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

Re: [Mpi-forum] MPI 3.0 Passed tickets
by Martin Schulz 29 May '12

29 May '12

Hi all, I tried sending this yesterday, but I don't think it made it out (sorry if this ends up being duplicated): I passed the changes for the tickets 204 and 266 (and the remaining parts of 207) to the respective chapter committees/authors (as a Latex diff to the current - as of yesterday - approved branch) for inclusion into the approved branch. Here is a list of the affected chapters: 204 Terms and Conventions Environment Appendix Language Bindings Summary 207 Environment 266 Environment Tools (replaces Profiling Interface Chapter) Appendix Language Bindings Summary Martin On May 29, 2012, at 2:49 PM, Richard Graham wrote: > Attached is the list of tickets that have passed all but the final 3.0 chapter vote, and have not yet been incorporated into the approved version – some were taken care of as this list was being created. I have included the ticket number, ticket owner, and the affected chapter. > > Authors: Please send me an update on the affected chapters, AND work with each of the chapter authors to make sure the changes get moved into the approved section. > > >>> As we discussed at the current meeting, THE GOAL IS TO GET ALL PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << > The goal is to let people start to read over the new version of the standard before the July meeting, so that we can get a draft out for comment 1-2 weeks after the July meeting, and allow for comments before the final vote in Sept. If you can’t make that date, please let me know ASAP, so we will find someone else who can do the work. > > Thanks, > Rich > <MPI 3.0 tickets for inclusion.xlsx>_______________________________________________ > 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] MPI 3.0 Passed tickets
by Fab Tillier 29 May '12

29 May '12

Ticket 265: All changes available in SVN URL: https://svn.mpi-forum.org/svn/mpi-forum-docs/trunk/working-groups/mpi-3/bw-…. I updated the ticket with the SVN URL, and tried merging the approved tree into it, but failed miserably... For some reason SVN isn't picking up on the history of that branch, will see if I can figure out what I'm doing wrong. Ticket 293: LaTeX patch attached to ticket, though with the wrong ticket number. Note that this ticket is likely to be superseded by the changes in ticket 278, so merging this into the approved branch is likely just setting ourselves up for merge conflicts later. Ticket 321: All changes available in SVN at URL: https://svn.mpi-forum.org/svn/mpi-forum-docs/trunk/working-groups/mpi-3/bw-…. SVN should make merging into the approved branch as straight forward as it gets. I updated the ticket with the SVN URL, and merged the approved branch (r1172) into it, so the merge back to approved should be simple. Affected chapters are chap-changes and chap-pt2pt. Ticket 279: already merged into approved by jsquyres 4 months ago. -Fab Richard Graham wrote on Mon, 28 May 2012 at 22:49:09 > Attached is the list of tickets that have passed all but the final 3.0 > chapter vote, and have not yet been incorporated into the approved > version - some were taken care of as this list was being created. I > have included the ticket number, ticket owner, and the affected chapter. > > Authors: Please send me an update on the affected chapters, AND work > with each of the chapter authors to make sure the changes get moved into > the approved section. > >>>> As we discussed at the current meeting, THE GOAL IS TO GET ALL > PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << > The goal is to let people start to read over the new version of the > standard before the July meeting, so that we can get a draft out for > comment 1-2 weeks after the July meeting, and allow for comments before > the final vote in Sept. If you can't make that date, please let me know > ASAP, so we will find someone else who can do the work. > > Thanks, > Rich >

