[Beowulf] MPI ABI

Toon Knapen toon.knapen at fft.be
Mon Oct 10 03:32:30 PDT 2005

Joachim Worringen wrote:

> Wrong - i.e., the value of MPI_COMM_WORLD is not defined in Fortran,
> either. This won't work if one MPI implementation sets
> MPI_COMM_WORLD to 35 and another expects 626.
> Of course, you are right for opaque datatypes like MPI_Group, but this
> is not sufficient for an ABI.

If only the 'named constants' are a problem: It's very easy to create a
few functions (like 'int mpi_comm_world()') that will return the correct
value for a given implementation.

Any other alternative (cfr. MorphMPI that was discussed some months ago)
is much more complicated and finally I'm not convinced it provides any
added value in respect to the solution mentioned above.

(I've never used the fortran interface of MPI BTW, always the C
interface. But I was wondering why we do not have the same problem with
our calls to BLAS: well for BLAS we always call the fortran interface
from C and C++)


More information about the Beowulf mailing list