[Mpi3-rma] [EXTERNAL] Re: request-based ops

Pavan Balaji balaji at mcs.anl.gov
Fri Jun 21 08:40:41 CDT 2013

Sorry, I'm joining in a bit late in the discussion.  Not allowing 
operations between the 'I' variant of the function and the completion 
call should be OK.  If we later want to change that, it can easily be 
done in a backward compatible way.  Also, if someone really want to 
bypass that restriction, he/she can create overlapping windows.

  -- Pavan

On 06/20/2013 09:49 AM, Barrett, Brian W wrote:
> On 6/20/13 7:39 AM, "Jim Dinan" <james.dinan at gmail.com
> <mailto:james.dinan at gmail.com>> wrote:
>     I agree with the general statement that the nonblocking call marks
>     the end of the epoch.  In the case of MPI_Win_test, it's trying to
>     close an exposure epoch -- not an access epoch -- so communication
>     operations from the origin process are not being completed by
>     synchronization (although, the test/wait synchronization could be
>     waiting for a self communication, that operation is completed by
>     start/complete not post/wait).  I think an MPI_Win_iwait operation
>     that uses the semantic you described should be identical to
>     MPI_Win_test.  Unless I'm missing something, which never happens
>     when discussing RMA.  :)
> You're right; I always get the exposure / access sides of generalized
> active target mixed up.  With that clarification, I think that
> MPI_Win_iwait and MPI_Win_test could be made equivalent fairly easily.
> Brian
> --
>    Brian W. Barrett
>    Scalable System Software Group
>    Sandia National Laboratories
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma

Pavan Balaji

More information about the mpiwg-rma mailing list