 rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago rebase? (#1)
* With the Intel compiler on Linux, prefer ifort for the final link step
icc has known problems with mixed-language builds that ifort can handle just fine. Fixes #1956
* Rename operands to put lda on the input/output constraint list
* Fix wrong constraints in inline assembly
for #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda4 as input and output. For #2009
* Fix inline assembly constraints
rework indices to allow marking argument lda as input and output.
* Fix inline assembly constraints
* Fix inline assembly constraints
* Fix inline assembly constraints in Bulldozer TRSM kernels
rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
* Correct range_n limiting
same bug as seen in #1388, somehow missed in corresponding PR #1389
* Allow multithreading TRMV again
revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)
* Fix error introduced during cleanup
* Reduce list of kernels in the dynamic arch build
to make compilation complete reliably within the 1h limit again
* init
* move fix to right place
* Fix missing -c option in AVX512 test
* Fix AVX512 test always returning false due to missing compiler option
* Make x86_32 imply NO_AVX2, NO_AVX512 in addition to NO_AVX
fixes #2033
* Keep xcode8.3 for osx BINARY=32 build
as xcode10 deprecated i386
* Make sure that AVX512 is disabled in 32bit builds
for #2033
* Improve handling of NO_STATIC and NO_SHARED
to avoid surprises from defining either as zero. Fixes #2035 by addressing some concerns from #1422
* init
* address warning introed with #1814 et al
* Restore locking optimizations for OpenMP case
restore another accidentally dropped part of #1468 that was missed in #2004 to address performance regression reported in #1461
* HiSilicon tsv110 CPUs optimization branch
add HiSilicon tsv110 CPUs optimization branch
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* add TARGET support for HiSilicon tsv110 CPUs
* Fix module definition conflicts between LAPACK and ReLAPACK
for #2043
* Do not compile in AVX512 check if AVX support is disabled
xgetbv is function depends on NO_AVX being undefined - we could change that too, but that combo is unlikely to work anyway
* ctest.c : add __POWERPC__ for PowerMac
* Fix crash in sgemm SSE/nano kernel on x86_64
Fix bug #2047.
Signed-off-by: Celelibi <celelibi@gmail.com>
* param.h : enable defines for PPC970 on DarwinOS
fixes:
gemm.c: In function 'sgemm_':
../common_param.h:981:18: error: 'SGEMM_DEFAULT_P' undeclared (first use in this function)
#define SGEMM_P SGEMM_DEFAULT_P
^
* common_power.h: force DCBT_ARG 0 on PPC970 Darwin
without this, we see
../kernel/power/gemv_n.S:427:Parameter syntax error
and many more similar entries
that relates to this assembly command
dcbt 8, r24, r18
this change makes the DCBT_ARG = 0
and openblas builds through to completion on PowerMac 970
Tests pass
* Make TARGET=GENERIC compatible with DYNAMIC_ARCH=1
for issue #2048
* make DYNAMIC_ARCH=1 package work on TSV110.
* make DYNAMIC_ARCH=1 package work on TSV110
* Add Intel Denverton
for #2048
* Add Intel Denverton
* Change 64-bit detection as explained in #2056
* Trivial typo fix
as suggested in #2022
* Disable the AVX512 DGEMM kernel (again)
Due to as yet unresolved errors seen in #1955 and #2029
* Use POSIX getenv on Cygwin
The Windows-native GetEnvironmentVariable cannot be relied on, as
Cygwin does not always copy environment variables set through Cygwin
to the Windows environment block, particularly after fork().
* Fix for #2063: The DllMain used in Cygwin did not run the thread memory
pool cleanup upon THREAD_DETACH which is needed when compiled with
USE_TLS=1.
* Also call CloseHandle on each thread, as well as on the event so as to not leak thread handles.
* AIX asm syntax changes needed for shared object creation
* power9 makefile. dgemm based on power8 kernel with following changes : 32x unrolled 16x4 kernel and 8x4 kernel using (lxv stxv butterfly rank1 update). improvement from 17 to 22-23gflops. dtrmm cases were added into dgemm itself
* Expose CBLAS interfaces for I?MIN and I?MAX
* Build CBLAS interfaces for I?MIN and I?MAX
* Add declarations for ?sum and cblas_?sum
* Add interface for ?sum (derived from ?asum)
* Add ?sum
* Add implementations of ssum/dsum and csum/zsum
as trivial copies of asum/zsasum with the fabs calls replaced by fmov to preserve code structure
* Add ARM implementations of ?sum
(trivial copies of the respective ?asum with the fabs calls removed)
* Add ARM64 implementations of ?sum
as trivial copies of the respective ?asum kernels with the fabs calls removed
* Add ia64 implementation of ?sum
as trivial copy of asum with the fabs calls removed
* Add MIPS implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add MIPS64 implementation of ?sum
as trivial copy of ?asum with the fabs replaced by mov to preserve code structure
* Add POWER implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmr to preserve code structure
* Add SPARC implementation of ?sum
as trivial copy of ?asum with the fabs replaced by fmov to preserve code structure
* Add x86 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add x86_64 implementation of ?sum
as trivial copy of ?asum with the fabs calls removed
* Add ZARCH implementation of ?sum
as trivial copies of the respective ?asum kernels with the ABS and vflpsb calls removed
* Detect 32bit environment on 64bit ARM hardware
for #2056, using same approach as #2058
* Add cmake defaults for ?sum kernels
* Add ?sum
* Add ?sum definitions for generic kernel
* Add declarations for ?sum
* Add -lm and disable EXPRECISION support on *BSD
fixes #2075
* Add in runtime CPU detection for POWER.
* snprintf define consolidated to common.h
* Support INTERFACE64=1
* Add support for INTERFACE64 and fix XERBLA calls
1. Replaced all instances of "int" with "blasint"
2. Added string length as "hidden" third parameter in calls to fortran XERBLA
* Correct length of name string in xerbla call
* Avoid out-of-bounds accesses in LAPACK EIG tests
see https://github.com/Reference-LAPACK/lapack/issues/333
* Correct INFO=4 condition
* Disable reallocation of work array in xSYTRF
as it appears to cause memory management problems (seen in the LAPACK tests)
* Disable repeated recursion on Ab_BR in ReLAPACK xGBTRF
due to crashes in LAPACK tests
* sgemm/strmm
* Update Changelog with changes from 0.3.6
* Increment version to 0.3.7.dev
* Increment version to 0.3.7.dev
* Misc. typo fixes
Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`
* Correct argument of CPU_ISSET for glibc <2.5
fixes #2104
* conflict resolve
* Revert reference/ fixes
* Revert Changelog.txt typos
* Disable the SkyLakeX DGEMMITCOPY kernel as well
as a stopgap measure for https://github.com/numpy/numpy/issues/13401 as mentioned in #1955
* Disable DGEMMINCOPY as well for now
#1955
* init
* Fix errors in cpu enumeration with glibc 2.6
for #2114
* Change two http links to https
Closes #2109
* remove redundant code #2113
* Set up CI with Azure Pipelines
[skip ci]
* TST: add native POWER8 to CI
* add native POWER8 testing to
Travis CI matrix with ppc64le
os entry
* Update link to IBM MASS library, update cpu support status
* first try migrating one of the arm builds from travis
* fix tabbing in azure commands
* Update azure-pipelines.yml
take out offending lines (although stolen from https://github.com/conda-forge/opencv-feedstock azure-pipelines fiie)
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* Update azure-pipelines.yml
* DOC: Add Azure CI status badge
* Add ARMV6 build to azure CI setup (#2122)
using aytekinar's Alpine image and docker script from the Travis setup
[skip ci]
* TST: Azure manylinux1 & clean-up
* remove some of the steps & comments
from the original Azure yml template
* modify the trigger section to use
develop since OpenBLAS primarily uses
this branch; use the same batching
behavior as downstream projects NumPy/
SciPy
* remove Travis emulated ARMv6 gcc build
because this now happens in Azure
* use documented Ubuntu vmImage name for Azure
and add in a manylinux1 test run to the matrix
[skip appveyor]
* Add NO_AFFINITY to available options on Linux, and set it to ON
to match the gmake default. Fixes second part of #2114
* Replace ISMIN and ISAMIN kernels on all x86_64 platforms (#2125)
* Mark iamax_sse.S as unsuitable for MIN due to issue #2116
* Use iamax.S rather than iamax_sse.S for ISMIN/ISAMIN on all x86_64 as workaround for #2116
* Move ARMv8 gcc build from Travis to Azure
* Move ARMv8 gcc build from Travis to Azure
* Update .travis.yml
* Test drone CI
* install make
* remove sudo
* Install gcc
* Install perl
* Install gfortran and add a clang job
* gfortran->gcc-gfortran
* Switch to ubuntu and parallel jobs
* apt update
* Fix typo
* update yes
* no need of gcc in clang build
* Add a cmake build as well
* Add cmake builds and print options
* build without lapack on cmake
* parallel build
* See if ubuntu 19.04 fixes the ICE
* Remove qemu armv8 builds
* arm32 build
* Fix typo
* TST: add SkylakeX AVX512 CI test
* adapt the C-level reproducer code for some
recent SkylakeX AVX512 kernel issues, provided
by Isuru Fernando and modified by Martin Kroeker,
for usage in the utest suite
* add an Intel SDE SkylakeX emulation utest run to
the Azure CI matrix; a custom Docker build was required
because Ubuntu image provided by Azure does not support
AVX512VL instructions
* Add option USE_LOCKING for single-threaded build with locking support
for calling from concurrent threads
* Add option USE_LOCKING for single-threaded build with locking support
* Add option USE_LOCKING for SMP-like locking in USE_THREAD=0 builds
* Add option USE_LOCKING but keep default settings intact
* Remove unrelated change
* Do not try ancient PGI hacks with recent versions of that compiler
should fix #2139
* Build and run utests in any case, they do their own checks for fortran availability
* Add softfp support in min/max kernels
fix for #1912
* Revert "Add softfp support in min/max kernels"
* Separate implementations of AMAX and IAMAX on arm
As noted in #1912 and comment on #1942, the combined implementation happens to "do the right thing" on hardfp, but cannot return both value and index on softfp where they would have to share the return register
* Ensure correct output for DAMAX with softfp
* Use generic kernels for complex (I)AMAX to support softfp
* improved zgemm power9 based on power8
* upload thread safety test folder
* hook up c++ thread safety test (main Makefile)
* add c++ thread test option to Makefile.rule
* Document NO_AVX512
for #2151
* sgemm pipeline improved, zgemm rewritten without inner packs, ABI lxvx v20 fixed with vs52
* Fix detection of AVX512 capable compilers in getarch
21eda8b5 introduced a check in getarch.c to test if the compiler is capable of
AVX512. This check currently fails, since the used __AVX2__ macro is only
defined if getarch itself was compiled with AVX2/AVX512 support. Make sure this
is the case by building getarch with -march=native on x86_64. It is only
supposed to run on the build host anyway.
