August 2023 - mpiwg-hybridpm - mailman3.cels.anl.gov

Meeting Canceled
by Jim Dinan 30 Aug '23

30 Aug '23

Hi All, As there is no topic for this week, let's cancel and if you can use the time for MPI 4.1 work. Please let me know if there are topics you'd like to schedule for any of our future meetings. Cheers, ~Jim.

1 0

Meeting Canceled
by Jim Dinan 16 Aug '23

16 Aug '23

Hi All, Sorry about the late notice. There's no meeting today. See you next week. Jim

1 0

Meeting Today
by Jim Dinan 09 Aug '23

09 Aug '23

Hi All, Edgar will lead a discussion today on MPI object handles and GPUs. Cheers, ~Jim.

1 0

Re: [mpiwg-hybridpm] Can MPI processes request different threading levels?
by Joachim Jenke 02 Aug '23

02 Aug '23

Thanks everyone for the lively discussion and pointing me to the relevant parts of the standard. We scanned the document to find the relevant text but could not find it. Based on the comments, I think we are good with our approach. The clarification for session init already answered our potential follow-up question :) As a background: we are currently integrating session support in MUST, so that all tool-internal communication executes in a separate session. As part of this effort, we move to MPMD mode and stop launching application code at the tool processes, which execute their run-loop in MPI_Init called from the application and never return to the application. To avoid deadlocking, if the application calls MPI_Init, we need to propagate this information to the tool processes and execute MPI_Init there as well. So we were wondering whether we should select the threading level appropriate for the tool processes or whether we need to replicate the call from the application processes. Our take-away is that we are fine with just calling MPI_Init, as the tool processes at the moment are purely single-threaded and won't use the world model anyways. Best Joachim Am 01.08.23 um 12:39 schrieb Holmes, Daniel John via mpiwg-hybridpm: > It is hinted at in the normative sentence: > "Different processes in MPI_COMM_WORLD may require different levels of thread support." > (MPI-4.1 section 11.2.1, page 492 lines 3-4) > > We must read "may require different" as "are permitted to set \mpiarg{required} to different" > > Note also that the thread support level can be requested via MPI_SESSION_INIT, which is a local procedure -- implying that the values given as INFO will apply only to the local process in the scope of the session. > > Best wishes, > Dan. > > -----Original Message----- > From: mpiwg-hybridpm <mpiwg-hybridpm-bounces(a)lists.mpi-forum.org> On Behalf Of Joseph Schuchart via mpiwg-hybridpm > Sent: Tuesday, August 1, 2023 11:24 AM > To: mpiwg-hybridpm(a)lists.mpi-forum.org > Cc: Joseph Schuchart <schuchart(a)icl.utk.edu> > Subject: Re: [mpiwg-hybridpm] Can MPI processes request different threading levels? > > Funny that this is inside an advice to implementors (which users like Joachim are not expected to read) when it really should be part of the normative text. > > Cheers > Joseph > > On 8/1/23 11:09, Jeff Hammond via mpiwg-hybridpm wrote: >> >> It is addressed here (in the text for MPI_INIT_THREAD): >> >> Advice to implementors. If provided is not MPI_THREAD_SINGLE then the >> MPI library should not invoke C or Fortran library calls that are not >> thread safe, e.g., in an environment where malloc is not thread safe, >> then malloc should not be used by the MPI library. >> >> Some implementors may want to use different MPI libraries for >> different levels of thread support. They can do so using dynamic >> linking and selecting which library will be linked when >> MPI_INIT_THREAD is invoked. If this is not possible, then >> optimizations for lower levels of thread support will occur only when >> the level of thread support required is specified at link time. >> >> *Note that required need not be the same value on all processes of >> MPI_COMM_WORLD.* (End of advice to implementors.) >> >> Jeff >> >> On Tue, Aug 1, 2023 at 11:47 AM Joachim Jenke via mpiwg-hybridpm >> <mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: >>> >>> Hello, >>> >>> I'm not sure whether this is the right group to ask, but it is >>> related to hybrid execution :) >>> >>> Must all MPI processes collectively call MPI_Init or MPI_Init_thread >>> - or can some processes call one and other processes call the other >> function? >>> >>> Also, can the processes request different thread-support levels? >>> >>> Our use case is MPMD execution, where some processes run >>> multi-threaded and other single-threaded. >>> >>> Thanks, >>> Joachim >>> >>> >>> -- >>> Dr. rer. nat. Joachim Jenke >>> >>> IT Center >>> Group: High Performance Computing >>> Division: Computational Science and Engineering RWTH Aachen >>> University Seffenter Weg 23 D 52074 Aachen (Germany) >>> Tel: +49 241 80- 24765 >>> Fax: +49 241 80-624765 >>> jenke(a)itc.rwth-aachen.de >>> www.itc.rwth-aachen.de <http://www.itc.rwth-aachen.de> >>> _______________________________________________ >>> mpiwg-hybridpm mailing list >>> mpiwg-hybridpm(a)lists.mpi-forum.org >>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm >> >> >> >> -- >> Jeff Hammond >> jeff.science(a)gmail.com >> http://jeffhammond.github.io/ >> >> _______________________________________________ >> mpiwg-hybridpm mailing list >> mpiwg-hybridpm(a)lists.mpi-forum.org >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm -- Dr. rer. nat. Joachim Jenke IT Center Group: High Performance Computing Division: Computational Science and Engineering RWTH Aachen University Seffenter Weg 23 D 52074 Aachen (Germany) Tel: +49 241 80- 24765 Fax: +49 241 80-624765 jenke(a)itc.rwth-aachen.de www.itc.rwth-aachen.de

1 0

Meeting Today
by Jim Dinan 02 Aug '23

02 Aug '23

Hi All, We will meet today and continue discussion on the Memory Allocation Kinds Side Document. Cheers, Jim

1 0

Re: [mpiwg-hybridpm] Can MPI processes request different threading levels?
by Jeff Hammond 01 Aug '23

01 Aug '23

Nothing prohibits a use case where MPI_Init + process-specific environment variables and/or MPI_Init_thread can initialize MPI with varying thread support, and I think the MPMD use case clearly demands it. Implementations are of course capable of (1) rounding up the thread level or (2) returning an inadequate value of provided, if a precise implementation of the application requirement cannot be satisfied. On Tue, Aug 1, 2023 at 1:39 PM Michael Knobloch via mpiwg-hybridpm < mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: > I'm not even sure it fully answers the question. This statement says > that MPI processes can request different requirements in > MPI_Init_thread, but it doesn't say whether all processes must be > initialized with a call to the same routine or if some can be > initialized with MPI_Init (and assume MPI_THREAD_SINGLE?) and some with > MPI_Init_thread, which was the original question. > > -Michael > > On 01.08.23 12:24, Joseph Schuchart via mpiwg-hybridpm wrote: > > Funny that this is inside an advice to implementors (which users like > > Joachim are not expected to read) when it really should be part of the > > normative text. > > > > Cheers > > Joseph > > > > On 8/1/23 11:09, Jeff Hammond via mpiwg-hybridpm wrote: > >> > >> It is addressed here (in the text for MPI_INIT_THREAD): > >> > >> Advice to implementors. If provided is not MPI_THREAD_SINGLE then the > >> MPI library should not invoke C or Fortran library calls that are not > >> thread safe, e.g., in an environment where malloc is not thread safe, > >> then malloc should not be used by the MPI library. > >> > >> Some implementors may want to use different MPI libraries for > >> different levels of thread support. They can do so using dynamic > >> linking and selecting which library will be linked when > >> MPI_INIT_THREAD is invoked. If this is not possible, then > >> optimizations for lower levels of thread support will occur only when > >> the level of thread support required is specified at link time. > >> > >> *Note that required need not be the same value on all processes of > >> MPI_COMM_WORLD.* (End of advice to implementors.) > >> > >> Jeff > >> > >> On Tue, Aug 1, 2023 at 11:47 AM Joachim Jenke via mpiwg-hybridpm > >> <mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: > >> > > >> > Hello, > >> > > >> > I'm not sure whether this is the right group to ask, but it is related > >> > to hybrid execution :) > >> > > >> > Must all MPI processes collectively call MPI_Init or MPI_Init_thread - > >> > or can some processes call one and other processes call the other > >> function? > >> > > >> > Also, can the processes request different thread-support levels? > >> > > >> > Our use case is MPMD execution, where some processes run > >> multi-threaded > >> > and other single-threaded. > >> > > >> > Thanks, > >> > Joachim > >> > > >> > > >> > -- > >> > Dr. rer. nat. Joachim Jenke > >> > > >> > IT Center > >> > Group: High Performance Computing > >> > Division: Computational Science and Engineering > >> > RWTH Aachen University > >> > Seffenter Weg 23 > >> > D 52074 Aachen (Germany) > >> > Tel: +49 241 80- 24765 > >> > Fax: +49 241 80-624765 > >> > jenke(a)itc.rwth-aachen.de > >> > www.itc.rwth-aachen.de <http://www.itc.rwth-aachen.de> > >> > _______________________________________________ > >> > mpiwg-hybridpm mailing list > >> > mpiwg-hybridpm(a)lists.mpi-forum.org > >> > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > >> > >> > >> > >> -- > >> Jeff Hammond > >> jeff.science(a)gmail.com > >> http://jeffhammond.github.io/ > >> > >> _______________________________________________ > >> mpiwg-hybridpm mailing list > >> mpiwg-hybridpm(a)lists.mpi-forum.org > >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > > > _______________________________________________ > > mpiwg-hybridpm mailing list > > mpiwg-hybridpm(a)lists.mpi-forum.org > > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > -- > Michael Knobloch > Juelich Supercomputing Centre (JSC) > Institute for Advanced Simulation (IAS) > Telefon: +49 2461 61-3546 > Telefax: +49 2461 61-6656 > > > > > ------------------------------------------------------------------------------------------------ > > ------------------------------------------------------------------------------------------------ > Forschungszentrum Juelich GmbH > 52425 Juelich > Sitz der Gesellschaft: Juelich > Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 > Vorsitzender des Aufsichtsrats: MinDir Stefan Müller > Geschaeftsfuehrung: Prof. Dr. Astrid Lambrecht (Vorsitzende), > Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens > > ------------------------------------------------------------------------------------------------ > > ------------------------------------------------------------------------------------------------ > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/

1 0

Re: [mpiwg-hybridpm] Can MPI processes request different threading levels?
by Holmes, Daniel John 01 Aug '23

01 Aug '23

MPI-4.0 section 11.2.1, page 488 lines 32-33 and 47-48: "When using the World Model, MPI is initialized by calling either MPI_INIT or MPI_INIT_THREAD." and "In the World Model, an MPI program must contain exactly one call to an MPI initialization routine: MPI_INIT or MPI_INIT_THREAD." Mix-and-match. Best wishes, Dan. -----Original Message----- From: mpiwg-hybridpm <mpiwg-hybridpm-bounces(a)lists.mpi-forum.org> On Behalf Of Michael Knobloch via mpiwg-hybridpm Sent: Tuesday, August 1, 2023 11:39 AM To: mpiwg-hybridpm(a)lists.mpi-forum.org Cc: Michael Knobloch <m.knobloch(a)fz-juelich.de> Subject: Re: [mpiwg-hybridpm] Can MPI processes request different threading levels? I'm not even sure it fully answers the question. This statement says that MPI processes can request different requirements in MPI_Init_thread, but it doesn't say whether all processes must be initialized with a call to the same routine or if some can be initialized with MPI_Init (and assume MPI_THREAD_SINGLE?) and some with MPI_Init_thread, which was the original question. -Michael On 01.08.23 12:24, Joseph Schuchart via mpiwg-hybridpm wrote: > Funny that this is inside an advice to implementors (which users like > Joachim are not expected to read) when it really should be part of the > normative text. > > Cheers > Joseph > > On 8/1/23 11:09, Jeff Hammond via mpiwg-hybridpm wrote: >> >> It is addressed here (in the text for MPI_INIT_THREAD): >> >> Advice to implementors. If provided is not MPI_THREAD_SINGLE then the >> MPI library should not invoke C or Fortran library calls that are not >> thread safe, e.g., in an environment where malloc is not thread safe, >> then malloc should not be used by the MPI library. >> >> Some implementors may want to use different MPI libraries for >> different levels of thread support. They can do so using dynamic >> linking and selecting which library will be linked when >> MPI_INIT_THREAD is invoked. If this is not possible, then >> optimizations for lower levels of thread support will occur only when >> the level of thread support required is specified at link time. >> >> *Note that required need not be the same value on all processes of >> MPI_COMM_WORLD.* (End of advice to implementors.) >> >> Jeff >> >> On Tue, Aug 1, 2023 at 11:47 AM Joachim Jenke via mpiwg-hybridpm >> <mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: >> > >> > Hello, >> > >> > I'm not sure whether this is the right group to ask, but it is >> > related to hybrid execution :) >> > >> > Must all MPI processes collectively call MPI_Init or >> > MPI_Init_thread - or can some processes call one and other >> > processes call the other >> function? >> > >> > Also, can the processes request different thread-support levels? >> > >> > Our use case is MPMD execution, where some processes run >> multi-threaded >> > and other single-threaded. >> > >> > Thanks, >> > Joachim >> > >> > >> > -- >> > Dr. rer. nat. Joachim Jenke >> > >> > IT Center >> > Group: High Performance Computing >> > Division: Computational Science and Engineering RWTH Aachen >> > University Seffenter Weg 23 D 52074 Aachen (Germany) >> > Tel: +49 241 80- 24765 >> > Fax: +49 241 80-624765 >> > jenke(a)itc.rwth-aachen.de >> > www.itc.rwth-aachen.de <http://www.itc.rwth-aachen.de> >> > _______________________________________________ >> > mpiwg-hybridpm mailing list >> > mpiwg-hybridpm(a)lists.mpi-forum.org >> > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm >> >> >> >> -- >> Jeff Hammond >> jeff.science(a)gmail.com >> http://jeffhammond.github.io/ >> >> _______________________________________________ >> mpiwg-hybridpm mailing list >> mpiwg-hybridpm(a)lists.mpi-forum.org >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm -- Michael Knobloch Juelich Supercomputing Centre (JSC) Institute for Advanced Simulation (IAS) Telefon: +49 2461 61-3546 Telefax: +49 2461 61-6656 ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Stefan Müller Geschaeftsfuehrung: Prof. Dr. Astrid Lambrecht (Vorsitzende), Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ _______________________________________________ mpiwg-hybridpm mailing list mpiwg-hybridpm(a)lists.mpi-forum.org https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm

1 0

Re: [mpiwg-hybridpm] Can MPI processes request different threading levels?
by Holmes, Daniel John 01 Aug '23

