| ec0d0771 | 17-Aug-2011 |
Matthew G Knepley <knepley@gmail.com> |
Fixed up MatSetValuesBatch() and now KSP ex4 verifies
Hg-commit: a7329fdf61e83d066cba0b54f0e193636ad4d3f4 |
| 37868618 | 17-Aug-2011 |
Matthew G Knepley <knepley@gmail.com> |
Added MatSetValuesBatch()
- The idea is to extract parallelism here, analogous to MatSetValuesBlocked()
- The default implementation just does a loop over MatSetValues() calls
Hg-commit: 5b195adbe6e
Added MatSetValuesBatch()
- The idea is to extract parallelism here, analogous to MatSetValuesBlocked()
- The default implementation just does a loop over MatSetValues() calls
Hg-commit: 5b195adbe6eb86917bc786cd1579c07684d7579e
show more ...
|
| 4a8d9b6d | 16-Aug-2011 |
Matthew G Knepley <knepley@gmail.com> |
Removed EXTERN_C from set values batch functions
Hg-commit: 1ca9c7a5d2a41f469ebba5019d180a7bc2a57ecd |
| 7d6a0e61 | 15-Aug-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
updated discussion of PETSc with threads
fixed MATSEQPTHREADAIJ to MATSEQAIJPTHREAD
Hg-commit: 264131f1283f2e9b3d7687b1b961f6b25c50057e |
| b3936283 | 12-Aug-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 023d6c6af2b31621eff92a00274d614b53ecfb4b |
| 8f1a2a5e | 12-Aug-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed sqrt() to PetscSqrtReal() to work properly with __float128
changed some %G to use %14.12e and (double) to work with __float128
Hg-commit: a8faa4e64896cf76f3f64094484ef5e06761417b |
| 2291e4fd | 11-Aug-2011 |
Jed Brown <jed@59A2.org> |
ibdiagvalid value should have type PetscBool
Hg-commit: 12888e04885d98800c5a4527e9945695be68ec26 |
| 34fc4b71 | 11-Aug-2011 |
Jed Brown <jed@59A2.org> |
Remove unused variable
Hg-commit: d0cc4cc46d4b4e01436ebfc114a9910808b3ce04 |
| 8a813f54 | 11-Aug-2011 |
Jed Brown <jed@59A2.org> |
Spell "PetscImaginaryPart" correctly
Hg-commit: 2bec26de86df93c86fe764f929b8887dc21b5c8a |
| 2c990fa1 | 09-Aug-2011 |
Hong Zhang <hzhang@mcs.anl.gov> |
bugfix of MatView() for factored seqaij and seqsbaij matrices
Hg-commit: b2a05b2919dfa8cfa9deeaf89b795711aa2a0691 |
| bbead8a2 | 04-Aug-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
updated PCPBJACOBI to work for AIJ matrices with set block size also
Hg-commit: 0e020f5b49d3e02ffdf403ba777ccbc6d0077bb9 |
| 7beb759e | 23-Jul-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 3588111ae446783418e6d80584c2c73dd6251041 |
| 63db405b | 23-Jul-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
unprotected pthreadclasses stuff
changed petscthread.py to not set pthreadclasses if not asked for, I don't fucking understand buildsystem
Hg-commit: 78fe98403c3fd0d24b12154743e831052982380c |
| 700485dd | 22-Jul-2011 |
Satish Balay <balay@mcs.anl.gov> |
make sure the default [without] pthread build works on mac/linux
Hg-commit: 3808130f8ca50168716c9a605970cb69fb7beb7e |
| b50af74f | 22-Jul-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
fixed fix of seqpthread vector and seqaijpthread to compile with c++
Hg-commit: 7b1cb1002f2d2c2300226483b7873be08fdd17d5 |
| 4b83fb64 | 22-Jul-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
Brad's fixes to get the Pthread Vec and Mat classes to compile with C++
Hg-commit: 077de594679dd06aa3a59be67bd36f3e72830a9e |
| 1dd75c29 | 22-Jul-2011 |
Kerry Stevens <kstevens@spetjhiong.com> |
commit after merge
Hg-commit: 3bf01fac98a3d658df5a9ca709c44bb8f15f7102 |
| cfcfc605 | 22-Jul-2011 |
Kerry Stevens <kstevens@spetjhiong.com> |
added functionality to map threads to cores on NON thread pool case
Hg-commit: 449c14ae9417e8418bf2d8e3c6483d3439439f76 |
| ff34cdc8 | 22-Jul-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed PETSC_USE_PTHREAD_CLASS to PETSC_HAVE_PTHREADCLASS so that name matches pthreadclasses.py name and cmake makes sure that specific code gets compiled
Hg-commit: 0a01a8e82fe4f48e319b836eb8da0
changed PETSC_USE_PTHREAD_CLASS to PETSC_HAVE_PTHREADCLASS so that name matches pthreadclasses.py name and cmake makes sure that specific code gets compiled
Hg-commit: 0a01a8e82fe4f48e319b836eb8da000af0532501
show more ...
|
| 0f9369f8 | 13-Jul-2011 |
Mark F. Adams <mark.adams@columbia.edu> |
refactored gamg PC. Parallel code added, not complete. MIS interface made better for future migration to a matop.
Hg-commit: 018311eb3f493463af3eaf23b17b29cb816f2ca6 |
| df17f0d5 | 18-Jul-2011 |
Kerry Stevens <kstevens@spetjhiong.com> |
commit after merge from petsc-dev (MatMult_Add function)
Hg-commit: 95a831be5dc2ba75bb17311cad14d2f38977c0b7 |
| c370d283 | 18-Jul-2011 |
Kerry Stevens <kstevens@spetjhiong.com> |
committed after merging Stevens & Smith changes
Hg-commit: 4180409d5c98871cc6c7e567a401ce07f093e17f |
| 9e800a48 | 18-Jul-2011 |
Kerry Stevens <kstevens@spetjhiong.com> |
corrected a problem that prohibited NON-thread pool & sequential from running
Hg-commit: 8809389b99ffe444be7e29a598d4ec17a3a9d0b6 |
| ba61063d | 15-Jul-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
added PETSC_USE_PTHREAD_CLASSES protection
removed // comments
added protection of pthread_barrier does not exist
Hg-commit: 972a3b80c9a6f09bc54db7149d59da967e72545e |
| 51d315f7 | 15-Jul-2011 |
Kerry Stevens <kstevens@spetjhiong.com> |
added thread pool versions
Hg-commit: 77a23f76a8d8ac73a47e6098a1f3e3805e47baf0 |