Skip to Content.
Sympa Menu

charm - [charm] charmrun with ibverbs fails

charm AT lists.cs.illinois.edu

Subject: Charm++ parallel programming system

List archive

[charm] charmrun with ibverbs fails


Chronological Thread 
  • From: hemmendd AT union.edu (David Hemmendinger)
  • To: charm AT cs.uiuc.edu
  • Subject: [charm] charmrun with ibverbs fails
  • Date: Thu, 17 Nov 2011 16:15:37 -0500
  • List-archive: <http://lists.cs.uiuc.edu/pipermail/charm>
  • List-id: CHARM parallel programming system <charm.cs.uiuc.edu>

We have namd on an IBM cluster running xcat (RedHat Linux kernel
2.6.18-194), with Mellanox OFED 1.5.2-2.1.0. We've downloaded the
Linux-x86_64-ibverbs compiled version of namd 2.8 as well as others. The
TCP versions run, but the ibverbs version fails. Charmrun reports that
the program loads on the compute nodes, but then fails with the message:
CmiAbort("failed to change qp state to RTR").
I've tried recompiling charmrun, using the current version
6.2.1 (which I also obtained with namd, though it's called v 6.3.2).
I specified only "ibverbs" as the option to build, along with x86_64,
but get the same error.
Can anyone suggest what to fix? The only additional evidence
that I have is that when I try the ibverbs ibv_rc_pingpong test on
the compute nodes, I get the same error unless I specify an option
not documented in the man page for ibv_rc_pingpong: -g 0. Mellanox
say that this GID specification is a new option. Does that suggest
that something like it must be specified in running charmrun?
Thanks for any help!

David Hemmendinger
hemmendd AT union.edu
Professor Emeritus http://athena.union.edu/~hemmendd
Computer Science Dept. +1 518 346 4489
Union College, Schenectady, NY 12308 FAX: +1 518 388 6789



  • [charm] charmrun with ibverbs fails, David Hemmendinger, 11/17/2011

Archive powered by MHonArc 2.6.16.

Top of Page