- mpiwg-tools - mailman3.cels.anl.gov

Re: [mpiwg-tools] Questions for the tools community
by Marc-André Hermanns 17 Sep '19

17 Sep '19

Hi Jeff, sorry for the delay in responding to your email. As Martin also has some conflicts for this upcoming Thursday, we had the idea to push this to another week. How would that work for you? @Kathryn: Do you know from anyone at LLNL or other US sites who planned to come, who we need to inform if we shift to another week? I put Bill Williams (from TU Dresden) in CC, as he initially planned to join as well. If shifting the topic to a later date doesn't work out for you, you can certainly sync up with me, but I think first-hand information is always best in a discussion and it's always a pleasure to have you on the call ;-) Cheers, Marc-Andre On 16.09.19 15:05, Jeff Squyres (jsquyres) wrote: > Something changed and I'm sorry, I can't be there this Thursday. I'm sorry! > > Can I sync up with someone before Thursday to explain what I'm asking about? > > >> On Aug 29, 2019, at 11:25 AM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: >> >> Sure, I can be there. >> >> Thanks! >> >> >>> On Aug 29, 2019, at 11:05 AM, Marc-André Hermanns via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org> wrote: >>> >>> Hi Jeff, >>> >>> sorry about the silence. As Martin already said, your mail led to some >>> internal discussion and we wanted to reach out beyond the tools people >>> on the mailing list because it seems no-one on the list is currently >>> able to handle Fortran descriptors, but we failed to keep you informed. >>> >>> Would you be available on *Sep 19* for the tools call? That would give >>> us some time to contact some people outside the working group and >>> they'd have time to either give feedback or organize their schedules >>> in a way to attend the call. >>> >>> Cheers, >>> Marc-Andre >>> >>> On 29.08.19 16:53, Martin Schulz via mpiwg-tools wrote: >>>> Hi Jeff, >>>> >>>> Due to the summer break and vacation time, especially in Europe, we >>>> haven’t had a chance to address this - Marc-Andre had suggested a >>>> tools call on this for after EuroMPI and then we’ll know more (we need >>>> to get tool implementors to the table who are not part of the WG). >>>> >>>> Personally, though, I fear you are right with your assumption, but I >>>> am happy to be proven wrong. >>>> >>>> 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: schulzm(a)in.tum.de <mailto:[email protected]> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>>> On 29. Aug 2019, at 16:46, Jeff Squyres (jsquyres) via mpiwg-tools >>>>> <mpiwg-tools(a)lists.mpi-forum.org >>>>> <mailto:[email protected]>> wrote: >>>>> >>>>> Should I infer from the lack of reply that no tools properly handle >>>>> either Fortran descriptors or the descriptor-based symbols? >>>>> >>>>> >>>>> >>>>>> On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via >>>>>> mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org >>>>>> <mailto:[email protected]>> wrote: >>>>>> >>>>>> In Open MPI, we are finally working towards supporting Fortran >>>>>> descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). >>>>>> >>>>>> Per MPI-3.1 p613 table 17.1, this means ***changing all the >>>>>> back-end symbols*** that are used for PMPI-style interception. >>>>>> >>>>>> A concrete example: >>>>>> >>>>>> - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS >>>>>> - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS >>>>>> >>>>>> I.e., anywhere the application calls MPI_SEND, it will be >>>>>> transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the >>>>>> choice buffer passed will be a descriptor (which may therefore be a >>>>>> subarray, not a contiguous buffer), not a naked pointer. >>>>>> >>>>>> Here's my questions: >>>>>> >>>>>> 1. Do any tools support Fortran descriptors as choice buffers? >>>>>> 2. Do any tools intercept the symbols as described in MPI-3.1 table >>>>>> 17.1? >>>>>> >>>>>> -- >>>>>> Jeff Squyres >>>>>> jsquyres(a)cisco.com <mailto:[email protected]> >>>>>> >>>>>> _______________________________________________ >>>>>> mpiwg-tools mailing list >>>>>> mpiwg-tools(a)lists.mpi-forum.org >>>>>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >>>>> >>>>> >>>>> -- >>>>> Jeff Squyres >>>>> jsquyres(a)cisco.com <mailto:[email protected]> >>>>> >>>>> _______________________________________________ >>>>> mpiwg-tools mailing list >>>>> mpiwg-tools(a)lists.mpi-forum.org >>>>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >>>> >>>> >>>> _______________________________________________ >>>> mpiwg-tools mailing list >>>> mpiwg-tools(a)lists.mpi-forum.org >>>> 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 >>> hermanns(a)itc.rwth-aachen.de >>> www.itc.rwth-aachen.de >>> >>> _______________________________________________ >>> mpiwg-tools mailing list >>> mpiwg-tools(a)lists.mpi-forum.org >>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >> >> >> -- >> Jeff Squyres >> jsquyres(a)cisco.com >> > > -- 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 hermanns(a)itc.rwth-aachen.de www.itc.rwth-aachen.de

1 0

Re: [mpiwg-tools] Questions for the tools community
by Jeff Squyres (jsquyres) 16 Sep '19

16 Sep '19

Something changed and I'm sorry, I can't be there this Thursday. I'm sorry! Can I sync up with someone before Thursday to explain what I'm asking about? > On Aug 29, 2019, at 11:25 AM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: > > Sure, I can be there. > > Thanks! > > >> On Aug 29, 2019, at 11:05 AM, Marc-André Hermanns via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org> wrote: >> >> Hi Jeff, >> >> sorry about the silence. As Martin already said, your mail led to some >> internal discussion and we wanted to reach out beyond the tools people >> on the mailing list because it seems no-one on the list is currently >> able to handle Fortran descriptors, but we failed to keep you informed. >> >> Would you be available on *Sep 19* for the tools call? That would give >> us some time to contact some people outside the working group and >> they'd have time to either give feedback or organize their schedules >> in a way to attend the call. >> >> Cheers, >> Marc-Andre >> >> On 29.08.19 16:53, Martin Schulz via mpiwg-tools wrote: >>> Hi Jeff, >>> >>> Due to the summer break and vacation time, especially in Europe, we >>> haven’t had a chance to address this - Marc-Andre had suggested a >>> tools call on this for after EuroMPI and then we’ll know more (we need >>> to get tool implementors to the table who are not part of the WG). >>> >>> Personally, though, I fear you are right with your assumption, but I >>> am happy to be proven wrong. >>> >>> 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: schulzm(a)in.tum.de <mailto:[email protected]> >>> >>> >>> >>> >>> >>> >>> >>> >>>> On 29. Aug 2019, at 16:46, Jeff Squyres (jsquyres) via mpiwg-tools >>>> <mpiwg-tools(a)lists.mpi-forum.org >>>> <mailto:[email protected]>> wrote: >>>> >>>> Should I infer from the lack of reply that no tools properly handle >>>> either Fortran descriptors or the descriptor-based symbols? >>>> >>>> >>>> >>>>> On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via >>>>> mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org >>>>> <mailto:[email protected]>> wrote: >>>>> >>>>> In Open MPI, we are finally working towards supporting Fortran >>>>> descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). >>>>> >>>>> Per MPI-3.1 p613 table 17.1, this means ***changing all the >>>>> back-end symbols*** that are used for PMPI-style interception. >>>>> >>>>> A concrete example: >>>>> >>>>> - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS >>>>> - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS >>>>> >>>>> I.e., anywhere the application calls MPI_SEND, it will be >>>>> transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the >>>>> choice buffer passed will be a descriptor (which may therefore be a >>>>> subarray, not a contiguous buffer), not a naked pointer. >>>>> >>>>> Here's my questions: >>>>> >>>>> 1. Do any tools support Fortran descriptors as choice buffers? >>>>> 2. Do any tools intercept the symbols as described in MPI-3.1 table >>>>> 17.1? >>>>> >>>>> -- >>>>> Jeff Squyres >>>>> jsquyres(a)cisco.com <mailto:[email protected]> >>>>> >>>>> _______________________________________________ >>>>> mpiwg-tools mailing list >>>>> mpiwg-tools(a)lists.mpi-forum.org >>>>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >>>> >>>> >>>> -- >>>> Jeff Squyres >>>> jsquyres(a)cisco.com <mailto:[email protected]> >>>> >>>> _______________________________________________ >>>> mpiwg-tools mailing list >>>> mpiwg-tools(a)lists.mpi-forum.org >>>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >>> >>> >>> _______________________________________________ >>> mpiwg-tools mailing list >>> mpiwg-tools(a)lists.mpi-forum.org >>> 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 >> hermanns(a)itc.rwth-aachen.de >> www.itc.rwth-aachen.de >> >> _______________________________________________ >> mpiwg-tools mailing list >> mpiwg-tools(a)lists.mpi-forum.org >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools > > > -- > Jeff Squyres > jsquyres(a)cisco.com > -- Jeff Squyres jsquyres(a)cisco.com

1 0

Sep 12, 2019: No call today
by Marc-André Hermanns 12 Sep '19

12 Sep '19

Hi all, this is just a brief reminder that due to the on-going EuroMPI conference in Zürich we'll skip our regular call today. 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 hermanns(a)itc.rwth-aachen.de www.itc.rwth-aachen.de

1 0

Re: [mpiwg-tools] Questions for the tools community
by Jeff Squyres (jsquyres) 06 Sep '19

06 Sep '19

On Sep 6, 2019, at 6:56 AM, John DelSignore <John.DelSignore(a)roguewave.com<mailto:[email protected]>> wrote: TotalView definitely supports F90/F95 array and pointer variables, which under the covers are implemented as some sort of descriptor object. The descriptor formats vary from platform to platform and compiler to compiler, but usually a descriptor consists of an lower bound, upper bound, stride, pointer to the data, and perhaps flags. The debug info (DWARF, STABS, etc.) also varies, so it can be complicated, but we handle the debug information such that the user sees the right data. I'm not familiar with what you mean by "Fortran descriptors as choice buffers". I don't know much about Fortran 03/08 and I don't know if a "choice buffer" a Fortran thing or an MPI-specific Fortran thing. Googling the term seems to turn up hits that imply that it's an MPI/Fortran thing, but I'm not sure. A brief explanation would help. AFAICT, depending on what the compiler does with "Fortran descriptors as choice buffers", a tool that already supports Fortran descriptors might "just work", but without a concrete example I can't say for sure. Yes, I mean "choice buffers" in the MPI Forum sense of the word: a buffer that is the application's choice. For example, the send buffer in MPI_Send(...) is what we typically call a "choice buffer." In MPI-3.0, there's two kinds of Fortran support: a) if you support choice buffers as naked pointers (i.e., the way it has always been in MPI), or b) if you support the standardized Fortran descriptor (which includes things you mention: base, bound, stride, ...etc.). If the MPI implantation supports Fortran descriptors, applications can behave a little differently. Additionally, the MPI spec mandates that if an MPI implementation supports a), then it has to have back-end symbols A. But if an implementation supports b), then it has to have back-end symbols B (where A != B). This is because the treatment of the choice buffers in the back-end symbol for MPI_SEND differs between A and B. Since MPI-related tools have used interposition for years (e.g., to intercept calls by the application to MPI_Send and replace it with their own MPI_Send), and since very few MPI implementations have supported Fortran descriptors for choice buffers, most (all?) such tools are interposing symbols based on a). I am wondering if any interposition-type tools support b). I ask because Open MPI is investigating supporting Fortran descriptors, which means that we are looking at supporting b) + B. -- Jeff Squyres jsquyres(a)cisco.com<mailto:[email protected]>

1 0

Re: [mpiwg-tools] Questions for the tools community
by John DelSignore 06 Sep '19

