Rename some private function and add documentation to public API functions.

Fixing errors due to bad merge after cherry-picking earlier in the branch.

Removing all debug printouts and changing the DOF traversal based on blocked variables to be consistent.

Remove old definitions related to DA.

Fixing examples to the new DMCreateMatrix API change.

Fixing the CreateMatrix to not get a MatType from user (PETSc API change). Set this directly on the DM now.

Making sure that the interface is const correct when dealing with MOAB objects on the getter methods.

First level modifications in porting barry/ex10.c to FEM. Added case to read tungsten_tiny.txt

Adding an extra argument to box mesh creation in order to specify the domain bounds. Modifying examples accordingly.

Adding a private routine to make sure that the correct NNZ pattern is set in the matrix so that it is usable right away. The user can set the correct values based on physics without changing the NNZ

Adding a private routine to make sure that the correct NNZ pattern is set in the matrix so that it is usable right away. The user can set the correct values based on physics without changing the NNZ pattern determined by DM. And fix memory leaks.

show more ...

Updating mbutil to the new moab interface change.

Fixed a bug in file read and updating the PSI example and syncing with Barrys changes in commit: petsc:psi-brian-wirth/935bd20.

Renaming ex2.cxx to ex2.c to be consistent with other examples. Also removing many debug outputs and fixing the ftype=1,2 residual functions for parallel runs.

Adding a public API method DMMoabGetOffset to get the local offset when dealing with global spaces for Vec/Mat.

Some fixes for different ways of computing the residual. The GlobalBlocked route doesnt work in parallel yet.

Updating the PSI example for Jay.

1) Modified and matched closely with Barrys latest version of the advection-reaction solver ex10.c in his wirth-fusion-materials branch.
2) The Jacobian from ex10 s

Updating the PSI example for Jay.

1) Modified and matched closely with Barrys latest version of the advection-reaction solver ex10.c in his wirth-fusion-materials branch.
2) The Jacobian from ex10 still doesnt work and so use only -snes_mf to run.
3) "runs" in parallel but does not give same answer as serial.. Needs testing.
4) Only FD now. Need to formulate FE version eventually.

show more ...

First draft of DMMoab based version of the advection-diffusion-reaction example ex10.c. There are still some differences in the temporal convergence of both the examples which needs to be verified ne

First draft of DMMoab based version of the advection-diffusion-reaction example ex10.c. There are still some differences in the temporal convergence of both the examples which needs to be verified next.

show more ...

Adding a new API function to set a field name. This is more of a utility function that follows the DMDA design.

Rename nfields->numFields and fields->fieldNames to be consistent with PetscSection. Need to make a move eventually to this structure to handle the internal field and point DOF.

Adding a variable to store ghosted elements. Change the public API to retrieve local and ghosted vertices/element sizes.

Fixes to make sure the MOAB based tag_iterate functions will work correctly even when local entities are non-contiguous. No more restrictions for either the MOAB or native vector in DMMoab now.

Renaming one of the cached variables for local ID storage and making it field dependent instead of just point dependent.

A very important fix for the ghosted vector updates.
The parallel bugs in incorrect residual evaluations in ex2.cxx were two-fold.

1) The ghost updates were incorrect.
2) The VecScatter context was

A very important fix for the ghosted vector updates.
The parallel bugs in incorrect residual evaluations in ex2.cxx were two-fold.

1) The ghost updates were incorrect.
2) The VecScatter context was wrongly formed. Perhaps a VecCopy might suffice for local Global-Local transfers. Optimize later.
3) The BC evaluation had a bug due to the way we were accessing the bit vector with EntityHandle static casts. This could be
a problem in evaluating field DOF data also. Currently not an issue but need to check if this can be optimized.
4) ex2.cxx runs correctly in serial and parallel. Preliminary profiling against ts/examples/tutorials/ex25.cxx shows that the FEM version against FD version is about 2-3 times slower in TSFunctionEval. This needs to be profiled further and corrected.
5) Now support both native PETSc and native MOAB-tag based Vec depending on whether mesh entities or contiguous.

Parallel bugs fixed. Optimizations upcoming to improve speedup.

show more ...

Necessary changes to examples and tests.

d6f5831: Modifying ex4.c to the new interface changes.
c084c7f: Fix example 1 to use the new interface.
da41520: Several changes to the 1-d ex2.cxx driver. T

Necessary changes to examples and tests.

d6f5831: Modifying ex4.c to the new interface changes.
c084c7f: Fix example 1 to use the new interface.
da41520: Several changes to the 1-d ex2.cxx driver. Trying out different methods of assembling the FEM operators since the default idea seems to be about 2 times slower than DMDA assembly.

show more ...

Adding the first draft capability to allow both MOAB and native Vec with DMMoab. Also setting the necessary Mapping data so that SetLocal and SetBlockedLocal can be used.