July 2012 - mpi-forum - mailman3.cels.anl.gov

Re: [Mpi-forum] Add MPI_COMM_SPLIT_TYPE to Comm Info?
by William Gropp 20 Jul '12

20 Jul '12

Good catch - MPI_COMM_SPLIT_TYPE clearly belongs in this list. Bill William Gropp Director, Parallel Computing Institute Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign On Jul 19, 2012, at 12:36 PM, Adam T. Moody wrote: > Hi Darius, > I think we want to include MPI_COMM_SPLIT_TYPE in the list of functions that apply an info object to a communicator in the communicator info section 6.4.4. This at first feels a little strange, since globally you end up with multiple communicators after the call. However, each process should end up belonging to at most one new comm, so the info would naturally apply to the comm the process belongs to. > -Adam > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] Final document of the night
by William Gropp 20 Jul '12

20 Jul '12

Fixed, though a little differently (the errata text was from MPI2.2). Bill William Gropp Director, Parallel Computing Institute Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign On Jul 19, 2012, at 12:50 PM, Jim Dinan wrote: > The sentence introducing the participants doesn't parse (page xxix, line 7): > > The following list includes some of the active participants who attended MPI-3 Forum meetings and in the e-mail discussions of the errata items and are not mentioned above. > > New text (add participated and comma): > > The following list includes some of the active participants who attended MPI-3 Forum meetings and _participated_ in the e-mail discussions of the errata items_,_ and are not mentioned above. > > ~Jim. > > On 7/19/12 11:18 AM, William Gropp wrote: >> Got it. I'll add it after I hear that there are no more changes. >> >> Bill >> >> William Gropp >> Director, Parallel Computing Institute >> Deputy Director for Research >> Institute for Advanced Computing Applications and Technologies >> Paul and Cynthia Saylor Professor of Computer Science >> University of Illinois Urbana-Champaign >> >> >> >> On Jul 19, 2012, at 10:24 AM, Jeff Hammond wrote: >> >>> On Thu, Jul 19, 2012 at 9:12 AM, William Gropp <wgropp(a)illinois.edu> wrote: >>>> I've updated the credits list - please check. >>> >>> Please add "Jed Brown" to the list of active participants credited. >>> >>> Thanks, >>> >>> Jeff >>> >>> -- >>> Jeff Hammond >>> Argonne Leadership Computing Facility >>> University of Chicago Computation Institute >>> jhammond(a)alcf.anl.gov / (630) 252-5381 >>> http://www.linkedin.com/in/jeffhammond >>> https://wiki.alcf.anl.gov/parts/index.php/User:Jhammond >>> _______________________________________________ >>> mpi-forum mailing list >>> mpi-forum(a)lists.mpi-forum.org >>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] MPI_INFO_GET_ENV naming question
by William Gropp 20 Jul '12

20 Jul '12

I prefer MPI_INFO_ENV . Using GET in the name will be confusing. Bill William Gropp Director, Parallel Computing Institute Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign On Jul 19, 2012, at 12:25 PM, Schulz, Martin wrote: > Hi all, > > During the chapter review for Chapter 8, the environment chapter, a question came up at the forum, which we thought requires wider feedback. > > Ticket 313 introduced an info object into MPI 3.0 that allows users to query the environment, which is called MPI_INFO_GET_ENV. This name, however, contains a verb and hence follows the naming structure of an MPI function, but is a constant. Hence, should we really go with this name or should we rename this into a name without a verb, like MPI_INFO_ENV, to avoid confusion? People in the room were leaning towards the latter and this is clearly ticket 0 at this point, but we wanted to ask to make sure there is no unintended impact with that change. > > Comments? > Thanks, > > Martin > > > ________________________________________________________________________ > Martin Schulz, schulzm(a)llnl.gov, http://people.llnl.gov/schulzm > CASC @ Lawrence Livermore National Laboratory, Livermore, USA > > > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] Final document of the night
by William Gropp 20 Jul '12

20 Jul '12

Got them. Thanks. Bill William Gropp Director, Parallel Computing Institute Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign On Jul 19, 2012, at 11:43 AM, Mohror, Kathryn wrote: > Hi Bill, > > Please add Marty Itzkowitz to the list of contributors (affiliation Oracle). > > Also, in the list of primary authors: For Bronis, please change "Profiling" to "Tool Support". > > Thanks! > Kathryn > >> -----Original Message----- >> From: mpi-forum-bounces(a)lists.mpi-forum.org [mailto:mpi-forum- >> bounces(a)lists.mpi-forum.org] On Behalf Of William Gropp >> Sent: Thursday, July 19, 2012 9:19 AM >> To: Main MPI Forum mailing list >> Subject: Re: [Mpi-forum] Final document of the night >> >> Got it. I'll add it after I hear that there are no more changes. >> >> Bill >> >> William Gropp >> Director, Parallel Computing Institute >> Deputy Director for Research >> Institute for Advanced Computing Applications and Technologies >> Paul and Cynthia Saylor Professor of Computer Science >> University of Illinois Urbana-Champaign >> >> >> >> On Jul 19, 2012, at 10:24 AM, Jeff Hammond wrote: >> >>> On Thu, Jul 19, 2012 at 9:12 AM, William Gropp <wgropp(a)illinois.edu> >> wrote: >>>> I've updated the credits list - please check. >>> >>> Please add "Jed Brown" to the list of active participants credited. >>> >>> Thanks, >>> >>> Jeff >>> >>> -- >>> Jeff Hammond >>> Argonne Leadership Computing Facility >>> University of Chicago Computation Institute >>> jhammond(a)alcf.anl.gov / (630) 252-5381 >>> http://www.linkedin.com/in/jeffhammond >>> https://wiki.alcf.anl.gov/parts/index.php/User:Jhammond >>> _______________________________________________ >>> mpi-forum mailing list >>> mpi-forum(a)lists.mpi-forum.org >>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

EruoMPI 2012 - Call for Participation(!!)
by Jesper Larsson Träff 20 Jul '12

20 Jul '12

[Dear MPI Forum members - with all due apologies for possibly multiple copies of this etc.: hope to see many of you at the EuroMPI 2012 conference in September in Vienna - Best, Jesper] EuroMPI 2012 CALL FOR PARTICIPATION EuroMPI 2012 - the prime meeting for researchers, developers, and students in message-passing parallel computing with MPI (and related paradigms) - calls for *active participation* in the conference which will take place from September 23rd to September 26th in Vienna, Austria, at the Austrian Academy of Sciences. The conference will feature - 22 strong technical papers on aspects of MPI library implementation, applications using MPI, extensions to MPI, implementation on novel architectures, performance analysis etc. - 4 invited talks by leading experts, e.g. on MPI implementation and novel (HPC) architectures - 2 tutorials on performance analysis and advanced MPI - The IMUDI special session dedicated to improving MPI user and developer interaction with technical papers and an invited talk - An MPI Forum session on MPI 3.0 - A technically oriented vendor session - A selection of posters More details, including conference and hotel registration links, can be found at www.eurompi2012.org The (early) registration fee of 500 Euro covers the full program, printed proceedings in the Springer LNCS series, all lunches and coffee breaks, and the social program in Vienna. The early registration fee will be valid until July 30th. Tutorial participation is in addition, and tutorials can be attended independently of the conference. General Chair: Jack Dongarra Program chairs and local organizers: Siegfried Benkner, University of Vienna Jesper Larsson Trï¿½ff, Vienna University of Technology

1 0

Re: [Mpi-forum] Sept MPI Forum meeting
by Richard Graham 20 Jul '12

20 Jul '12

Jesper, Thanks a lot for all your help here. We will proceed with the reduced schedule - Sept. 20-21st. Rich -----Original Message----- From: mpi-forum-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Jesper Larsson Träff Sent: Friday, July 20, 2012 3:11 AM To: Torsten Hoefler Cc: Main MPI Forum mailing list; Cheryl Kovacs Subject: Re: [Mpi-forum] Sept MPI Forum meeting Hi all, Rich requested the week before EuroMPI looong time ago (for 4 days), and we have found room etc. for that period. After EuroMPI is close to semester start here in Austria, that is, it will be much more difficult now to get a seminar room for that period. Of course, the overlap with PaCT is very unfortunate (and these overlaps are often a problem for EuroMPI - something we could discuss at the "Business Meeting" at the conference!) All I wanted to say is: please make up your mind soon and let us know what we should plan for Best, Jesper On Thu, Jul 19, 2012 at 07:21:34PM +0200, Torsten Hoefler wrote: > Rich, > > That sounds fine. The current dates overlap 100% with PACT, which > prevents me from attending the MPI Forum. I would appreciate to have > it after EuroMPI. But I understand if other reasons disallow a > re-scheduling or if it's to late to make such a change. I just thought > I'd bring this conflict up :-). > > Thanks, > Torsten > > On Wed, 2012-07-18 at 23:17 +0000, Richard Graham wrote: > > In the meeting today, we discussed changing the schedule for the > > Sept meeting in Vienna. Since we are at a transition stage between > > MPI 3.0 and “MPI-next”, we decided to reduce the meetings from a 3 > > day meeting to a 2 day meeting - Sept 20-21, 2012. If someone > > knows of a reason to go back to a three day agenda, please let us > > know ASAP. The proposed agenda includes: > > > > > > > > Thursday – Sept 20th > > > > **************** > > > > 9-10 : Discussion of outside comments on the MPI 3.0 draft > > > > 10-12 : MPI 3.0: chapter votes > > > > 12-1 : lunch > > > > 1-5 : MPI-Next: organizational structure and areas of focus > > > > > > > > Friday – Sept 21 > > > > ************ > > > > 9-9:30 : MPI 3.0 - final vote > > > > 9:30 -12 : Working group time > > > > 12-1 lunch > > > > 1-3 : Working group time > > > > 3-5 : MPI-Next: organizational structure and areas of focus – cont’d > > > > > > > > If there are other topics that should be added to the agenda, please > > let me know ASAP. > > > > > > > > Thanks, > > > > Rich > > > > > > > > > > > > > > > > _______________________________________________ > > mpi-forum mailing list > > mpi-forum(a)lists.mpi-forum.org > > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > > _______________________________________________ mpi-forum mailing list mpi-forum(a)lists.mpi-forum.org http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] Indices
by Rolf Rabenseifner 20 Jul '12

