| 4a217a95 | 29-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Fixed compiler problems with complex |
| 5743f1d7 | 28-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Fix DMPlexComputeCellGeometry() to handle lower dimensional cells - Now I call DMPlexGetLabelValue() for depth each time, which might be too expensive. We could stash the depth label pointer |
| 73868372 | 28-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Fixed error in projection from 3D to 2D - Need to check for case where normal is already z |
| 02a80efe | 28-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Adding boundary integration - Added a boundary integration input to DMPlexSetFEMIntegration(), this also takes cell normals - Added an integrateBdResidualFEM() function to DMPlex - f0 and f1
DMPlex: Adding boundary integration - Added a boundary integration input to DMPlexSetFEMIntegration(), this also takes cell normals - Added an integrateBdResidualFEM() function to DMPlex - f0 and f1 for boundary integration take the cell normal as an argument (right now not used) -
show more ...
|
| 1885f254 | 28-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Fixed cell inversion for TetGen to commute with mesh interpolation |
| 607a6623 | 26-Mar-2013 |
Barry Smith <bsmith@mcs.anl.gov> |
removed path argument to XXXInitializePackage() and XXXRegister() |
| 471ecdac | 25-Mar-2013 |
Jed Brown <jed@59A2.org> |
segbuffer: add PetscSegBufferGetInts() to encourage use of PETSC_RESTRICT |
| 137cf7b6 | 25-Mar-2013 |
Jed Brown <jed@59A2.org> |
segbuffer: create wrapper structure
The old structure exposed the linked list property of the implementation. Now we have a simple wrapper structure so that implementation detail is not leaked. This
segbuffer: create wrapper structure
The old structure exposed the linked list property of the implementation. Now we have a simple wrapper structure so that implementation detail is not leaked. This will also make it possible to implement some thread-safe operations.
show more ...
|
| 63089020 | 22-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'knepley/plex-surface-integration' into knepley/plex
# Please enter a commit message to explain why this merge is necessary, # especially if it merges an updated upstream into a topic b
Merge branch 'knepley/plex-surface-integration' into knepley/plex
# Please enter a commit message to explain why this merge is necessary, # especially if it merges an updated upstream into a topic branch. # # Lines starting with '#' will be ignored, and an empty message aborts # the commit.
show more ...
|
| 652b88e8 | 22-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Added integration over boundary to DMPlexComputeResidualFEM() - Added quadBd, f0/1BdFuncs to PetscFEM struct - Needs to be tested |
| 184d77ed | 20-Mar-2013 |
Jed Brown <jed@59A2.org> |
DMGetLocalToGlobalMapping_*: normalize names from DMCreateLocalToGlobalMapping_* |
| 77b76215 | 19-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'jed/plex-scalable-partition-closure' of bitbucket.org:petsc/petsc into knepley/plex |
| 656e37a6 | 19-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex ex7: Orientations now working properly - Added check for inverted cells - Added check that vertex cones match for interpolated and non-interpolated meshes - Fixed hex and tet tests - Support r
DMPlex ex7: Orientations now working properly - Added check for inverted cells - Added check that vertex cones match for interpolated and non-interpolated meshes - Fixed hex and tet tests - Support reading ExodusII files and use mesh generators
show more ...
|
| 2e1b13c2 | 19-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Changed orientation convention and fixed many bugs - Tetrahedra now all have outward facing normals - We have to flip input from TetGen and ExodusII - Hexahedra now all have outward facing
DMPlex: Changed orientation convention and fixed many bugs - Tetrahedra now all have outward facing normals - We have to flip input from TetGen and ExodusII - Hexahedra now all have outward facing normals - We have to flip input from ExodusII - Fixed bug with orientation -1 (now encoded as -faceSize) - Fixed parallel bug with DMPlexInterpolate() - Fixed determinant calculation for Tet and Hex - Fixed bug with reversing orientation, now must also start on the opposite vertex - This solution in GetTransitiveClosure() is suboptimal, should change the definition of negative orientation
show more ...
|
| cb3cb6d9 | 19-Mar-2013 |
Jed Brown <jed@59A2.org> |
DMPlexCreatePartitionClosure: replace quadratic algorithm
This implementation uses a PetscBT to avoid duplicates and incrementally packs a PetscSegArray with the points needed by each rank.
This fu
DMPlexCreatePartitionClosure: replace quadratic algorithm
This implementation uses a PetscBT to avoid duplicates and incrementally packs a PetscSegArray with the points needed by each rank.
This function used to dominate run-time when loading meshes in parallel and is now a negligible fraction of run time.
show more ...
|
| 923cbbb6 | 19-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Fix default |J| for DMPlexComputeCellGeometry() |
| 139a35cc | 18-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Fix for cell geometry (unbelievable screwup) |
| 705246ef | 18-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
Merge branch 'master' of bitbucket.org:petsc/petsc into knepley/plex |
| bf16daa6 | 18-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex ex7: Check that |J| is nonzero for each cell in interpolated mesh |
| 49dc4407 | 18-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Allow DMPlexComputeCellGeometry() to work with interpolated meshes |
| 80330477 | 18-Mar-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex: Added prototype for DMPlexCopyCoordinates(), and docs for that and DMPlexInterpolate() |
| 22566864 | 18-Mar-2013 |
Jed Brown <jed@59A2.org> |
DMPlexDistribute: fix label distribution
This old code was not advancing to the next label when it finds "depth" (which is not distributed), thus it just compared "depth" multiple times and never ch
DMPlexDistribute: fix label distribution
This old code was not advancing to the next label when it finds "depth" (which is not distributed), thus it just compared "depth" multiple times and never checked the other label names.
show more ...
|
| 37045cdd | 18-Mar-2013 |
Jed Brown <jed@59A2.org> |
DMPlexVTKWriteAll_VTU: fix writing vertex data |
| 0059ad2a | 17-Mar-2013 |
Satish Balay <balay@mcs.anl.gov> |
dmplex: doc fixes [nightlybuilds] |
| 878cb397 | 17-Mar-2013 |
Satish Balay <balay@mcs.anl.gov> |
doc: nightlybuild fixes |