06 Sep '19

Hi Jeff, I'm not sure I fully understand your question. TotalView definitely supports F90/F95 array and pointer variables, which under the covers are implemented as some sort of descriptor object. The descriptor formats vary from platform to platform and compiler to compiler, but usually a descriptor consists of an lower bound, upper bound, stride, pointer to the data, and perhaps flags. The debug info (DWARF, STABS, etc.) also varies, so it can be complicated, but we handle the debug information such that the user sees the right data. I'm not familiar with what you mean by "Fortran descriptors as choice buffers". I don't know much about Fortran 03/08 and I don't know if a "choice buffer" a Fortran thing or an MPI-specific Fortran thing. Googling the term seems to turn up hits that imply that it's an MPI/Fortran thing, but I'm not sure. A brief explanation would help. AFAICT, depending on what the compiler does with "Fortran descriptors as choice buffers", a tool that already supports Fortran descriptors might "just work", but without a concrete example I can't say for sure. To answer your second question, I'm not 100% sure what you mean by "intercept", but I'm fairly sure that TotalView does not intercept symbols as described in MPI-3.1 table 17.1. Cheers, John D. On 08/29/19 10:46, Jeff Squyres (jsquyres) via mpiwg-tools wrote: Should I infer from the lack of reply that no tools properly handle either Fortran descriptors or the descriptor-based symbols? On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org><mailto:[email protected]> wrote: In Open MPI, we are finally working towards supporting Fortran descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). Per MPI-3.1 p613 table 17.1, this means ***changing all the back-end symbols*** that are used for PMPI-style interception. A concrete example: - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS I.e., anywhere the application calls MPI_SEND, it will be transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the choice buffer passed will be a descriptor (which may therefore be a subarray, not a contiguous buffer), not a naked pointer. Here's my questions: 1. Do any tools support Fortran descriptors as choice buffers? 2. Do any tools intercept the symbols as described in MPI-3.1 table 17.1? -- Jeff Squyres jsquyres(a)cisco.com<mailto:[email protected]> _______________________________________________ mpiwg-tools mailing list mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools

1 0

Face-to-face WG time today / No regular call
by Marc-André Hermanns 05 Sep '19

05 Sep '19

Dear all, this is a short reminder that today is the first day of the face-to-face meeting at ETHZ and we're having some WG time in the afternoon. It starts at 1 p.m. UTC+1 (Zürich time) so unfortunately quite early for most people in the US. See details for the F2F agenda at: https://www.mpi-forum.org/meetings/2019/09/agenda There is a special WebEx setup for this. Please see the details at: https://github.com/mpi-forum/mpi-standard/wiki/MPI-Forum-Webex-Information There will be *NO* regular call with the regular Webex (as they'd overlap anyway). 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 hermanns(a)itc.rwth-aachen.de www.itc.rwth-aachen.de

1 0

Re: [mpiwg-tools] Questions for the tools community
by Jeff Squyres (jsquyres) 29 Aug '19

29 Aug '19

Sure, I can be there. Thanks! > On Aug 29, 2019, at 11:05 AM, Marc-André Hermanns via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org> wrote: > > Hi Jeff, > > sorry about the silence. As Martin already said, your mail led to some > internal discussion and we wanted to reach out beyond the tools people > on the mailing list because it seems no-one on the list is currently > able to handle Fortran descriptors, but we failed to keep you informed. > > Would you be available on *Sep 19* for the tools call? That would give > us some time to contact some people outside the working group and > they'd have time to either give feedback or organize their schedules > in a way to attend the call. > > Cheers, > Marc-Andre > > On 29.08.19 16:53, Martin Schulz via mpiwg-tools wrote: >> Hi Jeff, >> >> Due to the summer break and vacation time, especially in Europe, we >> haven’t had a chance to address this - Marc-Andre had suggested a >> tools call on this for after EuroMPI and then we’ll know more (we need >> to get tool implementors to the table who are not part of the WG). >> >> Personally, though, I fear you are right with your assumption, but I >> am happy to be proven wrong. >> >> 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: schulzm(a)in.tum.de <mailto:[email protected]> >> >> >> >> >> >> >> >> >>> On 29. Aug 2019, at 16:46, Jeff Squyres (jsquyres) via mpiwg-tools >>> <mpiwg-tools(a)lists.mpi-forum.org >>> <mailto:[email protected]>> wrote: >>> >>> Should I infer from the lack of reply that no tools properly handle >>> either Fortran descriptors or the descriptor-based symbols? >>> >>> >>> >>>> On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via >>>> mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org >>>> <mailto:[email protected]>> wrote: >>>> >>>> In Open MPI, we are finally working towards supporting Fortran >>>> descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). >>>> >>>> Per MPI-3.1 p613 table 17.1, this means ***changing all the >>>> back-end symbols*** that are used for PMPI-style interception. >>>> >>>> A concrete example: >>>> >>>> - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS >>>> - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS >>>> >>>> I.e., anywhere the application calls MPI_SEND, it will be >>>> transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the >>>> choice buffer passed will be a descriptor (which may therefore be a >>>> subarray, not a contiguous buffer), not a naked pointer. >>>> >>>> Here's my questions: >>>> >>>> 1. Do any tools support Fortran descriptors as choice buffers? >>>> 2. Do any tools intercept the symbols as described in MPI-3.1 table >>>> 17.1? >>>> >>>> -- >>>> Jeff Squyres >>>> jsquyres(a)cisco.com <mailto:[email protected]> >>>> >>>> _______________________________________________ >>>> mpiwg-tools mailing list >>>> mpiwg-tools(a)lists.mpi-forum.org >>>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >>> >>> >>> -- >>> Jeff Squyres >>> jsquyres(a)cisco.com <mailto:[email protected]> >>> >>> _______________________________________________ >>> mpiwg-tools mailing list >>> mpiwg-tools(a)lists.mpi-forum.org >>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >> >> >> _______________________________________________ >> mpiwg-tools mailing list >> mpiwg-tools(a)lists.mpi-forum.org >> 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 > hermanns(a)itc.rwth-aachen.de > www.itc.rwth-aachen.de > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools -- Jeff Squyres jsquyres(a)cisco.com

