An optimized BLAS library based on GotoBLAS2
OpenBLAS is an optimized BLAS library based on GotoBLAS2 1.13 BSD version. The project is supported by the Lab of Parallel Software and Computational Science, ISCAS. http://www.rdcps.ac.cn
-
1
derived packages
- Download package
-
Checkout Package
osc -A https://api.opensuse.org checkout openSUSE:Leap:15.0/openblas && cd $_
- Create Badge
Refresh
Refresh
Source Files
Filename | Size | Changed |
---|---|---|
README.HPC.SUSE | 0000000889 889 Bytes | |
README.SUSE | 0000000851 851 Bytes | |
_constraints | 0000000171 171 Bytes | |
_multibuild | 0000000180 180 Bytes | |
c_xerbla_no-void-return.patch | 0000000367 367 Bytes | |
fix-arm64-cpuid-return.patch | 0000000584 584 Bytes | |
openblas-0.2.20.tar.gz | 0011637301 11.1 MB | |
openblas-noexecstack.patch | 0000001189 1.16 KB | |
openblas-s390.patch | 0000001060 1.04 KB | |
openblas.changes | 0000017660 17.2 KB | |
openblas.spec | 0000013522 13.2 KB |
Latest Revision
Ludwig Nussel (lnussel_factory)
accepted
request 574427
from
Leap Reviewbot (leaper)
(revision 5)
Automatic request from SUSE:SLE-15:GA by UpdateCrawler
Comments 1
Hello, there. I've found a pretty nasty bug with the pthreads version of the OpenBLAS package on the main repos. The TL;DR is that if you happen to link it in (in my case, it was brought in through OpenCV) it kills OpenMP for the entire thing, as if it had defined
omp_get_num_procs()
andomp_get_max_threads()
to always return 1.I've been trying to chase the bug down to suggest a fix and while doing it, noticed that the updated version (0.3.4) of the package in the science factory repo does not have the problem.
I'm looking for the right channel to bring this information to such that the bug can be worked out of the main upstream repositories. Would this be it?
I've been going in more detail in a thread on the forums ( https://forums.opensuse.org/showthread.php/534152-OpenBLAS-(-openblas_pthreads0-)-disables-OpenMP-for-whatever-it-s-linked-into )
If this is not it, could you point me in the right direction. I'm willing to bang my head against this a little more to help zero in a smooth fix