64.cpp
Martin Kroeker [Sat, 15 Aug 2020 11:31:28 +0000 (13:31 +0200)]
Update gemm64.cpp

4 years agoAdd trivial gemm test for multithread consistency
Martin Kroeker [Sat, 15 Aug 2020 11:30:29 +0000 (13:30 +0200)]
Add trivial gemm test for multithread consistency

4 years agoAdd a dedicated POWER9 build to the Travis CI (#2774)
Martin Kroeker [Wed, 12 Aug 2020 21:08:38 +0000 (23:08 +0200)]
Add a dedicated POWER9 build to the Travis CI (#2774)

* Add dedicated POWER9 build (using new syntax to ensure it runs as a P9-only containerized job rather than a VM that
might end up on P8 hardware half of the time)
* Bump gcc version for POWER9 build

4 years agoMerge pull request #2765 from martin-frbg/issue2760
Martin Kroeker [Tue, 11 Aug 2020 20:40:17 +0000 (22:40 +0200)]
Merge pull request #2765 from martin-frbg/issue2760

Add memory barrier to the PPC blas_lock implementation for Linux

4 years agoMerge pull request #2773 from martin-frbg/issue2770
Martin Kroeker [Tue, 11 Aug 2020 19:02:55 +0000 (21:02 +0200)]
Merge pull request #2773 from martin-frbg/issue2770

Fix Makefiles still mishandling NO_CBLAS=0 and NO_LAPACKE=0

4 years agoMerge pull request #2772 from mhillenibm/s390x_gemm_tuning
Martin Kroeker [Tue, 11 Aug 2020 16:14:09 +0000 (18:14 +0200)]
Merge pull request #2772 from mhillenibm/s390x_gemm_tuning

s390x: GEMM tuning for z14

4 years agoFix mishandling of NO_CBLAS=0 and NO_LAPACKE=0
Martin Kroeker [Tue, 11 Aug 2020 11:40:40 +0000 (13:40 +0200)]
Fix mishandling of NO_CBLAS=0 and NO_LAPACKE=0

4 years agofix another source of NO_CBLAS=0 surprise
Martin Kroeker [Tue, 11 Aug 2020 11:27:19 +0000 (13:27 +0200)]
fix another source of NO_CBLAS=0 surprise

4 years agoMerge pull request #76 from xianyi/develop
Martin Kroeker [Tue, 11 Aug 2020 11:25:12 +0000 (13:25 +0200)]
Merge pull request #76 from xianyi/develop

rebase

4 years agos390x/SGEMM: adjust default P and Q to multiples of M
Marius Hillenbrand [Tue, 11 Aug 2020 10:55:59 +0000 (12:55 +0200)]
s390x/SGEMM: adjust default P and Q to multiples of M

We recently changed the register blocking for SGEMM on s390x to 16x4.
However, we did not adjust Q to a multiple of 16 and thus fell back to
the 8x4 kernel at each block's margin, without need. Adjust P and Q to
multiples of 16 to employ the faster 16x4 kernel for complete full-sized
blocks.

Signed-off-by: Marius Hillenbrand <mhillen@linux.ibm.com>
4 years agos390x: Factor out small block sizes for SGEMM/DGEMM on z14
Marius Hillenbrand [Tue, 11 Aug 2020 10:55:53 +0000 (12:55 +0200)]
s390x: Factor out small block sizes for SGEMM/DGEMM on z14

For small register blockings that are too small to fill up vector
registers with column vectors, we currently use a generic code block.
Replace that with instantiations of the generic code as individual
functions, so that the compiler can optimize each one separately.

Signed-off-by: Marius Hillenbrand <mhillen@linux.ibm.com>
4 years agos390x: Optimize SGEMM/DGEMM blocks for z14 with explicit loop unrolling/interleaving
Marius Hillenbrand [Tue, 11 Aug 2020 10:55:42 +0000 (12:55 +0200)]
s390x: Optimize SGEMM/DGEMM blocks for z14 with explicit loop unrolling/interleaving

Improve performance of SGEMM and DGEMM on z14 and z15 by unrolling and
interleaving the inner loop of the SGEMM 16x4 and DGEMM 8x4 blocks.
Specifically, we explicitly interleave vector register loads and
computation of two iterations.

