Hi all,
The Semantic Terms Working Group would like to request
- a re-reading of issue #96 "Semantic Terms"
- a no-no-vote for the changes since Albuquerque added for
-- removing changes to MPI-3.1 that were never intended:
- correct definition of local and non-local
- not defining service/inquiry procedures as blocking/nonblocking
-- handling special cases correctly (MPROBE/MRECV, IPROBE/PROBE)
- a full reading if the no-no-vote fails
- a first vote if the no-no-vote passes.
The terms chapter is missing many terms we would like to see added for MPI-4
* Issue: https://github.com/mpi-forum/mpi-issues/issues/96
* PR: https://github.com/mpi-forum/mpi-standard/pull/116
* PDF:
https://github.com/mpi-forum/mpi-issues/files/4145580/mpi-report-issue96-2020-02-02-annotated.pdf
Current status of #96:
- Had a complete reading at Albuquerque
- Is strongly needed to achieve consistency for
- nonblocking collectives (in since MPI-3.0 and inconsistent to semantic
terms)
- nonblocking collectives (voted in for MPI-4.0 but inconsistent to
semantic terms)
- Major problems from Albuquerque are resolved, see above.
- "Perfect is the enemy of good enough"
We hope that we are now really good enough for MPI-4.0 ;-)
And we know that we are not perfect, e.g., we opened doors for one-sided
or partitinoed communications, but not more.
Thanks and kind regards
Rolf
On 1/22/20 1:35 PM, Martin Schulz via mpi-forum wrote:
> (Sorry, some may get this a second time due address autocompletion problems)
>
> Hi all,
>
> The February MPI Forum meeting is getting closer. Logistics and
> registration is online at the usual place:
>
> https://www.mpi-forum.org/meetings/2020/02/logistics
>
> Please register as soon as you can.
>
> Note, due to the holiday on Monday, the meeting is this time from
> Tuesday 2pm to Friday 12:30.
>
> As a consequence, the cutoff for ballots, readings, etc. is Tuesday
> February 4th. Please send all announcements to the list. If there are
> any additional requests for plenary sessions, please let me know as
> well, with the connected WG chair in CC.
>
> Also, as we are wrapping up MPI 4.0 and I will (hopefully) not get many
> requests directly connected with MPI 4.0, I will try to make time for
> WGs again. Hence, WG-chairs, please send me any time requests for the
> forum meeting for WG discussions.
>
> Thanks!
>
> Martin
>
>
> —
> Prof. Dr. Martin Schulz, Chair of Computer Architecture and Parallel Systems
> Department of Informatics, TU-Munich, Boltzmannstraße 3, D-85748 Garching
> Member of the Board of Directors at the Leibniz Supercomputing Centre (LRZ)
> Email: schu...@in.tum.de <mailto:schu...@in.tum.de>
>
>
>
> _______________________________________________
> mpi-forum mailing list
> mpi-forum@lists.mpi-forum.org
> https://lists.mpi-forum.org/mailman/listinfo/mpi-forum
>
_______________________________________________
mpi-forum mailing list
mpi-forum@lists.mpi-forum.org
https://lists.mpi-forum.org/mailman/listinfo/mpi-forum
_______________________________________________
mpi-forum mailing list
mpi-forum@lists.mpi-forum.org
https://lists.mpi-forum.org/mailman/listinfo/mpi-forum
