December 2015 - mpiwg-p2p - mailman3.cels.anl.gov

Re: [mpiwg-p2p] Fwd: MPI P2P "messages may overtake" assertion
by Jim Dinan 15 Dec '15

15 Dec '15

It may be hard for tools or existing analysis techniques to provide assistance to users when this mode is selected. However, there is adequate support through the MPI interface to detect when this mode has been selected and restrict the tool's behavior -- i.e. a tool may have to ignore traffic on communicators where this info key is active. As Jeff pointed out, tools can even prevent the application from providing this info key to the library. I don't really see any necessary fix to the proposal. We could add an advice to users to remind them that they should ensure tools are compatible with the info keys. And the reverse advice to tools writers that they should check info keys for compatibility. ~Jim. On Mon, Dec 14, 2015 at 10:59 AM, Daniel Holmes <dholmes(a)epcc.ed.ac.uk> wrote: > Hi Jeff, > > ... tools can just intercept the info operations and strip this key if > they can't support it. > > Agreed: this should have been response 4 on my list (Markus and I > discussed that option too). > > Cheers, > Dan. > > > On 14/12/2015 15:54, Jeff Hammond wrote: > > Remember that one of the use cases of allow_overtaking is applications > that have exact matching, in which case allow_overtaking is a way of > turning off a feature that isn't used, in order to get a high-performing > message queue implementation. In the exact matching case, tools will have > no problem matching up sends and recvs. > > If tools cannot handle MPI_THREAD_MULTIPLE already, then I don't really > care if they can't support this assertion either. > > And in any case, such tools can just intercept the info operations and > strip this key if they can't support it. > > Jeff > > On Mon, Dec 14, 2015 at 7:44 AM, Daniel Holmes <dholmes(a)epcc.ed.ac.uk> > wrote: > >> Hi all, >> >> Markus has reminded me that we discussed the effect of the >> mpi_assert_allow_overtaking assertion on tools (see below). >> >> The basic problem is that message matching would be non-deterministic and >> it would be impossible for a tool to show the user which receive operation >> satisfied which send operation without internally using some sort of >> sequence number for each send/receive operation. As Markus mentions, this >> is a similar problem to the MPI_THREAD_MULTIPLE "logically concurrent" >> issue, with which tools already struggle. >> >> My responses were: >> 1) the user asked for this behaviour so the tool could simply gracefully >> give up the linking function and just state the information it knows >> 2) the tool could hook all send and receive operations and piggy-back a >> sequence number into the message header >> 3) the tool could hook all send and receive operations and serialise them >> to prevent overtaking >> I don't know the viability/acceptability of any of these approaches. >> >> Preferably before the formal reading at the next face-to-face meeting, we >> should discuss the possibility that this assertion will disrupt tools and >> ascertain if we can/want to do anything to mitigate that disruption. >> >> Cheers, >> Dan. >> >> -------- Forwarded Message -------- Subject: MPI P2P "messages may >> overtake" assertion Date: Mon, 14 Dec 2015 16:24:44 +0100 From: Markus >> Geimer <m.geimer(a)fz-juelich.de> <m.geimer(a)fz-juelich.de> To: >> dholmes(a)epcc.ed.ac.uk CC: Marc-André Hermanns >> <m.a.hermanns(a)fz-juelich.de> <m.a.hermanns(a)fz-juelich.de> >> >> Hi Daniel, >> >> Do you remember our conversation after the MPI BoF at SC'15 about the >> proposed "messages may overtake" assertion and that it may cause severe >> problems for tools, as correct message matching is very important for >> many of them? In fact, the issue here is similar to properly supporting >> MPI_THREAD_MULTIPLE, which most tools don't do either right now... >> >> Meanwhile I talked to Marc-Andre Hermanns from the Tools WG -- which >> kick-started a discussion on this end. But I guess this topic also >> requires some involvement of the P2P WG. For example, it would be good >> if you could summarize the ideas you had in mind how tools could handle >> this assertion. Thanks! >> >> That said, I don't think there is any urgency for discussions, as >> people are slowly leaving for the Christmas break (e.g., I'm already >> out of office). But we should definitely follow up on this early next >> year. >> >> Merry Christmas, >> Markus >> >> -- >> Dr. Markus Geimer >> Juelich Supercomputing Centre >> Institute for Advanced Simulation >> Forschungszentrum Juelich GmbH >> 52425 Juelich, Germany >> >> Phone: +49-2461-61-1773 >> Fax: +49-2461-61-6656 >> E-Mail: m.geimer(a)fz-juelich.de >> WWW: http://www.fz-juelich.de/jsc >> >> >> ------------------------------------------------------------------------------------------------ >> ------------------------------------------------------------------------------------------------ >> Forschungszentrum Juelich GmbH >> 52425 Juelich >> Sitz der Gesellschaft: Juelich >> Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 >> Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher >> Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), >> Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, >> Prof. Dr. Sebastian M. Schmidt >> ------------------------------------------------------------------------------------------------ >> ------------------------------------------------------------------------------------------------ >> >> >> >> >> -- >> Dan Holmes >> Applications Consultant in HPC Research >> EPCC, The University of Edinburgh >> James Clerk Maxwell Building >> The Kings Buildings >> Peter Guthrie Tait Road >> Edinburgh >> EH9 3FD >> T: +44(0)131 651 3465 >> E: dholmes(a)epcc.ed.ac.uk >> >> *Please consider the environment before printing this email.* >> >> >> >> >> The University of Edinburgh is a charitable body, registered in >> Scotland, with registration number SC005336. >> >> _______________________________________________ >> mpiwg-p2p mailing list >> mpiwg-p2p(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-p2p >> > > > > -- > Jeff Hammond > jeff.science(a)gmail.com > http://jeffhammond.github.io/ > > > -- > Dan Holmes > Applications Consultant in HPC Research > EPCC, The University of Edinburgh > James Clerk Maxwell Building > The Kings Buildings > Peter Guthrie Tait Road > Edinburgh > EH9 3FD > T: +44(0)131 651 3465 > E: dholmes(a)epcc.ed.ac.uk > > *Please consider the environment before printing this email.* > > > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > mpiwg-p2p mailing list > mpiwg-p2p(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-p2p >

1 0

Re: [mpiwg-p2p] Next meeting 18th Jan 2016
by Jeff Hammond 14 Dec '15

14 Dec '15

Can someone tell me the best replacement for the following pattern if freeing active requests is made erroneous? void send_fire_and_forget_ack(int target, MPI_Comm comm) { MPI_Request request; MPI_Isend(NULL, 0, MPI_BYTE, target, special_tag, comm, &request); MPI_Request_free(&request); return; } The reason to not use MPI_Send here is it can block on resource exhaustion. Note that this is a real usage model. One of my applications send fire-and-forget acks and it was a performance issue on Blue Gene/Q. Because we send a payload - it is unclear if this is actually necessary - I replaced MPI_Send with MPI_Bsend to solve the problem (yes, Bsend really does help here), but we could instead use Isend+Request_free (which would be easy since our payload is trivial and could be a global const). The reason I bring this up even though Bsend works for us is that I would much rather deprecate Bsend that make freeing active requests erroneous. Jeff On Mon, Dec 14, 2015 at 7:55 AM, Daniel Holmes <dholmes(a)epcc.ed.ac.uk> wrote: > Hi all, > > The next point-to-point teleconference will be Monday 18th January 2016 at > 11am Central US via webex. > > I will send out another reminder nearer the time. > > Tentative agenda: > > MPI-4.0, issue #11: add communicator assertions (remember implications for > tools!) > MPI-4.0, issue #26: deprecate cancel for send requests > > MPI-4.0, issue TBD: add ARECV/FSEND > MPI-4.0, issue TBD: deal with minor correction to MPI_WAIT raised by Bill > and Rajeev (a la Bill's email 29/06/2015) > * MPI-4.0, issue TBD: deprecate request free for active receive requests > (arising from Dec 2015 meeting)* > MPI-4.0, issue TBD: define batch requests (a la Pavan's email of > 14/12/2015) > > * MPI-Next, issue TBD: disallow request free for active receive requests > (i.e. upgrade advice to erroneous)* > MPI-Next, issue #27: disallow cancel for send requests (i.e. upgrade > advice to erroneous) > > Cheers, > Dan. > > -- > Dan Holmes > Applications Consultant in HPC Research > EPCC, The University of Edinburgh > James Clerk Maxwell Building > The Kings Buildings > Peter Guthrie Tait Road > Edinburgh > EH9 3FD > T: +44(0)131 651 3465 > E: dholmes(a)epcc.ed.ac.uk > > *Please consider the environment before printing this email.* > > > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > mpiwg-p2p mailing list > mpiwg-p2p(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-p2p > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/

1 0

Re: [mpiwg-p2p] Fwd: MPI P2P "messages may overtake" assertion
by Daniel Holmes 14 Dec '15

14 Dec '15

Hi Jeff, > ... tools can just intercept the info operations and strip this key if > they can't support it. Agreed: this should have been response 4 on my list (Markus and I discussed that option too). Cheers, Dan. On 14/12/2015 15:54, Jeff Hammond wrote: > Remember that one of the use cases of allow_overtaking is applications > that have exact matching, in which case allow_overtaking is a way of > turning off a feature that isn't used, in order to get a > high-performing message queue implementation. In the exact matching > case, tools will have no problem matching up sends and recvs. > > If tools cannot handle MPI_THREAD_MULTIPLE already, then I don't > really care if they can't support this assertion either. > > And in any case, such tools can just intercept the info operations and > strip this key if they can't support it. > > Jeff > > On Mon, Dec 14, 2015 at 7:44 AM, Daniel Holmes <dholmes(a)epcc.ed.ac.uk > <mailto:[email protected]>> wrote: > > Hi all, > > Markus has reminded me that we discussed the effect of the > mpi_assert_allow_overtaking assertion on tools (see below). > > The basic problem is that message matching would be > non-deterministic and it would be impossible for a tool to show > the user which receive operation satisfied which send operation > without internally using some sort of sequence number for each > send/receive operation. As Markus mentions, this is a similar > problem to the MPI_THREAD_MULTIPLE "logically concurrent" issue, > with which tools already struggle. > > My responses were: > 1) the user asked for this behaviour so the tool could simply > gracefully give up the linking function and just state the > information it knows > 2) the tool could hook all send and receive operations and > piggy-back a sequence number into the message header > 3) the tool could hook all send and receive operations and > serialise them to prevent overtaking > I don't know the viability/acceptability of any of these approaches. > > Preferably before the formal reading at the next face-to-face > meeting, we should discuss the possibility that this assertion > will disrupt tools and ascertain if we can/want to do anything to > mitigate that disruption. > > Cheers, > Dan. > > -------- Forwarded Message -------- > Subject: MPI P2P "messages may overtake" assertion > Date: Mon, 14 Dec 2015 16:24:44 +0100 > From: Markus Geimer <m.geimer(a)fz-juelich.de> > <mailto:[email protected]> > To: dholmes(a)epcc.ed.ac.uk <mailto:[email protected]> > CC: Marc-André Hermanns <m.a.hermanns(a)fz-juelich.de> > <mailto:[email protected]> > > > > Hi Daniel, > > Do you remember our conversation after the MPI BoF at SC'15 about the > proposed "messages may overtake" assertion and that it may cause severe > problems for tools, as correct message matching is very important for > many of them? In fact, the issue here is similar to properly supporting > MPI_THREAD_MULTIPLE, which most tools don't do either right now... > > Meanwhile I talked to Marc-Andre Hermanns from the Tools WG -- which > kick-started a discussion on this end. But I guess this topic also > requires some involvement of the P2P WG. For example, it would be good > if you could summarize the ideas you had in mind how tools could handle > this assertion. Thanks! > > That said, I don't think there is any urgency for discussions, as > people are slowly leaving for the Christmas break (e.g., I'm already > out of office). But we should definitely follow up on this early next > year. > > Merry Christmas, > Markus > > -- > Dr. Markus Geimer > Juelich Supercomputing Centre > Institute for Advanced Simulation > Forschungszentrum Juelich GmbH > 52425 Juelich, Germany > > Phone: +49-2461-61-1773 > Fax: +49-2461-61-6656 > E-Mail:[email protected] <mailto:[email protected]> > WWW:http://www.fz-juelich.de/jsc > > > ------------------------------------------------------------------------------------------------ > ------------------------------------------------------------------------------------------------ > Forschungszentrum Juelich GmbH > 52425 Juelich > Sitz der Gesellschaft: Juelich > Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 > Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher > Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), > Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, > Prof. Dr. Sebastian M. Schmidt > ------------------------------------------------------------------------------------------------ > ------------------------------------------------------------------------------------------------ > > > > -- > Dan Holmes > Applications Consultant in HPC Research > EPCC, The University of Edinburgh > James Clerk Maxwell Building > The Kings Buildings > Peter Guthrie Tait Road > Edinburgh > EH9 3FD > T: +44(0)131 651 3465 > E:[email protected] <mailto:[email protected]> > > *Please consider the environment before printing this email.* > > > > > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > mpiwg-p2p mailing list > mpiwg-p2p(a)lists.mpi-forum.org <mailto:[email protected]> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-p2p > > > > > -- > Jeff Hammond > jeff.science(a)gmail.com <mailto:[email protected]> > http://jeffhammond.github.io/ -- Dan Holmes Applications Consultant in HPC Research EPCC, The University of Edinburgh James Clerk Maxwell Building The Kings Buildings Peter Guthrie Tait Road Edinburgh EH9 3FD T: +44(0)131 651 3465 E: dholmes(a)epcc.ed.ac.uk *Please consider the environment before printing this email.* The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

1 0

Next meeting 18th Jan 2016
by Daniel Holmes 14 Dec '15

14 Dec '15

Hi all, The next point-to-point teleconference will be Monday 18th January 2016 at 11am Central US via webex. I will send out another reminder nearer the time. Tentative agenda: MPI-4.0, issue #11: add communicator assertions (remember implications for tools!) MPI-4.0, issue #26: deprecate cancel for send requests MPI-4.0, issue TBD: add ARECV/FSEND MPI-4.0, issue TBD: deal with minor correction to MPI_WAIT raised by Bill and Rajeev (a la Bill's email 29/06/2015) MPI-4.0, issue TBD: deprecate request free for active receive requests (arising from Dec 2015 meeting) MPI-4.0, issue TBD: define batch requests (a la Pavan's email of 14/12/2015) MPI-Next, issue TBD: disallow request free for active receive requests (i.e. upgrade advice to erroneous) MPI-Next, issue #27: disallow cancel for send requests (i.e. upgrade advice to erroneous) Cheers, Dan. -- Dan Holmes Applications Consultant in HPC Research EPCC, The University of Edinburgh James Clerk Maxwell Building The Kings Buildings Peter Guthrie Tait Road Edinburgh EH9 3FD T: +44(0)131 651 3465 E: dholmes(a)epcc.ed.ac.uk *Please consider the environment before printing this email.* The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

1 0

Re: [mpiwg-p2p] Fwd: MPI P2P "messages may overtake" assertion
by Jeff Hammond 14 Dec '15

14 Dec '15

Remember that one of the use cases of allow_overtaking is applications that have exact matching, in which case allow_overtaking is a way of turning off a feature that isn't used, in order to get a high-performing message queue implementation. In the exact matching case, tools will have no problem matching up sends and recvs. If tools cannot handle MPI_THREAD_MULTIPLE already, then I don't really care if they can't support this assertion either. And in any case, such tools can just intercept the info operations and strip this key if they can't support it. Jeff On Mon, Dec 14, 2015 at 7:44 AM, Daniel Holmes <dholmes(a)epcc.ed.ac.uk> wrote: > Hi all, > > Markus has reminded me that we discussed the effect of the > mpi_assert_allow_overtaking assertion on tools (see below). > > The basic problem is that message matching would be non-deterministic and > it would be impossible for a tool to show the user which receive operation > satisfied which send operation without internally using some sort of > sequence number for each send/receive operation. As Markus mentions, this > is a similar problem to the MPI_THREAD_MULTIPLE "logically concurrent" > issue, with which tools already struggle. > > My responses were: > 1) the user asked for this behaviour so the tool could simply gracefully > give up the linking function and just state the information it knows > 2) the tool could hook all send and receive operations and piggy-back a > sequence number into the message header > 3) the tool could hook all send and receive operations and serialise them > to prevent overtaking > I don't know the viability/acceptability of any of these approaches. > > Preferably before the formal reading at the next face-to-face meeting, we > should discuss the possibility that this assertion will disrupt tools and > ascertain if we can/want to do anything to mitigate that disruption. > > Cheers, > Dan. > > -------- Forwarded Message -------- Subject: MPI P2P "messages may > overtake" assertion Date: Mon, 14 Dec 2015 16:24:44 +0100 From: Markus > Geimer <m.geimer(a)fz-juelich.de> <m.geimer(a)fz-juelich.de> To: > dholmes(a)epcc.ed.ac.uk CC: Marc-André Hermanns <m.a.hermanns(a)fz-juelich.de> > <m.a.hermanns(a)fz-juelich.de> > > Hi Daniel, > > Do you remember our conversation after the MPI BoF at SC'15 about the > proposed "messages may overtake" assertion and that it may cause severe > problems for tools, as correct message matching is very important for > many of them? In fact, the issue here is similar to properly supporting > MPI_THREAD_MULTIPLE, which most tools don't do either right now... > > Meanwhile I talked to Marc-Andre Hermanns from the Tools WG -- which > kick-started a discussion on this end. But I guess this topic also > requires some involvement of the P2P WG. For example, it would be good > if you could summarize the ideas you had in mind how tools could handle > this assertion. Thanks! > > That said, I don't think there is any urgency for discussions, as > people are slowly leaving for the Christmas break (e.g., I'm already > out of office). But we should definitely follow up on this early next > year. > > Merry Christmas, > Markus > > -- > Dr. Markus Geimer > Juelich Supercomputing Centre > Institute for Advanced Simulation > Forschungszentrum Juelich GmbH > 52425 Juelich, Germany > > Phone: +49-2461-61-1773 > Fax: +49-2461-61-6656 > E-Mail: m.geimer(a)fz-juelich.de > WWW: http://www.fz-juelich.de/jsc > > > ------------------------------------------------------------------------------------------------ > ------------------------------------------------------------------------------------------------ > Forschungszentrum Juelich GmbH > 52425 Juelich > Sitz der Gesellschaft: Juelich > Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 > Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher > Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), > Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, > Prof. Dr. Sebastian M. Schmidt > ------------------------------------------------------------------------------------------------ > ------------------------------------------------------------------------------------------------ > > > > > -- > Dan Holmes > Applications Consultant in HPC Research > EPCC, The University of Edinburgh > James Clerk Maxwell Building > The Kings Buildings > Peter Guthrie Tait Road > Edinburgh > EH9 3FD > T: +44(0)131 651 3465 > E: dholmes(a)epcc.ed.ac.uk > > *Please consider the environment before printing this email.* > > > > > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > mpiwg-p2p mailing list > mpiwg-p2p(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-p2p > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/

