| ae1aaa3f | 17-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'knepley/feature-plex-replicate'
* knepley/feature-plex-replicate:
Plex: Now DMPlexGetRedundantDM() has more sensible behavior on 1 proc |
| 02cee238 | 17-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'knepley/feature-snes-stokes-test'
* knepley/feature-snes-stokes-test:
Plex: Drawing Vecs now works for multiple fields and components
SNES ex69: Started mantle test
SNES ex69: Fi
Merge branch 'knepley/feature-snes-stokes-test'
* knepley/feature-snes-stokes-test:
Plex: Drawing Vecs now works for multiple fields and components
SNES ex69: Started mantle test
SNES ex69: Fixed test
SNES ex69: Started a composite viscostiy test
SNES ex69: Updated BC specification
SNES ex69: Must use -pc_use_amat with FieldSplit
Plex+Draw: Improve viewer - Nice title - Allow override of min/max scale - Allow P0 and P1 (just quick and dirty)
Plex+VTK: Set boundary values with the time specified by the OutputSequence number
Plex: Now only local vectors are used with the Draw viewer - Set boundary values with the time specified by the OutputSequence number
Plex: Fixed raw array access in DMPlexVecGetClosure()
Doc+SNES: I think the option should also be with the monitor function
TS ex47: Advection with Plex - I want to see if I can use MG
show more ...
|
| 1b82c3eb | 17-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Plex+Gmsh: Allow files with no newline at the end |
| b358f029 | 17-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'master' into knepley/fix-plex-examples-cleanup
* master: (156 commits)
Doc: Added headers for documentation sections
Bib: Small corrections
doc: update example makefiles with del
Merge branch 'master' into knepley/fix-plex-examples-cleanup
* master: (156 commits)
Doc: Added headers for documentation sections
Bib: Small corrections
doc: update example makefiles with deleted sourcefile info
doc: fix manual build errors
doc: fix SOURCEH for some fortran includes that got moved
doc: fix broken manualpages - bug introduced by 1b266c9
MatMissingDiagonal() is actually not a collective operation
PetscFECreateDefault(): allow 0th order quadrature
fix gmakegentest.py to ignore *.xx files that are not source code but are in the files
increase table input size to (A->cmap->n + a->B->cmap->N) for MatTransposeMatMultSymbolic_MPIAIJ_MPIAIJ()
gtest: add TEST and RM clean output for Satish
gtest: set PATH for windows DLL tests, and enable -Dpetsc_EXPORTS only during (dll) build - and not for examples
gtest: use target-specific var to deduplicate FC rules for modules
gtest: for 'allgtest' and 'allgtest-tap' - use V=0 instead of 'grep -v CC' etc..
gtest: quiet absoft compiler commands [so they can be filtered].
remove incorrect requires from more test files, remove output from -fp_trap example in test because it cannot be machine independent
gtest: windows fixes. add filter to remove filename from compiler stdout. And fix filters to use petscdiff
Removed generated requires in the wrong place from test examples.
Fix test output for additional information printed about computed eigenvalues
gtest: fix ex5f.F to use 'format' so that the output from ifort matches the one from gfortran
...
show more ...
|
| 63eb7d2f | 16-Jan-2017 |
Satish Balay <balay@mcs.anl.gov> |
doc: update example makefiles with deleted sourcefile info |
| e328ff09 | 16-Jan-2017 |
Toby Isaac <tisaac@uchicago.edu> |
DMPlexTranferVecTree_Inject(): fix indexing for multiple fields |
| 0b6a6f68 | 13-Jan-2017 |
Toby Isaac <tisaac@uchicago.edu> |
PetscFECreateDefault(): allow 0th order quadrature
I meant to specify that only negative quadrature orders would be ignored. |
| ccaf501d | 13-Jan-2017 |
SanderA <sander.arens@ugent.be> |
Plex+HDF5: also output topology for visualization for uninterpolated meshes
In /topology/cells the block size is always 1 and point numbers are output.
What we really need for visualization is a blo
Plex+HDF5: also output topology for visualization for uninterpolated meshes
In /topology/cells the block size is always 1 and point numbers are output.
What we really need for visualization is a block size which tells us the number of vertices per cell
and the vertex numbers (between 0 and the total number of vertices).
show more ...
|
| 68dbc166 | 13-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Plex: Now DMPlexGetRedundantDM() has more sensible behavior on 1 proc |
| 6497092e | 11-Jan-2017 |
Barry Smith <bsmith@mcs.anl.gov> |
fix unused variables etc in examples detected by clang static analyzer.