* c_check: Unlink correct file
* power9 zgemm ztrmm optimized
* conflict resolve
* Add gfortran workaround for ABI violations in LAPACKE
for #2154 (see gcc bug 90329)
* Add gfortran workaround for ABI violations
for #2154 (see gcc bug 90329)
* Add gfortran workaround for potential ABI violation
for #2154
* Update fc.cmake
* Remove any inadvertent use of -march=native from DYNAMIC_ARCH builds
from #2143, -march=native precludes use of more specific options like -march=skylake-avx512 in individual kernels, and defeats the purpose of dynamic arch anyway.
* Avoid unintentional activation of TLS code via USE_TLS=0
fixes #2149
* Do not force gcc options on non-gcc compilers
fixes compile failure with pgi 18.10 as reported on OpenBLAS-users
* Update Makefile.x86_64
* Zero ecx with a mov instruction
PGI assembler does not like the initialization in the constraints.
* Fix mov syntax
* new sgemm 8x16
* Update dtrmm_kernel_16x4_power8.S
* PGI compiler does not like -march=native
* Fix build on FreeBSD/powerpc64.
Signed-off-by: Piotr Kubaj <pkubaj@anongoth.pl>
* Fix build for PPC970 on FreeBSD pt. 1
FreeBSD needs DCBT_ARG=0 as well.
* Fix build for PPC970 on FreeBSD pt.2
FreeBSD needs those macros too.
* cgemm/ctrmm power9
* Utest needs CBLAS but not necessarily FORTRAN
* Add mingw builds to Appveyor config
* Add getarch flags to disable AVX on x86
(and other small fixes to match Makefile behaviour)
* Make disabling DYNAMIC_ARCH on unsupported systems work
needs to be unset in the cache for the change to have any effect
* Mingw32 needs leading underscore on object names
(also copy BUNDERSCORE settings for FORTRAN from the corresponding Makefile)
6 years ago |
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860 |
- ### AMAX ###
-
- ifndef SAMAXKERNEL
- SAMAXKERNEL = amax.S
- endif
-
- ifndef DAMAXKERNEL
- DAMAXKERNEL = amax.S
- endif
-
- ifndef QAMAXKERNEL
- QAMAXKERNEL = amax.S
- endif
-
- ifndef CAMAXKERNEL
- CAMAXKERNEL = zamax.S
- endif
-
- ifndef ZAMAXKERNEL
- ZAMAXKERNEL = zamax.S
- endif
-
- ifndef XAMAXKERNEL
- XAMAXKERNEL = zamax.S
- endif
-
- ### AMIN ###
-
- ifndef SAMINKERNEL
- SAMINKERNEL = amin.S
- endif
-
- ifndef DAMINKERNEL
- DAMINKERNEL = amin.S
- endif
-
- ifndef QAMINKERNEL
- QAMINKERNEL = amin.S
- endif
-
- ifndef CAMINKERNEL
- CAMINKERNEL = zamin.S
- endif
-
- ifndef ZAMINKERNEL
- ZAMINKERNEL = zamin.S
- endif
-
- ifndef XAMINKERNEL
- XAMINKERNEL = zamin.S
- endif
-
- ### MAX ###
-
- ifndef SMAXKERNEL
- SMAXKERNEL = max.S
- endif
-
- ifndef DMAXKERNEL
- DMAXKERNEL = max.S
- endif
-
- ifndef QMAXKERNEL
- QMAXKERNEL = max.S
- endif
-
- ### MIN ###
-
- ifndef SMINKERNEL
- SMINKERNEL = min.S
- endif
-
- ifndef DMINKERNEL
- DMINKERNEL = min.S
- endif
-
- ifndef QMINKERNEL
- QMINKERNEL = min.S
- endif
-
- ### IAMAX ###
-
- ifndef ISAMAXKERNEL
- ISAMAXKERNEL = iamax.S
- endif
-
- ifndef IDAMAXKERNEL
- IDAMAXKERNEL = iamax.S
- endif
-
- ifndef IQAMAXKERNEL
- IQAMAXKERNEL = iamax.S
- endif
-
- ifndef ICAMAXKERNEL
- ICAMAXKERNEL = izamax.S
- endif
-
- ifndef IZAMAXKERNEL
- IZAMAXKERNEL = izamax.S
- endif
-
- ifndef IXAMAXKERNEL
- IXAMAXKERNEL = izamax.S
- endif
-
- ### IAMIN ###
-
- ifndef ISAMINKERNEL
- ISAMINKERNEL = iamin.S
- endif
-
- ifndef IDAMINKERNEL
- IDAMINKERNEL = iamin.S
- endif
-
- ifndef IQAMINKERNEL
- IQAMINKERNEL = iamin.S
- endif
-
- ifndef ICAMINKERNEL
- ICAMINKERNEL = izamin.S
- endif
-
- ifndef IZAMINKERNEL
- IZAMINKERNEL = izamin.S
- endif
-
- ifndef IXAMINKERNEL
- IXAMINKERNEL = izamin.S
- endif
-
- ### IMAX ###
-
- ifndef ISMAXKERNEL
- ISMAXKERNEL = iamax.S
- endif
-
- ifndef IDMAXKERNEL
- IDMAXKERNEL = iamax.S
- endif
-
- ifndef IQMAXKERNEL
- IQMAXKERNEL = iamax.S
- endif
-
- ### IMIN ###
-
- ifndef ISMINKERNEL
- ISMINKERNEL = iamin.S
- endif
-
- ifndef IDMINKERNEL
- IDMINKERNEL = iamin.S
- endif
-
- ifndef IQMINKERNEL
- IQMINKERNEL = iamin.S
- endif
-
- ### ASUM ###
-
- ifndef SASUMKERNEL
- SASUMKERNEL = asum.S
- endif
-
- ifndef DASUMKERNEL
- DASUMKERNEL = asum.S
- endif
-
- ifndef CASUMKERNEL
- CASUMKERNEL = zasum.S
- endif
-
- ifndef ZASUMKERNEL
- ZASUMKERNEL = zasum.S
- endif
-
- ifndef QASUMKERNEL
- QASUMKERNEL = asum.S
- endif
-
- ifndef XASUMKERNEL
- XASUMKERNEL = zasum.S
- endif
-
- ### AXPY ###
-
- ifndef SAXPYKERNEL
- SAXPYKERNEL = axpy.S
- endif
-
- ifndef DAXPYKERNEL
- DAXPYKERNEL = axpy.S
- endif
-
- ifndef CAXPYKERNEL
- CAXPYKERNEL = zaxpy.S
- endif
-
- ifndef ZAXPYKERNEL
- ZAXPYKERNEL = zaxpy.S
- endif
-
- ifndef QAXPYKERNEL
- QAXPYKERNEL = axpy.S
- endif
-
- ifndef XAXPYKERNEL
- XAXPYKERNEL = zaxpy.S
- endif
-
- ### COPY ###
-
- ifndef SCOPYKERNEL
- SCOPYKERNEL = copy.S
- endif
-
- ifndef DCOPYKERNEL
- DCOPYKERNEL = copy.S
- endif
-
- ifndef CCOPYKERNEL
- CCOPYKERNEL = zcopy.S
- endif
-
- ifndef ZCOPYKERNEL
- ZCOPYKERNEL = zcopy.S
