- mpiwg-sessions - mailman3.cels.anl.gov

Re: [mpiwg-sessions] [EXTERNAL] RE: Looks like we do have a topic to discuss soon
by Pritchard Jr., Howard 18 Jul '24

18 Jul '24

https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/9… From: mpiwg-sessions <mpiwg-sessions-bounces(a)lists.mpi-forum.org> on behalf of MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Reply-To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Date: Friday, July 12, 2024 at 11:26 AM To: "Holmes, Daniel John" <daniel.john.holmes(a)intel.com>, MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Cc: "Pritchard Jr., Howard" <howardp(a)lanl.gov> Subject: Re: [mpiwg-sessions] [EXTERNAL] RE: Looks like we do have a topic to discuss soon Hi Dan, I think we should add “This procedure is local” as you suggest. I’ll open an issue/pr pair maybe today or on Monday. Thanks. Howard From: "Holmes, Daniel John" <daniel. john. holmes@ intel. com> Date: Friday, July 12, 2024 ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd Hi Dan, I think we should add “This procedure is local” as you suggest. I’ll open an issue/pr pair maybe today or on Monday. Thanks. Howard From: "Holmes, Daniel John" <daniel.john.holmes(a)intel.com> Date: Friday, July 12, 2024 at 9:53 AM To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Cc: "Pritchard Jr., Howard" <howardp(a)lanl.gov> Subject: [EXTERNAL] RE: Looks like we do have a topic to discuss soon Hi Howard, The MPI_SESSION_INIT procedure is listed as “l” for local and (2) meaning not related to an MPI operation in the huge semantics table in Appendix A (MPI-4.1, p 881, line 29). This is correct but quite well hidden. Perhaps we could add “This procedure is local.” in normative text just after the binding? We do elucidate the point, using an example (Example 11.9) where two MPI objects (communicators) are created using the same session in one MPI process but in different sessions in another MPI process. This example cannot be correct unless the creation/instantiation of a session is NOT collective and does NOT synchronise. We could also include one or more pictures, like the ones we’ve used in the WG when talking about bubbles. The text art in Example 11.9 is not the most pleasing visual expression of the concept. Best wishes, Dan. From: mpiwg-sessions <mpiwg-sessions-bounces(a)lists.mpi-forum.org> On Behalf Of Pritchard Jr., Howard via mpiwg-sessions Sent: Friday, July 12, 2024 4:32 PM To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Cc: Pritchard Jr., Howard <howardp(a)lanl.gov> Subject: [mpiwg-sessions] Looks like we do have a topic to discuss soon HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4. 2. There seems to be more ambiguity about what is implied by the wording in section 11. 3 concerning MPI_SESSION_INIT – Session instantiation ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4.2. There seems to be more ambiguity about what is implied by the wording in section 11.3 concerning MPI_SESSION_INIT – Session instantiation is intended to be a lightweight operation than many of us may have realized. We don’t explicitly state whether this call is in any sense synchronizing with other processes. We’ve talked about this so long, for instance in terms of asymmetry with MPI_SESSION_FINALIZE, that we didn’t consider that an implementor may construe that MPI_SESSION_INIT has the same implicit synchronization behavior as MPI_INIT. We may want to add something in an errata to clarify what lightweight means. Again I will not be able to call in on Monday but perhaps if others want to discuss? The meeting will be automatically recorded BTW. Thanks, Howar ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_4180174387]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/www.instagram.…>[signature_937381201]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/twitter.com/Lo…>[signature_35314574]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/www.linkedin.c…>[signature_2948438747]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/www.facebook.c…>

1 0

Re: [mpiwg-sessions] [EXTERNAL] RE: Looks like we do have a topic to discuss soon
by Pritchard Jr., Howard 12 Jul '24

