I spoke with Ralph last night (he's still on vacation); there has been a change
of plan.
Unfortunately, the HEAD of the v1.8 branch is currently too unstable for
immediate release (and we really need to do an immediate v1.8.1 release). MTT
is showing a bunch of MPI-layer failures. It does not
I need the flatten datatype call for handling true rdma in the one-sided code
as well. Is there a plan to implement this feature soon?
-Nathan
From: devel [devel-boun...@open-mpi.org] on behalf of George Bosilca
[bosi...@icl.utk.edu]
Sent: Monday, April 2
Per http://www.open-mpi.org/community/lists/devel/2014/04/14572.php and per
discussion on the call this morning, 1.8.1rc1 has been posted:
http://www.open-mpi.org/software/ompi/v1.8/
The only code change is to fix the critical fix regarding session directory
behavior, and to disallow (by de
On Apr 22, 2014, at 1:01 PM, "Jeff Squyres (jsquyres)"
wrote:
> The only code change is to fix the critical fix regarding session directory
> behavior, and to disallow (by default) root from executing mpirun.
...to fix the critical *bug*...
:-)
--
Jeff Squyres
jsquy...@cisco.com
For corpora
Per discussion on the call today, I'm going to start RM approving some of the
most obvious / clearly-needed CMRs that are now pending for v1.8.2. This
mainly includes CMRs that have the potential to improve MTT scores. For
example: compiler errors, obvious bugs, etc.
Other CMRs can wait until
It looks like the standard does not explicitly forbid that combination
of assertions so I will remove the check. Be aware that assertion will
effectively be a no-op because you are saying no operation proceed or
succeed the fence call making the call pointless.
-Nathan
On Mon, Apr 21, 2014 at 02:
Both these issues are fixed in the trunk and are scheduled for the
1.8. The commit you need to check is r28584 and the corresponding
ticket is
https://svn.open-mpi.org/trac/ompi/ticket/4554
George.
On Mon, Apr 21, 2014 at 8:45 AM, Lisandro Dalcin wrote:
> It seems the implementation of MPI_Ad