| 2442b659 | 06-Jul-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
Register Sieve vector
Hg-commit: 40b9805b32ee7b669897764d2f5f4dc5ccbdfa51 |
| 765467ad | 06-Jul-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
Added Sieve vector
Hg-commit: dde9d6b56ac371e8cefe89efcf54a30e14dd3d3e |
| a4a7e6be | 06-Jul-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
Just comments
Hg-commit: e902be44bbd61a41103caef4bc5cd4394137cb76 |
| 49a1719e | 05-Jul-2006 |
Satish Balay <balay@mcs.anl.gov> |
remove references to RANDOM_DEFAULT
Hg-commit: 3a92de1c2f2faa407092f3bc61cebe9554d1b669 |
| 7c617552 | 30-Jun-2006 |
Satish Balay <balay@mcs.anl.gov> |
c99 complex example fixes
Hg-commit: 98653f66515cde4922652e4a96e0a1ea4ea5a966 |
| 5f481a85 | 30-Jun-2006 |
Satish Balay <balay@mcs.anl.gov> |
more c99 complex fixes
Hg-commit: fb4c43bce3651f2e8c064cf6e9858cd58c7ea692 |
| 1f6afcd2 | 29-Jun-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
Do not allow logging of negative flops
Hg-commit: 98fae6a1268711b61c4a71bca4a0c96f0b66fc58 |
| e6b907ac | 28-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
changed Mat_SeqAIJ, BAIJ and SBAIJ structs to shared a common block of records. Updated code to use this.
Hg-commit: 65f31362560702f05a5b46398c031eb13cbcbf08 |
| f53c201f | 26-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
auto commit
Hg-commit: 837674bc383b1823de55b4f030b3d2580d5f5d65 |
| f32d5d43 | 25-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
minor improvements to web pages partially optimised MatMatMult for AIJ * Dense -> Dense
Hg-commit: 49fdde9b92ba89a6e7ae6774d68fd55a15003b38 |
| b8f8c88e | 24-Jun-2006 |
Hong Zhang <hzhang@mcs.anl.gov> |
Optimize MatFDColoringApply() to reduce VecScatters
MatFDColoringCreate_MPIAI(): add impl for ctype == IS_COLORING_LOCAL skip calling MatFDColoringApply_SeqAIJ(). Use MatFDColoringApply(). add impl
Optimize MatFDColoringApply() to reduce VecScatters
MatFDColoringCreate_MPIAI(): add impl for ctype == IS_COLORING_LOCAL skip calling MatFDColoringApply_SeqAIJ(). Use MatFDColoringApply(). add impl when the input x1=local ghosted vector significant num of VecScatter() is reduced add 'ISColoringType ctype' into MatFDColoring struct skip DAGlobalToLocalBegin/End when the input X=local ghosted vector ISColoringGetIS(): add impl for ctype == IS_COLORING_LOCAL
Hg-commit: 754948e31e75209c086573e54b573c1a36fd912f
show more ...
|
| 8e10ea92 | 24-Jun-2006 |
Satish Balay <balay@mcs.anl.gov> |
fix memory leak in VecCreateGhostWithArray()
Hg-commit: 7b9e5021bb5b30f08a45cc57ff30be19195c8c55 |
| ce72d2ba | 24-Jun-2006 |
Satish Balay <balay@mcs.anl.gov> |
use c comments [not c++]
Hg-commit: 46d23ed5a9eb233a57ec7d0fb67f5539192edc84 |
| 724c17b2 | 21-Jun-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
Fixed compiler warnings
Hg-commit: e3253e5dcad46ea296b1efac4be3651957f5db28 |
| 910ba992 | 21-Jun-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
Fixed cmpiler warnings
Hg-commit: c65647627c18b808ed35b9d5f9013608b89ec7d9 |
| ad08c97f | 21-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
added back PetscMap routines from previous PETSc releases added new pub
Hg-commit: 0674202596768765aa92a70cdab405334e52d307 |
| 6ce558ae | 21-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
vecgreateghost now auto sets VecSetLocalToGlobalMapping() hg sucks, I don't remember what I changed
Hg-commit: 80900d4fa4bf867df1dc3867199aa585a0afb76c |
| 19e29e70 | 16-Jun-2006 |
Satish Balay <balay@mcs.anl.gov> |
compile fixes in nightly builds
Hg-commit: 52cdb2ddaa600c35e9e6c0dc2a6a5d951399e476 |
| cf502942 | 13-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