- endif
-
- ifndef QCOPYKERNEL
- QCOPYKERNEL = copy.S
- endif
-
- ifndef XCOPYKERNEL
- XCOPYKERNEL = zcopy.S
- endif
-
- ### DOT ###
-
- ifndef SDOTKERNEL
- SDOTKERNEL = dot.S
- endif
-
- ifndef DDOTKERNEL
- DDOTKERNEL = dot.S
- endif
-
- ifndef CDOTKERNEL
- CDOTKERNEL = zdot.S
- endif
-
- ifndef ZDOTKERNEL
- ZDOTKERNEL = zdot.S
- endif
-
- ifndef QDOTKERNEL
- QDOTKERNEL = dot.S
- endif
-
- ifndef XDOTKERNEL
- XDOTKERNEL = zdot.S
- endif
-
- ### NRM2 ###
-
- ifndef SNRM2KERNEL
- SNRM2KERNEL = nrm2.S
- endif
-
- ifndef DNRM2KERNEL
- DNRM2KERNEL = nrm2.S
- endif
-
- ifndef QNRM2KERNEL
- QNRM2KERNEL = nrm2.S
- endif
-
- ifndef CNRM2KERNEL
- CNRM2KERNEL = znrm2.S
- endif
-
- ifndef ZNRM2KERNEL
- ZNRM2KERNEL = znrm2.S
- endif
-
- ifndef XNRM2KERNEL
- XNRM2KERNEL = znrm2.S
- endif
-
- ### ROT ###
-
- ifndef SROTKERNEL
- SROTKERNEL = rot.S
- endif
-
- ifndef DROTKERNEL
- DROTKERNEL = rot.S
- endif
-
- ifndef QROTKERNEL
- QROTKERNEL = rot.S
- endif
-
- ifndef CROTKERNEL
- CROTKERNEL = zrot.S
- endif
-
- ifndef ZROTKERNEL
- ZROTKERNEL = zrot.S
- endif
-
- ifndef XROTKERNEL
- XROTKERNEL = zrot.S
- endif
-
- ### SCAL ###
-
- ifndef SSCALKERNEL
- SSCALKERNEL = scal.S
- endif
-
- ifndef DSCALKERNEL
- DSCALKERNEL = scal.S
- endif
-
- ifndef CSCALKERNEL
- CSCALKERNEL = zscal.S
- endif
-
- ifndef ZSCALKERNEL
- ZSCALKERNEL = zscal.S
- endif
-
- ifndef QSCALKERNEL
- QSCALKERNEL = scal.S
- endif
-
- ifndef XSCALKERNEL
- XSCALKERNEL = zscal.S
- endif
-
- ### SUM ###
-
- ifndef SSUMKERNEL
- SSUMKERNEL = sum.S
- endif
-
- ifndef DSUMKERNEL
- DSUMKERNEL = sum.S
- endif
-
- ifndef CSUMKERNEL
- CSUMKERNEL = zsum.S
- endif
-
- ifndef ZSUMKERNEL
- ZSUMKERNEL = zsum.S
- endif
-
- ifndef QSUMKERNEL
- QSUMKERNEL = sum.S
- endif
-
- ifndef XSUMKERNEL
- XSUMKERNEL = zsum.S
- endif
-
- ### SWAP ###
-
- ifndef SSWAPKERNEL
- SSWAPKERNEL = swap.S
- endif
-
- ifndef DSWAPKERNEL
- DSWAPKERNEL = swap.S
- endif
-
- ifndef CSWAPKERNEL
- CSWAPKERNEL = zswap.S
- endif
-
- ifndef ZSWAPKERNEL
- ZSWAPKERNEL = zswap.S
- endif
-
- ifndef QSWAPKERNEL
- QSWAPKERNEL = swap.S
- endif
-
- ifndef XSWAPKERNEL
- XSWAPKERNEL = zswap.S
- endif
-
- ### GEMV ###
-
- ifndef SGEMVNKERNEL
- SGEMVNKERNEL = gemv_n.S
- endif
-
- ifndef SGEMVTKERNEL
- SGEMVTKERNEL = gemv_t.S
- endif
-
- ifndef DGEMVNKERNEL
- DGEMVNKERNEL = gemv_n.S
- endif
-
- ifndef DGEMVTKERNEL
- DGEMVTKERNEL = gemv_t.S
- endif
-
- ifndef CGEMVNKERNEL
- CGEMVNKERNEL = zgemv_n.S
- endif
-
- ifndef CGEMVTKERNEL
- CGEMVTKERNEL = zgemv_t.S
- endif
-
- ifndef ZGEMVNKERNEL
- ZGEMVNKERNEL = zgemv_n.S
- endif
-
- ifndef ZGEMVTKERNEL
- ZGEMVTKERNEL = zgemv_t.S
- endif
-
- ifndef QGEMVNKERNEL
- QGEMVNKERNEL = gemv_n.S
- endif
-
- ifndef QGEMVTKERNEL
- QGEMVTKERNEL = gemv_t.S
- endif
-
- ifndef XGEMVNKERNEL
- XGEMVNKERNEL = zgemv_n.S
- endif
-
- ifndef XGEMVTKERNEL
- XGEMVTKERNEL = zgemv_t.S
- endif
-
- ifndef SCABS_KERNEL
- SCABS_KERNEL = cabs.S
- endif
-
- ifndef DCABS_KERNEL
- DCABS_KERNEL = cabs.S
- endif
-
- ifndef QCABS_KERNEL
- QCABS_KERNEL = cabs.S
- endif
-
- ifndef LSAME_KERNEL
- LSAME_KERNEL = lsame.S
- endif
-
- ### AXPBY ###
-
- ifndef SAXPBYKERNEL
- SAXPBYKERNEL = ../arm/axpby.c
- endif
-
- ifndef DAXPBYKERNEL
- DAXPBYKERNEL = ../arm/axpby.c
- endif
-
- ifndef CAXPBYKERNEL
- CAXPBYKERNEL = ../arm/zaxpby.c
- endif
-
- ifndef ZAXPBYKERNEL
- ZAXPBYKERNEL = ../arm/zaxpby.c
- endif
-
- SBLASOBJS += \
- samax_k$(TSUFFIX).$(SUFFIX) samin_k$(TSUFFIX).$(SUFFIX) smax_k$(TSUFFIX).$(SUFFIX) smin_k$(TSUFFIX).$(SUFFIX) \
- isamax_k$(TSUFFIX).$(SUFFIX) isamin_k$(TSUFFIX).$(SUFFIX) ismax_k$(TSUFFIX).$(SUFFIX) ismin_k$(TSUFFIX).$(SUFFIX) \
- sasum_k$(TSUFFIX).$(SUFFIX) ssum_k$(TSUFFIX).$(SUFFIX) saxpy_k$(TSUFFIX).$(SUFFIX) scopy_k$(TSUFFIX).$(SUFFIX) \
- sdot_k$(TSUFFIX).$(SUFFIX) sdsdot_k$(TSUFFIX).$(SUFFIX) dsdot_k$(TSUFFIX).$(SUFFIX) \
- snrm2_k$(TSUFFIX).$(SUFFIX) srot_k$(TSUFFIX).$(SUFFIX) sscal_k$(TSUFFIX).$(SUFFIX) sswap_k$(TSUFFIX).$(SUFFIX) \
- saxpby_k$(TSUFFIX).$(SUFFIX)
-
- DBLASOBJS += \
- damax_k$(TSUFFIX).$(SUFFIX) damin_k$(TSUFFIX).$(SUFFIX) dmax_k$(TSUFFIX).$(SUFFIX) dmin_k$(TSUFFIX).$(SUFFIX) \
- idamax_k$(TSUFFIX).$(SUFFIX) idamin_k$(TSUFFIX).$(SUFFIX) idmax_k$(TSUFFIX).$(SUFFIX) idmin_k$(TSUFFIX).$(SUFFIX) \
- dasum_k$(TSUFFIX).$(SUFFIX) daxpy_k$(TSUFFIX).$(SUFFIX) dcopy_k$(TSUFFIX).$(SUFFIX) ddot_k$(TSUFFIX).$(SUFFIX) \
- dnrm2_k$(TSUFFIX).$(SUFFIX) drot_k$(TSUFFIX).$(SUFFIX) dscal_k$(TSUFFIX).$(SUFFIX) dswap_k$(TSUFFIX).$(SUFFIX) \
- daxpby_k$(TSUFFIX).$(SUFFIX) dsum_k$(TSUFFIX).$(SUFFIX)
-
- QBLASOBJS += \
- qamax_k$(TSUFFIX).$(SUFFIX) qamin_k$(TSUFFIX).$(SUFFIX) qmax_k$(TSUFFIX).$(SUFFIX) qmin_k$(TSUFFIX).$(SUFFIX) \
- iqamax_k$(TSUFFIX).$(SUFFIX) iqamin_k$(TSUFFIX).$(SUFFIX) iqmax_k$(TSUFFIX).$(SUFFIX) iqmin_k$(TSUFFIX).$(SUFFIX) \