Note that this change only adds one C function, since SGEMM 16x4 and
DGEMM 8x4 actually map to the same C code: they both hold intermediate
results in a 4x4 grid of vector registers, and the C implementation is
built around that.

Signed-off-by: Marius Hillenbrand <mhillen@linux.ibm.com>
4 years agoMerge pull request #2764 from martin-frbg/lapacktests
Martin Kroeker [Mon, 10 Aug 2020 11:27:51 +0000 (13:27 +0200)]
Merge pull request #2764 from martin-frbg/lapacktests

Fix array overruns in the LIN part of the LAPACK testsuite

4 years agoAdd memory barrier to the blas_lock implementation for Linux
Martin Kroeker [Sun, 9 Aug 2020 17:17:04 +0000 (19:17 +0200)]
Add memory barrier to the blas_lock implementation for Linux

as recommended by cparrott73 in #2760

4 years ago Fix use of unallocated array in workspace query and wrong type of argument to xSCAL
Martin Kroeker [Sun, 9 Aug 2020 11:02:27 +0000 (13:02 +0200)]
 Fix use of unallocated array in workspace query and wrong type of argument to xSCAL

4 years agoExpand TAU array as SGEMQR/DGEMQR read elements 2 and 3
Martin Kroeker [Sun, 9 Aug 2020 10:59:20 +0000 (12:59 +0200)]
Expand TAU array as SGEMQR/DGEMQR read elements 2 and 3

4 years agoCreate Jenkinsfile for OSUOSL PowerCI
Martin Kroeker [Sat, 8 Aug 2020 16:05:20 +0000 (18:05 +0200)]
Create Jenkinsfile for OSUOSL PowerCI

4 years agoMerge pull request #2761 from RajalakshmiSR/Makefile_err
Martin Kroeker [Sat, 8 Aug 2020 10:20:04 +0000 (12:20 +0200)]
Merge pull request #2761 from RajalakshmiSR/Makefile_err

Remove extra symbol in Makefile

4 years agoRemove extra symbol in Makefile
Rajalakshmi Srinivasaraghavan [Fri, 7 Aug 2020 20:27:44 +0000 (15:27 -0500)]
Remove extra symbol in Makefile

While trying out different unroll values, noted that
make failed due to this extra symbol.

4 years agoMerge pull request #2758 from martin-frbg/undef_shift
Martin Kroeker [Mon, 3 Aug 2020 21:30:26 +0000 (23:30 +0200)]
Merge pull request #2758 from martin-frbg/undef_shift

Fix GCC ubsan warnings in x86_64 complex dot and gemv_t kernels

4 years agoMerge pull request #2757 from martin-frbg/cmake64
Martin Kroeker [Sun, 2 Aug 2020 21:05:21 +0000 (23:05 +0200)]
Merge pull request #2757 from martin-frbg/cmake64

Fix lapack-tests linking to a suffixed libopenblas in cmake builds

4 years agoMultiply by 2 instead of left-shifting a potentially negative number
Martin Kroeker [Sun, 2 Aug 2020 16:29:56 +0000 (18:29 +0200)]
Multiply by 2 instead of left-shifting a potentially negative number

fixes GCC ubsan warning in the BLAS tests

4 years agoMultiply instead of doing a left shift of a potentially negative number
Martin Kroeker [Sun, 2 Aug 2020 16:27:40 +0000 (18:27 +0200)]
Multiply instead of doing a left shift of a potentially negative number

fixes GCC ubsan report in the BLAS tests

4 years agoMultiply by two instead of left-shifting one place
Martin Kroeker [Sun, 2 Aug 2020 16:25:09 +0000 (18:25 +0200)]
Multiply by two instead of left-shifting one place

fixes GCC ubsan report of "left shift of negative value -2" in the BLAS tests

4 years agoMultiply by two rather than left shift by one place
Martin Kroeker [Sun, 2 Aug 2020 16:22:31 +0000 (18:22 +0200)]
Multiply by two rather than left shift by one place

fixes GCC ubsan report of "left shift of negative value -2" in the BLAS tests

4 years agoApply current library name suffix
Martin Kroeker [Sun, 2 Aug 2020 15:58:33 +0000 (17:58 +0200)]
Apply current library name suffix

4 years agoApply library name suffix to openblas if any
Martin Kroeker [Sun, 2 Aug 2020 15:57:12 +0000 (17:57 +0200)]
Apply library name suffix to openblas if any

4 years agoMerge pull request #75 from xianyi/develop
Martin Kroeker [Sun, 2 Aug 2020 15:50:06 +0000 (17:50 +0200)]
Merge pull request #75 from xianyi/develop

rebase

4 years agoMerge pull request #2753 from martin-frbg/issue2751
Martin Kroeker [Sun, 2 Aug 2020 13:32:46 +0000 (15:32 +0200)]
Merge pull request #2753 from martin-frbg/issue2751

Add SYMBOLPREFIX and/or SYMBOLSUFFIX to cblas prototypes

4 years agoAdd SYMBOLPREFIX and/or -SUFFIX to cblas.h if needed
Martin Kroeker [Sun, 2 Aug 2020 09:20:08 +0000 (11:20 +0200)]
Add SYMBOLPREFIX and/or -SUFFIX to cblas.h if needed

4 years agoImprove substitution rules for SYMBOLPREFIX and -SUFFIX addition
Martin Kroeker [Sat, 1 Aug 2020 15:06:03 +0000 (17:06 +0200)]
Improve substitution rules for SYMBOLPREFIX and -SUFFIX addition

4 years agoMerge pull request #2756 from martin-frbg/issue2755
Martin Kroeker [Sat, 1 Aug 2020 13:19:02 +0000 (15:19 +0200)]
Merge pull request #2756 from martin-frbg/issue2755

Protect against inadvertent activation of USE_CUDA

4 years agoProtect against inadvertent activation of USE_CUDA
Martin Kroeker [Sat, 1 Aug 2020 10:31:39 +0000 (12:31 +0200)]
Protect against inadvertent activation of USE_CUDA

4 years agoAdd SYMBOLPREFIX and/or SYMBOLSUFFIX to cblas prototypes
Martin Kroeker [Fri, 31 Jul 2020 14:03:33 +0000 (16:03 +0200)]
Add SYMBOLPREFIX and/or SYMBOLSUFFIX to cblas prototypes

4 years agoMerge pull request #2752 from kadler/cpuid_aix
Martin Kroeker [Fri, 31 Jul 2020 10:52:24 +0000 (12:52 +0200)]
Merge pull request #2752 from kadler/cpuid_aix

Use systemcfg APIs for CPU detection on AIX

4 years agoUse systemcfg APIs for CPU detection on AIX
Kevin Adler [Fri, 31 Jul 2020 01:52:16 +0000 (20:52 -0500)]
Use systemcfg APIs for CPU detection on AIX

AIX libc already provides ready access to an integer that contains a bit
identifying the CPU it's running on, so there's no need to call a
program and grep its output. Additionally, prtconf is not available in
the PASE runtime, which provides an AIX emulation layer on the IBM i
operating system.

The AIX systemcfg.h also provides macro definitions like POWER_8,
POWER_9, etc for all the bits defining the CPUs as well as macros like
__power_8(), __power_9_andup() that return booleans, but I did not use
them. Since these macros depend on the level of the OS in which it is
built, they may not be defined and instead the associated hex literals
are used directly.

4 years agoFix inadvertent version number reversal to 0.3.9.dev caused by #2710
Martin Kroeker [Thu, 30 Jul 2020 09:40:52 +0000 (11:40 +0200)]
Fix inadvertent version number reversal to 0.3.9.dev caused by #2710

4 years agoMerge pull request #2749 from martin-frbg/make_ppc
Martin Kroeker [Thu, 30 Jul 2020 09:35:53 +0000 (11:35 +0200)]
Merge pull request #2749 from martin-frbg/make_ppc

Reorganize OpenMP build options for POWER and allow compiling for POWER9 with old gcc

4 years agoMerge pull request #2750 from RajalakshmiSR/dgemv_p10
Martin Kroeker [Thu, 30 Jul 2020 08:13:19 +0000 (10:13 +0200)]
Merge pull request #2750 from RajalakshmiSR/dgemv_p10

dgemv optimization for POWER10

4 years agodgemv optimization for POWER10
Rajalakshmi Srinivasaraghavan [Wed, 29 Jul 2020 23:59:32 +0000 (18:59 -0500)]
dgemv optimization for POWER10

Making use of new vector pair POWER10 instructions in dgemv_n and dgemv_t.
Also adding a new block 4x128 to make use of Matrix-Multiply Assist (MMA)
feature introduced in POWER ISA v3.1.  Tested on simulator and there
are no new test failures.

