Merge branch 'barry/2025-07-30/fix-matshellsetoperationarg' into 'main'

Complete usage of PetscVoidFn and PetscErrorCodeFn and introduce PetscFortranCallbackFn

See merge request petsc/petsc!8615

Change use of void (*)(void) and void (**)(void) in the PETSc API to PetscVoidFn * and PetscVoidFn ** or PetscErrorCodeFn * or PetscErrorCodeFn **

Merge branch 'barry/2025-08-29/improve-ts-pseudo-performance' into 'main'

Optimize TSStep_Pseudo() for most common use case

See merge request petsc/petsc!8675

Add `-petsc_viewer_stdout_format formatname` to allow initializing the format of `PETSC_VIEWER_STDOUT_()`

Make PetscStackView() a public function

Merge branch 'barry/2025-08-31/fix-hypre-default-setting-reset-to-same' into 'main'

Fixes the bug with hypre whan user changes to parameters are ignored

See merge request petsc/petsc!8678

Merge branch 'barry/2025-09-04/deprecate-mpiu-bool' into 'main'

Deprecate MPIU_BOOL which used to be for PETSc's enum Bool but now is just MPI_C_BOOL

See merge request petsc/petsc!8691

Add PetscOptionsBool3()

Deprecate MPIU_BOOL which used to be for PETSc' enum Bool but now is just MPI_C_BOOL so no longer needs its own name

Merge branch 'add-tao-get-constraints' into 'main'

Add getters for TAO constraint routines

See merge request petsc/petsc!8619

Add test for TAO constraint getters, and update changes/dev notes

Merge branch 'MarDiehl/use-bool-instead-of-enum' into 'main'

use bool instead of enum

See merge request petsc/petsc!8528

use standard bool instead of enum

PetscBool is equivalent to _Bool/bool in C

For C standards prior to C23, true = 1 and false = 0 are defined in <stdbool.h>.
For C23, true and false are real keywor

use standard bool instead of enum

PetscBool is equivalent to _Bool/bool in C

For C standards prior to C23, true = 1 and false = 0 are defined in <stdbool.h>.
For C23, true and false are real keywords.

The Fortran equivalent is logical(C_BOOL) with C_BOOL from
ISO_C_binding.

The C and the C++ standard do not specify the size of Booleans, but they
typically have a size of 1 Byte with b00000000 for false and b00000001 for true.
When using compilers from the same vendor, it is sane to assume that
_Bool/bool in C and bool in C++ are fully compatible. But that is not
guaranteed by the standards

show more ...

ensure that Fortran Bool is interoperable

The kind C_BOOL for logical from ISO_C_binding only defines the size of
the logical as 1 Byte. It does not define which values (bit patterns) are
used to en

ensure that Fortran Bool is interoperable

The kind C_BOOL for logical from ISO_C_binding only defines the size of
the logical as 1 Byte. It does not define which values (bit patterns) are
used to encode true and false. In stdbool.h from C, true is defined as 1
and false as 0, but these values are not guaranteed by the Fortran standard.

There are (at least) two contemporary Fortran compilers that use a
difference convention: Intel Fortran (ifort and ifx) and NVIDIA Fortran
(nvfortran, former PGI). Both represent true as all bits set, i.e. -1 for
a signed integer or 255 for an unsigned integer.

Intel Fortran can be made interoperable with the "-standard-semantics" or
"-fpscomp logicals". The latter is now set by default to ensure
interoperability while avoiding effects such as modified name mangling
rules and performance degradations due to the former.

NVIDIA Fortran can be made interoperable with the "-Munixlogical" flag.
This is set manually in the "linux-cuda-single-cxx" pipeline. This is
now set by default.

show more ...

Merge branch 'sneslinesearch-improvements' into 'main'

SNESLinesearch: Minor improvements

See merge request petsc/petsc!8630

SNESLinesearch: Minor improvements

- add missing configuration options in the documentation of the line searches (some of the available options were not listed), and consistently list their respecti

SNESLinesearch: Minor improvements

- add missing configuration options in the documentation of the line searches (some of the available options were not listed), and consistently list their respective default values

- add proper convergence criteria for the secant method in the L2 line search based on ltol and atol (otherwise, the secant method is mostly running for the maximum number of iterations)

- fix a small bug in the CP line search, where steptol * lambda was used to check for convergence based on the change of lambda instead of the ltol parameter

- rename sneslinesearch->maxstep to sneslinesearch->maxlambda in the options database to be consistent with its purpose in the algorithms

- rename SNESLineSearchL2 to SNESLineSearchSecant to better represent its underlying approach

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

Merge branch 'knepley/test-run-opts' into 'main'

Test: Added EXTRA_OPTIONS_INITiAL to allow overridding of test options

See merge request petsc/petsc!8643

Test: Added EXTRA_OPTIONS_INITIAL to allow overridding of test options
- EXTRA_OPTIONS returns to its initial meaning

configure: disable Fortran bindings of external packages by default

Merge branch 'zach/extern-plex-vec-get-oriented-closure' into 'main'

plex: Make DMPlexVecGetOrientedClosure externally visible

See merge request petsc/petsc!8637

plex: Make DMPlexVecGetOrientedClosure externally visible

Merge branch 'danofinn/ts-add-avg-dg' into 'main'

TSDiscgrad: Adding average DG and -ts_discgrad_type option; Swarm: Changes to field swarm field functions

See merge request petsc/petsc!7920

TSDiscgrad: Adding average DG and -ts_discgrad_type option
- Also fixed automatic particle number determination

FOO

Merge branch 'knepley/fix-ksp-monitor-view' into 'main'

KSP: Add HDF5 monitor

See merge request petsc/petsc!8624