- qasum_k$(TSUFFIX).$(SUFFIX) qaxpy_k$(TSUFFIX).$(SUFFIX) qcopy_k$(TSUFFIX).$(SUFFIX) qdot_k$(TSUFFIX).$(SUFFIX) \
- qnrm2_k$(TSUFFIX).$(SUFFIX) qrot_k$(TSUFFIX).$(SUFFIX) qscal_k$(TSUFFIX).$(SUFFIX) qswap_k$(TSUFFIX).$(SUFFIX) \
- qsum_k$(TSUFFIX).$(SUFFIX)
-
- CBLASOBJS += \
- camax_k$(TSUFFIX).$(SUFFIX) camin_k$(TSUFFIX).$(SUFFIX) icamax_k$(TSUFFIX).$(SUFFIX) icamin_k$(TSUFFIX).$(SUFFIX) \
- casum_k$(TSUFFIX).$(SUFFIX) caxpy_k$(TSUFFIX).$(SUFFIX) caxpyc_k$(TSUFFIX).$(SUFFIX) ccopy_k$(TSUFFIX).$(SUFFIX) \
- cdotc_k$(TSUFFIX).$(SUFFIX) cdotu_k$(TSUFFIX).$(SUFFIX) cnrm2_k$(TSUFFIX).$(SUFFIX) csrot_k$(TSUFFIX).$(SUFFIX) \
- cscal_k$(TSUFFIX).$(SUFFIX) cswap_k$(TSUFFIX).$(SUFFIX) caxpby_k$(TSUFFIX).$(SUFFIX) csum_k$(TSUFFIX).$(SUFFIX)
-
- ZBLASOBJS += \
- zamax_k$(TSUFFIX).$(SUFFIX) zamin_k$(TSUFFIX).$(SUFFIX) izamax_k$(TSUFFIX).$(SUFFIX) izamin_k$(TSUFFIX).$(SUFFIX) \
- zasum_k$(TSUFFIX).$(SUFFIX) zaxpy_k$(TSUFFIX).$(SUFFIX) zaxpyc_k$(TSUFFIX).$(SUFFIX) zcopy_k$(TSUFFIX).$(SUFFIX) \
- zdotc_k$(TSUFFIX).$(SUFFIX) zdotu_k$(TSUFFIX).$(SUFFIX) znrm2_k$(TSUFFIX).$(SUFFIX) zdrot_k$(TSUFFIX).$(SUFFIX) \
- zscal_k$(TSUFFIX).$(SUFFIX) zswap_k$(TSUFFIX).$(SUFFIX) zaxpby_k$(TSUFFIX).$(SUFFIX) zsum_k$(TSUFFIX).$(SUFFIX)
-
- XBLASOBJS += \
- xamax_k$(TSUFFIX).$(SUFFIX) xamin_k$(TSUFFIX).$(SUFFIX) ixamax_k$(TSUFFIX).$(SUFFIX) ixamin_k$(TSUFFIX).$(SUFFIX) \
- xasum_k$(TSUFFIX).$(SUFFIX) xaxpy_k$(TSUFFIX).$(SUFFIX) xaxpyc_k$(TSUFFIX).$(SUFFIX) xcopy_k$(TSUFFIX).$(SUFFIX) \
- xdotc_k$(TSUFFIX).$(SUFFIX) xdotu_k$(TSUFFIX).$(SUFFIX) xnrm2_k$(TSUFFIX).$(SUFFIX) xqrot_k$(TSUFFIX).$(SUFFIX) \
- xscal_k$(TSUFFIX).$(SUFFIX) xswap_k$(TSUFFIX).$(SUFFIX) xsum_k$(TSUFFIX).$(SUFFIX)
-
- ### AMAX ###
-
-
-
-
- $(KDIR)samax_k$(TSUFFIX).$(SUFFIX) $(KDIR)samax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SAMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)damax_k$(TSUFFIX).$(SUFFIX) $(KDIR)damax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DAMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)qamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)qamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QAMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)camax_k$(TSUFFIX).$(SUFFIX) $(KDIR)camax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CAMAXKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)zamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)zamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZAMAXKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)xamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)xamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XAMAXKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- ### AMIN ###
-
- $(KDIR)samin_k$(TSUFFIX).$(SUFFIX) $(KDIR)samin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SAMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)damin_k$(TSUFFIX).$(SUFFIX) $(KDIR)damin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DAMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)qamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)qamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QAMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)camin_k$(TSUFFIX).$(SUFFIX) $(KDIR)camin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CAMINKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)zamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)zamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZAMINKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)xamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)xamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XAMINKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- ### MAX ###
-
- $(KDIR)smax_k$(TSUFFIX).$(SUFFIX) $(KDIR)smax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -UUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)dmax_k$(TSUFFIX).$(SUFFIX) $(KDIR)dmax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -UUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)qmax_k$(TSUFFIX).$(SUFFIX) $(KDIR)qmax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -UUSE_ABS -UUSE_MIN $< -o $@
-
- ### MIN ###
-
- $(KDIR)smin_k$(TSUFFIX).$(SUFFIX) $(KDIR)smin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -UUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)dmin_k$(TSUFFIX).$(SUFFIX) $(KDIR)dmin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -UUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)qmin_k$(TSUFFIX).$(SUFFIX) $(KDIR)qmin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -UUSE_ABS -DUSE_MIN $< -o $@
-
-
- ### IAMAX ###
-
