Merge branch 'barry/2022-04-14/fix-rebalance' into 'main'

Major repairs for DMPlexRebalanceSharedVertices()

See merge request petsc/petsc!5234

Fix rebalancing of ghost pointers for DMPLEX

/spend 25h

Commit-type: optimization, bug-fix, style-fix, feature

MATAIJCUSPARSE: add option -mat_factor_bind_factorization <host,device>

MATAIJCUSPARSE: share device buffers used in ILU0 or ICC0

Merge branch 'jczhang/fix-dpcpp-compilation-warnings' into 'main'

Fix some compilation warnings with the new Intel dpcpp compiler

See merge request petsc/petsc!5380

Added -mpi_linear_solver_server that allows using MPI parallel KSP solvers from sequential or OpenMPI code

Uses -ksp_type preonly -pc_type mpi The inner KSP has a prefix of -mpi

-mpi_linear_solver_

Added -mpi_linear_solver_server that allows using MPI parallel KSP solvers from sequential or OpenMPI code

Uses -ksp_type preonly -pc_type mpi The inner KSP has a prefix of -mpi

-mpi_linear_solver_server_view gives a very short summary of the solves

Commit-type: feature
/spend 15h

show more ...

Merge branch 'wence/petsc-check-do-while' into 'main'

Wrap PetscCheck and PetscAssert in do { } while (0)

Closes #1199

See merge request petsc/petsc!5291

Fix a bunch of 'set but not used' compilation warnings

MATAIJCUSPARSE: fix some typos that cause compile errors

MATAIJCUSPARSE: support symbolic and numeric ILU0 and ICC0 with natural ordering on device

MATAIJCUSPARSE: function renaming to mimic cusparse APIs

MATAIJCUSPARSE: minor style changes

MATAIJ: cosmetic changes

MATAIJCUSPARSE: minor style changes

Move create and filter graph from PCGAMG to AMTAIJ

Wrap PetscCheck and PetscAssert in do { } while (0)

These macros look like functions, so they should behave like
functions. Fix up the resulting invalid syntax due to the previous
behaviour of these

Wrap PetscCheck and PetscAssert in do { } while (0)

These macros look like functions, so they should behave like
functions. Fix up the resulting invalid syntax due to the previous
behaviour of these macros being "if (...)
statement_without_semicolon". Fixes #1199.

show more ...

Change if () { PetscCall() } three liner and friends to one liners

for i in `git ls-files | grep "\.[ch]$"` ; do sed 's?\$?ZZZ?g' $i | tr '\n' '$' | sed 's?\([ ]*\)if (\([-;,.\*+=a-z0-9A-Z_>]*\)) {\

Change if () { PetscCall() } three liner and friends to one liners

for i in `git ls-files | grep "\.[ch]$"` ; do sed 's?\$?ZZZ?g' $i | tr '\n' '$' | sed 's?\([ ]*\)if (\([-;,.\*+=a-z0-9A-Z_>]*\)) {\$[ ]*PetscCall(\([- ._+=a-z0-9A-Z>*,()]*);\)\$[ ]*}\$?\1if (\2) PetscCall(\3$?g' | tr '$' '\n' | sed 's?ZZZ?$?g' > $i.joe ; mv $i.joe $i ; done

for i in `git ls-files | grep "\.[hc]$"` ; do sed 's?\$?ZZZ?g' $i | tr '\n' '$' | sed 's?\([ }else]*\)if (\([-;,.\*+=a-z0-9A-Z_>]*\)) {\$[ ]*PetscCall(\([- ._+=a-z0-9A-Z>*,()]*);\)\$\([ ]*\)} \([- ._+=a-z0-9A-Z>*,()]*);\)\$?\1if (\2) PetscCall(\3$\4\5$?g' | tr '$' '\n' | sed 's?ZZZ?$?g' > $i.joe ; mv $i.joe $i ; done

Yes, really ugly but Barry still cannot master awk

Commit-type: housekeeping

show more ...

Change matrix factor database options that occur in KSP/PC to always use KSP/PC options prefix

Add MatSetOptionsPrefixFactor() and MatAppendOptionsPrefixFactor() to allow controlling the options pre

Change matrix factor database options that occur in KSP/PC to always use KSP/PC options prefix

Add MatSetOptionsPrefixFactor() and MatAppendOptionsPrefixFactor() to allow controlling the options prefix
used by factors created from this matrix

Change MatSetOptionsPrefix() to no longer affect the options prefix used by factors created from this matrix

As part of the above work the handling of the factor matrix options is now done in code in the factor matrix
not in the code that originates the factor matrix

Update examples output to the new approach where the factored matrix prefix is set from the KSP/PC

Much of this work was performed by Hong Zhang

Commit-type: usability
/spend 5h

show more ...

Add notes to MatDestroy() that destroys of matrices may not
go through this routine.

I wasted several hours debugging a simple memory leak due to this
design.

/spend 3h

PetscObjectFunctionCompose: clean up composed functions at Destroy time

Merge remote-tracking branch 'origin/release'

add MatGetDiagonal to Kokkos

Kokkos: guard a statement with kokkos version

Fix for use of macros with arguments when macro was used without arguments

E:\FGH\Git\Windows\petsc-3.17.1\src\mat\impls\aij\seq\aijmkl\aijmkl.c(148): warning C4003: not enough arguments for functio

Fix for use of macros with arguments when macro was used without arguments

E:\FGH\Git\Windows\petsc-3.17.1\src\mat\impls\aij\seq\aijmkl\aijmkl.c(148): warning C4003: not enough arguments for function-like macro invocation 'mkl_sparse_x_create_csr'
E:\FGH\Git\Windows\petsc-3.17.1\src\mat\impls\aij\seq\aijmkl\aijmkl.c(148): error C2059: syntax error: ','
E:\FGH\Git\Windows\petsc-3.17.1\src\mat\impls\aij\seq\aijmkl\aijmkl.c(177): warning C4003: not enough arguments for function-like macro invocation 'mkl_sparse_x_export_csr'

/spend 10m
Reported-by: Henrik Büsing <Henrik.Buesing@fgh-ma.de>

show more ...

Merge remote-tracking branch 'origin/release'