August 2011 - mpiwg-hybridpm - mailman3.cels.anl.gov

Re: [Mpi3-hybridpm] MPI_Comm_split_type
by Pavan Balaji 26 Aug '11

26 Aug '11

On 08/26/2011 02:12 AM, Supalov, Alexander wrote: > Thanks. The advice to implementers mentioned an Info argument (see > 23/6). There appears to be no MPI_Info argument in the syntax (see > 22/26). Is this a remnant of the ol'times, or a glint of the future? > > Actually, having an MPI_Info here would be very tempting, as one > could, e.g., ask the new communicator ranks to be reordered to honour > cache hierarchy, or latency, etc. With the time we could then also > agree on some keys/effects common to all implementations. Oops, the info argument was always the plan. I somehow messed up the prototype. Thanks for catching it. We did have a detailed discussion on whether we can predefine some info arguments, but there was no consensus. So we decided to leave it as implementation specific. If you have any ideas, please feel free to propose them. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [Mpi3-hybridpm] MPI_Comm_split_type
by Supalov, Alexander 26 Aug '11

26 Aug '11

Thanks. The advice to implementers mentioned an Info argument (see 23/6). There appears to be no MPI_Info argument in the syntax (see 22/26). Is this a remnant of the ol'times, or a glint of the future? Actually, having an MPI_Info here would be very tempting, as one could, e.g., ask the new communicator ranks to be reordered to honour cache hierarchy, or latency, etc. With the time we could then also agree on some keys/effects common to all implementations. -----Original Message----- From: mpi3-hybridpm-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Pavan Balaji Sent: Friday, August 26, 2011 2:09 AM To: mpi3-hybridpm(a)lists.mpi-forum.org Subject: Re: [Mpi3-hybridpm] MPI_Comm_split_type Folks, Ron and I made some corrections to the MPI_Comm_split_type definition. Attached is a new version of the chapter with these changes. Comments before the next telecon are welcome, so we can minimize the time we spend on this during the telecon. Thanks, -- Pavan On 08/23/2011 08:23 PM, Pavan Balaji wrote: > > Attached is the updated context chapter with the new > MPI_Comm_split_type function (marked in red). Please look through it > to make sure we didn't miss anything. For the official reading, we'll > drop the MPI_COMM_TYPE_PROCESS constant; the endpoints proposal will add it back. > > -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji -------------------------------------------------------------------------------------- Intel GmbH Dornacher Strasse 1 85622 Feldkirchen/Muenchen, Deutschland Sitz der Gesellschaft: Feldkirchen bei Muenchen Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer Registergericht: Muenchen HRB 47456 Ust.-IdNr./VAT Registration No.: DE129385895 Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052

1 0

Re: [Mpi3-hybridpm] Telecon
by Pavan Balaji 25 Aug '11

25 Aug '11

All, The meeting minutes for the last telecon have been uploaded on the wiki: https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/MPI3Hybrid/notes-2011-08-… -- Pavan On 08/15/2011 07:14 PM, Pavan Balaji wrote: > > We'll have our next telecon on Thursday (08/18) at 3pm central time. The > telecon information is on the wiki. > > https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/MPI3Hybrid > > -- Pavan > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [Mpi3-hybridpm] MPI_Comm_split_type
by Pavan Balaji 25 Aug '11

25 Aug '11

Folks, Ron and I made some corrections to the MPI_Comm_split_type definition. Attached is a new version of the chapter with these changes. Comments before the next telecon are welcome, so we can minimize the time we spend on this during the telecon. Thanks, -- Pavan On 08/23/2011 08:23 PM, Pavan Balaji wrote: > > Attached is the updated context chapter with the new MPI_Comm_split_type > function (marked in red). Please look through it to make sure we didn't > miss anything. For the official reading, we'll drop the > MPI_COMM_TYPE_PROCESS constant; the endpoints proposal will add it back. > > -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

[Mpi3-hybridpm] MPI_Comm_split_type
by Pavan Balaji 23 Aug '11

23 Aug '11

Attached is the updated context chapter with the new MPI_Comm_split_type function (marked in red). Please look through it to make sure we didn't miss anything. For the official reading, we'll drop the MPI_COMM_TYPE_PROCESS constant; the endpoints proposal will add it back. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [Mpi3-hybridpm] telcon today
by Pavan Balaji 23 Aug '11

23 Aug '11

Hi Steve, No, we'll have a call next week, not this week. -- Pavan On 08/23/2011 10:39 AM, Steve Oyanagi wrote: > Sorry, I don't remember from the last meeting. Is there a > telecon today at 11:00 CDT (like in about 20 minutes)? > - Steve > _______________________________________________ > Mpi3-hybridpm mailing list > Mpi3-hybridpm(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-hybridpm -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

[Mpi3-hybridpm] telcon today
by Steve Oyanagi 23 Aug '11

23 Aug '11

Sorry, I don't remember from the last meeting. Is there a telecon today at 11:00 CDT (like in about 20 minutes)? - Steve

1 0

Re: [Mpi3-hybridpm] Ticket 217 - a.k.a. Helper Threads
by Douglas Miller 19 Aug '11

19 Aug '11

I have a modified External Interfaces chapter for Ticket 217 "Helper Threads". Both LaTex and PDF are now attached to ticket 217, for review and discussion at the next meeting. The SVN copy of the chapter HAS NOT been updated. https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/217 thanks, _______________________________________________ Douglas Miller BlueGene Messaging Development IBM Corp., Rochester, MN USA Bldg 030-2 A401 dougmill(a)us.ibm.com Douglas Miller/Rochester/IBM

1 0

[Mpi3-hybridpm] MPI_Win_Allocate_shared proposal draft
by Torsten Hoefler 16 Aug '11

16 Aug '11

Hi all, Jim and I updated the MPI_Win_Allocate_shared proposal draft with the feedback from the last Forum. We will discuss if the draft satisfies the needs of the hybrid WG tomorrow at the Hybrid telecon. However, we would also like the RMA group to check the draft, and it's integration in the One Sided Chapter, carefully. Today's draft is at https://svn.mpi-forum.org/trac/mpi-forum-web/attachment/wiki/MPI3Hybrid/win… You can find the changes quickly if you search for ticket 284 in the draft. Thank you very much, Jim & 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

Re: [Mpi3-hybridpm] Telecon
by Pavan Balaji 16 Aug '11

16 Aug '11

On 08/16/2011 06:06 PM, Snir, Marc wrote: > you can circulate the file I sent you -- I wont have time for further changes. Attached. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0