1 0

Re: [Mpi-forum] MPI 3.0 Passed tickets
by Richard Graham 29 May '12

29 May '12

It means that I could not tell from a quick look at the ticket which chapters were impacted, and would like the ticket author to send back this information. Also, w/o a detailed description of what changes need to go into the standard (perhaps via a patch), the changes will not make it into the standard. Rich -----Original Message----- From: mpi-forum-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Bronis R. de Supinski Sent: Tuesday, May 29, 2012 10:30 PM To: Main MPI Forum mailing list Subject: Re: [Mpi-forum] MPI 3.0 Passed tickets Rich: What does it mean when you did not list anything in the chapter column? It looks like 229 touches everything but, as Bill notes with 287, we lack a clean description of what needs to be implemented in each chapter. I believe 266 only touches the tools/profiling chapter but again nothing is listed. Martin has finally gotten me the updates just today as we had agreed he would some time ago. I do not know when I will be able to review and commit them but probably not this week. Otherwise, I believe everything that touches the chpaters that I own (ei and prof/tools) has been implemented. The ticket authors really need to step up on this stuff. Bronis On Mon, 28 May 2012, Richard Graham wrote: > Attached is the list of tickets that have passed all but the final 3.0 chapter vote, and have not yet been incorporated into the approved version – some were taken care of as this list was being created. I have included the ticket number, ticket owner, and the affected chapter. > > Authors: Please send me an update on the affected chapters, AND work with each of the chapter authors to make sure the changes get moved into the approved section. > >>>> As we discussed at the current meeting, THE GOAL IS TO GET ALL >>>> PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << > The goal is to let people start to read over the new version of the standard before the July meeting, so that we can get a draft out for comment 1-2 weeks after the July meeting, and allow for comments before the final vote in Sept. If you can’t make that date, please let me know ASAP, so we will find someone else who can do the work. > > Thanks, > Rich >

1 0

Re: [Mpi-forum] MPI 3.0 Passed tickets
by Bronis R. de Supinski 29 May '12

29 May '12

Rich: What does it mean when you did not list anything in the chapter column? It looks like 229 touches everything but, as Bill notes with 287, we lack a clean description of what needs to be implemented in each chapter. I believe 266 only touches the tools/profiling chapter but again nothing is listed. Martin has finally gotten me the updates just today as we had agreed he would some time ago. I do not know when I will be able to review and commit them but probably not this week. Otherwise, I believe everything that touches the chpaters that I own (ei and prof/tools) has been implemented. The ticket authors really need to step up on this stuff. Bronis On Mon, 28 May 2012, Richard Graham wrote: > Attached is the list of tickets that have passed all but the final 3.0 chapter vote, and have not yet been incorporated into the approved version – some were taken care of as this list was being created. I have included the ticket number, ticket owner, and the affected chapter. > > Authors: Please send me an update on the affected chapters, AND work with each of the chapter authors to make sure the changes get moved into the approved section. > >>>> As we discussed at the current meeting, THE GOAL IS TO GET ALL PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << > The goal is to let people start to read over the new version of the standard before the July meeting, so that we can get a draft out for comment 1-2 weeks after the July meeting, and allow for comments before the final vote in Sept. If you can’t make that date, please let me know ASAP, so we will find someone else who can do the work. > > Thanks, > Rich >

1 0

Migration of Forum SVN and Trac services
by Jeff Squyres 29 May '12

29 May '12

Heads up: Indiana University would like to migrate the Forum's Subversion and Trac services to new servers sometime in the immediate future. The purpose of the migration is to retire some (really) old server hardware. Delivered SVN and Trac performance probably won't be significantly impacted. Doing the migration while you are all at the Japan meeting -- and actively using both SVN and Trac -- is obviously a bad idea. *** Would it be acceptable to move everything on Friday morning, US Eastern time (June 1, 2012)? I *assume* that you'll all be flying back at that point, or asleep, or otherwise signed out for the weekend. -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Re: [Mpi-forum] why (int)0x4c000101 ?
by Jeff Squyres 29 May '12

29 May '12

You should probably ask on the MPICH mailing list -- this is the mailing list for the general MPI Forum itself. On May 27, 2012, at 10:42 AM, ï¿½ï¿½ï¿½ï¿½ ï¿½ï¿½ wrote: > when I read mpich2-1.4.1 code, I saw code as follow: > > Code: > 1 > 2 > 3 > 4 > 5 > 6 > 7 > 8 > 9 > 10 > 11 > 12 > 13 > 14 > 15 > 16 > 17 > 18 > typedef int MPI_Datatype; > #define MPI_CHAR ((MPI_Datatype)0x4c000101) > #define MPI_SIGNED_CHAR ((MPI_Datatype)0x4c000118) > #define MPI_UNSIGNED_CHAR ((MPI_Datatype)0x4c000102) > #define MPI_BYTE ((MPI_Datatype)0x4c00010d) > #define MPI_WCHAR ((MPI_Datatype)0x4c00040e) > #define MPI_SHORT ((MPI_Datatype)0x4c000203) > #define MPI_UNSIGNED_SHORT ((MPI_Datatype)0x4c000204) > #define MPI_INT ((MPI_Datatype)0x4c000405) > #define MPI_UNSIGNED ((MPI_Datatype)0x4c000406) > #define MPI_LONG ((MPI_Datatype)0x4c000407) > #define MPI_UNSIGNED_LONG ((MPI_Datatype)0x4c000408) > #define MPI_FLOAT ((MPI_Datatype)0x4c00040a) > #define MPI_DOUBLE ((MPI_Datatype)0x4c00080b) > #define MPI_LONG_DOUBLE ((MPI_Datatype)0x4c000c0c) > #define MPI_LONG_LONG_INT ((MPI_Datatype)0x4c000809) > #define MPI_UNSIGNED_LONG_LONG ((MPI_Datatype)0x4c000819) > #define MPI_LONG_LONG MPI_LONG_LONG_INT > > Anyone can tell me what is the meaning of those hex values? and why does the author use them? > why do they define "MPI_CHAR" as ((MPI_Datatype)0x4c000101) instead of "char"? > thx > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Re: [Mpi-forum] MPI 3.0 Passed tickets
by Torsten Hoefler 29 May '12

29 May '12

Hi all, Merged tickets for my chapters (Colls, Topology, RMA): Topology: #140 #258 #158 #162 #259 #299 Collectives: #140 #265 #312 RMA: #140 #270 #298 #308 all in approved/MPI-3.0 . I also added comments to all tickets that I merged. Author fields are not always set correctly, I suggest we re-discuss the purpose/usage of those fields. All the Best, Torsten On Tue, May 29, 2012 at 03:14:06PM +0900, William Gropp wrote: > For my chapter (groups and context): > 140 is done > 182 is done > 207 is done > 219 is done > 254 is done > (in most cases, other chapters also need to be modified before the ticket > is complete). > 287 probably belongs in the groups and context chapter, but there is no > clean description of the change in the ticket, only the nearly 700 page > modified draft standard (did we really read and vote on the entire 681 > page document?). > This is probably a good time to remind everyone that we need a clean > description of the change; producing a PDF of the entire document doesn't > provide a workable way to communicate the change to the Forum, as > realistically, no one is going to read through all of those pages. Please > provide the LaTeX diff and enough of the chapter(s) rendered as PDF to > enable review, or clearly indicate the full scope (ranges of pages) of the > change. > 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 May 29, 2012, at 2:49 PM, Richard Graham wrote: > > Attached is the list of tickets that have passed all but the final 3.0 > chapter vote, and have not yet been incorporated into the approved > version � some were taken care of as this list was being created. I > have included the ticket number, ticket owner, and the affected chapter. > > Authors: Please send me an update on the affected chapters, AND work > with each of the chapter authors to make sure the changes get moved into > the approved section. > > >>> As we discussed at the current meeting, THE GOAL IS TO GET ALL > PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << > The goal is to let people start to read over the new version of the > standard before the July meeting, so that we can get a draft out for > comment 1-2 weeks after the July meeting, and allow for comments before > the final vote in Sept. If you can�t make that date, please let me know > ASAP, so we will find someone else who can do the work. > > Thanks, > Rich > <MPI 3.0 tickets for > inclusion.xlsx>_______________________________________________ > mpi-forum mailing list > [1]mpi-forum(a)lists.mpi-forum.org > [2]http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > References > > Visible links > 1. mailto:[email protected] > 2. 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 -- ### 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] MPI 3.0 Passed tickets
by William Gropp 29 May '12

29 May '12

For my chapter (groups and context): 140 is done 182 is done 207 is done 219 is done 254 is done (in most cases, other chapters also need to be modified before the ticket is complete). 287 probably belongs in the groups and context chapter, but there is no clean description of the change in the ticket, only the nearly 700 page modified draft standard (did we really read and vote on the entire 681 page document?). This is probably a good time to remind everyone that we need a clean description of the change; producing a PDF of the entire document doesn't provide a workable way to communicate the change to the Forum, as realistically, no one is going to read through all of those pages. Please provide the LaTeX diff and enough of the chapter(s) rendered as PDF to enable review, or clearly indicate the full scope (ranges of pages) of the change. 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 May 29, 2012, at 2:49 PM, Richard Graham wrote: > Attached is the list of tickets that have passed all but the final 3.0 chapter vote, and have not yet been incorporated into the approved version – some were taken care of as this list was being created. I have included the ticket number, ticket owner, and the affected chapter. > > Authors: Please send me an update on the affected chapters, AND work with each of the chapter authors to make sure the changes get moved into the approved section. > > >>> As we discussed at the current meeting, THE GOAL IS TO GET ALL PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << > The goal is to let people start to read over the new version of the standard before the July meeting, so that we can get a draft out for comment 1-2 weeks after the July meeting, and allow for comments before the final vote in Sept. If you can’t make that date, please let me know ASAP, so we will find someone else who can do the work. > > Thanks, > Rich > <MPI 3.0 tickets for inclusion.xlsx>_______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

MPI 3.0 Passed tickets
by Richard Graham 29 May '12

29 May '12

Attached is the list of tickets that have passed all but the final 3.0 chapter vote, and have not yet been incorporated into the approved version - some were taken care of as this list was being created. I have included the ticket number, ticket owner, and the affected chapter. Authors: Please send me an update on the affected chapters, AND work with each of the chapter authors to make sure the changes get moved into the approved section. >>> As we discussed at the current meeting, THE GOAL IS TO GET ALL PASSED CHANGES INTO THE APPROVED REPOSITORY BY JUNE 15. << The goal is to let people start to read over the new version of the standard before the July meeting, so that we can get a draft out for comment 1-2 weeks after the July meeting, and allow for comments before the final vote in Sept. If you can't make that date, please let me know ASAP, so we will find someone else who can do the work. Thanks, Rich

1 0

Re: [Mpi-forum] why (int)0x4c000101 ?
by Dave Goodell 27 May '12

27 May '12

To the OP's original question, this isn't really the right venue to discuss this sort of thing. This list is for discussing MPI standardization issues, not implementation issues for specific MPI implementations. If you wish to discuss this sort of thing further, please do so on mpich-discuss(a)mcs.anl.gov or users(a)open-mpi.org. On May 28, 2012, at 12:51 AM GMT+09:00, Jeff Hammond wrote: > One will observe that information appears to be coded in those values. > You see that the third-to-last hex-digit indicates the number of bytes > in the type, with the exception of long double, which uses c=12 when > the type is actually 16 bytes. I don't know why this is, but it > doesn't really matter since there is only one built-in type of this > size. Inspecting the first- and second-to-list hex-digits indicates > that signed and unsigned types often (but not always) differ by only > one bit. FWIW, it's encoding the type size in the second least significant byte. That is, the value given by ((type & 0x0000ff00) >> 8). As for "long double", on some platforms this actually is an 80-bit "double extended" floating point type corresponding to the actual width of x86 floating point registers. 80 bits is only 10 bytes (0xa in hexadecimal), but compilers store it in a 12-byte slot in order to keep all of the values 4-byte aligned. -Dave

1 0