Merge branch 'jczhang/2026-01-29/hypre-umpire-pool-size' into 'main'

Add option -hypre_umpire_device_pool_size <size in MiB>

See merge request petsc/petsc!8984

HYPRE: add option -hypre_umpire_device_pool_size <size in MiB>

When hypre is configured with Umpire (recommended by hypre), the default Umpire pool size on device is set
to 4 GB per process by hypre

HYPRE: add option -hypre_umpire_device_pool_size <size in MiB>

When hypre is configured with Umpire (recommended by hypre), the default Umpire pool size on device is set
to 4 GB per process by hypre which could lead to out-of-memory when one maps multiple MPI processes to one GPU.

show more ...

Merge branch 'jczhang/2025-10-27/update-kokkos-5.0' into 'main'

Update to Kokkos-5.0

See merge request petsc/petsc!8815

Kokkos: update to v5.0.0

CI: remove kokkos from linux-cuda-pkgs (in favor of linux-kokkos-cuda) as v5 breaks with many external pkgs,
CI: continue to use kokkos v4.7.01 for hip builds - as they curr

Kokkos: update to v5.0.0

CI: remove kokkos from linux-cuda-pkgs (in favor of linux-kokkos-cuda) as v5 breaks with many external pkgs,
CI: continue to use kokkos v4.7.01 for hip builds - as they currently use rocm-5.4.3

Co-authored-by: Satish Balay <balay@mcs.anl.gov>

show more ...

Merge branch 'jczhang/2025-12-10/add-caliper' into 'main'

Add Caliper, a profiling package used by Hypre

See merge request petsc/petsc!8895

Packages: add Caliper, a profiling package used by Hypre

Merge branch 'balay/hypre-3.0' into 'main'

configure: HYPRE v3.0 and compatible MFEM

See merge request petsc/petsc!8688

Add Umpire support

CI: CUDA+umpire can trigger errors due to default 4G/mpi-task GPU RAM usage, so disable for linux-cuda-pkgs, linux-cuda-latest

> CUDA ERROR (code = 1, invalid argument) at memory

Add Umpire support

CI: CUDA+umpire can trigger errors due to default 4G/mpi-task GPU RAM usage, so disable for linux-cuda-pkgs, linux-cuda-latest

> CUDA ERROR (code = 1, invalid argument) at memory.c:136
> terminate called after throwing an instance of 'thrust::system::system_error'
> what(): reduce failed to synchronize: cudaErrorIllegalAddress: an illegal memory access was encountered

Co-authored-by: Pierre Jolivet <pierre@joliv.et>
Co-authored-by: Satish Balay <balay@mcs.anl.gov>

show more ...

Merge branch 'jolivet/disable-fortran-bindings' into 'main'

configure: disable Fortran bindings of external packages by default

See merge request petsc/petsc!8647

configure: disable Fortran bindings of external packages by default

Merge branch 'jolivet/update-packages' into 'main'

configure: update CMake/SuiteSparse

See merge request petsc/petsc!8184

configure: CHOLMOD requires OpenMP for CUDA builds

add --with-suitesparse-cuda: default to auto-enable/disable (wrt --with-cuda --with-openmp dependencies) unless user explicitly specifies this opti

configure: CHOLMOD requires OpenMP for CUDA builds

add --with-suitesparse-cuda: default to auto-enable/disable (wrt --with-cuda --with-openmp dependencies) unless user explicitly specifies this option

show more ...

Merge branch 'add_tetgen_triangle_exe' into 'main'

configure: add -download-triangle-build-exec, -download-tetgen-build-exec options to enable build of binaries

See merge request petsc/petsc!7839

configure: add -download-triangle-build-exec, -download-tetgen-build-exec options to enable build of binaries

Merge branch 'strumpack' into 'main'

Update the interface to STRUMPACK

See merge request petsc/petsc!6819

Update the interface to STRUMPACK

STRUMPACK provides a sparse direct solver (but unlike fi SuperLU, it is based on the multifrontal method). STRUMPACK also provides preconditioners based on approxim

Update the interface to STRUMPACK

STRUMPACK provides a sparse direct solver (but unlike fi SuperLU, it is based on the multifrontal method). STRUMPACK also provides preconditioners based on approximate multifrontal LU factorization. In the precoditioners larger dense blocks in the sparse LU factors are compressed using rank-structured matrix approximations (or using lossy/lossless compression using the ZFP library).

See https://github.com/pghysels/STRUMPACK https://portal.nersc.gov/project/sparse/strumpack/

This commit adds different preconditioning options. Previously only the HSS (hierarchically Semi-Separable) format was supported. Now we also have BLR (Block Low Rank), HODLR (Hierarchically Off-Diagonal) with the option to replace low-rank with butterfly, Lossy and Lossless compression (through ZFP). The new methods work considerably better than the older HSS code. When using

-pc_type lu -pc_factor_mat_solver_type strumpack

the solver behaves as a direct solver. When using

-pc_type ilu -pc_factor_mat_solver_type strumpack

it will work as a preconditioner, using BLR compression. This should be a robust preconditioner for a wide range of problems. One can tune the compression tolerance, and the minimum block (separator) size for compression using:

-mat_strumpack_compression_rel_tol 1e-3 -mat_strumpack_compression_min_sep_size 500

You can also select the compression type explicitly

-pc_type ilu -pc_factor_mat_solver_type strumpack -mat_strumpack_compression LOSSY

For now we recommend BLR, as it seems to work best, and does not require additional external dependencies (it is also what MUMPS implements).

Also updates the test src/ksp/ksp/tutorials/ex52.c.

Also adds support for the solve with multiple RHS.

Also adds an option for the GEOMETRIC fill-reducing ordering which performs nested dissection on a regular nx x ny x nz grid with nc degrees of freedom per grid-point and a stencil (potentially a wider stencil). It is assumed that the matrix is in the natural (lexicographical?) ordering. This can make reordering much faster for regular grid problems.

Also adds installation scripts STRUMPACK dependencies SLATE, ZFP and ButterflyPACK. SLATE is a scalapack replacement with GPU capabilities. ZFP is used for floating point compression. ButterflyPACK is used for the HODLR and HODBF compression schemes.

show more ...

Merge branch 'stefanozampini/mfem-hip' into 'main'

Some changes when working on enabling MFEM + PETSC + HIP

See merge request petsc/petsc!6651

CI: tweak tests for devices

activate HYPRE tests in arch-ci-linux-cuda-pkgs

Merge remote-tracking branch 'origin/release'

Merge branch 'jczhang/2023-04-11/fix-sfkok-stream/release' into 'release'

Let SF Kokkos backend use an execution space made from PetscDefaultCuda/HipStream

See merge request petsc/petsc!6317

Let SF Kokkos backend use an execution space made from PetscDefaultCuda/HipStream

Merge branch 'jacobf/2023-02-08/the-great-gcovening' into 'main'

The Great GCOVening

See merge request petsc/petsc!6040

Drop gcov suffix from old gcov CI jobs

Merge branch 'adams/pcamgx-testfix' into 'main'

CI: add --download-amgx to linux-cuda-pkgs

See merge request petsc/petsc!5548

CI: add --download-amgx to linux-cuda-pkgs
And fix related issues