- $(KDIR)isamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)isamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ISAMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)idamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)idamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IDAMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)iqamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)iqamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IQAMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)icamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)icamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ICAMAXKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)izamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)izamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IZAMAXKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)ixamax_k$(TSUFFIX).$(SUFFIX) $(KDIR)ixamax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IXAMAXKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -DUSE_ABS -UUSE_MIN $< -o $@
-
- ### IAMIN ###
-
- $(KDIR)isamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)isamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ISAMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)idamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)idamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IDAMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)iqamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)iqamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IQAMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)icamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)icamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ICAMINKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)izamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)izamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IZAMINKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)ixamin_k$(TSUFFIX).$(SUFFIX) $(KDIR)ixamin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IXAMINKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -DUSE_ABS -DUSE_MIN $< -o $@
-
- ### IMAX ###
-
- $(KDIR)ismax_k$(TSUFFIX).$(SUFFIX) $(KDIR)ismax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ISMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -UUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)idmax_k$(TSUFFIX).$(SUFFIX) $(KDIR)idmax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IDMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -UUSE_ABS -UUSE_MIN $< -o $@
-
- $(KDIR)iqmax_k$(TSUFFIX).$(SUFFIX) $(KDIR)iqmax_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IQMAXKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -UUSE_ABS -UUSE_MIN $< -o $@
-
- ### IMIN ###
-
- $(KDIR)ismin_k$(TSUFFIX).$(SUFFIX) $(KDIR)ismin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ISMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -UUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)idmin_k$(TSUFFIX).$(SUFFIX) $(KDIR)idmin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IDMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -UUSE_ABS -DUSE_MIN $< -o $@
-
- $(KDIR)iqmin_k$(TSUFFIX).$(SUFFIX) $(KDIR)iqmin_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(IQMINKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -UUSE_ABS -DUSE_MIN $< -o $@
-
- ### ASUM ###
- $(KDIR)sasum_k$(TSUFFIX).$(SUFFIX) $(KDIR)sasum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SASUMKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)dasum_k$(TSUFFIX).$(SUFFIX) $(KDIR)dasum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DASUMKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qasum_k$(TSUFFIX).$(SUFFIX) $(KDIR)qasum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QASUMKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)casum_k$(TSUFFIX).$(SUFFIX) $(KDIR)casum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CASUMKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)zasum_k$(TSUFFIX).$(SUFFIX) $(KDIR)zasum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZASUMKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)xasum_k$(TSUFFIX).$(SUFFIX) $(KDIR)xasum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XASUMKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE $< -o $@
-
- ### SUM ###
- $(KDIR)ssum_k$(TSUFFIX).$(SUFFIX) $(KDIR)ssum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SSUMKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)dsum_k$(TSUFFIX).$(SUFFIX) $(KDIR)dsum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DSUMKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qsum_k$(TSUFFIX).$(SUFFIX) $(KDIR)qsum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QSUMKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)csum_k$(TSUFFIX).$(SUFFIX) $(KDIR)csum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CSUMKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)zsum_k$(TSUFFIX).$(SUFFIX) $(KDIR)zsum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZSUMKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)xsum_k$(TSUFFIX).$(SUFFIX) $(KDIR)xsum_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XSUMKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE $< -o $@
-
- ### AXPY ###
- $(KDIR)saxpy_k$(TSUFFIX).$(SUFFIX) $(KDIR)saxpy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SAXPYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)daxpy_k$(TSUFFIX).$(SUFFIX) $(KDIR)daxpy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DAXPYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qaxpy_k$(TSUFFIX).$(SUFFIX) $(KDIR)qaxpy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QAXPYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)caxpy_k$(TSUFFIX).$(SUFFIX) $(KDIR)caxpy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CAXPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCOMPLEX -UCONJ -UDOUBLE $< -o $@
-
- $(KDIR)zaxpy_k$(TSUFFIX).$(SUFFIX) $(KDIR)zaxpy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZAXPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UCONJ -DDOUBLE $< -o $@
-
- $(KDIR)xaxpy_k$(TSUFFIX).$(SUFFIX) $(KDIR)xaxpy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XAXPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UCONJ -DXDOUBLE $< -o $@
-
- $(KDIR)caxpyc_k$(TSUFFIX).$(SUFFIX) $(KDIR)caxpyc_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CAXPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCONJ -UDOUBLE $< -o $@
-
- $(KDIR)zaxpyc_k$(TSUFFIX).$(SUFFIX) $(KDIR)zaxpyc_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZAXPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCONJ -DDOUBLE $< -o $@
-
- $(KDIR)xaxpyc_k$(TSUFFIX).$(SUFFIX) $(KDIR)xaxpyc_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XAXPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCONJ -DXDOUBLE $< -o $@
-
- $(KDIR)scopy_k$(TSUFFIX).$(SUFFIX) $(KDIR)scopy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SCOPYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DC_INTERFACE $< -o $@
-
- $(KDIR)dcopy_k$(TSUFFIX).$(SUFFIX) $(KDIR)dcopy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DCOPYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE -DC_INTERFACE $< -o $@
-
- $(KDIR)qcopy_k$(TSUFFIX).$(SUFFIX) $(KDIR)qcopy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QCOPYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE -DC_INTERFACE $< -o $@
-
- $(KDIR)ccopy_k$(TSUFFIX).$(SUFFIX) $(KDIR)ccopy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CCOPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -DC_INTERFACE $< -o $@
-
- $(KDIR)zcopy_k$(TSUFFIX).$(SUFFIX) $(KDIR)zcopy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZCOPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -DC_INTERFACE $< -o $@
-
- $(KDIR)xcopy_k$(TSUFFIX).$(SUFFIX) $(KDIR)xcopy_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XCOPYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -DC_INTERFACE $< -o $@
-
- $(KDIR)ddot_k$(TSUFFIX).$(SUFFIX) $(KDIR)ddot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qdot_k$(TSUFFIX).$(SUFFIX) $(KDIR)qdot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)sdot_k$(TSUFFIX).$(SUFFIX) $(KDIR)sdot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- ifdef DSDOTKERNEL
-
- $(KDIR)dsdot_k$(TSUFFIX).$(SUFFIX) $(KDIR)dsdot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DSDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DDSDOT $< -o $@
-
- $(KDIR)sdsdot_k$(TSUFFIX).$(SUFFIX) $(KDIR)sdsdot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DSDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DDSDOT $< -o $@
-
- else
-
- $(KDIR)dsdot_k$(TSUFFIX).$(SUFFIX) $(KDIR)dsdot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DDSDOT $< -o $@
-
- $(KDIR)sdsdot_k$(TSUFFIX).$(SUFFIX) $(KDIR)sdsdot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SDOTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE -DDSDOT $< -o $@
-
- endif
-
- $(KDIR)zdotu_k$(TSUFFIX).$(SUFFIX) $(KDIR)zdotu_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZDOTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -UCONJ $< -o $@
-
- $(KDIR)zdotc_k$(TSUFFIX).$(SUFFIX) $(KDIR)zdotc_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZDOTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE -DCONJ $< -o $@
-
- $(KDIR)xdotu_k$(TSUFFIX).$(SUFFIX) $(KDIR)xdotu_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XDOTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -UCONJ $< -o $@
-
- $(KDIR)xdotc_k$(TSUFFIX).$(SUFFIX) $(KDIR)xdotc_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XDOTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE -DCONJ $< -o $@
-
- $(KDIR)cdotu_k$(TSUFFIX).$(SUFFIX) $(KDIR)cdotu_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CDOTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -UCONJ $< -o $@
