On Wed, Oct 17, 2012 at 4:31 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:
> Well if it always happens with OpenMPI then you should just turn off > one-sided for Open MPI? Sure, but we have to remember to remove the check when/if they fix the bug. It's kinda like deciding whether something will work based on the current username or name of a compiler. It's not what we usually do which is to test for features. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20121017/fa4aa11a/attachment.html>
