Both. It originally meant we need to get the fix out ASAP. Follow-up
e-mail (Louis sent privately to me, I responded just now to the users
list to publish/archive the info) suggests that our fixes don't fix his
problem. So, the fixes are insufficient, or he's encountering a
different problem.
Jeff Squyres wrote:
...just catching up after the holidays...
Just to ensure I understand: does this mean that the sm issue is *not* yet
resolved? Or does it mean that it *is* resolved on the 1.4 branch and we need
to get 1.4.1 out ASAP?
On Jan 4, 2010, at 10:17 AM, Eugene Loh wrote:
Open MPI wrote:
Comment(by eugene):
Another thread has been attributed to this problem:
http://www.open-mpi.org/community/lists/users/2010/01/11674.php [[BR]]
''Dual quad core Opteron hangs on Bcast.''
In this case, repeated broadcasts are hanging with Fedora FC11.
Looks like we're getting the trac 2043 hang with FC11. The problem
appears to be widespread. Is there anything we should do to fix this
particular distribution? Talk to the Fedora people?
