- mpiwg-coll - mailman3.cels.anl.gov

Re: [mpiwg-coll] [mpiwg-persistence] Issue of embiggenment for Ticket #25--Persistent Collectives
by Wesley Bland 26 Mar '20

26 Mar '20

One thing that would be very helpful would be to make a clear line in the pull request where everything before that line is what was voted on originally and everything after is what was necessary for embiggenment. Since this almost entirely will show up in the LaTeX (with a few changes in the normative text too), it would be helpful to keep the unrelated changes (e.g. whitespace, comments, etc.) to a minimum to make it easy for people to see what’s changed at a glance. Thanks, Wes > On Mar 25, 2020, at 10:50 PM, Skjellum, Anthony via mpiwg-persistence <mpiwg-persistence(a)lists.mpi-forum.org> wrote: > > Martin, it is fine to cover this as you suggest in the virtual meeting; hopefully we will do this via ticket 0 in the end. > I agree this has to come after Bill's email issues. > > Thank you > Tony > > Anthony Skjellum, PhD > Professor of Computer Science and Chair of Excellence > Director, SimCenter > University of Tennessee at Chattanooga (UTC) > tony-skjellum(a)utc.edu <mailto:[email protected]> [or skjellum(a)gmail.com <mailto:[email protected]>] > cell: 205-807-4968 > > From: mpiwg-persistence <mpiwg-persistence-bounces(a)lists.mpi-forum.org> on behalf of Martin Schulz via mpiwg-persistence <mpiwg-persistence(a)lists.mpi-forum.org> > Sent: Wednesday, March 25, 2020 10:36 PM > To: Anthony Skjellum <skjellum(a)gmail.com> > Cc: MPI-3 Collective Subgroup Discussions <mpiwg-coll(a)lists.mpi-forum.org>; MPI WG Persistence working Group <mpiwg-persistence(a)lists.mpi-forum.org>; William Gropp <wgropp(a)illinois.edu> > Subject: Re: [mpiwg-persistence] Issue of embiggenment for Ticket #25--Persistent Collectives > > Hi Tony, > > (Adding Bill and Wesley) > > We also talked about this today and we agree, that this should be done sooner than later, but we do need the other changes from Bill’s email taken care of first by the chapter committees. Only this will allow for easy merges - the goal is to get this done by next week (please help push the chapter committees you are working with). > > As for the procedure - we technically have handled that as a ticket 0 in past, but this has gotten larger. I would suggest we reserve part of a virtual meeting for that in which you can run through the changes. Based on the amount of reactions we can then decide whether this is truly ticket 0, or we need a vote (which IMHO would be an errata vote). > > Does this make sense and help? > Thanks! > > 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 25. Mar 2020, at 15:44, Anthony Skjellum <skjellum(a)gmail.com <mailto:[email protected]>> wrote: >> >> Dear Martin, we are at this time considering when we need to do the embiggenment work on the persistent collective operations. We think it should >> be done now. >> >> How should we proceed? >> >> 1) We propose to open a new ticket and new PR. Do you agree? >> 2) Is this strictly a chapter-author update? >> 3) Is this something to be voted? Errata? full ticket? >> >> Please advise us. >> >> Thank you in advance, >> Tony (on behalf of the WGs) >> >> -- >> Anthony Skjellum, PhD >> skjellum(a)gmail.com <mailto:[email protected]> >> Cell: +1-205-807-4968 >> >> > > _______________________________________________ > mpiwg-persistence mailing list > mpiwg-persistence(a)lists.mpi-forum.org <mailto:[email protected]> > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-persistence <https://lists.mpi-forum.org/mailman/listinfo/mpiwg-persistence>

1 0

Re: [mpiwg-coll] Issue of embiggenment for Ticket #25--Persistent Collectives
by Martin Schulz 25 Mar '20

25 Mar '20

Hi Tony, (Adding Bill and Wesley) We also talked about this today and we agree, that this should be done sooner than later, but we do need the other changes from Bill’s email taken care of first by the chapter committees. Only this will allow for easy merges - the goal is to get this done by next week (please help push the chapter committees you are working with). As for the procedure - we technically have handled that as a ticket 0 in past, but this has gotten larger. I would suggest we reserve part of a virtual meeting for that in which you can run through the changes. Based on the amount of reactions we can then decide whether this is truly ticket 0, or we need a vote (which IMHO would be an errata vote). Does this make sense and help? Thanks! 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 25. Mar 2020, at 15:44, Anthony Skjellum <skjellum(a)gmail.com> wrote: > > Dear Martin, we are at this time considering when we need to do the embiggenment work on the persistent collective operations. We think it should > be done now. > > How should we proceed? > > 1) We propose to open a new ticket and new PR. Do you agree? > 2) Is this strictly a chapter-author update? > 3) Is this something to be voted? Errata? full ticket? > > Please advise us. > > Thank you in advance, > Tony (on behalf of the WGs) > > -- > Anthony Skjellum, PhD > skjellum(a)gmail.com <mailto:[email protected]> > Cell: +1-205-807-4968 > >

1 0

Issue of embiggenment for Ticket #25--Persistent Collectives
by Anthony Skjellum 25 Mar '20

25 Mar '20

Dear Martin, we are at this time considering *when* we need to do the embiggenment work on the persistent collective operations. We think it should be done now. How should we proceed? 1) We propose to open a new ticket and new PR. Do you agree? 2) Is this strictly a chapter-author update? 3) Is this something to be voted? Errata? full ticket? Please advise us. Thank you in advance, Tony (on behalf of the WGs) -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0

No meeting today
by Anthony Skjellum 18 Mar '20

18 Mar '20

Hi, if you get a duplicate cancellation, sorry. We will resume (with some normalcy) this meeting next week. I simply have to cancel today to deal with emerging planning issues at my center. I hope all is well with you all. Thanks, Tony -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0

Meeting at 10am Eastern Wednesday
by Anthony Skjellum 10 Mar '20

10 Mar '20

We will meet Wednesday to discuss current work and proposals for MPI 4.x -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0

Re: [mpiwg-coll] No meeting today: Persistent/Collective
by Wozniak, Justin M. 22 Jan '20

22 Jan '20

Hi Could you send me the call info for this? Thanks On 1/22/20 8:33 AM, Anthony Skjellum via mpiwg-coll wrote: Hi, I do not have any agenda items. We will meet for sure next week. Please attend the Forum's virtual meeting @ 10am Central / 11am Eastern US time. Thank you, Tony -- Anthony Skjellum, PhD skjellum(a)gmail.com<mailto:[email protected]> Cell: +1-205-807-4968 _______________________________________________ mpiwg-coll mailing list mpiwg-coll(a)lists.mpi-forum.org<mailto:[email protected]> https://lists.mpi-forum.org/mailman/listinfo/mpiwg-coll -- Justin M Wozniak

1 0

No meeting today: Persistent/Collective
by Anthony Skjellum 22 Jan '20

22 Jan '20

Hi, I do not have any agenda items. We will meet for sure next week. Please attend the Forum's virtual meeting @ 10am Central / 11am Eastern US time. Thank you, Tony -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0

Canceling today's meeting---Persistent and Collective Joint meeting
by Anthony Skjellum 15 Jan '20

15 Jan '20

Hi, I apologize; today's meeting @ 10:00am Eastern will not meet. We will meet next week at 10am Eastern. Thank you Tony Skjellum -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0

No meeting today; next meeting January 8
by Anthony Skjellum 18 Dec '19

18 Dec '19

Hi, the persistent and collective WGs will meet next on January 8. We need to go over the tickets that have emerged that are related to these efforts, which may need to be reviewed more by the group :-) Key outcome: Partitioned Comm. had its successful reading in ABQ! Happy Holidays. Tony -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0

Re: [mpiwg-coll] [Mpi-forum] Error/gap in MPI_NEIGHBOR_ALLTOALL/ALLGATHER
by Anthony Skjellum 29 Sep '19

29 Sep '19

All, please see: https://github.com/mpi-forum/mpi-issues/issues/153 Thanks, Tony On Fri, Sep 27, 2019 at 6:37 PM Anthony Skjellum via mpi-forum < mpi-forum(a)lists.mpi-forum.org> wrote: > Rolf let’s open a Ticket > > Anthony Skjellum, PhD > 205-807-4968 > > > > On Sep 27, 2019, at 6:09 PM, Rolf Rabenseifner via mpi-forum < > mpi-forum(a)lists.mpi-forum.org> wrote: > > > > Dear MPI collective WG, > > > > you may try to resolve the problem with a maybe wrong > > MPI specification for MPI_NEIGHBOR_ALLTOALL/ALLGATHER > > > > Dear MPI Forum member, > > > > you may own/use an MPI implementation that implements > > MPI_NEIGHBOR_ALLTOALL/ALLGATHER > > with race conditions if #nprocs in one dimension is > > only 1 or 2 and periodic==true > > > > The problem was reported as a bug of the OpenMPI library > > by Simone Chiochetti from DICAM at the University of Trento, > > but seems to be a bug in the MPI specification, > > or at least an advice to implementors is missing. > > > > I produced a set of animated slides. > > Please look at them in presentation mode with animation. > > > > Have fun with a problem that clearly prevents the use > > of MPI_NEIGHBOR_... routines with cyclic boundary condition > > if one wants to verify that mpirun -np 1 is doing > > the same as the sequential code. > > > > Best regards > > Rolf > > > > -- > > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)hlrs.de . > > High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530 . > > University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832 . > > Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner . > > Nobelstr. 19, D-70550 Stuttgart, Germany . . . . (Office: Room 1.307) . > > <neighbor_mpi-3_bug.pptx> > > _______________________________________________ > > mpi-forum mailing list > > mpi-forum(a)lists.mpi-forum.org > > https://lists.mpi-forum.org/mailman/listinfo/mpi-forum > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpi-forum > -- Anthony Skjellum, PhD skjellum(a)gmail.com Cell: +1-205-807-4968

1 0