...

Also for PVAR start, what if there are no handles associated with a session and MPI_T_PVAR_ALL_HANDLES is used? Do we return SUCCESS or ERR_PVAR_NO_STARTSTOP?

Hm. I think this is a separate ticket, yes? And I would think returning the error code would be more helpful than success. However, I can see the other side too, that success makes sense, since nothing erroneous actually happened.

Here is what the text says: If the constant MPI_T_PVAR_ALL_HANDLES is passed in handle, the MPI implementation attempts to start all variables within the session identified by the parameter session for which handles have been allocated. For me this reads as returning SUCCESS since all variables, which had handles allocated, have been reset. Granted, we could be more explicit. (the same applies to stop and reset as well) Martin

...

...

and 387

Looks good. :-)

...

due to lack of time (I blame Fortran!)

Long live Fortran 2028!

:-)

Kathryn

...

-- Jeff Squyres [email protected] For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

_______________________________________________ mpiwg-tools mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools

______________________________________________________________ Kathryn Mohror, [email protected], http://people.llnl.gov/mohror1 CASC @ Lawrence Livermore National Laboratory, Livermore, CA, USA

_______________________________________________ mpiwg-tools mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools