[mvapich-discuss] Possible RMA passive mode deadlock

sreeram potluri potluri at cse.ohio-state.edu
Fri Feb 18 13:06:14 EST 2011


Jim,

Thank you for reporting the issue. Thanks for narrowing it down and
providing a test case too. I will take a look at this and get back to you
soon.

Sreeram Potluri

On Fri, Feb 18, 2011 at 12:39 PM, James Dinan <dinan at mcs.anl.gov> wrote:

> Hi,
>
> I've been trying to pin down a deadlock problem in my application and I
> think I've narrowed it down to a small MPI test that locks up under MVAPICH2
> 1.5 through the latest 1.6 release candidate.  I've tried the same test on
> MPICH2 version 1.2.1p1 and 1.3.2p1 and it does not appear to have this
> problem.
>
> The deadlock appears to occur when multiple passive mode put operations
> with exclusive locks contend with one another.  I've attached my test case;
> this test locks up for me on our Fusion cluster with 4+ processes and 1
> process per node.  I've also attached a stack trace for each process at the
> point when things lock up.
>
> Any suggestions on how to proceed?
>
> Thanks,
>  ~Jim.
>
> _______________________________________________
> mvapich-discuss mailing list
> mvapich-discuss at cse.ohio-state.edu
> http://mail.cse.ohio-state.edu/mailman/listinfo/mvapich-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.cse.ohio-state.edu/pipermail/mvapich-discuss/attachments/20110218/c60a7c9f/attachment-0001.html


More information about the mvapich-discuss mailing list