Thank you so much for your ultra fast answer Pierre!
Nicolas
De : pie...@joliv.et
Envoyé : mardi 1 juillet 2025 18:58
À : TARDIEU Nicolas
Cc : petsc-users@mcs.anl.gov
Objet : Re: [petsc-users] New Fortran API to PCFieldSplitGetSubKSP
Expéditeur externe : Vér
Dear Nicolas,
> On 1 Jul 2025, at 6:31 PM, TARDIEU Nicolas via petsc-users
> wrote:
>
> Dear PETSc team,
>
> I am having hard time upgrading our code's interface to the new Fortran
> API...
> For instance, I lack a Fortran example of the use of PCFieldSplitGetSubKSP.
> The doc says "You m
Dear PETSc team,
I am having hard time upgrading our code's interface to the new Fortran API...
For instance, I lack a Fortran example of the use of PCFieldSplitGetSubKSP. The
doc says "You must pass in a KSP array that is large enough to contain all the
KSPs" but the interface file is :
sub
Hi Pierre and Barry,
Thanks for your answers and sorry for my reaction time (a bit overwhelmed...).
Yes I confirm that I already tried `MAT_FORCE_DIAGONAL_ENTRIES` following a
Pierre suggestion, but it did not help...
I understand from your messages that this option indeed needs a fix. Do yo
It's probably already been done, but if not, you should run under Valgrind.
"Random" crashes are usually an indication of memory corruption.
> On Jul 1, 2025, at 4:16 AM, Klaij, Christiaan via petsc-users
> wrote:
>
> It's been a while, in the meantime we did upgrade the OS and the
> co
It's been a while, in the meantime we did upgrade the OS and the
compilers but the problem still persists.
Can it be that one must call the PetscLogEventRegister and the
PetscLogEventBegin/End on all procs? Currently we don't do so,
think of some small system being build and solved. This event is
