To follow up for the web archives:
Chris helped us out by tracking down the problem to bad command line
parsing in mpirun (a.k.a. orterun) and submitted a patch to fix it.
I've committed the fix on both the trunk and and the v1.0 branch.
I'll make a 1.0.2a4 shortly that contains this (and other) fixes.
Thanks Chris!
On Jan 12, 2006, at 5:59 PM, Chris Gottbrath wrote:
Tom,
Just saw your note so I thought I would
check this out.
I did see the behavior you describe with
/bin/bash when I did
mpirun -tv -np 4 a.out
TV shouldn't be crashing like that. Please
feel free to report that as a bug to Etnus.
Meanwhile, I do have success with
totalview mpirun -a -np 4 a.out
I don't think I have any OpenMPI
specific tcl scripts in my .totalview
dir.
If you continue to have trouble
give us a shout here at Etnus.
We didn't have OpenMPI integrated
into our testing system so it
isn't listed as supported but
it should be possible to get at
least process aquisition working
between TV 7.1.0 and OpenMPI.
Cheers,
Chris
--
Chris Gottbrath
Partner Technologies Engineer Etnus, LLC
chris.gottbr...@etnus.com http://www.etnus.com/
Voice: 508-652-7700 x7735 Fax: 508-652-7787
_______________________________________________
users mailing list
us...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users
--
{+} Jeff Squyres
{+} The Open MPI Project
{+} http://www.open-mpi.org/