4 years agoSeparate OpenMP handling and allow compilation of Power9 code with older gcc
Martin Kroeker [Wed, 29 Jul 2020 23:14:08 +0000 (01:14 +0200)]
Separate OpenMP handling and allow compilation of Power9 code with older gcc

4 years agoMerge pull request #74 from xianyi/develop
Martin Kroeker [Wed, 29 Jul 2020 23:04:09 +0000 (01:04 +0200)]
Merge pull request #74 from xianyi/develop

rebase

4 years agoMerge pull request #2741 from martin-frbg/issue2739
Martin Kroeker [Wed, 29 Jul 2020 08:01:14 +0000 (10:01 +0200)]
Merge pull request #2741 from martin-frbg/issue2739

Adjust A53 SGEMM parameters to reflect recent switch to 8x8 kernel

4 years agoMerge pull request #2744 from martin-frbg/issue2738
Martin Kroeker [Tue, 28 Jul 2020 17:32:04 +0000 (19:32 +0200)]
Merge pull request #2744 from martin-frbg/issue2738

Add AMD Renoir/Matisse cpu autodetection and preliminary support for Zen3

4 years agoMerge pull request #2740 from RajalakshmiSR/clang-power
Martin Kroeker [Tue, 28 Jul 2020 16:15:25 +0000 (18:15 +0200)]
Merge pull request #2740 from RajalakshmiSR/clang-power

Fix compilation issues with clang on POWER

4 years agoPut hint to use git develop rather than master branch in README
Martin Kroeker [Tue, 28 Jul 2020 14:22:41 +0000 (14:22 +0000)]
Put hint to use git develop rather than master branch in README

4 years agoAdd AMD Renoir/Matisse and preliminary support for Zen3 as Zen2
Martin Kroeker [Tue, 28 Jul 2020 13:53:17 +0000 (13:53 +0000)]
Add AMD Renoir/Matisse and preliminary support for Zen3 as Zen2

also support AMD family 22 Jaguar/Puma as Bobcat

4 years agoAdd AMD Renoir models and preliminary support for ZEN3 as ZEN2
Martin Kroeker [Tue, 28 Jul 2020 13:45:23 +0000 (13:45 +0000)]
Add AMD Renoir models and preliminary support for ZEN3 as ZEN2

also remap erroneous family 16 entry to BOBCAT and reclaim erroneous family 25 "Barcelona" for Zen3

4 years agomissing braces
Martin Kroeker [Mon, 27 Jul 2020 20:19:22 +0000 (20:19 +0000)]
missing braces

4 years agoAdjust A53 SGEMM parameters to reflect move to 8x8 kernel
Martin Kroeker [Mon, 27 Jul 2020 19:54:46 +0000 (19:54 +0000)]
Adjust A53 SGEMM parameters to reflect move to 8x8 kernel

4 years agoFix compilation issues with clang on POWER
Rajalakshmi Srinivasaraghavan [Mon, 27 Jul 2020 19:11:07 +0000 (14:11 -0500)]
Fix compilation issues with clang on POWER

As gcc defaults to -malign-power, removing that option. Also
adding -fno-integrated-as to use GNU assembler for powerpc
assembly optimization files. Fixed other compilation errors
reported in dgemv_t.c file.

