[petsc-dev] New MPICH

[Matthew Knepley]

On Tue, Jul 29, 2014 at 3:28 PM, Satish Balay <balay at mcs.anl.gov> wrote:

> What do you mean? SEGV at run time? something else?
>

No SEGV, valgrind clean, just failure codes from MPI that cannot be googled
(large integers).


> mpich used in petsc hasn't changed in a while.
>
> [and master nightlybuilds are fine.. (you'll have to look at next build
> logs though..)]
>
> I'll suggest a 'git clean -f -d -x' - and a proper  rebuild....


Will do. It only happens in optimized, so maybe its another Apple compiler
bug.

   Matt


>
> satish
>
>
> On Tue, 29 Jul 2014, Matthew Knepley wrote:
>
> > I reconfigured, which brought down a new MPICH, and now for optimized
> builds
> > on my Mac, all MPI_I*() functions die. Does anyone else have this?
> >
> >    Matt
> >
> >
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20140729/d7bcd83d/attachment.html>