DMPlex: use PetscSectionGet/RestorePointSyms() in DMPlexVecGet/SetClosure()

You can see that we no longer use the cone orientations directly: they
are used by the PetscSectionSym, but once the permu

DMPlex: use PetscSectionGet/RestorePointSyms() in DMPlexVecGet/SetClosure()

You can see that we no longer use the cone orientations directly: they
are used by the PetscSectionSym, but once the permutations/flips have
been returned, they are used.

The downside is that instead of just reading/writing backwards if o < 0,
we have indirect referencing through the permutation arrays. But, e.g.,
for 2D, where the only permutation is equivalent to reading/writing
backward:

- there is only one permutation array, so hopefully it stays in cache
and computing the indirection is not as bad as the worst-case for
pointer chasing, and
- the out-or-order access is always on the element array (small,
hopefully already in cache), not on the vector array (large, in-order
access could improve cache performance)

show more ...

DMPlex: added DMPlexGet/RestoreCompressedClosure()

This construction appeared a lot, so I wrapped it up.

DMPlex: added DMPlexGetPointDualSpaceFEM()

Right now we could just get a dual space for a point based on its
height/depth, but I figured it couldn't hurt to have the interface
provide individualized

DMPlex: added DMPlexGetPointDualSpaceFEM()

Right now we could just get a dual space for a point based on its
height/depth, but I figured it couldn't hurt to have the interface
provide individualized dual spaces for each point, just in case someone
wants to proceed with prismatic elements or hp methods.

show more ...

DMPlex: set PetscSectionSym in DMPlexCreateSectionInitial()

This just creates a PetscSectionSym from PetscDualSpaceGetSymmetries(),
with the proviso that if a section field has multiple components,

DMPlex: set PetscSectionSym in DMPlexCreateSectionInitial()

This just creates a PetscSectionSym from PetscDualSpaceGetSymmetries(),
with the proviso that if a section field has multiple components, we go
ahead and expand the symmetries for each component. These expanded
symmetries are redundant, but they allow us to avoid double-loops (over
nodes and field components), when reading/writing with permutations
and/or flips.

show more ...

Merge branch 'master' into stefano_zampini/feature-pcbddc-saddlepoint

Conflicts:
src/docs/website/documentation/changes/dev.html
src/ksp/ksp/examples/tutorials/ex59.c
src/ksp/pc/impls/bddc/bddc.c

Merge branch 'master' into stefano_zampini/feature-pcbddc-saddlepoint

Conflicts:
src/docs/website/documentation/changes/dev.html
src/ksp/ksp/examples/tutorials/ex59.c
src/ksp/pc/impls/bddc/bddc.c
src/ksp/pc/impls/bddc/bddcfetidp.c
src/ksp/pc/impls/bddc/bddcprivate.c
src/ksp/pc/impls/bddc/bddcschurs.c
src/mat/impls/is/matis.c
src/mat/impls/is/matis.h

show more ...

Merge branch 'knepley/feature-plex-3d-sem-order'

* knepley/feature-plex-3d-sem-order:
Plex ex6: Added 3D tests
Plex: Add in 3D spectral ordering for closure

Plex: Add in 3D spectral ordering for closure

Merge branch 'knepley/feature-plex-particle-example'

* knepley/feature-plex-particle-example: (103 commits)
PetscTypes: * Added newly types into fortran definition
DMSwarmMigrate: printf fix for

Merge branch 'knepley/feature-plex-particle-example'

* knepley/feature-plex-particle-example: (103 commits)
PetscTypes: * Added newly types into fortran definition
DMSwarmMigrate: printf fix for __float128
fix c++ comments for c89
fix scanf warnings using PetscInt_FMT
Sys+DM: Fixing 64-bit format strings
Swarm: Fixed 64-bit integer problems
Swarm: Fixed PetscFree() usage
Swarm: Correct format in scanf
Plex: Small cleanup
Plex: Fixes to draw for complex
Plex: Fixed type for GetNeighbors
Swarm: Cleanup of Dave's code - Guy has a kid and think he doesn't have to play by the rules anymore
SNES ex63: Stokes example with Particles
Swarm: Fixed prototypes
Swarm: Added missing PetscFunctionBegin
Swarm: Fixed DMGetLocalVecTor_Swarm()
Swarm: Added DMSwarmCreate/DestroyLocalVectorFromField()
Swarm: Added missing prototype
Plex: Since we are using an SF, ranks that are not found default to 0, not -1 - You need to check the index for the negative value
Plex: Added a new point location type to remove entries which are not found - This was the default, but its not what DMSwarm wanted and its non-intuitive I agree
...

Conflicts:
src/dm/impls/plex/plexcreate.c
src/dm/interface/dm.c

show more ...

Merge branch 'tisaac/dm-feature-adapt-label'

Adds an optional interface for adapting a DM by a specified label. Good
support in DMP4est; DMPlex will at least refine (I need to see if
Triangle/Tetge

Merge branch 'tisaac/dm-feature-adapt-label'

Adds an optional interface for adapting a DM by a specified label. Good
support in DMP4est; DMPlex will at least refine (I need to see if
Triangle/Tetgen support point-deletion coarsening); no DMDA
implementation, but I might put in one that at least detects uniform
coarsening/refinement.

* tisaac/dm-feature-adapt-label:
Plex tests ex20: add names to output meshes
DMPlex tests ex20: fix bad dev/null'd output in makefile test
DMPlex: gotta root out that c99 before it corrupts our kids
DMPlex test ex20: added triangle and ctetgen regression tests for DMLabelAdapt()
DMRefine_Plex_Label(): move DMPlexRestoreTransitiveClosure() after loop
DMPlex: added test ex20 for DMAdaptLabel()
DMPlex: put ifdef guards around static function only used when packages present.
DMFOREST: fix initialization
DMPlex: fix __FUNC__ of recent change
DMAdaptFlag: added DM_ADAPT_DETERMINE
DMPlex: added DMAdaptLabel implementation for DMPlex
DMAdaptLabel: change second argument to actual label
DM: added DMAdaptLabel() interface

show more ...

DMPlex: added DMAdaptLabel implementation for DMPlex

It just sets max volumes based on flags and calls the existing
refinement options.

Merge branch 'stefano_zampini/dmplex-matis-support-2'

Merge branch 'tisaac/dm-label-use-is'

* tisaac/dm-label-use-is:
IS: fix ISLocate(): ex7 passes
IS: added test for ISLocate()
IS: added ISLocate implementations for block and stride
ISGeneral

Merge branch 'tisaac/dm-label-use-is'

* tisaac/dm-label-use-is:
IS: fix ISLocate(): ex7 passes
IS: added test for ISLocate()
IS: added ISLocate implementations for block and stride
ISGeneral: add locate implementation
DMLabel: use ISLocate()
IS: added ISLocate() interface and default implementation
change regression logs for increased number of index sets
DMLabel: added DMLabelSetStratumIS()
DMLabel: switch internal storage from array to IS

show more ...

Merge branch 'tisaac/dm-boundary-in-ds'

* tisaac/dm-boundary-in-ds:
DMPlex: skip bcs whose labels are present

DMPlex: skip bcs whose labels are present

Merge branch 'tisaac/dm-boundary-in-ds'

Boundary conditions are better conceptually matched to the PetscDS,
which holds the equations, than to the DM. This makes it easier to
share the same problem

Merge branch 'tisaac/dm-boundary-in-ds'

Boundary conditions are better conceptually matched to the PetscDS,
which holds the equations, than to the DM. This makes it easier to
share the same problem between different meshes, such as in a hierarchy.