4 years agoMerge pull request #2737 from ashwinyes/add_thunderx3_target
Martin Kroeker [Mon, 27 Jul 2020 13:19:47 +0000 (15:19 +0200)]
Merge pull request #2737 from ashwinyes/add_thunderx3_target

ARM64: Add THUNDERX3T110 Target

4 years agoARM64: Add THUNDERX3T110 Target
Ashwin Sekhar T K [Thu, 11 Jun 2020 11:12:49 +0000 (04:12 -0700)]
ARM64: Add THUNDERX3T110 Target

4 years agoMerge pull request #2735 from martin-frbg/move_potrf
Martin Kroeker [Sun, 26 Jul 2020 17:54:11 +0000 (19:54 +0200)]
Merge pull request #2735 from martin-frbg/move_potrf

Move potrf_parallel.c from lapack/getrf to lapack/potrf where it belongs

4 years agoMerge pull request #2734 from RajalakshmiSR/p10_fix
Martin Kroeker [Sat, 25 Jul 2020 07:02:32 +0000 (09:02 +0200)]
Merge pull request #2734 from RajalakshmiSR/p10_fix

Fix to store results in correct order for POWER10 GEMM kernels

4 years agoUse _Atomic instead of volatile where available (file moved from ../getrf)
Martin Kroeker [Sat, 25 Jul 2020 06:52:24 +0000 (08:52 +0200)]
Use _Atomic instead of volatile where available (file moved from ../getrf)

must have misplaced this in ../getrf when I made that change in March 2018 (40160ff)
the only changes since then were
RFC : Add half precision gemm for bfloat16 in OpenBLAS Rajalakshmi Srinivasaraghavan
Rajalakshmi Srinivasaraghavan committed on 14 Apr 2020 as 7ebbb50

    Change _STDC_VERSION__ to __STDC_VERSION__
Zhiyong Dang committed on 11 May 2018 as 3716267

4 years agoDelete potrf_parallel.c (moving it to ../potrf)
Martin Kroeker [Sat, 25 Jul 2020 06:42:39 +0000 (06:42 +0000)]
Delete potrf_parallel.c (moving it to ../potrf)

4 years agoFix to store results in correct order for POWER10 GEMM kernels
Rajalakshmi Srinivasaraghavan [Sat, 25 Jul 2020 04:08:11 +0000 (23:08 -0500)]
Fix to store results in correct order for POWER10 GEMM kernels

There is a recent compiler change in __builtin_mma_disassemble_acc() which
affects the order of storing result in POWER10. Also removing new LDFLAG
-mno-power10-stub as it is handled by linker automatically.

4 years agoMerge pull request #2720 from martin-frbg/issue2694
Martin Kroeker [Fri, 24 Jul 2020 21:19:45 +0000 (23:19 +0200)]
Merge pull request #2720 from martin-frbg/issue2694

WIP Further fixes for 32bit POWER8

4 years agoTypo fix
Martin Kroeker [Fri, 24 Jul 2020 16:04:58 +0000 (16:04 +0000)]
Typo fix

4 years agoRegroup the 32 and 64bit sections and restore 64bit CAXPY
Martin Kroeker [Fri, 24 Jul 2020 10:13:46 +0000 (10:13 +0000)]
Regroup the 32 and 64bit sections and restore 64bit CAXPY

4 years agoMerge pull request #2721 from martin-frbg/p8align
Martin Kroeker [Fri, 24 Jul 2020 09:06:20 +0000 (11:06 +0200)]
Merge pull request #2721 from martin-frbg/p8align

Fix alignment errors in the power8 saxpy kernel

4 years agoMerge pull request #2731 from martin-frbg/pgippc
Martin Kroeker [Fri, 24 Jul 2020 09:05:16 +0000 (11:05 +0200)]
Merge pull request #2731 from martin-frbg/pgippc

Fixes for compilation on POWER with PGI compilers

4 years agoUse OPENBLAS_MAKE_COMPLEX_FLOAT on PPC only
Martin Kroeker [Thu, 23 Jul 2020 20:40:13 +0000 (20:40 +0000)]
Use OPENBLAS_MAKE_COMPLEX_FLOAT on PPC only

4 years agoAdd ifdefs around call to altivec microkernel
Martin Kroeker [Thu, 23 Jul 2020 18:30:42 +0000 (18:30 +0000)]
Add ifdefs around call to altivec microkernel

4 years agoTypo fix
Martin Kroeker [Thu, 23 Jul 2020 17:34:56 +0000 (17:34 +0000)]
Typo fix

4 years agoRewrite assignment to complex for better portability
Martin Kroeker [Thu, 23 Jul 2020 15:10:59 +0000 (17:10 +0200)]
Rewrite assignment to complex for better portability