12 Jul '24

Hi Dan, I think we should add “This procedure is local” as you suggest. I’ll open an issue/pr pair maybe today or on Monday. Thanks. Howard From: "Holmes, Daniel John" <daniel.john.holmes(a)intel.com> Date: Friday, July 12, 2024 at 9:53 AM To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Cc: "Pritchard Jr., Howard" <howardp(a)lanl.gov> Subject: [EXTERNAL] RE: Looks like we do have a topic to discuss soon Hi Howard, The MPI_SESSION_INIT procedure is listed as “l” for local and (2) meaning not related to an MPI operation in the huge semantics table in Appendix A (MPI-4.1, p 881, line 29). This is correct but quite well hidden. Perhaps we could add “This procedure is local.” in normative text just after the binding? We do elucidate the point, using an example (Example 11.9) where two MPI objects (communicators) are created using the same session in one MPI process but in different sessions in another MPI process. This example cannot be correct unless the creation/instantiation of a session is NOT collective and does NOT synchronise. We could also include one or more pictures, like the ones we’ve used in the WG when talking about bubbles. The text art in Example 11.9 is not the most pleasing visual expression of the concept. Best wishes, Dan. From: mpiwg-sessions <mpiwg-sessions-bounces(a)lists.mpi-forum.org> On Behalf Of Pritchard Jr., Howard via mpiwg-sessions Sent: Friday, July 12, 2024 4:32 PM To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Cc: Pritchard Jr., Howard <howardp(a)lanl.gov> Subject: [mpiwg-sessions] Looks like we do have a topic to discuss soon HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4. 2. There seems to be more ambiguity about what is implied by the wording in section 11. 3 concerning MPI_SESSION_INIT – Session instantiation ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4.2. There seems to be more ambiguity about what is implied by the wording in section 11.3 concerning MPI_SESSION_INIT – Session instantiation is intended to be a lightweight operation than many of us may have realized. We don’t explicitly state whether this call is in any sense synchronizing with other processes. We’ve talked about this so long, for instance in terms of asymmetry with MPI_SESSION_FINALIZE, that we didn’t consider that an implementor may construe that MPI_SESSION_INIT has the same implicit synchronization behavior as MPI_INIT. We may want to add something in an errata to clarify what lightweight means. Again I will not be able to call in on Monday but perhaps if others want to discuss? The meeting will be automatically recorded BTW. Thanks, Howar ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_4180174387]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/www.instagram.…>[signature_937381201]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/twitter.com/Lo…>[signature_35314574]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/www.linkedin.c…>[signature_2948438747]<https://urldefense.com/v3/__https:/urldefense.us/v3/__https:/www.facebook.c…>

1 0

Re: [mpiwg-sessions] Looks like we do have a topic to discuss soon
by Holmes, Daniel John 12 Jul '24

12 Jul '24

Hi Howard, The MPI_SESSION_INIT procedure is listed as “l” for local and (2) meaning not related to an MPI operation in the huge semantics table in Appendix A (MPI-4.1, p 881, line 29). This is correct but quite well hidden. Perhaps we could add “This procedure is local.” in normative text just after the binding? We do elucidate the point, using an example (Example 11.9) where two MPI objects (communicators) are created using the same session in one MPI process but in different sessions in another MPI process. This example cannot be correct unless the creation/instantiation of a session is NOT collective and does NOT synchronise. We could also include one or more pictures, like the ones we’ve used in the WG when talking about bubbles. The text art in Example 11.9 is not the most pleasing visual expression of the concept. Best wishes, Dan. From: mpiwg-sessions <mpiwg-sessions-bounces(a)lists.mpi-forum.org> On Behalf Of Pritchard Jr., Howard via mpiwg-sessions Sent: Friday, July 12, 2024 4:32 PM To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Cc: Pritchard Jr., Howard <howardp(a)lanl.gov> Subject: [mpiwg-sessions] Looks like we do have a topic to discuss soon HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4. 2. There seems to be more ambiguity about what is implied by the wording in section 11. 3 concerning MPI_SESSION_INIT – Session instantiation ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization. ZjQcmQRYFpfptBannerEnd HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4.2. There seems to be more ambiguity about what is implied by the wording in section 11.3 concerning MPI_SESSION_INIT – Session instantiation is intended to be a lightweight operation than many of us may have realized. We don’t explicitly state whether this call is in any sense synchronizing with other processes. We’ve talked about this so long, for instance in terms of asymmetry with MPI_SESSION_FINALIZE, that we didn’t consider that an implementor may construe that MPI_SESSION_INIT has the same implicit synchronization behavior as MPI_INIT. We may want to add something in an errata to clarify what lightweight means. Again I will not be able to call in on Monday but perhaps if others want to discuss? The meeting will be automatically recorded BTW. Thanks, Howar ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_4180174387]<https://urldefense.us/v3/__https:/www.instagram.com/losalamosnatlab/__;!!G_…>[signature_937381201]<https://urldefense.us/v3/__https:/twitter.com/LosAlamosNatLab__;!!G_uCfscf7…>[signature_35314574]<https://urldefense.us/v3/__https:/www.linkedin.com/company/los-alamos-natio…>[signature_2948438747]<https://urldefense.us/v3/__https:/www.facebook.com/LosAlamosNationalLab/__;…>

1 0

Looks like we do have a topic to discuss soon
by Pritchard Jr., Howard 12 Jul '24

12 Jul '24

HI Everyone, Well it looks like we have a topic we do need to discuss soon, possibly for an errata item for 4.2. There seems to be more ambiguity about what is implied by the wording in section 11.3 concerning MPI_SESSION_INIT – Session instantiation is intended to be a lightweight operation than many of us may have realized. We don’t explicitly state whether this call is in any sense synchronizing with other processes. We’ve talked about this so long, for instance in terms of asymmetry with MPI_SESSION_FINALIZE, that we didn’t consider that an implementor may construe that MPI_SESSION_INIT has the same implicit synchronization behavior as MPI_INIT. We may want to add something in an errata to clarify what lightweight means. Again I will not be able to call in on Monday but perhaps if others want to discuss? The meeting will be automatically recorded BTW. Thanks, Howar ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_4180174387]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_937381201]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_35314574]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_2948438747]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0

