Martin Kroeker
8a9d492af7
Add default for blas_omp_threads_local
1 year ago
Martin Kroeker
87d31af2ae
Add openblas_set_num_threads_local()
1 year ago
Martin Kroeker
e7a895e714
Add Apple M as NeoverseN1
1 year ago
Chris Sidebottom
dc20a78188
Use functionally equivalent dynamic targets
Similar to `drivers/other/dynamic.c`, I've looked for functionally
equivalent targets and mapped them in the default DYNAMIC_ARCH build.
Users can still build specific cores using DYNAMIC_LIST.
1 year ago
Mark Seminatore
6bd7c54af5
introduce MT_TRACE to clean up SMP_DEBUG code
1 year ago
Mark Seminatore
edac80d7e8
some cleanup, dynamically scale threads, add missing WIN_CASE defn
1 year ago
Mark Seminatore
4ebf814b42
fix bug failing to mark task as finished.
1 year ago
Mark Seminatore
5f51811728
try at new threading model
1 year ago
Shiyou Yin
1310a0931b
loongarch: Refine build control for loongarch64.
1. Use getauxval instead of cpucfg to test hardware capability.
2. Remove unnecessary code and option for compiler check in c_check.
1 year ago
Chip-Kerchner
d99aad8ee3
Fix older version of gcc - missing __has_builtin, cpuid and no support of P10.
1 year ago
Martin Kroeker
9b5f8eb33a
Fix empty function prototypes
1 year ago
Martin Kroeker
9324520d0e
typo fix
1 year ago
Martin Kroeker
ff6437f2d7
Add workaround for omp_get_max_threads hanging on FreeBSD with libomp from LLVM14
1 year ago
Chip-Kerchner
4eecccd49b
Fix __builtin_cpu_is for AIX.
1 year ago
Chip-Kerchner
5e31c57083
Only define __builtin_cpu_is and __builtin_cpu_supports if not present.
1 year ago
Chip-Kerchner
7dcb2d67f2
Have POWER7 return arch=POWER6.
1 year ago
Chip-Kerchner
c8882bd9d8
Remove POWER7 from cpu list.
1 year ago
Chip Kerchner
badfb2e60f
Merge branch 'develop' into XLC-AIX
1 year ago
Martin Kroeker
e12aaed13d
Fix unwanted fallthrough from Intel Family 6 to 15 in case of identification failure
1 year ago
Chip-Kerchner
880af052dd
Fix dynamic dispatch P9 for clang.
2 years ago
Chip-Kerchner
3655632611
Another small change.
2 years ago
Chip-Kerchner
36e08f6994
One more small change.
2 years ago
Chip-Kerchner
298bf1f240
Reduce differences.
2 years ago
Chip-Kerchner
71c6689af4
Fix dynamic dispatch to work for clang.
2 years ago
Chip-Kerchner
c60f9d9c08
Add missing CPU_POWER5.
2 years ago
Chip Kerchner
3cc72a3797
Only include cpu_id and cpu_supports in AIX and fix parameter types.
2 years ago
Chip-Kerchner
09212f84bf
Fix default case for cpu_is.
2 years ago
Chip-Kerchner
2d0b233425
Fix missing parens.
2 years ago
Chip-Kerchner
a8c90eb3ed
Added cpu_is
2 years ago
Chip-Kerchner
b677d0d5fd
Adding missing endif
2 years ago
Chip-Kerchner
e5dc376912
Remove duplicate defines.
2 years ago
Chip-Kerchner
10210748de
Revert PGI changes.
2 years ago
Chip-Kerchner
a922a07e61
Cleanup white spaces.
2 years ago
Chip-Kerchner
12130ee961
Remove tab.
2 years ago
Chip-Kerchner
eb738d9929
Minor changes.
2 years ago
Chip-Kerchner
48da98b2a7
Merge remote-tracking branch 'origin/develop' into XLC-AIX
2 years ago
Chip-Kerchner
3b1150fcee
Fix CPU identification to work on AIX.
2 years ago
Martin Kroeker
90f890ee67
fix improper function prototypes (empty parentheses) (USE_TLS branch)
2 years ago
Martin Kroeker
cf2174fb69
fix improper function prototypes (empty parentheses)
2 years ago
Martin Kroeker
c6b1d8e7a3
fix improper function prototypes (empty parentheses)
2 years ago
Martin Kroeker
c4bd4a2e5d
fix improper function prototypes (empty parentheses)
2 years ago
Martin Kroeker
7e939fb831
Fix handling of additional buffer structures in case of overflow
2 years ago
Tiziano Müller
6a611db560
memory: show correct number of max threads
2 years ago
Martin Kroeker
c2f4bdbbb4
Merge pull request #4163 from martin-frbg/issue4017
Rework OpenMP thread count limit handling
2 years ago
Martin Kroeker
9ff84dc3f2
remove unused status variable
2 years ago
Martin Kroeker
3326b924b3
remove status variable blas_num_threads_set; initialize openmp thread maximum on startup
2 years ago
Chris Sidebottom
f971ef55f2
Add ARMV8SVE to AArch64 Dynamic Dispatch
In order to enable support for future cores which have similar tunings
(in this case I'm doing this for the Arm(R) Neoverse(TM) V2 core), this generically detects SVE support and enables it. This should better manage the size and complexity of dynamic dispatch rather than just copy pasting the same parameters.
To make `ARMV8SVE` more representive of the common 128-bit SVE case,
I've split it and similar parameters from A64FX which has the wider
512-bit SVE.
2 years ago
Martin Kroeker
3bdcf3259d
Merge branch 'xianyi:develop' into issue4101
2 years ago
Martin Kroeker
b34f19a365
Ensure that a premature call to set_num_threads will not overwrite unrelated memory
2 years ago
Martin Kroeker
66904f8148
Ensure that a premature call will not overwrite unrelated memory
2 years ago