PMIx Monthly Meeting: Thursday, Sept. 9 (and other reminders)

10 views
Skip to first unread message

Josh Hursey

unread,
Sep 7, 2021, 9:55:59 AM9/7/21
to pmix-forum
We will have our next PMIx ASC Monthly Meeting on Thursday, Sept. 9, 2021 from 11-Noon US CT. 

The agenda is below:

I would like to spend a little time in this meeting reviewing the open tickets after we have working group updates.

Upcoming Important Dates:
 * PMIx ASC Monthly Meeting: Thursday, Sept. 9
 * PMIx ASC Officer Nominations Due: Tuesday, Sept. 21
 * PMIx ASC 4Q Agenda Items Due: Tuesday, Sept. 28
 * PMIx ASC 4Q Agenda Freeze: Tuesday, Oct. 12
 * PMIx ASC 2022 Dates: Tuesday, Oct. 12
   https://doodle.com/poll/88df7c6axymabkim
 * PMIx ASC 4Q Meeting: Tuesday, Oct. 26 & Thursday, Oct. 28


--
Josh Hursey
IBM Spectrum MPI Developer

Aurelien Bouteiller

unread,
Sep 9, 2021, 10:47:50 AM9/9/21
to Josh Hursey, pmix-forum
Josh I can’t attend I am following the MPI Forum meeting at that time

Best
Aurelien
> --
> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pmix-forum/0101017bc08b46ef-0eaaa83c-e0b8-4a05-831c-607e1deec0b2-000000%40us-west-2.amazonses.com.

Josh Hursey

unread,
Sep 9, 2021, 10:59:07 AM9/9/21
to pmix-forum
I'm also on the MPI Forum and plan to step out for a bit to run the PMIx meeting. Given that the MPI Forum is going to pull some folks away, we may try to keep today's PMIx meeting short.

I know we want to discuss two items today:
 - PMIx v4.1 release
 - Slices working group

We will push reviewing open issues until next month.


John DelSignore

unread,
Sep 9, 2021, 12:35:35 PM9/9/21
to Josh Hursey, pmix-forum, John DelSignore

As I mentioned on the call, having a publicly available, implementation-neutral PMIx header file that defined the “standard” PMIx API would be useful for tools that want to be PMIx implementation agnostic.

 

The OpenMP tools working group did this for OMPT and OMPD. The link to the OpenMP OMPT/OMPD header files is on GitHub here:

 

https://github.com/OpenMP/sources/tree/main/include

 

The omp-tools.h file defines the OMP 5.1 OMPT and OMPD APIs as described in the OMP Specification.

The ompd-types.h file defines implementation-specific values related to thread identifier kinds.

 

Cheers, John D.

To view this discussion on the web visit https://groups.google.com/d/msgid/pmix-forum/0101017bcb11c9fc-40c0f8d0-13d5-4b88-a972-8bfb808d971c-000000%40us-west-2.amazonses.com.

CAUTION: This email originated from outside of the organization. Do not click on links or open attachments unless you recognize the sender and know the content is safe.

 


This e-mail may contain information that is privileged or confidential. If you are not the intended recipient, please delete the e-mail and any attachments and notify us immediately.


Raffenetti, Kenneth J.

unread,
Sep 13, 2021, 3:35:07 PM9/13/21
to John DelSignore, Josh Hursey, pmix-forum
Is the goal a common ABI? Or would tools be looking for something more narrow? Poking around in the OpenPMIx headers, there's lots of struct definitions (and supporting macros) where any change could cause an ABI break.

Ken

On 9/9/21, 11:35 AM, "pmix-...@googlegroups.com on behalf of John DelSignore" <pmix-...@googlegroups.com on behalf of JDelS...@perforce.com> wrote:

As I mentioned on the call, having a publicly available, implementation-neutral PMIx header file that defined the “standard” PMIx API would be useful for tools that want to be PMIx implementation agnostic.

The OpenMP tools working group did this for OMPT and OMPD. The link to the OpenMP OMPT/OMPD header files is on GitHub here:

https://github.com/OpenMP/sources/tree/main/include

The omp-tools.h <https://github.com/OpenMP/sources/blob/main/include/omp-tools.h> file defines the OMP 5.1 OMPT and OMPD APIs as described in the OMP Specification.
The ompd-types.h <https://github.com/OpenMP/sources/blob/main/include/ompd-types.h> file defines implementation-specific values related to thread identifier kinds.
> https://doodle.com/poll/88df7c6axymabkim <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoodle.com%2Fpoll%2F88df7c6axymabkim&data=04%7C01%7Cjdelsignore%40perforce.com%7C7eba5dd2512448a550f508d973a25fc7%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637667964091908473%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=uA0k6FzQ6Z15JBtRlva9WN224ulxDLmINRY%2BPmRKDac%3D&reserved=0>
> * PMIx ASC 4Q Meeting: Tuesday, Oct. 26 & Thursday, Oct. 28
>
>
> --
> Josh Hursey
> IBM Spectrum MPI Developer
>
> --
> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com <mailto:pmix-forum%2Bunsu...@googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pmix-forum/0101017bc08b46ef-0eaaa83c-e0b8-4a05-831c-607e1deec0b2-000000%40us-west-2.amazonses.com <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F0101017bc08b46ef-0eaaa83c-e0b8-4a05-831c-607e1deec0b2-000000%2540us-west-2.amazonses.com&data=04%7C01%7Cjdelsignore%40perforce.com%7C7eba5dd2512448a550f508d973a25fc7%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637667964091918432%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=v3k%2FRy0JYmzLv55TUiB5tNiLjNstRf8G0tawHTyX%2BOo%3D&reserved=0>.