Reported-by: nightly builds
x |
| 586db2dd | 11-Jan-2017 |
Barry Smith <bsmith@mcs.anl.gov> |
remove __FUNCT__ definitions I missed before |
| d1df6f1d | 07-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Plex: Drawing Vecs now works for multiple fields and components |
| bde60d6a | 05-Jan-2017 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'knepley/feature-plex-ex-advection' into knepley/feature-snes-stokes-test
* knepley/feature-plex-ex-advection:
Plex+Draw: Improve viewer - Nice title - Allow override of min/max sca
Merge branch 'knepley/feature-plex-ex-advection' into knepley/feature-snes-stokes-test
* knepley/feature-plex-ex-advection:
Plex+Draw: Improve viewer - Nice title - Allow override of min/max scale - Allow P0 and P1 (just quick and dirty)
Plex+VTK: Set boundary values with the time specified by the OutputSequence number
Plex: Now only local vectors are used with the Draw viewer - Set boundary values with the time specified by the OutputSequence number
Plex: Fixed raw array access in DMPlexVecGetClosure()
Doc+SNES: I think the option should also be with the monitor function
TS ex47: Advection with Plex - I want to see if I can use MG
Conflicts:
src/dm/impls/plex/plex.c
show more ...
|
| 05decbb5 | 05-Jan-2017 |
Mark Adams <mfadams@lbl.gov> |
Plex ex2: Allow this to use an auto-generated mesh |
| 14640d07 | 05-Jan-2017 |
Mark Adams <mfadams@lbl.gov> |
Plex ex6: Removed unused variables |
| bb6937b7 | 05-Jan-2017 |
Mark Adams <mfadams@lbl.gov> |
Plex: Add some targets for tutorials |
| bdd7d0ca | 31-Dec-2016 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'mlange/feature-plex-med'
* mlange/feature-plex-med:
Plex+Fluent: Finally got warning working for me
Plex+Fluent: numFaces should not have been unsigned
fix sign-compare warning i
Merge branch 'mlange/feature-plex-med'
* mlange/feature-plex-med:
Plex+Fluent: Finally got warning working for me
Plex+Fluent: numFaces should not have been unsigned
fix sign-compare warning in arch-linux-pkgs-cxx-mlib_el6
plex+MED: fix -Wunused-variable
Plex+MED: Fix variable protection
Plex+MED: Fix unsigned ints in the Fluent format
Plex-MED: Adding test output
Plex-MED: Updating the MED version to download
Configure: Make MED configure with given HDF5 path
Plex-Med: Read, communicate and identify facets in parallel
Plex: Optionally return vertex SF from DMPlexCreateFromCellListParallel
Plex-Med: Switch to parallel facet reads from file
Plex: Cannot assume hash keys are sorted
Plex-Med: Adding partitioned reads of coordinates and cell-connectivity
Plex-Med: Apply facet IDs via vertex joins
Plex-Med: Build DM from cells and coordinated in the .med file
Plex-Med: Add skeleton reader for .med mesh files with simple tests
Plex-Med: Add .med mesh library to build system
Conflicts:
src/dm/impls/plex/makefile
show more ...
|
| 6ecf4022 | 30-Dec-2016 |
Barry Smith <bsmith@mcs.anl.gov> |
Merge branch 'balay/add-prefix-test' |
| 8ad9143b | 30-Dec-2016 |
Barry Smith <bsmith@mcs.anl.gov> |
f90impl.h needs to be available for users with --prefix installation
Reported-by: Satish Balay <balay@mcs.anl.gov> |
| b7be58f8 | 26-Dec-2016 |
Barry Smith <bsmith@mcs.anl.gov> |
Removed use of PETSC_NULL in a couple places that got overlooked in last commit
Reported-by: nightly tests |
| 390e1bf2 | 25-Dec-2016 |
Barry Smith <bsmith@mcs.anl.gov> |
Minor cleanup, remove some outdated code, remove unneeded constructs, improve manual pages slightly, add missing test case
Time: 2.5 hours |
| 339e3443 | 21-Dec-2016 |
Matthew G. Knepley <knepley@gmail.com> |
Plex+Draw: Improve viewer
- Nice title
- Allow override of min/max scale
- Allow P0 and P1 (just quick and dirty) |
| cf9ba882 | 21-Dec-2016 |
Matthew G. Knepley <knepley@gmail.com> |
Plex+VTK: Set boundary values with the time specified by the OutputSequence number |
| f13a32a3 | 21-Dec-2016 |
Matthew G. Knepley <knepley@gmail.com> |
Plex: Now only local vectors are used with the Draw viewer
- Set boundary values with the time specified by the OutputSequence number |
| 8a84db2d | 21-Dec-2016 |
Matthew G. Knepley <knepley@gmail.com> |
Plex: Fixed raw array access in DMPlexVecGetClosure() |