| dfff0f24 | 10-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
Redefine MKL complex types so things work when building PETSc with complex number support. |
| 87c2a1d7 | 09-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
Added MatDiagonalScale, MatDiagonalSet, and MatAXPY methods for SeqAIJMKL; fixed ops->scale not being reset in conversion from SeqAIJMKL to SeqAIJ. |
| ef76dfe8 | 09-Aug-2017 |
Jed Brown <jed@jedbrown.org> |
MatMPIAIJGetLocalMat: small doc fix |
| 3457812c | 05-Aug-2017 |
Satish Balay <balay@mcs.anl.gov> |
Merge remote-tracking branch 'origin/hzhang/reorg-matdestroysubmats' |
| 86e85357 | 03-Aug-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
for baij and dummy matrices |
| f68bb481 | 03-Aug-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
rename |
| d0118459 | 03-Aug-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
bugfix for MatCreateSubMatrix() for reuse and !sorted colomn IS |
| 3fa15762 | 02-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
The _SpMV2 versions of the SeqAIJMKL operations now check to see if mkl_sparse_optimize() has been called, and call it if needed. |
| 0fb991dc | 02-Aug-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
rename MatDestroy_SeqAIJ_Submatrices -> MatDestroySubMatrix_SeqAIJ |
| 8d3fe1b0 | 01-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
Ensured that MKL sparse inspector-executor routines are not called for matrices with no nonzero entries. |
| c5aa4e58 | 01-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
Minor edit: Fixed spacing for a function call. |
| 2b11a2d1 | 01-Aug-2017 |
Karl Rupp <me@karlrupp.net> |
Merge branch 'karlrupp/fix-MatDuplicate-GPU'
* karlrupp/fix-MatDuplicate-GPU:
Provides MatDuplicate() functionality for AIJCUSPARSE and AIJVIENNACL matrices.
Reported-by: Jose Roman |
| 969800c5 | 01-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
More fixes for transposed versions of MatMult/MatMultAdd for SeqAIJMKL; enabled those versions. |
| db63039f | 01-Aug-2017 |
Richard Tran Mills <rmills@rmills.org> |
Added MatScale_SeqAIJMKL(); fixed many errors found when testing with
rectangular matrices. |
| c3cca76e | 31-Jul-2017 |
Karl Rupp <me@karlrupp.net> |
ViennaCL: Added MatDuplicate()
Fix similar to MatDuplicate() for CUSPARSE implemented in
9ff858a8fa4d1b883cb00760b421121f0c50abc9 |
| 2d033e1f | 31-Jul-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
add MATOP_DESTROYSUBMATRICES to cleanup MatDestroySubMatrices()
Reported-by: Stefano Zampini <stefano.zampini@gmail.com> |
| c23be786 | 31-Jul-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
Merge branch 'hzhang/bugfix-RedundantMat/master' |
| 6008c479 | 31-Jul-2017 |
Hong Zhang <hzhang@mcs.anl.gov> |
Merge branch 'hzhang/bugfix-matdestroysubmats' |
| 6e369cd5 | 28-Jul-2017 |
Richard Tran Mills <rmills@rmills.org> |
Created a function that creates MKL SpMV2 matrix handles and performs optimization to eliminate code duplication. |
| 9b6041cc | 28-Jul-2017 |
Richard Tran Mills <rmills@rmills.org> |
Removed obsolete __FUNCT__ definitions. |
| 9ff858a8 | 28-Jul-2017 |
Karl Rupp <rupp@iue.tuwien.ac.at> |
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 ...
|
| f36dfe3f | 28-Jul-2017 |
Richard Tran Mills <rmills@rmills.org> |
Prevent MKL sparse inspector-executor routines from being called on
matrices with zero rows -- the routines don't handle this case. |
| 58678438 | 28-Jul-2017 |
Richard Tran Mills <rmills@rmills.org> |
Fixed nrows being passed to mkl_sparse_x_create_csr() for both nrows and
ncols. |
| f68ad4bd | 27-Jul-2017 |
Richard Tran Mills <rmills@rmills.org> |
Added error messages for failures in creation/copy of matrix handles or
mkl_sparse_optimize. |
| e9c94282 | 27-Jul-2017 |
Richard Tran Mills <rmills@rmills.org> |
Fixes following the outline of Barry's fixes to AIJPERM. |