-
- $(KDIR)cdotc_k$(TSUFFIX).$(SUFFIX) $(KDIR)cdotc_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CDOTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE -DCONJ $< -o $@
-
- $(KDIR)snrm2_k$(TSUFFIX).$(SUFFIX) $(KDIR)snrm2_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SNRM2KERNEL)
- $(CC) $(CFLAGS) -UCOMPLEX -c -UDOUBLE $< -o $@
-
- $(KDIR)dnrm2_k$(TSUFFIX).$(SUFFIX) $(KDIR)dnrm2_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DNRM2KERNEL)
- $(CC) $(CFLAGS) -UCOMPLEX -c -DDOUBLE $< -o $@
-
- $(KDIR)qnrm2_k$(TSUFFIX).$(SUFFIX) $(KDIR)qnrm2_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QNRM2KERNEL)
- $(CC) $(CFLAGS) -UCOMPLEX -c -DXDOUBLE $< -o $@
-
- $(KDIR)cnrm2_k$(TSUFFIX).$(SUFFIX) $(KDIR)cnrm2_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CNRM2KERNEL)
- $(CC) $(CFLAGS) -DCOMPLEX -c -UDOUBLE $< -o $@
-
- $(KDIR)znrm2_k$(TSUFFIX).$(SUFFIX) $(KDIR)znrm2_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZNRM2KERNEL)
- $(CC) $(CFLAGS) -DCOMPLEX -c -DDOUBLE $< -o $@
-
- $(KDIR)xnrm2_k$(TSUFFIX).$(SUFFIX) $(KDIR)xnrm2_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XNRM2KERNEL)
- $(CC) $(CFLAGS) -DCOMPLEX -c -DXDOUBLE $< -o $@
-
- $(KDIR)srot_k$(TSUFFIX).$(SUFFIX) $(KDIR)srot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SROTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)drot_k$(TSUFFIX).$(SUFFIX) $(KDIR)drot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DROTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qrot_k$(TSUFFIX).$(SUFFIX) $(KDIR)qrot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QROTKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)csrot_k$(TSUFFIX).$(SUFFIX) $(KDIR)csrot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CROTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)zdrot_k$(TSUFFIX).$(SUFFIX) $(KDIR)zdrot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZROTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)xqrot_k$(TSUFFIX).$(SUFFIX) $(KDIR)xqrot_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XROTKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)sscal_k$(TSUFFIX).$(SUFFIX) $(KDIR)sscal_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SSCALKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)dscal_k$(TSUFFIX).$(SUFFIX) $(KDIR)dscal_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DSCALKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qscal_k$(TSUFFIX).$(SUFFIX) $(KDIR)qscal_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QSCALKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)cscal_k$(TSUFFIX).$(SUFFIX) $(KDIR)cscal_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CSCALKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)zscal_k$(TSUFFIX).$(SUFFIX) $(KDIR)zscal_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZSCALKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)xscal_k$(TSUFFIX).$(SUFFIX) $(KDIR)xscal_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XSCALKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)sswap_k$(TSUFFIX).$(SUFFIX) $(KDIR)sswap_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SSWAPKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)dswap_k$(TSUFFIX).$(SUFFIX) $(KDIR)dswap_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DSWAPKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)qswap_k$(TSUFFIX).$(SUFFIX) $(KDIR)qswap_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(QSWAPKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)cswap_k$(TSUFFIX).$(SUFFIX) $(KDIR)cswap_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CSWAPKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)zswap_k$(TSUFFIX).$(SUFFIX) $(KDIR)zswap_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZSWAPKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)xswap_k$(TSUFFIX).$(SUFFIX) $(KDIR)xswap_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(XSWAPKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -DXDOUBLE $< -o $@
-
- $(KDIR)saxpby_k$(TSUFFIX).$(SUFFIX) $(KDIR)saxpby_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(SAXPBYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -UDOUBLE $< -o $@
-
- $(KDIR)daxpby_k$(TSUFFIX).$(SUFFIX) $(KDIR)daxpby_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(DAXPBYKERNEL)
- $(CC) -c $(CFLAGS) -UCOMPLEX -DDOUBLE $< -o $@
-
- $(KDIR)caxpby_k$(TSUFFIX).$(SUFFIX) $(KDIR)caxpby_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(CAXPBYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UCONJ -UDOUBLE $< -o $@
-
- $(KDIR)zaxpby_k$(TSUFFIX).$(SUFFIX) $(KDIR)zaxpby_k$(TPSUFFIX).$(PSUFFIX) : $(KERNELDIR)/$(ZAXPBYKERNEL)
- $(CC) -c $(CFLAGS) -DCOMPLEX -UCONJ -DDOUBLE $< -o $@
-
-
|