parallel fix for fieldsplit; did not properly generate parallel matrices
Hg-commit: 7b34c32e57e2de741cd34e48d3fabeb2175adc66 |
| 2cfcea29 | 11-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
petsc-maint added ISLocalToGlobalMappingGetInfo() for Fortran
Hg-commit: b3e5369834bf6ab54a546e192fb7fb9d479e5fc2 |
| e2ee2a47 | 08-Jun-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
petscbinaryread now supports complex numbers
Hg-commit: ce300ba47bdb0aee2b113b259846f90b9279e966 |
| ee66e2f8 | 01-Jun-2006 |
Satish Balay <balay@mcs.anl.gov> |
- remove vecscattercreatetozero_() etc as they are now autogenerated - update ignore list
Hg-commit: 1a80917643041791da2f04b77ad1e27b748e00bd |
| 493850f3 | 31-May-2006 |
Satish Balay <balay@mcs.anl.gov> |
Fortran stubs for VecScatterCreateToAll/Zero routines
Hg-commit: 135eb2257c9ab2ebea71db2fc556b6fabe1b0ec2 |
| 51678082 | 25-May-2006 |
Barry Smith <bsmith@mcs.anl.gov> |
bk-changeset-1.4328.1.1 bsmith@barrysmith-57.mcs.anl.gov|ChangeSet|20060525205417|51990 ChangeSet 1.4328.1.1 06/05/25 15:54:17 bsmith@barrysmith-57.mcs.anl.gov +4 -0 aij.c: no shared data non
bk-changeset-1.4328.1.1 bsmith@barrysmith-57.mcs.anl.gov|ChangeSet|20060525205417|51990 ChangeSet 1.4328.1.1 06/05/25 15:54:17 bsmith@barrysmith-57.mcs.anl.gov +4 -0 aij.c: no shared data nonsense *** empty log message *** pdvec.c, bvec2.c: name declared twice superlu.c: couple more superlu options into manual page
src/mat/impls/aij/seq/aij.c 1.257 06/05/25 14:38:08 bsmith@barrysmith-57.mcs.anl.gov +0 -28 no shared data nonsense
src/mat/impls/aij/seq/aij.c 1.256 06/05/24 06:47:34 bsmith@barrys-computer.local +27 -0 *** empty log message ***
src/vec/vec/impls/mpi/pdvec.c 1.76 06/05/23 16:55:11 bsmith@barrysmith-11.mcs.anl.gov +0 -1 name declared twice
src/vec/vec/impls/seq/bvec2.c 1.85 06/05/23 16:54:29 bsmith@barrysmith-11.mcs.anl.gov +0 -1 name declared twice
src/mat/impls/aij/seq/superlu/superlu.c 1.89 06/05/23 14:47:38 bsmith@barrysmith-57.mcs.anl.gov +4 -1 couple more superlu options into manual page
Hg-commit: bcdf806e8caaef5ccfdbb51f0f3cfc7ea7e29cb1
show more ...
|
| 0db6efe0 | 17-May-2006 |
Matthew Knepley <knepley@mcs.anl.gov> |
bk-changeset-1.4321 knepley@khan.(none)|ChangeSet|20060517210857|51261 ChangeSet 1.4321 06/05/17 16:08:57 knepley@khan.(none) +7 -0 Working mesh field output
src/vec/vec/impls/seq/bvec2.c
bk-changeset-1.4321 knepley@khan.(none)|ChangeSet|20060517210857|51261 ChangeSet 1.4321 06/05/17 16:08:57 knepley@khan.(none) +7 -0 Working mesh field output
src/vec/vec/impls/seq/bvec2.c 1.84 06/05/17 16:08:55 knepley@khan.(none) +56 -3 Fix VTK data output
src/vec/vec/impls/mpi/pdvec.c 1.75 06/05/17 16:08:55 knepley@khan.(none) +56 -3 Fix VTK data output
src/dm/mesh/sieve/Partitioner.hh 1.9 06/05/17 16:08:55 knepley@khan.(none) +112 -105 Need these typedefs to be in class - Means making the class templated Return overlap from distribute() Now have partition_Induced() which follows mesh partition Now all fields in the mesh are also distributed Lots of cleanup Need to fix unify Need to fix coordinates
src/dm/mesh/sieve/Mesh.hh 1.66 06/05/17 16:08:54 knepley@khan.(none) +32 -221 Added getFields() - This is just crap Fixed createSplitField() - Could probably be optimized Jettisoned more dead code
src/dm/mesh/sieve/CoSifter.hh 1.63 06/05/17 16:08:54 knepley@khan.(none) +8 -1 Fixed size check Added allocatePatches()
src/dm/mesh/examples/tutorials/makefile 1.44 06/05/17 16:08:54 knepley@khan.(none) +3 -0 Added target
src/dm/mesh/examples/tutorials/ex1.c 1.41 06/05/17 16:08:54 knepley@khan.(none) +1 -0 Set the object name
Hg-commit: a4e47c00885a968022074852379a6eca36d3ff91
show more ...
|