Change names of Mat_XXX product contexts to MatProductCtx_XXX for code maintainability

Update destroy callback of all MatProductCtx and MatShellSetMatProductOperation() to use PetscCtxDestroyFn

Implements MatGetDiagonal for CUDA to take place on device.

Reported-by: Pierre.LEDAC@cea.fr
Thanks-to: Junchao Zhang

Merge remote-tracking branch 'origin/release'

Minor fixes to MATXXXAIJCUSPARSE manual pages

Merge branch 'balay/fix-doc' into 'release'

doc: remove non-existing references

See merge request petsc/petsc!8748

Merge branch 'jolivet/fix-arrays' into 'release'

Fix error in MatTransposeMatMultSymbolic() in case of empty rows

See merge request petsc/petsc!8773

MatSetUp_(Seq|MPI)_Hash(): fix confusing -info message

[0] <mat:mpiaij> MatSetUp_MPI_Hash(): Using hash-based MatSetValues() for MATMPISBAIJ because no preallocation provided

checkbadSource: enforce proper style in makefiles

Merge remote-tracking branch 'origin/release'

Tiny improvements to MatSOR manual page

MatZeroEntries_SeqAIJCUSPARSE() was zeroing both the CPU and GPU array. Instead just zero the GPU array if it is available, save unneeded CPU operation.

Merge branch 'jczhang/2025-09-19/precision-mumps' into 'main'

Support different precisions with MUMPS

See merge request petsc/petsc!8715

Enable ScaLAPACK and MUMPS for other precisions

Fix -Wshorten-64-to-32

Avoid extra allocation

Fix illegal access of i1[n1 - 1] when n1 is zero.

Co-authored-by: Junchao Zhang <jczhang@mcs.anl.gov>

doc: remove non-existing references

Reported-by: Victor Eijkhout <eijkhout@tacc.utexas.edu>

Run fprettify

Merge branch 'stefanozampini/submat-zerorows' into 'main'

implement ZeroRowsLocal and ZeroRowsColumnsLocal for local references

See merge request petsc/petsc!8705

the etags parse got confused by the text Function cxxxx() in a comment in a CU file so remove it

Merge branch 'jolivet/fix-pchpddm-scatter' into 'main'

Minor fixes and cleanup

See merge request petsc/petsc!8702

Mat: add ops for zerorowscolumnslocal

Gather CUDA lib includes in petscdevice_cuda.h + disable cuSPARSE warnings using DISABLE_CUSPARSE_DEPRECATED

Defining DISABLE_CUSPARSE_DEPRECATED is Nvidia's recommended way of disabling cuSPARSE wa

Gather CUDA lib includes in petscdevice_cuda.h + disable cuSPARSE warnings using DISABLE_CUSPARSE_DEPRECATED

Defining DISABLE_CUSPARSE_DEPRECATED is Nvidia's recommended way of disabling cuSPARSE warnings
This avoids requiring -Wdeprecated-declarations - that can mask other (non-cuSPARSE) warnings

show more ...

Fix warnings due to ignoring return value type with nodiscard attribute

Remove more unneeded braces