[Mpi-forum] MPI "Allocate receive" proposal

Rusty Lusk lusk at mcs.anl.gov
Mon Aug 26 10:27:05 CDT 2013

It's not quite the same.  If your buffer allocation (via malloc) after a probe fails, you can decide what to do (e.g. free some other memory and try again.)  The message isn't lost, and you may be in an unhappy state but not an undefined state.  I like the idea except for the introduction of the undefined result.  Perhaps this could be fixed.


On Monday,Aug 26, 2013, at 10:21 AM, Jeff Squyres (jsquyres) wrote:

> On Aug 26, 2013, at 11:14 AM, Pavan Balaji <balaji at mcs.anl.gov> wrote:
>> This should probably be moved to the point-to-point WG, but I'll comment on this thread for now.
> Is that WG active?  If so, we'd be happy to move it to there.
>> Such a proposal was brought up at the very early stage of the MPI-2.1 process (I don't remember by whom).  The main concern at that point was what happens when the memory allocation fails.  Will the receive fail? What happens to the sent message?
> I think Dave answered that well: it's just like any other resource exhaustion issue with a plain vanilla MPI receive.
> -- 
> Jeff Squyres
> jsquyres at cisco.com
> For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/
> _______________________________________________
> mpi-forum mailing list
> mpi-forum at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

More information about the mpi-forum mailing list