--
You received this message because you are subscribed to the Google Groups "pmix-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com <mailto:pmix-forum%2Bunsu...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/pmix-forum/9DF84BDF-0A59-4C25-B3C5-5B3A136C492C%40icl.utk.edu <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F9DF84BDF-0A59-4C25-B3C5-5B3A136C492C%2540icl.utk.edu&data=04%7C01%7Cjdelsignore%40perforce.com%7C7eba5dd2512448a550f508d973a25fc7%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637667964091918432%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=a5rGogof94OijudozMe%2F9DmJ7zNoePHZoDV1%2FkpT4C0%3D&reserved=0>.







--
Josh Hursey

IBM Spectrum MPI Developer



--
You received this message because you are subscribed to the Google Groups "pmix-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pmix-forum/0101017bcb11c9fc-40c0f8d0-13d5-4b88-a972-8bfb808d971c-000000%40us-west-2.amazonses.com <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F0101017bcb11c9fc-40c0f8d0-13d5-4b88-a972-8bfb808d971c-000000%2540us-west-2.amazonses.com%3Futm_medium%3Demail%26utm_source%3Dfooter&data=04%7C01%7Cjdelsignore%40perforce.com%7C7eba5dd2512448a550f508d973a25fc7%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637667964091928391%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=%2BzWeYI5f2Tjw%2FhtLAersA8fuWwLai80P7td%2F4ABNaAc%3D&reserved=0>.


CAUTION: This email originated from outside of the organization. Do not click on links or open attachments unless you recognize the sender and know the content is safe.




This e-mail may contain information that is privileged or confidential. If you are not the intended recipient, please delete the e-mail and any attachments and notify us immediately.

--
You received this message because you are subscribed to the Google Groups "pmix-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pmix-forum/MN2PR20MB34292EF70B1F3EEC9250066CC3D59%40MN2PR20MB3429.namprd20.prod.outlook.com <https://groups.google.com/d/msgid/pmix-forum/MN2PR20MB34292EF70B1F3EEC9250066CC3D59%40MN2PR20MB3429.namprd20.prod.outlook.com?utm_medium=email&utm_source=footer>.


John DelSignore

unread,
Sep 14, 2021, 4:02:31 PM9/14/21
to Ralph Castain, Raffenetti, Kenneth J., Josh Hursey, pmix-forum, John DelSignore
Hi Ralph and Ken,

Sorry for the late reply, I was on vacation yesterday.

Ralph is correct, a goal for TotalView is to be completely implementation independent, such that a single binary distribution of the debugger can run on a variety of system configurations. In many cases, it uses dlopen() to adapt to the system configuration at runtime. For example, for OMP-related debugging features, it is possible for the debugger to dynamically load whatever OMPD library exists for the OMP implementation and call into regardless of how the OMPD library is implemented. OMPD accomplished this by specifying every aspect of the OMPD API: the value of named constants, the data types of values, the layout of structures, API versioning support, etc. OMP took it one step further, which is to provide C-language header files that fully define the OMPD API.

To answer Ken's questions... I guess I'm looking for the same kind of thing from PMIx, for at least for the parts of the PMIx API that the tools use. However, I'm not sure how easy it would be to draw a line between the "tools portion" of the PMIx API and "the rest of it", because I suspect that a tool might want to be able to use any arbitrary part of the PMIx API. Given that, I'm not sure how narrow or wide it should be. Regardless, I think what I'm looking for is that enough of the PMIx API to be specified in enough detail, such that a PMIx tool and a PMIx library implementation can be ABI compatible on a given platform. The vision is that a tool (like TotalView) could be built with a PMIx implementation-neutral header file, dynamically load a PMIx (tool) library, call into the library, and get predictable results.

Cheers, John D.

-----Original Message-----
From: Ralph Castain <r...@pmix.org>
Sent: Monday, September 13, 2021 3:52 PM
To: Raffenetti, Kenneth J. <raff...@mcs.anl.gov>
Cc: John DelSignore <JDelS...@perforce.com>; Josh Hursey <jjhu...@open-mpi.org>; pmix-forum <pmix-...@googlegroups.com>
Subject: Re: [pmix-forum] PMIx Monthly Meeting: Thursday, Sept. 9 (and other reminders)

IIUC, John was looking for the Standard to specify the structures adequately (padding, ordering, etc.) to ensure ABI compatibility across implementations. Other changes we kicked around were to enumerate the value of all constants (done in the posted RFC) and to provide dlopen-friendly prototypes for all the APIs (which is what his requested header file would do).

