| 823a26a4 | 29-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: OpenCL is hardcoded for P0 coefficients right now |
| 754551f4 | 29-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Added auxiliary fields to PetscFEIntegrateJacobian() - Hooked it up to DMPlexComputeJacobian() |
| f234cb8b | 29-Aug-2013 |
Satish Balay <balay@mcs.anl.gov> |
dmplex: windows compile fix |
| c012ea0a | 28-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex+PetscFE: Fix compile errors for complex - LAPACK calls still broken |
| f1ea0e2f | 28-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex+PetscFE: Looks like boundary integration is working - DMComputeResidualFEM() checks for feBd, and does bd integral if present - Normals are still not being calculated - Added PetscFEIntegrateB
DMPlex+PetscFE: Looks like boundary integration is working - DMComputeResidualFEM() checks for feBd, and does bd integral if present - Normals are still not being calculated - Added PetscFEIntegrateBdResidual() and a basic implementation - Added normals to PetscCellGeometry - Auxiliary fields not turned on in bd integrals
show more ...
|
| 4819305b | 27-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscDualSpace: Added 1D reference element |
| 9b98b1bc | 27-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PescFE: Corrected spelling, Fixed pointer check |
| b1e2fd0a | 24-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE+OpenCL: It appears that P0 variable coefficient is working - Also cleaned up kernel generation |
| 9a559087 | 24-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Changed PetscFEIntegrateResidual() to take auxiliary fields - Also changed the PetscFEM struct to take auxiliary elements - Added P_0 Lagrange element |
| e42b8645 | 23-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE+OpenCL: Fixed number of blocks - Now test should pass |
| a24be9e8 | 23-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE+OpenCL: Fixed support for f0 - Can still eliminate some code when f0 or f1 is unused, but that does not work right anyway right now |
| 06f8b4d4 | 21-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Added marshalling to the OpenCL implementation |
| 855cd083 | 21-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Added a real number type to the OpenCL implementation |
| 1333f600 | 20-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: OpenCL default back to CPU for non-conforming block sizes |
| f30c5766 | 19-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Now cell blocks are settable from options - Fixed bad merge in 8residual integration |
| 133ee763 | 19-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Added OpenCL implementation |
| 0e70cedf | 19-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Add interface for tile sizes - Accelerator implementation has multilevel blocking |
| d5dbada5 | 19-Aug-2013 |
Jed Brown <jedbrown@mcs.anl.gov> |
DMDT FEM: initialize variables due to -Wmaybe-uninitialized |
| 0c35b76e | 18-Aug-2013 |
Jed Brown <jedbrown@mcs.anl.gov> |
DMDT: move FEM space support to own file dtfe.c |
| fdf9dbd0 | 18-Aug-2013 |
Jed Brown <jedbrown@mcs.anl.gov> |
DMDT: use PetscClassId for class ids [needed for 64-bit-indices] |
| 0483ade4 | 15-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DMPlex+PetscFE: Moved integration functions from DMPlex to PetscFE - Now PetscFE type indicates the integration method - Added all machinery for PetscFE types |
| 4f2201c4 | 12-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Fix bug in PetscDualSpaceSetUp_Lagrange() |
| a319912f | 11-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: FEM Jacobian computation working, and now cache default tabulation - Added PetscFEIntegratJacobianChunk() - Changed interface in DMPlexComputeJacobianFEM() - Changed signature of integrateJa
PetscFE: FEM Jacobian computation working, and now cache default tabulation - Added PetscFEIntegratJacobianChunk() - Changed interface in DMPlexComputeJacobianFEM() - Changed signature of integrateJacobianFEM in DMPlex - Added DMPLEX_JacobianFEM event - Now PetscFE caches default tabulation
show more ...
|
| a0845e3a | 11-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
PetscFE: Large reorganization of FEM dispatch - WARNING: Now DMPlex and PetscFE have a circular dependence. We need to move the FEM methods to PetscFE - Changed FEMIntegrateResidualBatch() from to Pe
PetscFE: Large reorganization of FEM dispatch - WARNING: Now DMPlex and PetscFE have a circular dependence. We need to move the FEM methods to PetscFE - Changed FEMIntegrateResidualBatch() from to PetscFEIntegrateResidualChunk() - Changed DMPlexComputeResidualFEM() to use new interface - Changed signature of f_k and g_kl functions - Added DMPlexResidualFEM event - Put PetscQuadrature and PetscCellGeometry - Moved and changed PetscFEM structure
show more ...
|
| 552aa4f7 | 11-Aug-2013 |
Matthew G. Knepley <knepley@gmail.com> |
DT: Changed PetscDTGaussJacobiQuadrature() to take a PetscQuadrature object |