[Mpi3-rma] MPI-3 UNIFIED model clarification
james.dinan at gmail.com
Sat Aug 3 07:38:29 CDT 2013
An RMA origin process is only able to provide guarantees about what is done
by the network, not what is observed by a thread in the target process. My
feeling on this is that it's up to the user to ensure that their
architecture does what they want. If the user wants to deal with
replication, buffering, and reordering in the node in a portable way, they
should use window synchronization. If the user wants to leverage a
hardware feature, that's fine too. MPI doesn't define what should happen
if you don't synchronize at the target, so the latter case is not invalid
but also not portable.
On Fri, Aug 2, 2013 at 11:42 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:
> "Underwood, Keith D" <keith.d.underwood at intel.com> writes:
> > The point of UNIFIED was that we add a more useful model that worked
> > in *some* places.
> Keith, should the standard include an example of UNIFIED that is
> portable, but does not call MPI_Win_sync?
> If so, what memory model should be assumed? (C11 seems like a likely
> candidate, though I think the best quality documentation on these
> affairs is associated with the Linux kernel, and is readily implemented
> without language extensions.)
> mpi3-rma mailing list
> mpi3-rma at lists.mpi-forum.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the mpiwg-rma