Updated meeting link for Sessions working group
by Pritchard Jr., Howard 11 Jul '24

11 Jul '24

HI Everyone, A couple of things (well actually three). First, we are switching to Univ. Tenn. Knoxville Zoom link to make recordings easier (LANL has too many restrictions). Second, we are going to try meeting an hour earlier at 8 AM US central time starting this coming Monday 15th July 2024. Third, I will not be able to attend next week. I’ve updated the meeting time and info at https://urldefense.us/v3/__https://github.com/mpiwg-sessions/mpi-standard/w… Remember we meet fortnightly now not every week. Thanks Aurelien for helping to set up the Zoom! Howard ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_1995070345]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_3022563710]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_634360932]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_186724739]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0

Topics for Sessions WG meeting 1 July '24
by Pritchard Jr., Howard 26 Jun '24

26 Jun '24

Hi All, We will meet this coming Monday 1 July ’24. Topics I have to discuss * Follow-up from MPI Forum (not me, I was out but hopefully others attended) * Revisit meeting time * Localization options for session finalize functionality Let me know if you have other topics you’d like to discuss so I can get them on the agenda. Thanks, Howard ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_3803821726]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_3698530431]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_99779366]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_3271831942]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0

Sessions discussions for MPI forum next week
by Pritchard Jr., Howard 12 Jun '24

12 Jun '24

Hi All, It appears there’s almost nothing on the agenda for the Forum next week except for one vote and an MPI ABI discussion so there’s been a request for WG discussions. If anyone wants to set aside a block of time for WG discussions, please let Wes or Martin Schultz know. One item that might be useful to discuss is issue https://urldefense.us/v3/__https://github.com/mpiwg-sessions/sessions-issue… (need local variant of MPI_Session_finalize). Please update the issue if there are discussions about it next week. I will be on vacation on Monday and Tuesday and likely will not be able to attend on Wednesday. Thanks, Howard ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_2162946050]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_3523414902]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_2028895216]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_4076210444]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0

Cancel 6/17 WG meeting owing to MPI Forum
by Pritchard Jr., Howard 10 Jun '24

10 Jun '24

Hi All, Next week is the MPI Forum so canceling our scheduled WG meeting for Monday 17 June. Howard ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_791011986]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_1825484749]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_3540433980]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_2684012162]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0

Sessions WG webex - 3rd June '24
by Pritchard Jr., Howard 31 May '24

31 May '24

Hi All, I will not be able to attend next Monday’s Sessions Webex owing to medical appointment. However, please go ahead without me as I know a number of you wanted to discuss the ParTec approach to sessions restart. I did some updating of the Sessions wiki which I hope will be more readable/helpful. I added a section with recent publications by WG members relevant to Sessions, as well as a list of current “todos” (reflected in issues) for MPI 5.0. See https://urldefense.us/v3/__https://github.com/mpiwg-sessions/sessions-issue… Also, please feel free to open new issues at https://urldefense.us/v3/__https://github.com/mpiwg-sessions/sessions-issue… if there are Sessions related items you think we should try to resolve for MPI 5.0 or beyond! Thanks, Howard ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_2559021787]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_709264874]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_831542296]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_1220129132]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0

Re: [mpiwg-sessions] Topics for discussion at next sessions WG meeting
by Pritchard Jr., Howard 17 May '24

17 May '24

Hi All, I added a link to the ParTec presentation off of our wiki home page: https://urldefense.us/v3/__https://github.com/mpiwg-sessions/sessions-issue… See link under “current working documents” Howard From: "Pritchard Jr., Howard" <howardp(a)lanl.gov> Date: Thursday, May 16, 2024 at 12:38 PM To: MPI Sessions working group <mpiwg-sessions(a)lists.mpi-forum.org> Subject: Topics for discussion at next sessions WG meeting HI All, We have a sessions WG meeting scheduled for this coming Monday, May 20th. A couple of items are * Sanity check on PR - https://urldefense.us/v3/__https://github.com/mpi-forum/mpi-standard/pull/9… * Follow-up discussion of the sessions reinit concepts. Note this depends some on getting a copy of Sonja’s presentation. * Other ? Also, Monday is a holiday for some Europeans – Whit Sunday observed - so we might have low attendance. Howard ------- [signature_61897647] Howard Pritchard Research Scientist HPC-ENV Los Alamos National Laboratory howardp(a)lanl.gov<mailto:[email protected]> [signature_1080859930]<https://urldefense.us/v3/__https://www.instagram.com/losalamosnatlab/__;!!G… >[signature_678507511]<https://urldefense.us/v3/__https://twitter.com/LosAlamosNatLab__;!!G_uCfscf… >[signature_1311332218]<https://urldefense.us/v3/__https://www.linkedin.com/company/los-alamos-nati… >[signature_1733800376]<https://urldefense.us/v3/__https://www.facebook.com/LosAlamosNationalLab/__… >

1 0