LAM/MPI General User's Mailing List Archives

From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2004-10-07 12:30:43

• Next message: Jeff Squyres: "Re: LAM: seg fault in lamgrow"
• Previous message: Jeff Squyres: "Re: LAM: 7.1.1 and PBS libs"
• In reply to: David Cronk: "Re: LAM: Spawn, Faulttolerance, Communicators between Child processes"
• Next in thread: Vinod Kannan: "Re: LAM: Spawn, Faulttolerance, Communicators between Child processes"

On Oct 7, 2004, at 10:20 AM, David Cronk wrote:

> Why do you suggest the merge? This seems to add a layer of
> complication. Following the spawn the children are all part of their
> own MPI_COMM_WORLD. Just have the children do the successive splits
> with their MPI_COMM_WORLD.

<smacks head> Duh!

You're absolutely correct. That's a better solution. :-)

Thanks for pointing it out!

-- 
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/

• Next message: Jeff Squyres: "Re: LAM: seg fault in lamgrow"
• Previous message: Jeff Squyres: "Re: LAM: 7.1.1 and PBS libs"
• In reply to: David Cronk: "Re: LAM: Spawn, Faulttolerance, Communicators between Child processes"
• Next in thread: Vinod Kannan: "Re: LAM: Spawn, Faulttolerance, Communicators between Child processes"