| f24ae36c | 25-Jan-2011 |
Victor Minden <victorminden@gmail.com> |
added MatGetVecs_SeqAIJCUSP
Hg-commit: ff404b085e733a36537394ff258b78c6be0dfbfd |
| df3a31bc | 21-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
finished converting cuda to cusp in source code
Hg-commit: 7d2435fd6968de7d5739915853d92ec73d2998c0 |
| b7ef5f13 | 21-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed more file names from cuda to cusp
Hg-commit: c5000cb46896546bd89b0a3eaeeef63664d38ffc |
| e3d235bd | 21-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed names of directories that had cuda to cusp
changed names of some files that had cuda to cusp
Hg-commit: 7e265e74bd58a9e4b021e6c83e3146332a07fbdb |
| 8154be41 | 21-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed names of all GPU related functions, macros and variables that involve CUSP to use the string CUSP instead of the misleading CUDA
Hg-commit: 3264fa4049fd83eceb12c85feb980ef207360220 |
| 6ce1633c | 16-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
added -pc_fieldsplit_stokes that automatically sets up Stokes solver with fieldsplit
Hg-commit: 93cf08337e0b89b643012c89924a66d5eb23e435 |
| b3a44c85 | 15-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
update MatSeqAIJFindZeroRows() to MatFindNonzeroRows()
Hg-commit: fbda5d24d43b60060a19e9257c1b0ad376959dc0 |
| f8260c8f | 14-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed PCFactorSetMatSolverPackage() to NOT require the Mat before being called, now it just keeps a copy of the package name and uses it later
added PCFactorSetUpMatSolverPackage() actually calls M
changed PCFactorSetMatSolverPackage() to NOT require the Mat before being called, now it just keeps a copy of the package name and uses it later
added PCFactorSetUpMatSolverPackage() actually calls MatGetFactor() so there now exists a factor matrix and PCFactorGetMatrix() can be called and then options set from the program.
Hg-commit: 7f7727b0030448abcd98a3d596d64ed83143f965
show more ...
|
| 62298a1e | 08-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
vi rs solver checks for zero rows in matrix and removes all those rows and colums from the inactive set
zero rows come from equality constraint in Allen-Cahn equation
Hg-commit: 7da664b0275039a9742f
vi rs solver checks for zero rows in matrix and removes all those rows and colums from the inactive set
zero rows come from equality constraint in Allen-Cahn equation
Hg-commit: 7da664b0275039a9742f5f7bd675c31970f57482
show more ...
|
| b409243c | 06-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
changed PETSC_HAVE_CRAYC to PETSC_HAVE_CRAY_VECTOR to add vector pragmas only if Cray compiler supports them by calling isCrayVector
Hg-commit: 3dd5aeaa5f195ee35733d6d4b90c0a26469d4cbb |
| be6adb11 | 01-Jan-2011 |
Barry Smith <bsmith@mcs.anl.gov> |
fixed some unitialized variables warnings with MPI-uni
Hg-commit: 925a462303ee56a402b0355f3680130c52d4c99c |
| e3c5b3ba | 26-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
changed Matlab spelling to MATLAB some general MATLAB cleanup
Hg-commit: efcb24f191dfedc98b283d9e63e7859032e8ef35 |
| 2eff7a51 | 24-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
automatically free all created objects that have not been freed during run in PetscFinalize
Hg-commit: 2aad0a0317aabf395b96783e2383b8b5e0222e16 |
| b3da158b | 22-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
removed VecValid()
removed MatCreateSeqAIJFromMatlab() since can use MatSeqAIJFromMatlab()
Hg-commit: c03e8486f64adf871eb3631cafe7ee7a1dd3fe47 |
| 5ccb76cb | 20-Dec-2010 |
Hong Zhang <hzhang@mcs.anl.gov> |
Continue finish MatSuperluSetILUDropTol rename (fortran interface)
Use dynamic call for MatMumpsSetIcntl()
Hg-commit: c9d42c995e05ccfbea76c64a7669d0498c8bfdf3 |
| 77919271 | 19-Dec-2010 |
Victor Minden <victorminden@gmail.com> |
added logging for MatCUDACopyToGPU
Hg-commit: f827c00680d866871553bcb5d07af091b51da3bb |
| 790a96df | 19-Dec-2010 |
Jed Brown <jed@59A2.org> |
Finish MatSuperluSetILUDropTol rename (fortran interface), use a dynamic call
Hg-commit: 69cb3610609689661db9c9ca5bd995bb97289536 |
| f250808b | 18-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
MatSetMUMPSxxxx MatSetSuperluxxxx should be MatMUMPSSetxxxx and MatSuperluSetxxxx because of paradigm that all methods are classsubclasssubsubclassactions
Hg-commit: b151f2166bfadc04209d28ba1c4c0d87
MatSetMUMPSxxxx MatSetSuperluxxxx should be MatMUMPSSetxxxx and MatSuperluSetxxxx because of paradigm that all methods are classsubclasssubsubclassactions
Hg-commit: b151f2166bfadc04209d28ba1c4c0d8738fbd513
show more ...
|
| 538a8cf4 | 18-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: e63d4912341873b79ec38ad7dfe8372325a96ba9 |
| aeeaa5c7 | 18-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
Matlab interface supports A(:,:) for accessing Matlab matrix from PetscMat
Hg-commit: 75faa6f8e51d56723bac66d2be9a17eb89c30e62 |
| 8ece9e9a | 17-Dec-2010 |
Satish Balay <balay@mcs.anl.gov> |
fix superlu fortran interface include
Hg-commit: ccc4de9d529de24b99eec8193380485f2d9c0076 |
| d954db57 | 17-Dec-2010 |
Hong Zhang <hzhang@mcs.anl.gov> |
add procedual routine MatSetSuperluILUDropTol() as requested by
[petsc-users] no PCFactorSetUseDropTolerance() in 3.1-p6.
Hg-commit: e9e109e2a8992b12b2c6d698f53976d53f60dc8b |
| 33f85c2f | 16-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
fixed converting of SeqAIJ to Matlab matrix to work for 64 bit ints in Matlab
moved PetscLogDestroy() to after objects are destroyed because cookies are needed until all objects are destroyed
Hg-com
fixed converting of SeqAIJ to Matlab matrix to work for 64 bit ints in Matlab
moved PetscLogDestroy() to after objects are destroyed because cookies are needed until all objects are destroyed
Hg-commit: 7ce1802937be80f7a305aff5cc37ede3d49baad6
show more ...
|
| cd6b891e | 14-Dec-2010 |
Barry Smith <bsmith@mcs.anl.gov> |
[petsc-maint #56731] Parallel assembly bug with compressedrow when pattern changes
recompute compressed row structure for each new matrix assembly, when requested so that compressed information is al
[petsc-maint #56731] Parallel assembly bug with compressedrow when pattern changes
recompute compressed row structure for each new matrix assembly, when requested so that compressed information is always up to date
Hg-commit: 8b3422309f64468fec0ab5079c31f1ed7cabd7f9
show more ...
|
| a0583221 | 10-Dec-2010 |
Victor Minden <victorminden@gmail.com> |
removed the commented out code that was the old cuda methodology
Hg-commit: c550dbcaf978f3d90202008b57d0d52cd16c4c26 |