* tisaac/dm-boundary-in-ds:
PetscDS: fix examples affected by moving boundaries
builder.py: skip empty targets
DMPlex: don't copy coordinates to self
SNES ex69: fix recent PetscDS changes
DMSetDS(): reference, then destroy
PetscDS: fix _Basic creation and destruction
PETSC_DEPRECATED in source files doesn't play well with fortran interfaces
DM: switch to using PetscDS boundary in examples
PetscDS: add missing PetscDSCopyBoundary, fix copy loops
DM: set PetscDS from options in DMSetFromOptions()
DM: protect against changing prob when populating boundary
DM, PetscDS: remove redundant boundary representations
DM: transfer boundary methods to PetscDS
PetscDS: start creating parallel for DMBoundary

show more ...

DMLabel: added DMLabelSetStratumIS()

Merge branch 'knepley/feature-plex-closure-reorder'

* knepley/feature-plex-closure-reorder:
IS+Plex: Changed interface to closure permutation to take an IS
IS+Plex: Changed interface to closure

Merge branch 'knepley/feature-plex-closure-reorder'

* knepley/feature-plex-closure-reorder:
IS+Plex: Changed interface to closure permutation to take an IS
IS+Plex: Changed interface to closure permutation to take an IS
Plex ex9: Can now test closure reordering - Also fixed some small bugs
Plex ex6: Example that shows how to set and retrieve values with the spectral element ordering
Plex: Allow closure permutation in PlexVecGet/SetClosure() - This supports spectral element ordering - This does not seem to introduce any performance degradation
Plex: Added DMPlexCreateSpectralClosurePermutation()
PetscSection: Added support for closure permutation - This supports entry in the spectral element ordering

show more ...

Plex: Small cleanup

Plex: Fixes to draw for complex

PetscDS: add missing PetscDSCopyBoundary, fix copy loops

IS+Plex: Changed interface to closure permutation to take an IS

Plex: Allow closure permutation in PlexVecGet/SetClosure()
- This supports spectral element ordering
- This does not seem to introduce any performance degradation

Plex: Added DMPlexCreateSpectralClosurePermutation()

Merge branch 'tisaac/ts-ex11-amr'

We now have an example (ts/examples/tutorials/ex11.c) where we can
use either DMPlex or DMP4est to handle adaptivity when time stepping.
We have tried to minimize t

Merge branch 'tisaac/ts-ex11-amr'

We now have an example (ts/examples/tutorials/ex11.c) where we can
use either DMPlex or DMP4est to handle adaptivity when time stepping.
We have tried to minimize the amount of code that is specialized to the
different DMs, and in the future more work can be done to improve this
interface.

* tisaac/ts-ex11-amr: (62 commits)
DMP4est: type fix
Plex: Fixes for analyzer
DMP4est: update snes ex12 regressions
DMPlex: fix parent/child interpolation with numFields > 0
DMP4est: also compute anyChange for uniformly refined/coarsened meshes
TS ex11: fix command line for AMR without computed gradients
TS ex11: remove gradient reconstruction from adaptivity requirements
DMPlexTree: some compilers want this var initialized
Plex: Changed Get/RestoreFaceFields() and Get/RestoreFaceGeometry() to return the number of active faces - This fixes a long-standing bad access in FV, although the values were never actually used.
DMPlexTS: removed unused variables
DMPlex/DMP4est: small fixes for next warnings and errors
SNES ex36 and ex52: These do not work anymore
DMPforestGetTransferSF_Point(): try to match parents
PetscViewerHDF5: destroy old file data if setting new filename
DMPlex + FV: remove leaky references in projection functions
Plex: Fixed incredibly stupid bug in Jacobian computation
PetscViewer HDF5: added get name method
DMPlexTree: do not skip out of child matrix loops early
PetscFV: correct field offsets when limiting gradients
DMP4estTree: ignore negative values when getting/setting vector values for transfer
...

show more ...

DMPLEX: disable generation of fortran stub for DMPlexGetSubdomainSection