[clblas] annotated tag v2.12 created (now d0c29b2)

Ghislain Vaillant ghisvail-guest at moszumanska.debian.org
Tue Jan 24 23:31:05 UTC 2017


This is an automated email from the git hooks/post-receive script.

ghisvail-guest pushed a change to annotated tag v2.12
in repository clblas.

        at  d0c29b2   (tag)
   tagging  1f3de2ae5582972f665c685b18ef0df43c1792bb (commit)
  replaces  upstream/v2.8+ds1
 tagged by  Kent Knox
        on  Wed Jan 18 11:06:03 2017 -0600

- Log -----------------------------------------------------------------
bugfix rollup release

Abhishek Shandilya (1):
      fix #265 - spelling errors in comments and print statements (#276)

Anton Lokhmotov (1):
      Only use the -m32 or -m64 compiler flags on x86.

Cyril Lashkevich (1):
      Fixed clang build error

David Tanner (14):
      Merge pull request #163 from hughperkins/fix-teardown
      Merge pull request #170 from hughperkins/fix-buggette-in-pull-163
      adding AutoGemm kernel selection logic for Fiji
      Merge pull request #202 from arrayfire/arrayfire-release-test
      Merge pull request #234 from haahh/update-travis
      proposed fix for gemm thread safety; using thread-local storage for kernel map using pre-C++-11 syntax
      re-submit after CI fix; removing dummy whitespace
      TLS for gcc 4.6
      fixing duplicate include and removing TODO note
      compiling kernels is now thread safe; not using global cl_kernel objects
      thread safety: no longer using global cl_kernel objects. thread safety is fixed pending customer verification
      fixed compareMatrices to use GTEST_FLOAT_EQ
      Merge pull request #249 from hughperkins/missing-kernelname
      Merge pull request #254 from hughperkins/gitignore-add-pyc

Ghislain Antony Vaillant (1):
      add missing include on stdlib.h

Hugh Perkins (10):
      Fix kernel crash on nvidia, caused by float4 alignemtn error, see https://github.com/clMathLibraries/clBLAS/issues/108 for more details
      fix 'array initializer must be an initializer list', https://github.com/clMathLibraries/clBLAS/issues/153
      Fix https://github.com/clMathLibraries/clBLAS/issues/159 , teardown/setup when using autogemm causes next call to gemm to fail (segfault)
      Fix catch-22 in build order, following pull-163, where init.c compile fails because AutoGemmClKernels.h hasnt been built yet
      fix hard-coding of opencl version to 2.0; fix 1d initialization of 2d arrays.
      Fix teardown of UserGemmClKernels
      Update gtest link to point to github
      fix error with missing KernelName variable
      Add .pyc files to .gitignore
      Make installing source tree optional

Ivan Vergiliev (2):
      Disable clang error on narrowing conversions.
      Point the CONTRIBUTING wiki links to the correct repository

J M Dieterich (1):
      Protect pragma in preprocessor macro by using _Pragma. clang 3.7 will not allow compilation of the code otherwise (found on FreeBSD-CURRENT).

Jakub Szuppe (1):
      Update .travis.yml and appveyor.yml

Jérôme Carretero (1):
      add missing dependency to pthread (using rwlock functions)

Kent Knox (19):
      Merge pull request #222 from dividiti/arm-support
      Merge pull request #235 from guacamoleo/develop
      Merge pull request #244 from tingxingdong/test
      Revert "fixed compareMatrices to use GTEST_FLOAT_EQ"
      Adding detection for boost 1.60
      Removed ::cerr wrt calling reference and clblas
      Removing the printing of unit test parameters
      Device selection for test-correctness and test-functional
      Merge pull request #258 from kknox/unit-test-improvements
      Merge pull request #252 from hughperkins/optional-install-src
      Removing the pedantic flag from gcc compiles
      Commenting out further #pragma warning messages
      Merge pull request #264 from kknox/fix-warnings
      Merge pull request #281 from IvanVergiliev/narrow-conversions
      Adding additional trsm samples
      Bump version to 2.12.0
      Update README with release notes
      Merge branch 'master' into 2.12
      Merge pull request #295 from kknox/2.12

