Hi Rolf, thanks for double checking this and for the hard work you put into the preparation of 4.1. You might be right that this behavior may imply changes either in an implementation or in a tool (as any observable behavior will eventually be relied upon by someone). I would therefore agree to a changelog and have opened a PR here: https://github.com/mpi-forum/mpi-standard/pull/836 Cheers, Marc-Andre On 29.06.23 16:51, Rolf Rabenseifner wrote:
Dear Marc-Andre and all,
I just checked all MPI-4.1 issues and related pull requests, whether (1) they have, (2) they have not, but should have, (3) they should not have and have not a change-log entry.
Problematic is category (2).
I detected the following Issues / PRs and sorted in (1-3):
(3) #160 PR-161 ERRATA: Clarify behavior for invalid MPI_T binding arguments
So here it is (3) -> (2)
(1) #443 PR-754 LIS entries in MPI_T chapter not reflecting IN vs. INOUT as defined in Term 2.3 chapter (3) #542 (PR-646) PR-795 #T Correct usage of the terms error and return codes in the tools chapter (3) #601 PR-734 Tools - Code examples improperly marked (3) #622 PR-734 Tools - Export "outcome is as if..." code fragments as proper examples (3) #674 PR-796 Tools chapter uses non-standard wildcard-notation for function names
As far as I understand, only Issue #160 / PR-161 can be problematic:
Are you sure that #160 does not cause - that an implementor of an MPI library has to fix something, or - that a user of the tools interface has to modify something in her/his tool, - or a third party, like trainiers have to change there course material?
If there is no such risk, then it is okay to be quiet, i.e. not to provide a change-log entry.
I personally expect that there are no such risks and therefore the decision "no change-log" was correct.
What is your answer?
See above. The issue was triggered by a user binding to MPI_COMM_WORLD "too early" (i.e., before MPI initialization) and therefore at least that user now needs to change his/her tool. (Even though that might have been done quite some time ago). If others disagree to this assessment, we can easily bin the Changelog, if not, we should wordsmith the Changelog entry as necessary in the July meeting. Cheers, Marc-Andre
----- Original Message -----
From: "MPI Tool WG" <[email protected]> To: "MPI Tool WG" <[email protected]> Cc: "Marc-Andre Hermanns" <[email protected]> Sent: Thursday, June 29, 2023 1:44:57 PM Subject: [mpiwg-tools] No call today (June 29, 2023)
Dear all,
due to a conflict with another meeting, I would cancel today's meeting.
As a teaser for future meetings: Martin brought back feedback/discussion notes from the Scalable Tools Workshop in Tahoe with several topics for us to work on.
https://dyninst.github.io/scalable_tools_workshop/petascale2023/assets/slide...
Cheers, Marc-Andre -- Dr. rer. nat. Marc-André Hermanns
IT Center Group: High Performance Computing Division: Computational Science and Engineering RWTH Aachen University Seffenter Weg 23 52074 Aachen Phone: +49 241 80-24381 [email protected] www.itc.rwth-aachen.de
Social Media Kanäle des IT Centers: https://blog.rwth-aachen.de/itc/ https://www.facebook.com/itcenterrwth https://www.linkedin.com/company/itcenterrwth https://twitter.com/ITCenterRWTH https://www.youtube.com/channel/UCKKDJJukeRwO0LP-ac8x8rQ
_______________________________________________ mpiwg-tools mailing list [email protected] https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools
-- Dr. rer. nat. Marc-André Hermanns IT Center Group: High Performance Computing Division: Computational Science and Engineering RWTH Aachen University Seffenter Weg 23 52074 Aachen Phone: +49 241 80-24381 [email protected] www.itc.rwth-aachen.de Social Media Kanäle des IT Centers: https://blog.rwth-aachen.de/itc/ https://www.facebook.com/itcenterrwth https://www.linkedin.com/company/itcenterrwth https://twitter.com/ITCenterRWTH https://www.youtube.com/channel/UCKKDJJukeRwO0LP-ac8x8rQ