I have some of those done in the "dlopen" branch of the MPIR shim (https://github.com/openpmix/mpir-to-pmix-guide/tree/dlopen) - easy enough to move those into a separate .h file and complete the set.
>> https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoodle.com%2Fpoll%2F88df7c6axymabkim&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263442964%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=RY1b6rXH1SjhwnwMRJViXjg4l%2B9AROaJoBG%2Fdb8aVRg%3D&amp;reserved=0 <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoodle.com%2Fpoll%2F88df7c6axymabkim&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263442964%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=RY1b6rXH1SjhwnwMRJViXjg4l%2B9AROaJoBG%2Fdb8aVRg%3D&amp;reserved=0>
>> * PMIx ASC 4Q Meeting: Tuesday, Oct. 26 & Thursday, Oct. 28
>>
>>
>> --
>> Josh Hursey
>> IBM Spectrum MPI Developer
>>
>> --
>> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com <mailto:pmix-forum%2Bunsu...@googlegroups.com>.
>> To view this discussion on the web visit https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F0101017bc08b46ef-0eaaa83c-e0b8-4a05-831c-607e1deec0b2-000000%2540us-west-2.amazonses.com&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263442964%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=RoPYxD2mCVhzAUjYYKelHF5GmTIekb09zyHfm3BQYl8%3D&amp;reserved=0 <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F0101017bc08b46ef-0eaaa83c-e0b8-4a05-831c-607e1deec0b2-000000%2540us-west-2.amazonses.com&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263442964%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=RoPYxD2mCVhzAUjYYKelHF5GmTIekb09zyHfm3BQYl8%3D&amp;reserved=0>.
>
> --
> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com <mailto:pmix-forum%2Bunsu...@googlegroups.com>.
> To view this discussion on the web visit https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F9DF84BDF-0A59-4C25-B3C5-5B3A136C492C%2540icl.utk.edu&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263442964%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=v1KlU6lZ5IZtTdoAlj%2FF0y7QG0yrP4QH5PiS8aoRWXA%3D&amp;reserved=0 <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F9DF84BDF-0A59-4C25-B3C5-5B3A136C492C%2540icl.utk.edu&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263452915%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=BSuOYUFktVdy2gO30AW6JWv79ImL%2FzjXjZi5LCIyq9E%3D&amp;reserved=0>.
>
>
>
>
>
>
>
> --
> Josh Hursey
>
> IBM Spectrum MPI Developer
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com.
> To view this discussion on the web visit https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F0101017bcb11c9fc-40c0f8d0-13d5-4b88-a972-8bfb808d971c-000000%2540us-west-2.amazonses.com&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263452915%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=9amJdf37g5jx1AMXexJW7Ri8dzqN7lpwI8rPuY8XvS4%3D&amp;reserved=0 <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F0101017bcb11c9fc-40c0f8d0-13d5-4b88-a972-8bfb808d971c-000000%2540us-west-2.amazonses.com%3Futm_medium%3Demail%26utm_source%3Dfooter&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263452915%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=evPYDAhufYlZ5xRwKuv8kEGh7%2F8cq07UotZACAIsU7w%3D&amp;reserved=0>.
>
>
> CAUTION: This email originated from outside of the organization. Do not click on links or open attachments unless you recognize the sender and know the content is safe.
>
>
>
>
> This e-mail may contain information that is privileged or confidential. If you are not the intended recipient, please delete the e-mail and any attachments and notify us immediately.
>
> --
> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com.
> To view this discussion on the web visit https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2FMN2PR20MB34292EF70B1F3EEC9250066CC3D59%2540MN2PR20MB3429.namprd20.prod.outlook.com&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263452915%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=aKHMxeTqUGiV8IeLsalg%2FM4rKBle6S%2FSUr5mHwY%2BiHA%3D&amp;reserved=0 <https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2FMN2PR20MB34292EF70B1F3EEC9250066CC3D59%2540MN2PR20MB3429.namprd20.prod.outlook.com%3Futm_medium%3Demail%26utm_source%3Dfooter&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263452915%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=i3m9ZPeoa%2F3AKBZZbboTx6raP9wDkl9GswFD%2Be6bwD4%3D&amp;reserved=0>.
>
>
> --
> You received this message because you are subscribed to the Google Groups "pmix-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pmix-forum+...@googlegroups.com.
> To view this discussion on the web visit https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fd%2Fmsgid%2Fpmix-forum%2F71BE1C8B-06D9-4FB4-9FA2-05CC9E77BBA1%2540mcs.anl.gov&amp;data=04%7C01%7CJDelSignore%40perforce.com%7C609e711841904d7deafe08d976efec45%7C95b666d19a7549ab95a38969fbcdc08c%7C0%7C0%7C637671595263452915%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=kHKATY1%2F0j41C81dLyhI3yK1RXFElL%2FiuXwpBS%2BIxt8%3D&amp;reserved=0.
Reply all
Reply to author
Forward
0 new messages