Mark Gates (2):
      x offset stored in offb, not offa, determines vectorization
      syr2: Y uses incy, not incx

Pavan Yalamanchili (9):
      Add caching mechanism based on context and device for gemm and trsm
      Work around for an nvidia bug when querying kernel function name
      Changing AutoGemm scripts to work with both python2 and python3
      Attempt to build from souce if build from binary fails
      Find the python interpreter using cmake
      Fixing issue with beta == 0 in AutoGemm kernels
      Fixing integer divides to make clBLAS work when building with python3
      Trtri kernel build options were hard coded to 2.0
      Fixing issues for when Beta == 0 in sgemm special cases

Shehzan Mohammed (3):
      Add cl_khr_fp64 when using double precision
      Fixing issue with beta == 0 in UserGemm kernels
      Fix barriers in dtrsm specialized kernels

Timmy (32):
      Update .travis.yml
      Update appveyor.yml
      bump develop branch version number to 2.9.0
      Update CMakeLists.txt
      Merge pull request #157 from TimmyLiu/develop
      merge master branch to develop branch. please only make pull requests to develop branch
      Merge pull request #174 from TimmyLiu/develop
      Merge pull request #175 from hughperkins/fix-teardown-userkernels-rebase-develop
      Merge pull request #171 from ghisvail/fix/missing-stdlib-include
      Merge pull request #115 from hughperkins/develop
      Merge pull request #178 from TimmyLiu/develop
      Merge pull request #181 from TimmyLiu/develop
      Merge pull request #183 from hughperkins/update-gtest-link
      Merge pull request #190 from notorca/master
      if BUILD_TEST is not set fortran compiler should not be required.
      Merge pull request #198 from tingxingdong/performanceData1
      Merge pull request #199 from arrayfire/arrayfire-release-test
      Merge pull request #200 from arrayfire/arrayfire-release-test
      attempts to fix travis ci build with missing fglrx
      previous commit failed to verify khrons' certificate. try dist trusty now
      typo fix
      not build client for now
      Merge pull request #209 from TimmyLiu/develop
      bump master branch version number to 2.10.0
      Merge pull request #210 from TimmyLiu/master
      bump develop version number to 2.11.0
      Merge pull request #213 from TimmyLiu/develop
      fix some exception hanlers. now test-functional all pass
      Merge pull request #214 from TimmyLiu/develop
      Merge pull request #221 from arrayfire/arrayfire-release-test
      Merge pull request #216 from iotamudelta/develop
      Merge pull request #242 from tingxingdong/test

tfauck (1):
      Support for altivec on powerpc64 P8 systems (#262)

tim (2):
      fix the compilation bug about c(z)dotc_
      fix the header accordingly

timmy (1):
      avoid removing userGemmClKernels.cc with make clean

timmyliu (4):
      find libblas.so correctly. use libblas.so for testing by default instead of libacml.so
      need gfortran for testing with netlib blas
      try to get gfortran working on travis ci
      remove version number

tingxingdong (11):
      Merge pull request #194 from guacamoleo/develop
      Merge pull request #193 from TimmyLiu/develop
      add perf data for fiji
      fix a hard coding bug
      put the numQueues to be 1
      fix a bug in gflops count
      Merge pull request #245 from guacamoleo/develop
      allow users to easily verify the gemm/trmm GPU results with the netlib cblas through client (#274)
      Merge pull request #282 from IvanVergiliev/fix-wiki-links-develop
      Merge pull request #283 from exmakhina/develop
      Merge pull request #290 from mgates3/develop

unknown (1):
      release event obj in sample code

-----------------------------------------------------------------------

No new revisions were added by this update.

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/clblas.git



More information about the debian-science-commits mailing list