What we don't have for proposal 1 yet is an implementation. Brian had indicated that he could create an interesting MPI_Win_alloc() implementation, but I don't know his timeline for doing that. Do we have other implementation volunteers? These seem easy, since it should be a copy and paste of code from unlock: MPI_Win_flush() MPI_Win_flushall() These seem a little harder, since they involve some actual new code: MPI_Win_lockall() MPI_Win_unlockall() MPI_Get_Accumulate() MPI_Compare_and_swap() Keith
-----Original Message----- From: [email protected] [mailto:mpi3-rma- [email protected]] On Behalf Of Rajeev Thakur Sent: Tuesday, May 25, 2010 3:22 PM To: [email protected] Subject: [Mpi3-rma] Proposal 2
There are less than 3 weeks to go before the next Forum meeting. Proposal 1 is more or less there, but we don't have a Proposal 2. We got a bit bogged down in allflushall, but other than that is there consensus on other topics to create a Proposal 2? Allflushall can be included in there and debated separately (with a time limit :-)).
Rajeev
_______________________________________________ mpi3-rma mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma