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

Re: [Mpi-forum] Feedback needed / Suggested Change to Chapter 8 (related to RMA)
by Jim Dinan 19 Jul '12

19 Jul '12

Same here. ~Jim. On 7/19/12 11:17 AM, William Gropp wrote: > This sounds good to me - deferring the discussion of the semantics to the RMA chapter makes the most sense. > > 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 11:05 AM, Schulz, Martin wrote: > >> Hi all, >> >> during the final chapter reads, the following issue came up in the section "Memory Allocation" in Chapter 8, MPI Environmental Management. Here is the context >> >>> \section{Memory Allocation} >>> \label{sec:misc-memalloc} >>> >>> In some systems, message-passing and remote-memory-access (\RMA/) operations >>> run faster when accessing specially allocated memory (e.g., memory that is >>> shared by the other processes in the communicating group on an SMP). \MPI/ >>> provides a mechanism for allocating and freeing such special memory. The use >>> of such memory for message-passing or \RMA/ is not mandatory, and this memory >>> can be used without restrictions as any other dynamically allocated memory. >> >> This is followed by the following sentence: >> >>> However, implementations may restrict the use of the \mpifunc{MPI\_WIN\_LOCK} >>> and \mpifunc{MPI\_WIN\_UNLOCK} functions to windows allocated in such memory >>> (see Section~\ref{sec:1sided-lock}.) >> >> This covers the restrictions for 2.2, but not for the new RMA chapter. Therefore, the RMA group (or a subset thereof) suggested to change that sentence to: >> >>> However, implementations may restrict the use of some RMA functionality as defined >>> in Section~\ref{sec:1sided-lock}. >> >> and leave the definition of the semantics to the RMA chapter. >> >> Please reply to the group if there are any objections or concerns. >> >> 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 > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >

1 0

Add MPI_COMM_SPLIT_TYPE to Comm Info?
by Adam T. Moody 19 Jul '12

19 Jul '12

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

1 0

Finishing Off MPI 3.0
by Richard Graham 19 Jul '12

19 Jul '12

Here is a summary of what we agreed to yesterday, with respect to getting MPI 3.0 out of the door. At this stage, most of the reviews are complete, but some changes are still needed. Chapter committees should finish off any additional reviews they think that are needed. Here is the timeline: - All changes to the Latex source are completed, reviewed, and checked into the approved branch - Thursday, July 26, 2012 at 5pm Pacific Daylight time (just need a time when we can cut a draft version) - Draft posted for comment - Friday, July 27, 2012. We will post instructions on how to comment, and make it clear that unless the changes are trivial, or a serious problem is detected, they may not make it into the 3.0 version, but into a follow-on version. - Comments are due by Sept 6, 2012 - We will review the comments and have a telecon to discuss follow up actions on approximately Sept 10th - Changes to 3.0, in response to comments, will be in place by Sept the 12th, with a revised document posted. - Final chapter votes on Thursday Sept 20th at the Forum meeting in Vienna - Final standard vote on Friday Sept, 21st at the Forum meeting in Vienna Rich

1 0

MPI_INFO_GET_ENV naming question
by Schulz, Martin 19 Jul '12

19 Jul '12

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

1 0

Re: [Mpi-forum] Sept MPI Forum meeting
by Torsten Hoefler 19 Jul '12

19 Jul '12

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

1 0

Indices
by Mohamad Chaarawi 19 Jul '12

19 Jul '12

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

1 0

Re: [Mpi-forum] Chapter Reviews
by Richard Graham 19 Jul '12

19 Jul '12

Pick a chapter - at this stage we have one reader for all the chapters. Rich -----Original Message----- From: mpi-forum-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Adam T. Moody Sent: Thursday, July 19, 2012 12:37 PM To: Main MPI Forum mailing list Subject: Re: [Mpi-forum] Chapter Reviews I'm available to read a chapter. What's the most recent pdf to use? Also, it seems that even "make cleandoc" is failing on r1472. -Adam Richard Graham wrote: >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 > > > >------------------------------------------------------------------------ > >_______________________________________________ >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] Final document of the night
by Mohror, Kathryn 19 Jul '12

19 Jul '12

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

1 0

Re: [Mpi-forum] Chapter Reviews
by Schulz, Martin 19 Jul '12

19 Jul '12

On Jul 19, 2012, at 11:36 AM, Adam T. Moody wrote: > I'm available to read a chapter. What's the most recent pdf to use? > > Also, it seems that even "make cleandoc" is failing on r1472. Sorry, my bad - fixed. Martin > -Adam > > Richard Graham wrote: > >> We are currently doing chapter reviews for the forum ï¿½ all tickets are committed and reviewed. If anyone has changes to make please coordinate this morning (Thursday 7/19/2012 ï¿½ 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 >> >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> 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 ________________________________________________________________________ Martin Schulz, schulzm(a)llnl.gov, http://people.llnl.gov/schulzm CASC @ Lawrence Livermore National Laboratory, Livermore, USA

1 0

Re: [Mpi-forum] Chapter Reviews
by Adam T. Moody 19 Jul '12

19 Jul '12

I'm available to read a chapter. What's the most recent pdf to use? Also, it seems that even "make cleandoc" is failing on r1472. -Adam Richard Graham wrote: >We are currently doing chapter reviews for the forum ï¿½ all tickets are committed and reviewed. If anyone has changes to make please coordinate this morning (Thursday 7/19/2012 ï¿½ 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 > > > >------------------------------------------------------------------------ > >_______________________________________________ >mpi-forum mailing list >mpi-forum(a)lists.mpi-forum.org >http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >

1 0