It’s weird that the tests are running OK on linux-pkgs-opt, but not on my naked
Linux VM.
If that’s not too much to ask, could you please fetch
jolivet/feature-hpddm-cleanup, ./configure --download-slepc --download-hpddm
And then, make -f gmakefile test globsearch='ksp_ksp_tutorials-ex76f_*'
That’s the stack
# [0]PETSC ERROR: #1 PCHPDDMSetUpNeumannOverlap_Private() line 505 in
/home/john/petsc/src/ksp/pc/impls/hpddm/hpddm.cxx
# [1]PETSC ERROR: #2 PCSetUp_HPDDM() line 624 in
/home/john/petsc/src/ksp/pc/impls/hpddm/hpddm.cxx
# [1]PETSC ERROR: #3 PCSetUp() line 894 in
/home/john/petsc/src/ksp/pc/interface/precon.c
# [1]PETSC ERROR: #4 KSPSetUp() line 376 in
/home/john/petsc/src/ksp/ksp/interface/itfunc.c
# [1]PETSC ERROR: #5 KSPSolve() line 703 in
/home/john/petsc/src/ksp/ksp/interface/itfunc.c
# [1]PETSC ERROR: #6 User provided function() line 0 in User file
data->setup is supposed to be NULL (PETSC_NULL_FUNCTION passed in
PCHPDDMSetAuxiliaryMat), but apparently it’s not (on my machine), since line
505 is in the if(data->setup) { } branch.
Thanks for your help,
Pierre
> On 17 Jan 2020, at 10:29 PM, Smith, Barry F. <[email protected]> wrote:
>
>
> Better to send more details. What function are you calling, what is the full
> error stack. Much quicker for us to help you then repeated email (you can
> never send too much information about a crash).
>
> PETSC_NULL_FUNCTION is the thing to use, but it is completely manual to for
> us to put them in the Fortran stubs, we have missed this case but we need to
> know what it is so we can fix it.
>
> Barry
>
>
>> On Jan 17, 2020, at 3:18 PM, Pierre Jolivet <[email protected]>
>> wrote:
>>
>>
>>
>>> On 13 Jan 2020, at 4:38 PM, Smith, Barry F. <[email protected]> wrote:
>>>
>>>
>>>
>>>> On Jan 13, 2020, at 9:32 AM, Pierre Jolivet <[email protected]>
>>>> wrote:
>>>>
>>>> Hello,
>>>> This is actually two separate questions, sorry.
>>>> 1) I’m looking for the Fortran equivalent of the following, but I couldn’t
>>>> get any help looking at the sources.
>>>> ierr = PetscOptionsBegin(PETSC_COMM_WORLD,"","","");CHKERRQ(ierr);
>>>> ierr = PetscOptionsFList("-mat_type","Matrix
>>>> type","MatSetType",MatList,deft,type,256,&flg);CHKERRQ(ierr);
>>>> ierr = PetscOptionsEnd();CHKERRQ(ierr);
>>>
>>> These are based on C macros and have some funny arguments so would require
>>> all custom Fortran stubs and interfaces. Sorry no one has written them.
>>
>> OK. No big deal.
>> Another question about Fortran: what’s the equivalent of a NULL
>> PetscErrorCode (*setup)(Mat, PetscReal, Vec, Vec, PetscReal, IS, void*)?
>> I’m passing a PETSC_NULL_FUNCTION but it doesn’t seem to do the job as I’m
>> getting segmentation faults in C (for some architectures), the code goes in
>> the if(setup) branch.
>>
>> I want to convert this:
>> https://gitlab.com/petsc/petsc/blob/fbda2a43629902c7e64367d3e3c13db2ceb5b740/src/ksp/ksp/examples/tutorials/ex76.c#L68
>> Into (hopefully something similar to) this:
>> https://gitlab.com/petsc/petsc/commit/fbda2a43629902c7e64367d3e3c13db2ceb5b740#42254f97b9989f77d6aa51f62644a282fc31f3fd_0_93
>>
>> Thanks in advance for bearing with me,
>> Pierre
>
