Merge branch 'origin/release' with a revert of !7720
MatShellGetScalingShifts: consolidate checks for MatShell subtypes
Add shared memory support for PCMPI using both shmget() and MPI_Win_allocate_shared()
Remove use of private() from OpenMP pragma's since Jed says they are unsafeThanks-to: Jed Brown
Add OpenMP directive to MatMult_SeqAIJ etcGives pretty good speedup on good memory-bandwidth Intel system and M2 MacReported-by: Yongzhong Li <yongzhong.li@mail.utoronto.ca>
Merge branch 'barry/2024-06-30/doc-vecsetinf' into 'main'Refactor VecSetInf() into VecFlag() and deprecate VecSetInf()See merge request petsc/petsc!7661
Refactor VecSetInf() into VecFlag() and deprecate VecSetInf()Reported-by: Stefano Zampini
Merge remote-tracking branch 'origin/release'
Fix error checking in DMDA for periodic cases where it cannot handle periodic boundary conditionsAlso add error checking to MatFDColoring() to check for bad coloring input.Also fix bug where call
Fix error checking in DMDA for periodic cases where it cannot handle periodic boundary conditionsAlso add error checking to MatFDColoring() to check for bad coloring input.Also fix bug where call to ISColoringGetIS() where the is already exists and the caller requests PETSC_OWN_POINTERReported-by: Richard Katz <richard.katz@earth.ox.ac.uk>
show more ...
Kokkos: try to always have an execution space argument
Mat: fix -Wmissing-field-initializers
This requires some changes to user FORTRAN codePass PETSC_NULL_XXX_ARRAY when passing a NULL as an argument that is expecting an arrayPass PETSC_NULL_ENUM when argument returns an enum instead of
This requires some changes to user FORTRAN codePass PETSC_NULL_XXX_ARRAY when passing a NULL as an argument that is expecting an arrayPass PETSC_NULL_ENUM when argument returns an enum instead of PETSC_NULL_INTEGERPass arrays (and not scalar values) when the argument is expecting an array; this means replace, for example, the argument v with [v]Use PetscObjectIsNull(obj) to check if the object is NULL, instead of obj == PETSC_NULL_XXXThe compiler will now automatically prevent you from using the wrong argument type for the first three bullets aboveThis will also require an update sowing with the new sowing branch this MR is using.
Ensure 64 bit integer when counting flops to accommodate large blocks.
Remove now redundant increase in mat nonzero state
Fix serious bugs in MatSetValuesBlocked for SeqBAIJ and MPIBAIJ when a reallocation is needed.Reported-by: Adrian Croucher <a.croucher@auckland.ac.nz>
Remove unneeded include <petsc/private/fortranimpl.h>
hip: support ROCm-5.7 through 6.1
Update source code removing all unneeded /*@C and associated manual stubs and interfaces
Use PetscStackCallExternalVoid() on CPardiso wrapper for clearer error message.Reported-by: Nils Friess
MATAIJCRL: fix for FPE
MatSolverTypeRegister: switch to PETSC_INTERN
12345678910>>...168