The standard already says: If a sender sends two messages in succession to the same destination, and both match the same receive, then this operation cannot receive the second message if the first one is still pending. If a receiver posts two receives in succession, and both match the same message, then the second receive operation cannot be satisfied by this message, if the first one is still pending. This requirement facilitates matching of sends to receives. It guarantees that message-passing code is deterministic, I think that already provides the appropriate guarantee. Two concurrent threads making MPI calls do not provide any enforcement of "?succession" and the standard points this out to be helpful. It should not really have been necessary to say this. Explicit temporal ordering of actions on otherwise concurrent threads by an application is common and it is not the job of the MPI standard to say how this is to be done. MPI should probably only say in any situation where "?succession" is ambiguous, MPI makes no guarantee of disambiguation. Where "succession" is clear cut, MPI will reflect the succession order. Dick Treumann - MPI Team IBM Systems & Technology Group Dept X2ZA / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601 Tele (845) 433-7846 Fax (845) 433-8363 mpi-22-bounces_at_[hidden] wrote on 03/27/2009 01:25:29 PM:

[image removed]

Re: [Mpi-22] Ordering guarantees and multiple threads

Greg Bronevetsky

to:

MPI 2.2, MPI 2.2

03/27/2009 01:27 PM

Sent by:

mpi-22-bounces_at_[hidden]

Please respond to "MPI 2.2"

At 09:35 AM 3/27/2009, Darius Buntinas wrote:

...

I've just created a ticket: https:// svn.mpi-forum.org/trac/mpi-forum-web/ticket/144 to suggest that we strengthen the ordering guarantees between sends (and receives) from different threads.

I think that the real question here is whether the potential benefit to the user outweighs the potential for optimization in a multithreaded implementation.

Please have a look at it and feel free to comment.

I like the spirit of this proposal but in practice it will be hard to specify. In particular, consider the sentence "two sends do not overlap if the first MPI_Send() returns before the second MPI_Send() is called". The word "before" here is only properly defined on a sequentially consistent memory. On any other memory model "before" is only defined by synchronization primitives supplied by the shared memory system. As such, I don't know how to be precise about something like "before" without talking about the shared memory system. This will be hard because we'll need to talk about concepts that all these systems share but for which they all provide different primitives and semantics. I don't know how the wider forum will react to such a specification. However, I agree that something like this is necessary.

Greg Bronevetsky Post-Doctoral Researcher 1028 Building 451 Lawrence Livermore National Lab (925) 424-5756 bronevetsky1_at_[hidden] http://greg.bronevetsky.com

_______________________________________________ mpi-22 mailing list mpi-22_at_[hidden] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-22

*