| #
6a6fc655
|
| 04-Jul-2012 |
Jed Brown <jed@59A2.org> |
Add const to many enum string lists
Hg-commit: c83e9a8fda58664955a3d65190eafbaf3a08ee6d
|
| #
cae2f346
|
| 04-Jul-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
Needed to make PETSC_RUNNING_ON_VALGRIND and hence the location of valgrind.h universal so that PetscCheckPointer() could be used anywhere and not be triggered when run under valgrind
Hg-commit: c65
Needed to make PETSC_RUNNING_ON_VALGRIND and hence the location of valgrind.h universal so that PetscCheckPointer() could be used anywhere and not be triggered when run under valgrind
Hg-commit: c6539cb758dd89a4e698360f117e72536e09cddc
show more ...
|
| #
a8b45ee7
|
| 04-Jul-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
added PetscCheckPointer() that uses sigaction and setjmp to check if a pointer points somewhere valid
Hg-commit: e9f69be1a6a58b4fa53793fc34fe1b574e1afc3f
|
| #
e5996a9a
|
| 15-May-2012 |
Paul Mullowney <paulm@txcorp.com> |
updated and merged from petsc-dev
Hg-commit: 4bdbb24181fe1c8ce13e9b82c1d499b152c3d14f
|
| #
5e765c61
|
| 06-May-2012 |
Jed Brown <jed@59A2.org> |
Use MPI_Init_thread() if available
Hg-commit: 356966214e842e95e7a4e0ad9bab369f74c57905
|
| #
924f458f
|
| 05-May-2012 |
Shri Abhyankar <abhyshr@mcs.anl.gov> |
Disable threadcomm unless PETSC_THREADCOMM_ACTIVE flag is set
Hg-commit: c3b4a0ef897275b07bf7a452d14718892c3f8c4b
|
| #
94c81401
|
| 05-May-2012 |
Shri Abhyankar <abhyshr@mcs.anl.gov> |
Initialize threadcomm in PetscInitialize()
Hg-commit: 6e720397eba0d75eb1822c13f48f29eb13399120
|
| #
f4de128a
|
| 05-May-2012 |
Shri Abhyankar <abhyshr@mcs.anl.gov> |
Test implementation of vector routines VecSet_Seq(), VecCreate_Seq(), VecAXPY_Seq(), and
VecDot_Seq() using the threadcomm interface.
Hg-commit: b78f79002f788fca8cf8fae57286bab682a0e628
|
| #
a1dac056
|
| 26-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 094d51dba926ed63684fca7208e94169d110be19
|
| #
2981ebdb
|
| 26-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
cleanup petscsys.h
remove some dead zope code
added & to section destroyed in DMDestroy_DA()
Hg-commit: 883cc780bf60b4959f61e5a756299945b4b0e965
|
| #
bd8b14e7
|
| 26-Apr-2012 |
Shri Abhyankar <abhyshr@mcs.anl.gov> |
Moved PetscThreadComm initialization to PetscInitialize(). Needs the option -use_threadcomm
Hg-commit: 9eb7da54044ac676dc56b5b34f0513a4c0e8e5aa
|
| #
6be1cbc2
|
| 24-Apr-2012 |
Matthew G Knepley <knepley@gmail.com> |
Merge
Hg-commit: 7e05a2b180c432ff273587a975dd35d18e53cafa
|
| #
10874c6b
|
| 24-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 1675b271e40142db39c5369e1cb1f55cd4b79291
|
| #
2414b3c7
|
| 24-Apr-2012 |
Jed Brown <jed@59A2.org> |
backout b30fbcc6d895 and 8a997a5a65d3, MPIU_SUM is necessary
Although MPI_SUM can (and must) be used with MPI_Accumulate, it cannot be used on derived data types with MPI_Allreduce.
Hg-commit: 1453
backout b30fbcc6d895 and 8a997a5a65d3, MPIU_SUM is necessary
Although MPI_SUM can (and must) be used with MPI_Accumulate, it cannot be used on derived data types with MPI_Allreduce.
Hg-commit: 145359318f0d3a8f2dca9591c3c17e298ae30733
show more ...
|
| #
9bccae41
|
| 23-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 719627a6cee54b4367ab263a30985c06b3d9399d
|
| #
a9207378
|
| 22-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
removed unneeded PetscInitializedMatlab()
Hg-commit: 3562debbf734b52d646a54974b6a2663d868ff1c
|
| #
945d1669
|
| 21-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
removed Matlab from name of PETSc functions used by Matlab and Julia and removed conditional compile for them
Hg-commit: 5ffefe13ed62057155dba8bb623b3d2a4db21202
|
| #
b0c94c62
|
| 20-Apr-2012 |
Jed Brown <jed@59A2.org> |
Do not need our own MPIU_SUM for complex, standard MPI_SUM is fine
Hg-commit: b30fbcc6d895ca9398230de8d0ed94a874aaf3fc
|
| #
ff9dbc5d
|
| 19-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
started adding Vec class
Hg-commit: 178ca803bbee7e0b3f99ae5b395bc30942cd626d
|
| #
7034d64c
|
| 19-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
started work on a Julia interface for PETSc
Hg-commit: 86bcee4fd95712cf04fbbcf5f5aee3796ce70914
|
| #
2adc2576
|
| 08-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: a7033f44b7153ddbc07db52896b096f5b5ac7e60
|
| #
ad74a23b
|
| 08-Apr-2012 |
Jed Brown <jed@59A2.org> |
merge PGMRES
Hg-commit: 2ed7d2b653d8d756e316e1b5a6232b04c6c99601
|
| #
8bf1f09c
|
| 07-Apr-2012 |
Shri Abhyankar <abhyshr@mcs.anl.gov> |
Removed unused flag PETSC_USE_PTHREAD
Hg-commit: f04ea9eabf50a1401c99e4b0884302b9ee768645
|
| #
f32c38ef
|
| 02-Apr-2012 |
Barry Smith <bsmith@mcs.anl.gov> |
removed __gierr and _ and __ error handling macros
Hg-commit: c5e26b6474802e7782241137c2819a23d99022d9
|
| #
205a32c2
|
| 17-Mar-2012 |
Jed Brown <jed@59A2.org> |
Namespace a bunch of global variables used for logging
Perhaps these should be placed in a smaller number of thread-local structures.
Note that I used the "petsc_" prefix for these because it's wha
Namespace a bunch of global variables used for logging
Perhaps these should be placed in a smaller number of thread-local structures.
Note that I used the "petsc_" prefix for these because it's what was
used by nearby variables. These are not public symbols, so it's
inconsequential (in the sense that it can be changed without impacting
source-level compatibility and does not appear in documentation), but
the symbols have extern linkage, so they still need to be namespaced.
Hg-commit: 5b2e3c0b1ea530838efefabde786d6294274b5d2
show more ...
|