On Jan 27, 2014, at 1:24 PM, Nathan Hjelm <hje...@lanl.gov> wrote: > On Mon, Jan 27, 2014 at 01:10:43PM -0800, Ralph Castain wrote: >> Nathan, I have no idea what you are talking about. What I do know is that >> you had me commit a patch to the trunk and v1.7.4 that caused multiple >> warnings about 32-bit issues to appear in both cases. George is reporting >> issues that look very much like the ones I'd expect based on those warnings. > > That patch for coll/ml was correct for 1.7.4 and no further action is > required. The two crashed George reported are in 1) the new vader (not > in 1.7.4), and 2) the coll/ml updates (also not in 1.7.4). Neither code > path exists in 1.7.4.
My bad - I rechecked 1.7.4 and found the warnings indeed do not exist there. > >> Release of 1.7.4 is still waiting for the patch you promised me last week to >> fix these problems. I don't give a rats a$$ about SGI at this stage - I just >> want to get your patch that fixes 1.7.4 so we can release the stupid thing!!! > > That patch did not cause any warning on 1.7.4. Only the trunk because it > conflicted with the update to coll/ml. That is not scheduled for 1.7.4 > so I am waiting on ORNL to make sure we do the right thing. > > As far as I am concerned 1.7.4 should be ready to go. Unless some new > 32-bit issue cropped up with coll/ml. Since MTT is suggested we are fine > with 32-bit I doubt that is the case. Brian is doing one last 32-bit smoke check to be sure as none of the MTT tests are running in that mode, so far as I'm aware. Thanks Ralph > > -Nathan > _______________________________________________ > devel mailing list > de...@open-mpi.org > http://www.open-mpi.org/mailman/listinfo.cgi/devel