1 0

Re: [mpiwg-tools] Questions for the tools community
by Marc-André Hermanns 29 Aug '19

29 Aug '19

Hi Jeff, sorry about the silence. As Martin already said, your mail led to some internal discussion and we wanted to reach out beyond the tools people on the mailing list because it seems no-one on the list is currently able to handle Fortran descriptors, but we failed to keep you informed. Would you be available on *Sep 19* for the tools call? That would give us some time to contact some people outside the working group and they'd have time to either give feedback or organize their schedules in a way to attend the call. Cheers, Marc-Andre On 29.08.19 16:53, Martin Schulz via mpiwg-tools wrote: > Hi Jeff, > > Due to the summer break and vacation time, especially in Europe, we > haven’t had a chance to address this - Marc-Andre had suggested a > tools call on this for after EuroMPI and then we’ll know more (we need > to get tool implementors to the table who are not part of the WG). > > Personally, though, I fear you are right with your assumption, but I > am happy to be proven wrong. > > 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: schulzm(a)in.tum.de <mailto:[email protected]> > > > > > > > > >> On 29. Aug 2019, at 16:46, Jeff Squyres (jsquyres) via mpiwg-tools >> <mpiwg-tools(a)lists.mpi-forum.org >> <mailto:[email protected]>> wrote: >> >> Should I infer from the lack of reply that no tools properly handle >> either Fortran descriptors or the descriptor-based symbols? >> >> >> >>> On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via >>> mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org >>> <mailto:[email protected]>> wrote: >>> >>> In Open MPI, we are finally working towards supporting Fortran >>> descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). >>> >>> Per MPI-3.1 p613 table 17.1, this means ***changing all the >>> back-end symbols*** that are used for PMPI-style interception. >>> >>> A concrete example: >>> >>> - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS >>> - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS >>> >>> I.e., anywhere the application calls MPI_SEND, it will be >>> transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the >>> choice buffer passed will be a descriptor (which may therefore be a >>> subarray, not a contiguous buffer), not a naked pointer. >>> >>> Here's my questions: >>> >>> 1. Do any tools support Fortran descriptors as choice buffers? >>> 2. Do any tools intercept the symbols as described in MPI-3.1 table >>> 17.1? >>> >>> -- >>> Jeff Squyres >>> jsquyres(a)cisco.com <mailto:[email protected]> >>> >>> _______________________________________________ >>> mpiwg-tools mailing list >>> mpiwg-tools(a)lists.mpi-forum.org >>> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools >> >> >> -- >> Jeff Squyres >> jsquyres(a)cisco.com <mailto:[email protected]> >> >> _______________________________________________ >> mpiwg-tools mailing list >> mpiwg-tools(a)lists.mpi-forum.org >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools > > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > 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 hermanns(a)itc.rwth-aachen.de www.itc.rwth-aachen.de

