Public bug reported:
Its kind of late but I do think it would be a good idea to upgrade openblas to
the latest available version 2.18
Compared to the current version 2.15 it includes a bunch of bugfixes, some
quite important performance improvements and better support for some amd,
powerpc and arm cpus.
As the meat of openblas is largely assembly and it is somewhat prone to having
bugs, especially in its legacy code paths, having the newest version will also
simplify future backporting of patches.
changelog:
OpenBLAS ChangeLog
====================================================================
Version 0.2.18
12-Apr-2016
common:
* If you set MAKE_NB_JOBS flag less or equal than zero,
make will be without -j.
x86/x86_64:
* Support building Visual Studio static library. (#813, Thanks,
theoractice)
* Fix bugs to pass buidbot CI tests (http://build.openblas.net)
ARM:
* Provide DGEMM 8x4 kernel for Cortex-A57 (Thanks, Ashwin Sekhar T K)
POWER:
* Optimize S and C BLAS3 on Power8
* Optimize BLAS2/1 on Power8
====================================================================
Version 0.2.17
20-Mar-2016
common:
* Enable BUILD_LAPACK_DEPRECATED=1 by default.
====================================================================
Version 0.2.16
15-Mar-2016
common:
* Avoid potential getenv segfault. (#716)
* Import LAPACK svn bugfix #142-#147,#150-#155
x86/x86_64:
* Optimize c/zgemv for AMD Bulldozer, Piledriver, Steamroller
* Fix bug with scipy linalg test.
ARM:
* Improve DGEMM for ARM Cortex-A57. (Thanks, Ashwin Sekhar T K)
POWER:
* Optimize D and Z BLAS3 functions for Power8.
====================================================================
Version 0.2.16.rc1
23-Feb-2016
common:
* Upgrade LAPACK to 3.6.0 version.
Add BUILD_LAPACK_DEPRECATED option in Makefile.rule to build
LAPACK deprecated functions.
* Add MAKE_NB_JOBS option in Makefile.
Force number of make jobs.This is particularly
useful when using distcc. (#735. Thanks, Jerome Robert.)
* Redesign unit test. Run unit/regression test at every build
(Travis-CI and Appveyor).
* Disable multi-threading for small size swap and ger. (#744. Thanks,
Jerome Robert)
* Improve small zger, zgemv, ztrmv using stack alloction (#727. Thanks,
Jerome Robert)
* Let openblas_get_num_threads return the number of active threads.
(#760. Thanks, Jerome Robert)
* Support illumos(OmniOS). (#749. Thanks, Lauri Tirkkonen)
* Fix LAPACK Dormbr, Dormlq bug. (#711, #713. Thanks, Brendan Tracey)
* Update scipy benchmark script. (#745. Thanks, John Kirkham)
x86/x86_64:
* Optimize trsm kernels for AMD Bulldozer, Piledriver, Steamroller.
* Detect Intel Avoton.
* Detect AMD Trinity, Richland, E2-3200.
* Fix gemv performance bug on Mac OSX Intel Haswell.
* Fix some bugs with CMake and Visual Studio
ARM:
* Support and optimize Cortex-A57 AArch64.
(#686. Thanks, Ashwin Sekhar TK)
* Fix Android build on ARMV7 (#778. Thanks, Paul Mustiere)
* Update ARMV6 kernels.
POWER:
* Fix detection of POWER architecture
(#684. Thanks, Sebastien Villemot)
** Affects: openblas (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1571811
Title:
FFe for openblas 2.18
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/openblas/+bug/1571811/+subscriptions
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
