[Mpi-forum] MPI Count proposal from today's meeting

Supalov, Alexander alexander.supalov at intel.com
Fri Jun 18 03:51:41 CDT 2010

Thanks. I agree that use of the MPI_UNDEFINED by the MPI_GET_COUNT may need additional review. Why do you think the ranks are "short int" entities?

-----Original Message-----
From: mpi-forum-bounces at lists.mpi-forum.org [mailto:mpi-forum-bounces at lists.mpi-forum.org] On Behalf Of N.M. Maclaren
Sent: Friday, June 18, 2010 9:30 AM
To: Main MPI Forum mailing list
Subject: Re: [Mpi-forum] MPI Count proposal from today's meeting

On Jun 17 2010, N.M. Maclaren wrote:
>On Jun 17 2010, Supalov, Alexander wrote:
>>Thanks. Why signed?
>    It is useful to be able to specify invalid or exceptional values as
>out-of-band ones, and that isn't possible for any positive integer.  The
>standard states that MPI_UNDEFINED is negative in one place, and that
>can be returned from MPI_Get_count.

Sorry about following up to myself, but I though about it later, and
this is a mathematically solid proof that either MPI_Count must be
signed or the specification of MPI_Get_count (at least) needs changing.

MPI_UNDEFINED has to be a value that will fit in a 'short' integer, as
it is heavily used for undefined ranks.  But ALL non-negative 'short'
integer values are also valid counts when using 'long' integers for
those, and it has to be something that can't be a valid count.  Q.E.D.

Nick Maclaren.

mpi-forum mailing list
mpi-forum at lists.mpi-forum.org
Intel GmbH
Dornacher Strasse 1
85622 Feldkirchen/Muenchen Germany
Sitz der Gesellschaft: Feldkirchen bei Muenchen
Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer
Registergericht: Muenchen HRB 47456 Ust.-IdNr.
VAT Registration No.: DE129385895
Citibank Frankfurt (BLZ 502 109 00) 600119052

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.

More information about the mpi-forum mailing list