1 0

Re: [mpiwg-tools] Questions for the tools community
by Jeff Squyres (jsquyres) 29 Aug '19

29 Aug '19

Gotcha. On Aug 29, 2019, at 10:53 AM, schulzm(a)in.tum.de<mailto:[email protected]> wrote: Hi Jeff, Due to the summer break and vacation time, especially in Europe, we haven’t had a chance to address this - Marc-Andre had suggested a tools call on this for after EuroMPI and then we’ll know more (we need to get tool implementors to the table who are not part of the WG). Personally, though, I fear you are right with your assumption, but I am happy to be proven wrong. 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: schulzm(a)in.tum.de<mailto:[email protected]> On 29. Aug 2019, at 16:46, Jeff Squyres (jsquyres) via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]>> wrote: Should I infer from the lack of reply that no tools properly handle either Fortran descriptors or the descriptor-based symbols? On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]>> wrote: In Open MPI, we are finally working towards supporting Fortran descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). Per MPI-3.1 p613 table 17.1, this means ***changing all the back-end symbols*** that are used for PMPI-style interception. A concrete example: - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS I.e., anywhere the application calls MPI_SEND, it will be transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the choice buffer passed will be a descriptor (which may therefore be a subarray, not a contiguous buffer), not a naked pointer. Here's my questions: 1. Do any tools support Fortran descriptors as choice buffers? 2. Do any tools intercept the symbols as described in MPI-3.1 table 17.1? -- Jeff Squyres jsquyres(a)cisco.com<mailto:[email protected]> _______________________________________________ mpiwg-tools mailing list mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools -- Jeff Squyres jsquyres(a)cisco.com<mailto:[email protected]> _______________________________________________ mpiwg-tools mailing list mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools -- Jeff Squyres jsquyres(a)cisco.com<mailto:[email protected]>

1 0

Re: [mpiwg-tools] Questions for the tools community
by schulzm＠in.tum.de 29 Aug '19

29 Aug '19

Hi Jeff, Due to the summer break and vacation time, especially in Europe, we haven’t had a chance to address this - Marc-Andre had suggested a tools call on this for after EuroMPI and then we’ll know more (we need to get tool implementors to the table who are not part of the WG). Personally, though, I fear you are right with your assumption, but I am happy to be proven wrong. 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: schulzm(a)in.tum.de > On 29. Aug 2019, at 16:46, Jeff Squyres (jsquyres) via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org> wrote: > > Should I infer from the lack of reply that no tools properly handle either Fortran descriptors or the descriptor-based symbols? > > > >> On Aug 25, 2019, at 11:31 AM, Jeff Squyres (jsquyres) via mpiwg-tools <mpiwg-tools(a)lists.mpi-forum.org> wrote: >> >> In Open MPI, we are finally working towards supporting Fortran descriptors for choice buffers (i.e., "TYPE(*), DIMENSION(..)"). >> >> Per MPI-3.1 p613 table 17.1, this means ***changing all the back-end symbols*** that are used for PMPI-style interception. >> >> A concrete example: >> >> - in mpif.h / the mpi module, MPI_SEND will become MPI_SEND_FTS >> - in the mpi_f08 module, MPI_SEND will become MPI_SEND_F08TS >> >> I.e., anywhere the application calls MPI_SEND, it will be transmogrified to either MPI_SEND_FTS or MPI_SEND_F08TS. And the choice buffer passed will be a descriptor (which may therefore be a subarray, not a contiguous buffer), not a naked pointer. >> >> Here's my questions: >> >> 1. Do any tools support Fortran descriptors as choice buffers? >> 2. Do any tools intercept the symbols as described in MPI-3.1 table 17.1? >> >> -- >> Jeff Squyres >> jsquyres(a)cisco.com >> >> _______________________________________________ >> mpiwg-tools mailing list >> mpiwg-tools(a)lists.mpi-forum.org >> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools > > > -- > Jeff Squyres > jsquyres(a)cisco.com > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-tools

1 0