I would certainly argue that the MPI_LOCK_ALL() question is not one that should derail the proposal. It works just fine for the purpose it was designed for. There are certainly less scalable ways that it COULD be used, but, then, how is that different from Alltoallv?

-----Original Message----- From: [email protected] [mailto:mpi3-rma- [email protected]] On Behalf Of Barrett, Brian W Sent: Tuesday, October 11, 2011 6:45 PM To: MPI 3.0 Remote Memory Access working group Subject: [Mpi3-rma] Reference implementation / October vote

All -

Given the agenda, have we officially punted on a vote in October? This seems bad to me unless we're going to start over; we have an implementation and our other requirements are met. If we're going to have a vote, we need to announce the implementation today/tomorrow. The chapter may not be perfect, but it's easily fixed by some very short additions in a follow-on ticket. Sorry I couldn't make the meeting Monday, but Torsten scheduled it while I was hosting a seminar speaker, which is one of those things I can't get out of...

Brian

-- Brian W. Barrett Dept. 1423: Scalable System Software Sandia National Laboratories

_______________________________________________ mpi3-rma mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma