Opened 5 years ago

Closed 5 years ago

#266 closed bug (fixed)

Fwd: [mpich-discuss] Problems with mpi spawn multiple

Reported by: William Gropp <wgropp@…> Owned by:
Priority: major Milestone:
Component: mpich Keywords:
Cc:

Description

There is a bug to fix here - the lack of a sensible error message
indicates that Comm_spawn_multiple (and possibly Comm_spawn) isn't
checking the arguments thoroughly.

Bill


Begin forwarded message:

> From: Rajeev Thakur <thakur@mcs.anl.gov>
> Date: November 3, 2008 7:46:04 PM CST
> To: "mpich-discuss@mcs.anl.gov" <mpich-discuss@mcs.anl.gov>
> Subject: Re: [mpich-discuss] Problems with mpi spawn multiple
> Reply-To: "mpich-discuss@mcs.anl.gov" <mpich-discuss@mcs.anl.gov>
>
> You can't pass MPI::ROOT to Spawn_multiple. You have to give the
> rank of the process in COMM_WORLD that is the root.
>
> Rajeev
>
> From: mpich-discuss-bounces@mcs.anl.gov [mailto:mpich-discuss-
bounces@mcs.anl.gov
> ] On Behalf Of fernando_luz
> Sent: Monday, November 03, 2008 2:35 PM
> To: mpich-discuss@mcs.anl.gov
> Subject: [mpich-discuss] Problems with mpi spawn multiple
>
> Hello,
>
> I have problems when i try use spwan multiple in mpich2 (1.0.7
> version)
>
> I receive this error when i execute my code:
>
> [0]Fatal error in MPI_Comm_spawn_multiple: Other MPI error, error
> stack:
> [0]MPI_Comm_spawn_multiple(152)..: MPI_Comm_spawn_multiple(count=2,
> cmds=0x816b080, argvs=(nil), maxprocs=0x816afc0, infos=0x81a1600,
> root=-3, MPI_COMM_WORLD, intercomm=0xbfa96f24, errors=(nil)) failed
> [0]MPID_Comm_spawn_multiple(56)..:
> [0]MPIDI_Comm_spawn_multiple(203):
> [0]MPID_Comm_accept(149).........:
> [0]MPIDI_Comm_accept(974)........: Unable to allocate -45393952
> bytes of memory for remote_translation (probably out of memory)
>
> job aborted:
> rank: node: exit code[: error message]
> 0: a53: -2: Fatal error in MPI_Comm_spawn_multiple: Other MPI error,
> error stack:
> MPI_Comm_spawn_multiple(152)..: MPI_Comm_spawn_multiple(count=2,
> cmds=0x816b080, argvs=(nil), maxprocs=0x816afc0, infos=0x81a1600,
> root=-3, MPI_COMM_WORLD, intercomm=0xbfa96f24, errors=(nil)) failed
> MPID_Comm_spawn_multiple(56)..:
> MPIDI_Comm_spawn_multiple(203):
> MPID_Comm_accept(149).........:
> MPIDI_Comm_accept(974)........: Unable to allocate -45393952 bytes
> of memory for remote_translation (probably out of memory)
>
>
> my code when I supose have a problem:
>
> universe_size = 3;
>
>   for (int i = 0; i < universe_size-1; i++){
>     strcpy(program_name[i], "worker_02");
>     information[i] = information[i].Create();
>     information[i].Set("wdir","/home/fernando_luz/");
>     information[i].Set("path","/home/fernando_luz/SVN/TPN3/
> casos_testes/02/worker_02/");
>     information[i].Set("host","10.2.7.53");
>     n_proc[i] = 1;
>   }
>   everyone = MPI::COMM_WORLD.Spawn_multiple(universe_size-1, (const
> char**) program_name, MPI::ARGVS_NULL, n_proc, information,
> MPI::ROOT);
> anybody have ideas?
>
> Thanks
>
> Fernando Luz

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




Attachments (1)

part0001.html (7.2 KB) - added by William Gropp 5 years ago.
Added by email2trac

Download all attachments as: .zip

Change History (3)

Changed 5 years ago by William Gropp

Added by email2trac

comment:1 Changed 5 years ago by William Gropp

  • id set to 266

This message has 1 attachment(s)

comment:2 Changed 5 years ago by thakur

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.