01 Aug '23

It is hinted at in the normative sentence: "Different processes in MPI_COMM_WORLD may require different levels of thread support." (MPI-4.1 section 11.2.1, page 492 lines 3-4) We must read "may require different" as "are permitted to set \mpiarg{required} to different" Note also that the thread support level can be requested via MPI_SESSION_INIT, which is a local procedure -- implying that the values given as INFO will apply only to the local process in the scope of the session. Best wishes, Dan. -----Original Message----- From: mpiwg-hybridpm <mpiwg-hybridpm-bounces(a)lists.mpi-forum.org> On Behalf Of Joseph Schuchart via mpiwg-hybridpm Sent: Tuesday, August 1, 2023 11:24 AM To: mpiwg-hybridpm(a)lists.mpi-forum.org Cc: Joseph Schuchart <schuchart(a)icl.utk.edu> Subject: Re: [mpiwg-hybridpm] Can MPI processes request different threading levels? Funny that this is inside an advice to implementors (which users like Joachim are not expected to read) when it really should be part of the normative text. Cheers Joseph On 8/1/23 11:09, Jeff Hammond via mpiwg-hybridpm wrote: > > It is addressed here (in the text for MPI_INIT_THREAD): > > Advice to implementors. If provided is not MPI_THREAD_SINGLE then the > MPI library should not invoke C or Fortran library calls that are not > thread safe, e.g., in an environment where malloc is not thread safe, > then malloc should not be used by the MPI library. > > Some implementors may want to use different MPI libraries for > different levels of thread support. They can do so using dynamic > linking and selecting which library will be linked when > MPI_INIT_THREAD is invoked. If this is not possible, then > optimizations for lower levels of thread support will occur only when > the level of thread support required is specified at link time. > > *Note that required need not be the same value on all processes of > MPI_COMM_WORLD.* (End of advice to implementors.) > > Jeff > > On Tue, Aug 1, 2023 at 11:47 AM Joachim Jenke via mpiwg-hybridpm > <mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: > > > > Hello, > > > > I'm not sure whether this is the right group to ask, but it is > > related to hybrid execution :) > > > > Must all MPI processes collectively call MPI_Init or MPI_Init_thread > > - or can some processes call one and other processes call the other > function? > > > > Also, can the processes request different thread-support levels? > > > > Our use case is MPMD execution, where some processes run > > multi-threaded and other single-threaded. > > > > Thanks, > > Joachim > > > > > > -- > > Dr. rer. nat. Joachim Jenke > > > > IT Center > > Group: High Performance Computing > > Division: Computational Science and Engineering RWTH Aachen > > University Seffenter Weg 23 D 52074 Aachen (Germany) > > Tel: +49 241 80- 24765 > > Fax: +49 241 80-624765 > > jenke(a)itc.rwth-aachen.de > > www.itc.rwth-aachen.de <http://www.itc.rwth-aachen.de> > > _______________________________________________ > > mpiwg-hybridpm mailing list > > mpiwg-hybridpm(a)lists.mpi-forum.org > > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > > > -- > Jeff Hammond > jeff.science(a)gmail.com > http://jeffhammond.github.io/ > > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm _______________________________________________ mpiwg-hybridpm mailing list mpiwg-hybridpm(a)lists.mpi-forum.org https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm

1 0

Re: [mpiwg-hybridpm] Can MPI processes request different threading levels?
by Michael Knobloch 01 Aug '23

01 Aug '23

I'm not even sure it fully answers the question. This statement says that MPI processes can request different requirements in MPI_Init_thread, but it doesn't say whether all processes must be initialized with a call to the same routine or if some can be initialized with MPI_Init (and assume MPI_THREAD_SINGLE?) and some with MPI_Init_thread, which was the original question. -Michael On 01.08.23 12:24, Joseph Schuchart via mpiwg-hybridpm wrote: > Funny that this is inside an advice to implementors (which users like > Joachim are not expected to read) when it really should be part of the > normative text. > > Cheers > Joseph > > On 8/1/23 11:09, Jeff Hammond via mpiwg-hybridpm wrote: >> >> It is addressed here (in the text for MPI_INIT_THREAD): >> >> Advice to implementors. If provided is not MPI_THREAD_SINGLE then the >> MPI library should not invoke C or Fortran library calls that are not >> thread safe, e.g., in an environment where malloc is not thread safe, >> then malloc should not be used by the MPI library. >> >> Some implementors may want to use different MPI libraries for >> different levels of thread support. They can do so using dynamic >> linking and selecting which library will be linked when >> MPI_INIT_THREAD is invoked. If this is not possible, then >> optimizations for lower levels of thread support will occur only when >> the level of thread support required is specified at link time. >> >> *Note that required need not be the same value on all processes of >> MPI_COMM_WORLD.* (End of advice to implementors.) >> >> Jeff >> >> On Tue, Aug 1, 2023 at 11:47 AM Joachim Jenke via mpiwg-hybridpm >> <mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: >> > >> > Hello, >> > >> > I'm not sure whether this is the right group to ask, but it is related >> > to hybrid execution :) >> > >> > Must all MPI processes collectively call MPI_Init or MPI_Init_thread - >> > or can some processes call one and other processes call the other >> function? >> > >> > Also, can the processes request different thread-support levels? >> > >> > Our use case is MPMD execution, where some processes run >> multi-threaded >> > and other single-threaded. >> > >> > Thanks, >> > Joachim >> > >> > >> > -- >> > Dr. rer. nat. Joachim Jenke >> > >> > IT Center >> > Group: High Performance Computing >> > Division: Computational Science and Engineering >> > RWTH Aachen University >> > Seffenter Weg 23 >> > D 52074 Aachen (Germany) >> > Tel: +49 241 80- 24765 >> > Fax: +49 241 80-624765 >> > jenke(a)itc.rwth-aachen.de >> > www.itc.rwth-aachen.de <http://www.itc.rwth-aachen.de> >> > _______________________________________________ >> > mpiwg-hybridpm mailing list >> > mpiwg-hybridpm(a)lists.mpi-forum.org >> > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm >> >> >> >> -- >> Jeff Hammond >> jeff.science(a)gmail.com >> http://jeffhammond.github.io/ >> >> _______________________________________________ >> mpiwg-hybridpm mailing list >> mpiwg-hybridpm(a)lists.mpi-forum.org >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm -- Michael Knobloch Juelich Supercomputing Centre (JSC) Institute for Advanced Simulation (IAS) Telefon: +49 2461 61-3546 Telefax: +49 2461 61-6656 ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Stefan Müller Geschaeftsfuehrung: Prof. Dr. Astrid Lambrecht (Vorsitzende), Karsten Beneke (stellv. Vorsitzender), Dr. Ir. Pieter Jansens ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------

1 0

Re: [mpiwg-hybridpm] Can MPI processes request different threading levels?
by Jeff Hammond 01 Aug '23

01 Aug '23

It's not normative because it's only telling implementers that they cannot assume something that is never written. The normative text permits the "required" argument to vary across processes, because it doesn't say it cannot. Jeff On Tue, Aug 1, 2023 at 1:24 PM Joseph Schuchart via mpiwg-hybridpm < mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: > Funny that this is inside an advice to implementors (which users like > Joachim are not expected to read) when it really should be part of the > normative text. > > Cheers > Joseph > > On 8/1/23 11:09, Jeff Hammond via mpiwg-hybridpm wrote: > > > > It is addressed here (in the text for MPI_INIT_THREAD): > > > > Advice to implementors. If provided is not MPI_THREAD_SINGLE then the > > MPI library should not invoke C or Fortran library calls that are not > > thread safe, e.g., in an environment where malloc is not thread safe, > > then malloc should not be used by the MPI library. > > > > Some implementors may want to use different MPI libraries for > > different levels of thread support. They can do so using dynamic > > linking and selecting which library will be linked when > > MPI_INIT_THREAD is invoked. If this is not possible, then > > optimizations for lower levels of thread support will occur only when > > the level of thread support required is specified at link time. > > > > *Note that required need not be the same value on all processes of > > MPI_COMM_WORLD.* (End of advice to implementors.) > > > > Jeff > > > > On Tue, Aug 1, 2023 at 11:47 AM Joachim Jenke via mpiwg-hybridpm > > <mpiwg-hybridpm(a)lists.mpi-forum.org> wrote: > > > > > > Hello, > > > > > > I'm not sure whether this is the right group to ask, but it is related > > > to hybrid execution :) > > > > > > Must all MPI processes collectively call MPI_Init or MPI_Init_thread - > > > or can some processes call one and other processes call the other > > function? > > > > > > Also, can the processes request different thread-support levels? > > > > > > Our use case is MPMD execution, where some processes run multi-threaded > > > and other single-threaded. > > > > > > Thanks, > > > Joachim > > > > > > > > > -- > > > Dr. rer. nat. Joachim Jenke > > > > > > IT Center > > > Group: High Performance Computing > > > Division: Computational Science and Engineering > > > RWTH Aachen University > > > Seffenter Weg 23 > > > D 52074 Aachen (Germany) > > > Tel: +49 241 80- 24765 > > > Fax: +49 241 80-624765 > > > jenke(a)itc.rwth-aachen.de > > > www.itc.rwth-aachen.de <http://www.itc.rwth-aachen.de> > > > _______________________________________________ > > > mpiwg-hybridpm mailing list > > > mpiwg-hybridpm(a)lists.mpi-forum.org > > > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > > > > > > > -- > > Jeff Hammond > > jeff.science(a)gmail.com > > http://jeffhammond.github.io/ > > > > _______________________________________________ > > mpiwg-hybridpm mailing list > > mpiwg-hybridpm(a)lists.mpi-forum.org > > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > > _______________________________________________ > mpiwg-hybridpm mailing list > mpiwg-hybridpm(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-hybridpm > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/

1 0