CUDA: Fixed compilation problems with complex.Reported and patch provided by: Greg Meyer <gregory.meyer@gmail.com> via petsc-maint
GPUs: Allow ViennaCL and VECCUDA to be used concurrently.This also gets rid of the weird veccuda.py.
CUDA: Fixed most test errors due to MatMultTranspose*
change MatSolverType to be a const char* typedef to match other XXTypeCommit-type: style-fixReported-by: Franck Houssen <franck.houssen@inria.fr>
Merge branch 'barry/matsolverpackage-matsolvertype'
Fix name MatSolverPackage since it is better to be consistent and use the terminology Type.Commit-type: style-fix, documentation
Rename some private functionsRename some private functions which start with XXX_, where XXX is the name of a main PETSc class,and mark static if needbe.
CUSPARSE: Fix MatDuplicate()MatDuplicate_SeqAIJ() used to be called for SEQAIJCUSPARSE matrices,which did not set up the GPU datastructures.This commit provides a dedicated MatDuplicate_SeqAIJCUS
CUSPARSE: Fix MatDuplicate()MatDuplicate_SeqAIJ() used to be called for SEQAIJCUSPARSE matrices,which did not set up the GPU datastructures.This commit provides a dedicated MatDuplicate_SeqAIJCUSPARSE() implementation.Reported-by: Jose Roman
show more ...
Reduce Vec type checks on CUDA/CUSP/VIENNACL
Add several Vec type checks on CUDA/CUSP/VIENNACL
Add macros PetscCheckTypeName(s)
Merge remote-tracking branch 'origin/jose/cusparse-sethandle'
MatCUSPARSESetHandle: do not destroy handle to set
Fix MatMult/MatMultTranspose SeqAIJCUSPARSE
Remove the use and definition of __FUNCT__ throughout the codeSince all modern C/C++ compilers provide this functionality we no longer need to provide it manually in PETScTime: 1.5 hoursThanks-t
Remove the use and definition of __FUNCT__ throughout the codeSince all modern C/C++ compilers provide this functionality we no longer need to provide it manually in PETScTime: 1.5 hoursThanks-to: Andreas Mang <andreas@ices.utexas.edu>
Fix performance issue in MatAIJCUSPARSE after 34d6c7aIn MatSeqAIJCUSPARSECopyToGPU the case of simplified copy must alsoset the GPU consistency flag
Miscellaneous changes in AIJCUSPARSE code- Some operations need synchronization of data from CPU to GPU, e.g. MatScale- Add assembly operation, so that lvec is set to type CUDA- In MatSeqAIJCUSPA
Miscellaneous changes in AIJCUSPARSE code- Some operations need synchronization of data from CPU to GPU, e.g. MatScale- Add assembly operation, so that lvec is set to type CUDA- In MatSeqAIJCUSPARSECopyToGPU do not recompute indices every time- Operations MatMultAdd* need to get the array with ReadWrite- In MatScale, do not change coherence flag if alpha=1.0- VecDuplicate_MPICUDA must make sure that the GPU array exists- Fix null pointer in VecScatterCUDAIndicesCreate_PtoP
Add const to VecCUDAGetArrayRead() to match with VecGetArrayRead()
Fix use of not initialized variable
Fix VECCUDA compilation with complex and single precision
Enable complex scalars with VECCUDA
Use VECCUDA with MATAIJCUSPARSE
Fix visibility issues in CUSP and VECCUDA
Begin renaming the PetscOptions objects for the future where PetscOptions will not be a singleton
fix violations of PETSc style guide: Usage of SETERRQ and NULL
1...<<11121314151617