[petsc-dev] Error with DMCompositeGetISLocalToGlobalMappings

[Jed Brown]

1. Call DMSetUp(packer) once you are done adding DMs. I will make it
automatic in petsc-dev.

2. Call PetscSubcommDestroy() to free unused memory.

3. I suggest running your experiments with less than 64 procs.

On Fri, Feb 3, 2012 at 18:32, <domenico.borzacchiello at univ-st-etienne.fr>wrote:

> Hi,
> I'm getting a segmentation violation error when I use
> DMCompositeGetISLocalToGlobalMappings on a DMComposite defined redundantly
> on subsets of the PETSC_COMM_WORLD. I could reproduce the error in the
> small code you can find here attached. It creates a PetsSubcomm from
> PETSC_COMM_WORLD, then it defines a DMDA on the Subcomm->comm and finally
> it packs it into a DMComposite.
> I get this error whenever I call DMCompositeGetISLocalToGlobalMapping on a
> DMComposite defined in such a way.
>
> I would be glad if you could give it a look.
> Thank you
> Domenico
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120203/88e75821/attachment.html>