I believe this ticket exists because an implementer asked if any outcome was allowed, including the machine catching on fire. It seems useful to clarify what we mean by outcome, since we use this term throughout the text. ~Jim. On Fri, Feb 6, 2015 at 12:04 PM, Jeff Hammond <[email protected]> wrote:
I think the chapter committee should handle this. I think the current text is obvious enough that no one is going to screw it up, but that the clarification has value and thus should be implemented.
Best,
Jeff
On Fri, Feb 6, 2015 at 8:43 AM, Jim Dinan <[email protected]> wrote:
Hi All,
The following ticket is marked as MPI 3.0 errata, but appears not to have gone forward for a vote: https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/458
Did we either abandon this or decide that it should be handled by the chapter committee? It seems likely that it was the latter and I forgot to update the ticket. Just want to make sure it doesn't get lost in the shuffle.
Thanks, ~Jim.
---8<---
Here is the proposed change.
Existing Text (Pg. 418, line 44):
The outcome of concurrent conflicting accesses to the same memory locations is undefined...
Proposed Change
The resulting data values, or outcome, of concurrent conflicting accesses to the same memory locations is undefined...
_______________________________________________ mpiwg-rma mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-rma
-- Jeff Hammond [email protected] http://jeffhammond.github.io/ _______________________________________________ mpiwg-rma mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-rma