20 Jul '12

Dear chapter authors of Coll, inqiry, binding, appLang, I fixed it. svn r1486. Here the list: Function index - lower case letters: ==================================== Unintended: ----------- chap-coll - Text in Section 5.13, Examples 5.29 - 5.33 DONE: \cfunc + \mpifuncindex or simply make upper case MPI_Bcast, 218 MPI_Ibarrier, 218 MPI_Recv, 219 MPI_Send, 219 MPI_Wait, 219 MPI_Waitall, 220 chap-inquiry - Text in Section 8.7, Examples 8.7 + 8.9 DONE: \cfunc + \mpifuncindex or simply make upper case MPI_Cancel, 360 MPI_Finalize, 360 chap-binding / Fortran - Text in Section 17.1.12 DONE: mpifunc --> const MPI_Request, 628 chap-appLang - A.1.3 - Subsection on mpi_f08 callback prototypes DONE: \ffunc + \mpifuncindex or simply make upper case MPI_Comm_create_keyval, 680 MPI_Op_create, 680 MPI_Win_create_keyval, 680 MPI_Comm_create_errhandler, 681 MPI_File_create_errhandler, 681 MPI_Grequest_start, 681 MPI_Type_create_keyval, 681 MPI_Win_create_errhandler, 681 MPI_Register_datarep, 682 Intended (therefore unchanged, as we had in MPI-2.1 and 2.2): --------- mpiexec, 358, 362, 364, 364, 488 mpirun, 363, 364 Best regards Rolf ----- Original Message ----- > From: "William Gropp" <wgropp(a)illinois.edu> > To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> > Sent: Thursday, July 19, 2012 8:21:15 PM > Subject: Re: [Mpi-forum] Indices > Yes, these need to be fixed. > > Mixed case in the function index means either the C binding name (or > Fortran08 name) was used where the language-independent name should be > used, or \func or \mpifunc were used for a language-binding-specific > name (which is incorrect macro usage). Chapter authors need to fix > this. There as also some non-functions in the function index; those > must be fixed. > > Example index entries need to be checked; I added a bunch to context > as many examples were missing the index entries (made with \exindex). > > I think it is ok for removed and deprecated names to show up in the > index. > > Bill > > > William Gropp > Director, Parallel Computing Institute > Deputy Director for Research > Institute for Advanced Computing Applications and Technologies > Paul and Cynthia Saylor Professor of Computer Science > University of Illinois Urbana-Champaign > > > > On Jul 19, 2012, at 12:17 PM, Mohamad Chaarawi wrote: > > > Hi All, > > > > After reviewing the indices, I found a few issues: > > > > - MPI_Type_struct and MPI_TYPE_EXTENT show up in the example > > indices. This should not happen as they are removed. > > - I'm not sure if removed and deprecated routines should show up in > > the Function Index, but right now, they do. > > - similarly in the MPI Callback Function Prototype Index > > - similarly in Constant and Predefined Handle Index > > - in the function index, many routines show up twice (caps and no > > caps) and some show up only with no caps.. All should be caps (I > > think). > > > > Do the chapter authors need to fix those in their chapters? > > > > Thanks, > > Mohamad > > _______________________________________________ > > mpi-forum mailing list > > mpi-forum(a)lists.mpi-forum.org > > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum -- 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: Allmandring 30)

1 0

Re: [Mpi-forum] Sept MPI Forum meeting
by Torsten Hoefler 20 Jul '12

20 Jul '12

Jesper, Sure, I mentioned the conflict a long time ago :-). But seeing that there was flexibility in the meeting length (3 -> 2 days) made me hope that there was other flexibility too. However, some people made international travel arrangements, so I retract my request :-). Thanks & Best, Torsten -- ### qreharg rug ebs fv crryF ------------- http://www.unixer.de/ ----- Torsten Hoefler | Performance Modeling and Simulation Lead Blue Waters Directorate | University of Illinois (UIUC) 1205 W Clark Street | Urbana, IL, 61801 NCSA Building | +01 (217) 244-7736

