8 months agoAdd bullet about MX MTL. v1.6 v1.6.2
jsquyres [Tue, 18 Sep 2012 10:02:09 +0000] rev 20751
Add bullet about MX MTL.

8 months agoFixes #3317: Fix MX MTL in v1.6 v1.6
jsquyres [Tue, 18 Sep 2012 10:01:00 +0000] rev 20750
Fixes #3317: Fix MX MTL in v1.6

Custom patch from
http://www.open-mpi.org/community/lists/users/2012/09/20202.php;
essentially reverting r26698 (from the trunk).

8 months agoSync NEWS with v1.6 branch.
jsquyres [Tue, 18 Sep 2012 10:03:09 +0000] rev 20749
Sync NEWS with v1.6 branch.

8 months agoFix the bind_downward code - it was incorrectly looking across the entire node instead of only looking below the locale to which the proc had been assigned. In other words, if the proc was mapped to a core, then the only hwthreads that should be considered for binding are those directly below that core. The binding algo was incorrectly looking at ALL hwthreads in that scenario, causing the proc to be bound to an HT outside of the mapped location.
rhc [Fri, 14 Sep 2012 22:01:19 +0000] rev 20748
Fix the bind_downward code - it was incorrectly looking across the entire node instead of only looking below the locale to which the proc had been assigned. In other words, if the proc was mapped to a core, then the only hwthreads that should be considered for binding are those directly below that core. The binding algo was incorrectly looking at ALL hwthreads in that scenario, causing the proc to be bound to an HT outside of the mapped location.

This now results in the procs being bound within their assigned location. It also causes us to use only the 0th HT on a core unless --use-hwthread-cpus has been specified (in which case, we use all the HTs in a core). Bind to core binds you to all HTs regardless - the --use-hwthread-cpus only impacts the oversubscribed determination and when binding to HT.

cmr:v1.7

8 months ago1.6.2rc3 was already released; on to rc4 v1.6
jsquyres [Fri, 14 Sep 2012 17:06:58 +0000] rev 20747
1.6.2rc3 was already released; on to rc4

8 months agoCMR #3287 missed a couple of changesets that had relevant pieces in them (r27245 and r27264). Fill in the blanks here. v1.7
rhc [Fri, 14 Sep 2012 16:49:26 +0000] rev 20746
CMR #3287 missed a couple of changesets that had relevant pieces in them (r27245 and r27264). Fill in the blanks here.

8 months agoRemove test program that doesn't exist on the v1.6 branch. v1.6
jsquyres [Fri, 14 Sep 2012 13:10:58 +0000] rev 20745
Remove test program that doesn't exist on the v1.6 branch.

8 months agoFixes #3321: Disable udcm in v1.7 v1.7
jsquyres [Thu, 13 Sep 2012 22:40:54 +0000] rev 20744
Fixes #3321: Disable udcm in v1.7

---svn-pre-commit-ignore-below---

r27335 [[BR]]
Per discussion with Ralph and Nathan, disable UDCM for now. It's
borked and needs some surgery to get back on its feet.

8 months agoFixes #3319: Move path_nfs test fix to v1.6 v1.6
jsquyres [Thu, 13 Sep 2012 22:33:03 +0000] rev 20743
Fixes #3319: Move path_nfs test fix to v1.6

---svn-pre-commit-ignore-below---

r27322 [[BR]]
If we get a filesystem type of "none", skip it.

8 months agoFixes #3320: Move path_nfs test fix to v1.7 v1.7
jsquyres [Thu, 13 Sep 2012 22:30:55 +0000] rev 20742
Fixes #3320: Move path_nfs test fix to v1.7

---svn-pre-commit-ignore-below---

r27322 [[BR]]
If we get a filesystem type of "none", skip it.