Merge remote-tracking branch 'origin/slurm-2.6'

Conflicts:
	NEWS
	doc/html/team.shtml

NEWS

@@ -203,9 +203,6 @@ documents those changes that are of interest to users and admins.
  -- Sched/backfill - Change default max_job_bf parameter from 50 to 100.
  -- Added -I|--item-extract option to sh5util to extract data item from series.
 
-* Changes in Slurm 2.6.7
-========================
-
 * Changes in Slurm 2.6.6
 ========================
  -- sched/backfill - Fix bug that could result in failing to reserve resources
@@ -243,6 +240,7 @@ documents those changes that are of interest to users and admins.
  -- Update documentation about QOS limits
  -- Retry task exit message from slurmstepd to srun on message timeout.
  -- Correction to logic reserving all nodes in a specified partition.
+ -- Added support for selecting AMD GPU by setting GPU_DEVICE_ORDINAL env var.
 
 * Changes in Slurm 2.6.5
 ========================

doc/html/team.shtml

@@ -35,6 +35,7 @@ Lead Slurm developers are:
 <li>Leith Bade (Australian National University)</li>
 <li>Troy Baer (The University of Tennessee, Knoxville)</li>
 <li>Susanne Balle (HP)</li>
+<li>Dominik Bartkiewicz (University of Warsaw, Poland)</li>
 <li>Ralph Bean (Rochester Institute of Technology)</li>
 <li>Alexander Bersenev (Institute of Mathematics and Mechanics, Russia)</li>
 <li>David Bigagli (SchedMD)</li>
@@ -187,6 +188,6 @@ Lead Slurm developers are:
 <!-- INDIVIDUALS, PLEASE KEEP IN ALPHABETICAL ORDER -->
 </ul>
 
-<p style="text-align:center;">Last modified 18 November 2013</p>
+<p style="text-align:center;">Last modified 5 February 2014</p>
 
 <!--#include virtual="footer.txt"-->

doc/html/topology.shtml

@@ -132,6 +132,16 @@ SwitchName=s3 Nodes=tux[12-15]
 SwitchName=s4 Switches=s[0-3]
 </pre>
 
+<p>Note that compute nodes on switches that lack a common parent switch can
+be used, but no job will span leaf switches without a common parent.
+For example, it is legal to remove the line "SwitchName=s4 Switches=s[0-3]"
+from the above topology.conf file.
+In that case, no job will span more than four compute nodes on any single leaf
+switch.
+This configuration can be useful if one wants to schedule multiple phyisical
+clusters as a single logical cluster under the control of a single slurmctld
+daemon.</p>
+
 <h2>User Options</h2>
 
 <p>For use with the topology/tree plugin, user can also specify the maximum
@@ -161,6 +171,6 @@ The value will be set component types listed in SLURM_TOPOLOGY_ADDR.
 Each component will be identified as either "switch" or "node".
 A period is used to separate each hardware component type.</p>
 
-<p style="text-align:center;">Last modified 13 January 2014</p>
+<p style="text-align:center;">Last modified 5 February 2014</p>
 
 <!--#include virtual="footer.txt"-->

src/plugins/gres/gpu/gres_gpu.c

@@ -250,6 +250,8 @@ extern void job_set_env(char ***job_env_ptr, void *gres_ptr)
 	if (dev_list) {
 		env_array_overwrite(job_env_ptr,"CUDA_VISIBLE_DEVICES",
 				    dev_list);
+		env_array_overwrite(job_env_ptr,"GPU_DEVICE_ORDINAL",
+				    dev_list);
 		xfree(dev_list);
 	}
 }
@@ -294,6 +296,8 @@ extern void step_set_env(char ***job_env_ptr, void *gres_ptr)
 	if (dev_list) {
 		env_array_overwrite(job_env_ptr,"CUDA_VISIBLE_DEVICES",
 				    dev_list);
+		env_array_overwrite(job_env_ptr,"GPU_DEVICE_ORDINAL",
+				    dev_list);
 		xfree(dev_list);
 	}
 }