4 years agoExclude altivec code paths if the compiler does not support them
Martin Kroeker [Thu, 23 Jul 2020 15:08:20 +0000 (17:08 +0200)]
Exclude altivec code paths if the compiler does not support them

4 years agoAvoid undefining NAME,CNAME etc for pgcc as it makes it ignore the new defininitions
Martin Kroeker [Thu, 23 Jul 2020 15:03:28 +0000 (17:03 +0200)]
Avoid undefining NAME,CNAME etc for pgcc as it makes it ignore the new defininitions

4 years agoMerge pull request #73 from xianyi/develop
Martin Kroeker [Thu, 23 Jul 2020 14:59:06 +0000 (16:59 +0200)]
Merge pull request #73 from xianyi/develop

rebase

4 years agoMerge pull request #2729 from martin-frbg/issue2728
Martin Kroeker [Wed, 22 Jul 2020 20:45:57 +0000 (22:45 +0200)]
Merge pull request #2729 from martin-frbg/issue2728

Unify BUFFER_SIZE settings for x86_64 again to fix DYNAMIC_ARCH crashes

4 years agoUnify BUFFER_SIZE settings for x86_64 again to fix potentially fatal mismatch in...
Martin Kroeker [Wed, 22 Jul 2020 17:30:55 +0000 (17:30 +0000)]
Unify BUFFER_SIZE settings for x86_64 again to fix potentially fatal mismatch in DYNAMIC_ARCH builds

4 years agoMerge pull request #2727 from wyphan/develop
Martin Kroeker [Tue, 21 Jul 2020 15:06:53 +0000 (17:06 +0200)]
Merge pull request #2727 from wyphan/develop

Patch for building on POWERPC with PGI compilers (was Patch for building on Summit)

4 years agoMerge pull request #2726 from martin-frbg/2725-2
Martin Kroeker [Tue, 21 Jul 2020 14:42:06 +0000 (16:42 +0200)]
Merge pull request #2726 from martin-frbg/2725-2

Add detection of stdatomic.h for cmake

4 years agoPatch for building on Summit
Wileam Phan [Tue, 21 Jul 2020 03:30:28 +0000 (23:30 -0400)]
Patch for building on Summit

4 years agoAdd trivial check for stdatomic.h
Martin Kroeker [Mon, 20 Jul 2020 22:52:09 +0000 (22:52 +0000)]
Add trivial check for stdatomic.h

4 years agoMerge pull request #72 from xianyi/develop
Martin Kroeker [Mon, 20 Jul 2020 22:49:12 +0000 (00:49 +0200)]
Merge pull request #72 from xianyi/develop

rebase

4 years agoMerge pull request #2725 from martin-frbg/ccheck_c11
Martin Kroeker [Sat, 18 Jul 2020 21:08:08 +0000 (23:08 +0200)]
Merge pull request #2725 from martin-frbg/ccheck_c11

Have c_check probe availability of C11 atomics support and stdatomic.h

4 years agoUpdate conditional for atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:19:59 +0000 (17:19 +0000)]
Update conditional for atomics to use HAVE_C11

4 years agoUpdate conditional for atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:14:50 +0000 (17:14 +0000)]
Update conditional for atomics to use HAVE_C11

4 years agoUpdate conditional for C11 atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:13:24 +0000 (17:13 +0000)]
Update conditional for C11 atomics to use HAVE_C11

4 years agoUpdate conditional for atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:09:56 +0000 (17:09 +0000)]
Update conditional for atomics to use HAVE_C11

4 years agoUpdate conditional for atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:09:01 +0000 (17:09 +0000)]
Update conditional for atomics to use HAVE_C11

4 years agoUpdate conditional for atomics to HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:07:38 +0000 (17:07 +0000)]
Update conditional for atomics to HAVE_C11

4 years agoUpdate conditional for atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:05:59 +0000 (17:05 +0000)]
Update conditional for atomics to use HAVE_C11

4 years agoUpdate conditional for atomics to use HAVE_C11
Martin Kroeker [Sat, 18 Jul 2020 17:03:31 +0000 (17:03 +0000)]
Update conditional for atomics to use HAVE_C11