[petsc-dev] Error in make test

Stefano Zampini stefano.zampini at gmail.com
Sun May 6 06:24:12 CDT 2012


I've just pulled and make test is broken for me. There's something messed
up with PetscThreadCommunicators.
Besides, I did not configure petsc-dev to use threads.
A revision which works for me is 23244.
I'm using openmpi.1.5.5.

Here is the traceback of the error

[0]PETSC ERROR: PetscCommDuplicate() line 200 in
/work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/tagm.c
[0]PETSC ERROR: PetscHeaderCreate_Private() line 51 in
/work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/inherit.c
[0]PETSC ERROR: SNESCreate() line 1293 in
/work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/snes/interface/snes.c
[0]PETSC ERROR: main() line 102 in src/snes/examples/tutorials/ex19.c


2012/5/5 Stefano Zampini <stefano.zampini at gmail.com>

> I reverted to 23244.
>
> 2012/5/5 Shri <abhyshr at mcs.anl.gov>
>
>> I've disabled the threadcomm code temporarily, so hopefully your code
>> should run as before now. Please pull and retry.
>>
>> Thanks,
>> Shri
>>
>> ------------------------------
>>
>> Same error on make test.
>>
>> Moreover, I got some mpi errors running my fortran codes and PETSc
>> doesn't show the traceback
>> I can only see that the problem occurs in PetscInitialize
>> Here is the cryptic output from MPI (openmpi.1.4.5)
>>
>> [neo155:25438] *** An error occurred in MPI_Attr_get
>> [neo155:25438] *** on communicator MPI COMMUNICATOR 3 DUP FROM 1
>> [neo155:25438] *** MPI_ERR_OTHER: known error not in list
>> [neo155:25438] *** MPI_ERRORS_ARE_FATAL (your MPI job will now abort)
>>
>>
>> 2012/5/5 Shri <abhyshr at mcs.anl.gov>
>>
>>> I pushed some changes to fix these errors. Can you please pull and retry.
>>>
>>>
>>> ------------------------------
>>>
>>> I had pulled some hours ago from
>>> http://petsc.cs.iit.edu/petsc/petsc-dev/rev/61898f35c333 and obtained
>>> the following error on make test
>>>
>>> [0]PETSC ERROR: PetscCommDuplicate() line 200 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/tagm.c
>>> [0]PETSC ERROR: PetscHeaderCreate_Private() line 51 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/inherit.c
>>> [0]PETSC ERROR: SNESCreate() line 1293 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/snes/interface/snes.c
>>> [0]PETSC ERROR: main() line 102 in src/snes/examples/tutorials/ex19.c
>>>
>>> I've just pulled and got the following error after typing make test
>>>
>>> [0]PETSC ERROR: PetscCommDestroy() line 246 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/tagm.c
>>> [0]PETSC ERROR: PetscThreadCommFinalizePackage() line 36 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/threadcomm/interface/dlregisthreadcomm.c
>>> [0]PETSC ERROR: PetscRegisterFinalizeAll() line 313 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/destroy.c
>>> [0]PETSC ERROR: PetscFinalize() line 1080 in
>>> /work/adz/zampini/MyWorkingCopyOfPetsc/petsc-dev/src/sys/objects/pinit.
>>>
>>> I did not configure for a threaded version of PETSc.
>>>
>>> --
>>> Stefano
>>>
>>>
>>>
>>
>>
>> --
>> Stefano
>>
>>
>>
>
>
> --
> Stefano
>



-- 
Stefano
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120506/d44e46f2/attachment.html>


More information about the petsc-dev mailing list