rm getredundantmatrix from matops; rename MatGetRedundantMatrix to MatCreateRedundantMatrix

merge MatCreateMPIAIJConcatenateSeqAIJSymbolic and xxxNumeric into MatCreateMPIMatConcatenateSeqMat_MPIAIJ()

MatGetRedundantMatrix() works for np=1

MatGetRedundantMatrix() replaces all MatGetRedundantMatrix_xxx()

fix block size; intend to unify MatGetRedundantMatrix_XXX to MatGetRedundantMatrix()

updated PCFactorSetUseInPlace() to toggle and added getter

Merge branch 'barry/fix-matgetdiagonalblock'

reduce dependancy on psubcomm for MatGetRedundantMatrix_MPIAIJ()

MatGetDiagonalBlock() can be called before matrix is assembled

rename MatCreateMPIAIJConcatenateSeqAIJ() to MatCreateMPIMatConcatenateSeqMat()

fix dropping of vecs in Fortran interface to MatCreateVecs
fix complex example that would not compile with c++

fix matgetvecs_ to matcreatevecs_

Merge branch 'knepley/fix-mat-view-bs'

* knepley/fix-mat-view-bs:
Mat: Correct MatView() for different row and column block sizes

added test to manual pages clarifying what row and column Mat blocksize actually mean
someone added them without any regard for how confusing it would be to users

MatGetVecs() and KSPGetVecs() replaced with MatCreateVecs() and KSPCreateVecs()
since they are creators not getters

Mat: Correct MatView() for different row and column block sizes

fix to MatView man page

add MatConvert_SeqAIJ_Elemental()

matlab: update matlab related files with the new share/petsc/matlab location

Merge branch 'jed/dmda-memusage'

removed some leftover localtoglobalmappingblock code

changed isams to issaws for code clarity

Merge branch 'jed/dmda-memusage'

added MatFindOffBlockDiagonalEntries() for MPIAIJ so that -pc_fieldsplit_detect_coupling can be used in parallel
optimized fieldsplit for 2 splits with multiplicative, less expensive residual computa

added MatFindOffBlockDiagonalEntries() for MPIAIJ so that -pc_fieldsplit_detect_coupling can be used in parallel
optimized fieldsplit for 2 splits with multiplicative, less expensive residual computation and remove unneeded MatGetSubMatrix() for that case
optimized MatGetSubMatrix() for request of entire matrix

show more ...

Merge branch 'knepley/feature-fe-dg-space'

* knepley/feature-fe-dg-space: (38 commits)
DMPlex: Added coordinate field info for generated mesh
DMPlex: Update to new DMPlexGetBoundary() interface

Merge branch 'knepley/feature-fe-dg-space'

* knepley/feature-fe-dg-space: (38 commits)
DMPlex: Added coordinate field info for generated mesh
DMPlex: Update to new DMPlexGetBoundary() interface
PetscFE: Fix complex casting
DMPlex: Fixup type mismatch in PetscFEIntegrate()
DMPlex ex62: Fixed test output
SNES ex62: Added test output
SNES ex62: Added tests for Q2-Q1 and Q2-P1disc - Added -show_error - Removed spurious refinement of quads - Fixed cast - CreatePressureNullspace() now uses DMPlexProjectFunction() - Replace SetupElement() with PetscFECreateDefault()
DMPlex ex3: Added tests for P1disc
SNES ex12: Updated to new PetscFECreateDefault()
PetscFE: Fixes for discontinuous elements
PetscSpace: Fixed polynomial space view
PetscFE: Changed PetscFECreateDefault() - Now continuity and reference cell shape are set from the command line
PetscFE: Added domain integration - Added PetscFEIntegrate(), only implemented for Basic - Added DMPlexComputeIntegralFEM() and DMPlexComputeL2FieldDiff() - This should replace the L2Diff functions
Mat+Doc: Fixed MatNullSpaceRemove()
SNES ex62: Updated to new PetscFECreateDefault()
SNES ex12: Updated to new PetscFECreateDefault() - 3D simplex moves the P_0 point from the centroid
DMPlex ex3: Updated for corrected cell dof
PetscFE: I think I finally fixed the cell dof
DMPlex ex3: Added test case for P1disc on a quad
DMPlex ex3: Added P1 discontinuous test - Something looks wrong with interpolation
...

show more ...