[petsc-dev] Bug in MatGetSubMatrix_SeqBAIJ

Barry Smith bsmith at mcs.anl.gov
Sun May 10 11:47:57 CDT 2015


   Better to make these reports a pull request even if you plan to merge it yourself rather than emailing a branch name; makes it easier for everyone else to track and see what you have done.

   If you intend the fix to go into maint you need to make the branch off of maint (not master) and then make the pull request off of maint.

  "is buggy" is also a bit vague a term. It would be better to explain at a high level what the bug was, looks like confusion between rows and column sizes or something.


> On May 10, 2015, at 6:51 AM, Stefano Zampini <stefano.zampini at gmail.com> wrote:
> The current code for MatGetSubMatrix_SeqBAIJ is buggy. I pushed a fix in stefano_zampini/matbaij-fixgetsubmatrix. If it is ok with you, than I'll merge to master (and maint?). 
> -- 
> Stefano

More information about the petsc-dev mailing list