Hi Jan, I think you identified the issue correctly in your last email, based on which symbols now appear in the libmpifort library compared to libfmpich in 3.1. I've made a ticket to make sure we fix this before the next release. http://trac.mpich.org/projects/mpich/ticket/2209 Ken On 12/03/2014 07:56 AM, Jan Bierbaum wrote:

On 28.11.2014 18:07, Jan Bierbaum wrote:

...

Is this a bug or are we missing something here? Any new insights? Did anybody manage to properly use the PMPI interface with a Fortran application and MPICH 3.1.3?

As a workaround I just use the Fortran wrapper from MPICH 3.1 now which still works fine with MPICH 3.1.3. But a proper fix would be better of course.

Regards, Jan _______________________________________________ discuss mailing list [email protected] To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

discuss mailing list [email protected] To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss