[petsc-dev] Small fix associating MatPtAP_seqaijperm_seqmaij_C

[Richard Tran Mills]

Hi Folks,

I didn't see a reply to this message, so pinging again: Does this need to
get merged into 'next' for testing? Can it just go directly into master? If
it needs to go into 'next', should I do this merge myself, or do we have
"gatekeepers" (e.g., Satish) who should do it?

--Richard

On Wed, Aug 16, 2017 at 3:40 PM, Richard Tran Mills <rtmills at anl.gov> wrote:

> Satish et al.,
>
> I noticed a failure when testing with SeqAIJPERM and have made a small fix
> that is on branch rmills/fix-seqaijperm_seqmaij. If such a small fix needs
> to be tested in next, then please merge to next for testing. Otherwise,
> please merge to master.
>
> Because it is so small, for convenience I have appended the git commit.
>
> --Richard
>
> rmills at Richards-MBP:tests(rmills/fix-seqaijperm_seqmaij=
> )arch-osx-clang-gfortran7-debug$ git show
> commit 1fde5765381347da5003c8b2045311701bd352b4 (HEAD ->
> rmills/fix-seqaijperm_seqmaij, origin/rmills/fix-seqaijperm_seqmaij)
> Author: Richard Tran Mills <rmills at rmills.org>
> Date:   Wed Aug 16 15:32:10 2017 -0700
>
>     Associated function for MatPtAP_seqaijperm_seqmaij_C.
>
> diff --git a/src/mat/impls/maij/maij.c b/src/mat/impls/maij/maij.c
> index 83a4c181fe..b2c1711bd5 100644
> --- a/src/mat/impls/maij/maij.c
> +++ b/src/mat/impls/maij/maij.c
> @@ -96,6 +96,7 @@ PetscErrorCode MatDestroy_SeqMAIJ(Mat A)
>    ierr = PetscFree(A->data);CHKERRQ(ierr);
>    ierr = PetscObjectComposeFunction((PetscObject)A,"MatConvert_
> seqmaij_seqaij_C",NULL);CHKERRQ(ierr);
>    ierr = PetscObjectComposeFunction((PetscObject)A,"MatPtAP_seqaij_
> seqmaij_C",NULL);CHKERRQ(ierr);
> +  ierr = PetscObjectComposeFunction((PetscObject)A,"MatPtAP_
> seqaijperm_seqmaij_C",NULL);CHKERRQ(ierr);
>    PetscFunctionReturn(0);
>  }
>
> @@ -3377,6 +3378,7 @@ PetscErrorCode  MatCreateMAIJ(Mat A,PetscInt dof,Mat
> *maij)
>        }
>        ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_
> seqmaij_seqaij_C",MatConvert_SeqMAIJ_SeqAIJ);CHKERRQ(ierr);
>        ierr = PetscObjectComposeFunction((PetscObject)B,"MatPtAP_seqaij_
> seqmaij_C",MatPtAP_SeqAIJ_SeqMAIJ);CHKERRQ(ierr);
> +      ierr = PetscObjectComposeFunction((PetscObject)B,"MatPtAP_
> seqaijperm_seqmaij_C",MatPtAP_SeqAIJ_SeqMAIJ);CHKERRQ(ierr);
>      } else {
>        Mat_MPIAIJ  *mpiaij = (Mat_MPIAIJ*)A->data;
>        Mat_MPIMAIJ *b;
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20170823/4724db97/attachment.html>