Merge branch 'barry/2022-03-30/fix-mpiu_allreduce' into 'main'

Fix mpiu_allreduce to always return PetscErrorCodes, never MPI error codes

See merge request petsc/petsc!5052

Fix mpiu_allreduce to always return PetscErrorCodes, never MPI error codes

Both for debug and optimized builds

Also removes shadow declaration errors from use of mpiu_allreduce

Commit-type: error-

Fix mpiu_allreduce to always return PetscErrorCodes, never MPI error codes

Both for debug and optimized builds

Also removes shadow declaration errors from use of mpiu_allreduce

Commit-type: error-checking
Reported-by: Lisandro Dalcin <dalcinl@gmail.com>

show more ...

Merge branch 'stefanozampini/fix-glvis-hosimplex' into 'main'

Fix high-order meshes viz

See merge request petsc/petsc!5037

DMPlexCreateCoordinateSpace: do not attach FE since projection will think that coordinates are up-to-date

DMSetCoordinates: allow NULL

Merge branch 'balay/doc-fixes' into 'main'

doc fixes

See merge request petsc/petsc!5028

doc fixes

CI get 'manualpages.err' on stdout

sed: -e expression #1, char 34: Invalid range end
Error! No level info in file:
Could not open raja.c

Merge branch 'jacobf/2022-02-23/variadic-chkerr' into 'main'

Variadic CHKERRQ()

See merge request petsc/petsc!4889

The great renaming:

- CHKERRQ() -> PetscCall()
- CHKERRV() -> PetscCallVoid()
- CHKERRMPI() -> PetscCallMPI()
- CHKERRABORT() -> PetscCallAbort()
- CHKERRCONTINUE() -> PetscCallContinue()
- CHKERRXX

The great renaming:

- CHKERRQ() -> PetscCall()
- CHKERRV() -> PetscCallVoid()
- CHKERRMPI() -> PetscCallMPI()
- CHKERRABORT() -> PetscCallAbort()
- CHKERRCONTINUE() -> PetscCallContinue()
- CHKERRXX() -> PetscCallThrow()
- CHKERRCXX() -> PetscCallCXX()
- CHKERRCUDA() -> PetscCallCUDA()
- CHKERRCUBLAS() -> PetscCallCUBLAS()
- CHKERRCUSPARSE() -> PetscCallCUSPARSE()
- CHKERRCUSOLVER() -> PetscCallCUSOLVER()
- CHKERRCUFFT() -> PetscCallCUFFT()
- CHKERRCURAND() -> PetscCallCURAND()
- CHKERRHIP() -> PetscCallHIP()
- CHKERRHIPBLAS() -> PetscCallHIPBLAS()
- CHKERRHIPSOLVER() -> PetscCallHIPSOLVER()
- CHKERRQ_CEED() -> PetscCallCEED()
- CHKERR_FORTRAN_VOID_FUNCTION() -> PetscCallFortranVoidFunction()
- CHKERRMKL() -> PetscCallMKL()
- CHKERRMMG() -> PetscCallMMG()
- CHKERRMMG_NONSTANDARD() -> PetscCallMMG_NONSTANDARD()
- CHKERRCGNS() -> PetscCallCGNS()
- CHKERRPTSCOTCH() -> PetscCallPTSCOTCH()
- CHKERRSTR() -> PetscCallSTR()
- CHKERRTC() -> PetscCallTC()

show more ...

fixes from recent linter change

chkerr and friends wrapped

Merge branch 'knepley-main-patch-53941' into 'main'

DM: Fix bad PetscCheck

See merge request petsc/petsc!5012

DM: Fix bad PetscCheck

Merge branch 'barry/2022-03-20/rm-dm-some-petsccheckfalse' into 'main'

Remove some PetscCheckFalse() from DM directories

See merge request petsc/petsc!4997

Remove some PetscCheckFalse() from DM directories

Fix several incorrect %d in formats

Commit-type: error-checking, house-keeping
/spend 45m

Merge branch 'knepley/feature-dm-aux-vec' into 'main'

DM: Add equation part to DM aux vector key

See merge request petsc/petsc!4995

DM: Add equation part to DM aux vector key

Merge branch 'jed/plex-coordinates-project' into 'main'

DMPlex: skip coordinate projection when FE space matches modulo quadrature

See merge request petsc/petsc!4891

DMPlex: skip coordinate projection when FE space matches modulo quadrature

This requires checking that dual spaces are equal so that an legit
projection is not required. For example, high order Gmsh

DMPlex: skip coordinate projection when FE space matches modulo quadrature

This requires checking that dual spaces are equal so that an legit
projection is not required. For example, high order Gmsh meshes have
equally spaced vertices while we prefer Gauss-Lobatto spacing, so that's
a case where a projection is required. (It could be made faster, but
that's for another day.)

show more ...

Merge branch 'jed/dm-prealloc-skip' into 'main'

DMSetMatrixPreallocateSkip: skip setup that will be discarded with MatSetPreallocationCOO()

See merge request petsc/petsc!4990

DMSetMatrixPreallocateSkip: skip setup that will be discarded with MatSetPreallocationCOO()

Analytic matrix preallocation runs on CPU and is relatively expensive.
It is entirely discarded when MatSe

DMSetMatrixPreallocateSkip: skip setup that will be discarded with MatSetPreallocationCOO()

Analytic matrix preallocation runs on CPU and is relatively expensive.
It is entirely discarded when MatSetPreallocationCOO() is called.

show more ...

Merge branch 'barry/2022-03-13/fix-some-manualpages/release' into 'main'

Improve generation of manual pages/fix a variety of manul pages basic formatting

See merge request petsc/petsc!4969

doc: Fix a variety of manual page formatting issues

/spend 1h

Merge branch 'jed/dm-block-size-gcd' into 'main'

DM: set local and global vector sizes based on gcd

See merge request petsc/petsc!4937

DM: set local and global vector sizes based on gcd

This enables having block size 3 for elasticity with Q3 elements, for
example. The alternative reduces to bs=1 in such cases because the block
size

DM: set local and global vector sizes based on gcd

This enables having block size 3 for elasticity with Q3 elements, for
example. The alternative reduces to bs=1 in such cases because the block
sizes don't match. Also use gcd for DMGetLocalToGlobalMapping.

show more ...