1 0

Fwd: MPI P2P "messages may overtake" assertion
by Daniel Holmes 14 Dec '15

14 Dec '15

Hi all, Markus has reminded me that we discussed the effect of the mpi_assert_allow_overtaking assertion on tools (see below). The basic problem is that message matching would be non-deterministic and it would be impossible for a tool to show the user which receive operation satisfied which send operation without internally using some sort of sequence number for each send/receive operation. As Markus mentions, this is a similar problem to the MPI_THREAD_MULTIPLE "logically concurrent" issue, with which tools already struggle. My responses were: 1) the user asked for this behaviour so the tool could simply gracefully give up the linking function and just state the information it knows 2) the tool could hook all send and receive operations and piggy-back a sequence number into the message header 3) the tool could hook all send and receive operations and serialise them to prevent overtaking I don't know the viability/acceptability of any of these approaches. Preferably before the formal reading at the next face-to-face meeting, we should discuss the possibility that this assertion will disrupt tools and ascertain if we can/want to do anything to mitigate that disruption. Cheers, Dan. -------- Forwarded Message -------- Subject: MPI P2P "messages may overtake" assertion Date: Mon, 14 Dec 2015 16:24:44 +0100 From: Markus Geimer <m.geimer(a)fz-juelich.de> To: dholmes(a)epcc.ed.ac.uk CC: Marc-André Hermanns <m.a.hermanns(a)fz-juelich.de> Hi Daniel, Do you remember our conversation after the MPI BoF at SC'15 about the proposed "messages may overtake" assertion and that it may cause severe problems for tools, as correct message matching is very important for many of them? In fact, the issue here is similar to properly supporting MPI_THREAD_MULTIPLE, which most tools don't do either right now... Meanwhile I talked to Marc-Andre Hermanns from the Tools WG -- which kick-started a discussion on this end. But I guess this topic also requires some involvement of the P2P WG. For example, it would be good if you could summarize the ideas you had in mind how tools could handle this assertion. Thanks! That said, I don't think there is any urgency for discussions, as people are slowly leaving for the Christmas break (e.g., I'm already out of office). But we should definitely follow up on this early next year. Merry Christmas, Markus -- Dr. Markus Geimer Juelich Supercomputing Centre Institute for Advanced Simulation Forschungszentrum Juelich GmbH 52425 Juelich, Germany Phone: +49-2461-61-1773 Fax: +49-2461-61-6656 E-Mail: m.geimer(a)fz-juelich.de WWW: http://www.fz-juelich.de/jsc ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, Prof. Dr. Sebastian M. Schmidt ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ -- Dan Holmes Applications Consultant in HPC Research EPCC, The University of Edinburgh James Clerk Maxwell Building The Kings Buildings Peter Guthrie Tait Road Edinburgh EH9 3FD T: +44(0)131 651 3465 E: dholmes(a)epcc.ed.ac.uk *Please consider the environment before printing this email.* The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

1 0