1 0

Re: [Mpi-forum] Chapter Reviews
by Richard Graham 20 Jul '12

20 Jul '12

Thanks !!! Pass the details on to the appropriate chapter committee. Rich From: Supalov, Alexander [mailto:[email protected]] Sent: Thursday, July 19, 2012 10:29 PM To: Main MPI Forum mailing list <mpi-forum(a)lists.mpi-forum.org> Subject: Re: [Mpi-forum] Chapter Reviews Done (see attached). What is the procedure to vet and implement these changes? By the way, whoever is going to review chapter 17, please look into 602/1-6: Compliant MPI-3 implementations providing a Fortran interface must provide provide one or both of the following: _ The USE mpi_f08 Fortran support method. _ The USE mpi and INCLUDE 'mpif.h' Fortran support methods. Double “provide” aside, this passage amounts to opening the door to a silent removal (not even deprecation) of the old Fortran interface. This cannot be the real intent of the Forum, I hope. Also, some advices to the implementors effectively suggest breaking backward compatibility on the old Fortran interface, if retained, up to and including a failed application build and changes to the application necessary as a consequence. Again, I think this is a bit too much of a good thing. From: mpi-forum-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Richard Graham Sent: Thursday, July 19, 2012 5:29 PM To: Main MPI Forum mailing list Subject: Re: [Mpi-forum] Chapter Reviews Feel free to take it – thanks !!!! Rich From: mpi-forum-bounces(a)lists.mpi-forum.org<mailto:[email protected]> [mailto:[email protected]]<mailto:[mailto:[email protected]]> On Behalf Of Supalov, Alexander Sent: Thursday, July 19, 2012 11:02 AM To: Main MPI Forum mailing list Subject: Re: [Mpi-forum] Chapter Reviews Hi Rich, It might help those interested if you described the applicable deadlines, the desired output of the review, and the resp. correction procedure, should it become necessary. I hate to see my favorite Appendix A orphaned. Anyone against my taking on it? Best regards. Alexander From: mpi-forum-bounces(a)lists.mpi-forum.org<mailto:[email protected]> [mailto:[email protected]]<mailto:[mailto:[email protected]]> On Behalf Of Richard Graham Sent: Thursday, July 19, 2012 4:45 PM To: mpi-forum(a)lists.mpi-forum.org<mailto:[email protected]> Subject: [Mpi-forum] Chapter Reviews We are currently doing chapter reviews for the forum – all tickets are committed and reviewed. If anyone has changes to make please coordinate this morning (Thursday 7/19/2012 – Chicago time) with the person who is doing the review. After these are done, responsibility will revert to the chapter committees. I will send out a separate note on the timelines we agreed to yesterday, but in a nutshell, all changes must be completed and reviewed by Thursday 7/26/2012, 5pm Pacific time, so that we can prepare a draft version to be posted for comment. If someone remote would like to take the responsibility for an unclaimed chapter, please let me know. Thanks, Rich Chapter Reviewer Front Matter Mohamad Chap 1 Brian B. Chap 2 Scott Chap 3 Anh Chap 4 Terry Chap 5 Nathan Chap 6 Darius Chap 7 Shinji Chap 8 Pavan Chap 9 Howard - done Chap 10 Hideyuki Chap 11 Howard Chap 12 Brian S. Chap 13 Brian S. Chap 14 Rich Chap 15 Chap 16 Chap 17 Appendix A Appendix B Jeff/Rolf Bib Indices Intel GmbH Dornacher Strasse 1 85622 Feldkirchen/Muenchen, Deutschland Sitz der Gesellschaft: Feldkirchen bei Muenchen Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer, Christian Lamprechter Registergericht: Muenchen HRB 47456 Ust.-IdNr./VAT Registration No.: DE129385895 Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052 Intel GmbH Dornacher Strasse 1 85622 Feldkirchen/Muenchen, Deutschland Sitz der Gesellschaft: Feldkirchen bei Muenchen Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer, Christian Lamprechter Registergericht: Muenchen HRB 47456 Ust.-IdNr./VAT Registration No.: DE129385895 Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052

1 0

AUTO: Rolf Riesen is on vacation (returning 07/08/2012)
by Rolf Riesen 20 Jul '12

20 Jul '12

I am out of the office until 07/08/2012. I'm on vacation (mostly) without email access. I'll respond to your email when I get back. Note: This is an automated response to your message "Re: [Mpi-forum] Chapter Reviews" sent on 20/7/2012 6:28:54. This is the only notification you will receive while this person is away.

1 0