February 2011 - mpi-forum - mailman3.cels.anl.gov

MPI_UNDEFINED
by Jeff Squyres 04 Feb '11

04 Feb '11

Question about MPI_UNDEFINED (and all the constants in the top table in MPI-2.2 p515): its type in C is listed as (const int) or unnamed (enum). But in Open MPI, we have this value in mpi.h as: #define MPI_UNDEFINED -32766 So does MPICH2. A #define of an integer constant will work just fine with other const int's, but it technically isn't a const int or unnamed enum. Are we wrong? Or should I propose a ticket to change the wording of this table from: C type: const int (or unnamed enum) to C type: const int, unnamed enum, or integer constant (the same issue doesn't exist for the C++ constants -- even though the point is moot with them, since they're deprecated -- because they're in the MPI:: namespace, and therefore really should be const int's or unnamed enum's) -- 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] Cleaning up the MPI Forum web site
by William Gropp 04 Feb '11

04 Feb '11

On the CMS, I'm moving the files to Illinois, and I plan to use the standard system here (I'll know what that is in a day or two :) ). Bill On Feb 4, 2011, at 9:50 AM, Jeff Squyres wrote: > On Feb 4, 2011, at 10:46 AM, William Gropp wrote: > >> The MPI Forum web site (www.mpi-forum.org) dates from the MPI-2 >> effort and contains a lot of stuff that is of limited use (for >> example, the email archives for the MPI 2 process). I'd like to >> clean up the web site (retaining copies of everything), limiting >> the web site to the documents, description of the Forum and >> processes, and pointers to the ongoing efforts. My questions are >> >> (1) Do you agree with removing (though retaining a copy) the >> archives from MPI-2.0. Note that there are no MPI-1 archives, and >> the MPI-2.0 archives contain only the emails of some of the >> discussions. I'd rather not include incomplete archives, and at >> this point, I don't think they have much value. > > Sounds good to me. > >> (2) I propose a minimal reformat of the page - nothing fancy, but a >> change. Is that ok, or should the pages stay the same? > > Is it worth moving to a CMS of some kind (e.g., Drupal)? I don't > think we'd care about 90% of the features of a CMS -- i.e., the web > site can still be just as simple as it is now -- but it might be > nice to not have to worry about HTML, etc. > > -- > Jeff Squyres > jsquyres(a)cisco.com > For corporate legal information go to: > http://www.cisco.com/web/about/doing_business/legal/cri/ > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum William Gropp Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign

1 0

Re: [Mpi-forum] Cleaning up the MPI Forum web site
by Jeff Squyres 04 Feb '11

04 Feb '11

On Feb 4, 2011, at 10:46 AM, William Gropp wrote: > The MPI Forum web site (www.mpi-forum.org) dates from the MPI-2 effort and contains a lot of stuff that is of limited use (for example, the email archives for the MPI 2 process). I'd like to clean up the web site (retaining copies of everything), limiting the web site to the documents, description of the Forum and processes, and pointers to the ongoing efforts. My questions are > > (1) Do you agree with removing (though retaining a copy) the archives from MPI-2.0. Note that there are no MPI-1 archives, and the MPI-2.0 archives contain only the emails of some of the discussions. I'd rather not include incomplete archives, and at this point, I don't think they have much value. Sounds good to me. > (2) I propose a minimal reformat of the page - nothing fancy, but a change. Is that ok, or should the pages stay the same? Is it worth moving to a CMS of some kind (e.g., Drupal)? I don't think we'd care about 90% of the features of a CMS -- i.e., the web site can still be just as simple as it is now -- but it might be nice to not have to worry about HTML, etc. -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Cleaning up the MPI Forum web site
by William Gropp 04 Feb '11

04 Feb '11

The MPI Forum web site (www.mpi-forum.org) dates from the MPI-2 effort and contains a lot of stuff that is of limited use (for example, the email archives for the MPI 2 process). I'd like to clean up the web site (retaining copies of everything), limiting the web site to the documents, description of the Forum and processes, and pointers to the ongoing efforts. My questions are (1) Do you agree with removing (though retaining a copy) the archives from MPI-2.0. Note that there are no MPI-1 archives, and the MPI-2.0 archives contain only the emails of some of the discussions. I'd rather not include incomplete archives, and at this point, I don't think they have much value. (2) I propose a minimal reformat of the page - nothing fancy, but a change. Is that ok, or should the pages stay the same? I will produce a draft version for inspection before making the change. This is part of moving the material to a new server, in part in reaction to the recent access outage. Let me know what you think. Bill William Gropp Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign

1 0

Re: [Mpi-forum] Feb 7-9 MPI Forum meeting
by Graham, Richard L. 03 Feb '11

03 Feb '11

Yes ----- Original Message ----- From: Bronis R. de Supinski [mailto:[email protected]] Sent: Thursday, February 03, 2011 08:51 PM To: Graham, Richard L. Cc: Main MPI Forum mailing list <mpi-forum(a)lists.mpi-forum.org>; Torsten Hoefler <htor(a)illinois.edu> Subject: Re: [Mpi-forum] Feb 7-9 MPI Forum meeting Rich: Are we going to make this agenda changes? Thanks, Bronis On Tue, 1 Feb 2011, Bronis R. de Supinski wrote: > > Rich: > > Re: >> This is fine with me - Do realize that the RMA session on Monday afternoon >> is plenary, in preparation for the formal reading in one of the very near >> meetings ? > > Yes. I do not see a way to accommodate all of the > plenaries without one guaranteed late evening. It > would be nice to have the option to have one not > go late (my proposal does that for Tuesday for > those of us not involved in the RMA WG). > > Thanks, > > Bronis > > >> >> Rich >> >> >> On 2/1/11 12:26 PM, "Bronis R. de Supinski" <bronis(a)llnl.gov> wrote: >> >> >> >> Rich and Torsten: >> >> How about we switch the sparse collectives plenary >> session with the profiling interfaces and external >> interfaces sessions but also move the persistence >> working group to Tuesday morning at the same time >> as th first RMA working group session (instead of >> at the same time as the second RMA WG session)? >> >> That would allow many of us to finish at 4PM Tuesday... >> >> Bronis >> >> >> >> On Tue, 1 Feb 2011, Torsten Hoefler wrote: >> >>> Bronis, >>>> Re: >>>>> A couple of items: >>>>> - I just posted a revised agenda for Monday, to accommodate a special >>>>> request. If this is a problem for someone, please let me know. >>>> >>>> It seems really bizarre to have the RMA plenary session >>>> Monday afternoon so that we have plenaries all day Monday >>>> only to have only the RMA working group session from 9-11AM >>>> on Tuesday. The prior schedule allows many of us to finish >>>> at 4 or 5PM on Monday. I think that schedule is highly preferable. >>> I had the special request to move the RMA plenary to the first day. >>> The reason is that I have to be in Washington on Tuesday and will only >>> be around on Monday afternoon. Actually, it would be great if the Sparse >>> Collective reading would also be on Monday instead of Tuesday (unless >>> somebody else wants to read it). I don't need to attend the Profiling >>> Interface and External Interfaces (even though they sound very >>> interesting). >>> >>> I apologize for any inconvenience. >>> >>> Thanks & Best, >>> Torsten >>> >>> -- >>> bash$ :(){ :|:&};: --------------------- 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 >>> >> _______________________________________________ >> 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] Feb 7-9 MPI Forum meeting
by Bronis R. de Supinski 03 Feb '11

03 Feb '11

Rich: Are we going to make this agenda changes? Thanks, Bronis On Tue, 1 Feb 2011, Bronis R. de Supinski wrote: > > Rich: > > Re: >> This is fine with me - Do realize that the RMA session on Monday afternoon >> is plenary, in preparation for the formal reading in one of the very near >> meetings ? > > Yes. I do not see a way to accommodate all of the > plenaries without one guaranteed late evening. It > would be nice to have the option to have one not > go late (my proposal does that for Tuesday for > those of us not involved in the RMA WG). > > Thanks, > > Bronis > > >> >> Rich >> >> >> On 2/1/11 12:26 PM, "Bronis R. de Supinski" <bronis(a)llnl.gov> wrote: >> >> >> >> Rich and Torsten: >> >> How about we switch the sparse collectives plenary >> session with the profiling interfaces and external >> interfaces sessions but also move the persistence >> working group to Tuesday morning at the same time >> as th first RMA working group session (instead of >> at the same time as the second RMA WG session)? >> >> That would allow many of us to finish at 4PM Tuesday... >> >> Bronis >> >> >> >> On Tue, 1 Feb 2011, Torsten Hoefler wrote: >> >>> Bronis, >>>> Re: >>>>> A couple of items: >>>>> - I just posted a revised agenda for Monday, to accommodate a special >>>>> request. If this is a problem for someone, please let me know. >>>> >>>> It seems really bizarre to have the RMA plenary session >>>> Monday afternoon so that we have plenaries all day Monday >>>> only to have only the RMA working group session from 9-11AM >>>> on Tuesday. The prior schedule allows many of us to finish >>>> at 4 or 5PM on Monday. I think that schedule is highly preferable. >>> I had the special request to move the RMA plenary to the first day. >>> The reason is that I have to be in Washington on Tuesday and will only >>> be around on Monday afternoon. Actually, it would be great if the Sparse >>> Collective reading would also be on Monday instead of Tuesday (unless >>> somebody else wants to read it). I don't need to attend the Profiling >>> Interface and External Interfaces (even though they sound very >>> interesting). >>> >>> I apologize for any inconvenience. >>> >>> Thanks & Best, >>> Torsten >>> >>> -- >>> bash$ :(){ :|:&};: --------------------- 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 >>> >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> >> >

1 0

Travel next week
by Jeff Squyres 03 Feb '11

03 Feb '11

The US is getting pummeled with a massive snow storm right now. It will be interesting to see how many of us make it next week! Best of luck in your travels... -- 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] Feb 7-9 MPI Forum meeting
by Bronis R. de Supinski 01 Feb '11

01 Feb '11

Rich: Re: > This is fine with me - Do realize that the RMA session on Monday > afternoon is plenary, in preparation for the formal reading in one of > the very near meetings ? Yes. I do not see a way to accommodate all of the plenaries without one guaranteed late evening. It would be nice to have the option to have one not go late (my proposal does that for Tuesday for those of us not involved in the RMA WG). Thanks, Bronis > > Rich > > > On 2/1/11 12:26 PM, "Bronis R. de Supinski" <bronis(a)llnl.gov> wrote: > > > > Rich and Torsten: > > How about we switch the sparse collectives plenary > session with the profiling interfaces and external > interfaces sessions but also move the persistence > working group to Tuesday morning at the same time > as th first RMA working group session (instead of > at the same time as the second RMA WG session)? > > That would allow many of us to finish at 4PM Tuesday... > > Bronis > > > > On Tue, 1 Feb 2011, Torsten Hoefler wrote: > >> Bronis, >>> Re: >>>> A couple of items: >>>> - I just posted a revised agenda for Monday, to accommodate a special >>>> request. If this is a problem for someone, please let me know. >>> >>> It seems really bizarre to have the RMA plenary session >>> Monday afternoon so that we have plenaries all day Monday >>> only to have only the RMA working group session from 9-11AM >>> on Tuesday. The prior schedule allows many of us to finish >>> at 4 or 5PM on Monday. I think that schedule is highly preferable. >> I had the special request to move the RMA plenary to the first day. >> The reason is that I have to be in Washington on Tuesday and will only >> be around on Monday afternoon. Actually, it would be great if the Sparse >> Collective reading would also be on Monday instead of Tuesday (unless >> somebody else wants to read it). I don't need to attend the Profiling >> Interface and External Interfaces (even though they sound very >> interesting). >> >> I apologize for any inconvenience. >> >> Thanks & Best, >> Torsten >> >> -- >> bash$ :(){ :|:&};: --------------------- 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 >> > _______________________________________________ > 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] Feb 7-9 MPI Forum meeting
by Graham, Richard L. 01 Feb '11

01 Feb '11

This is fine with me - Do realize that the RMA session on Monday afternoon is plenary, in preparation for the formal reading in one of the very near meetings ? Rich On 2/1/11 12:26 PM, "Bronis R. de Supinski" <bronis(a)llnl.gov> wrote: Rich and Torsten: How about we switch the sparse collectives plenary session with the profiling interfaces and external interfaces sessions but also move the persistence working group to Tuesday morning at the same time as th first RMA working group session (instead of at the same time as the second RMA WG session)? That would allow many of us to finish at 4PM Tuesday... Bronis On Tue, 1 Feb 2011, Torsten Hoefler wrote: > Bronis, >> Re: >>> A couple of items: >>> - I just posted a revised agenda for Monday, to accommodate a special >>> request. If this is a problem for someone, please let me know. >> >> It seems really bizarre to have the RMA plenary session >> Monday afternoon so that we have plenaries all day Monday >> only to have only the RMA working group session from 9-11AM >> on Tuesday. The prior schedule allows many of us to finish >> at 4 or 5PM on Monday. I think that schedule is highly preferable. > I had the special request to move the RMA plenary to the first day. > The reason is that I have to be in Washington on Tuesday and will only > be around on Monday afternoon. Actually, it would be great if the Sparse > Collective reading would also be on Monday instead of Tuesday (unless > somebody else wants to read it). I don't need to attend the Profiling > Interface and External Interfaces (even though they sound very > interesting). > > I apologize for any inconvenience. > > Thanks & Best, > Torsten > > -- > bash$ :(){ :|:&};: --------------------- 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 > _______________________________________________ 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] Feb 7-9 MPI Forum meeting
by Torsten Hoefler 01 Feb '11

01 Feb '11

On Tue, Feb 01, 2011 at 09:26:09AM -0800, Bronis R. de Supinski wrote: > > Rich and Torsten: > > How about we switch the sparse collectives plenary > session with the profiling interfaces and external > interfaces sessions but also move the persistence > working group to Tuesday morning at the same time > as th first RMA working group session (instead of > at the same time as the second RMA WG session)? > > That would allow many of us to finish at 4PM Tuesday... I don't mind. We could also move the sparse collectives reading to the next meeting, which would give us more time for reviews. Thanks & Best, Torsten -- bash$ :(){ :|:&};: --------------------- 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