Conflicts:
	NEWS
	doc/html/team.shtml

Set GPU_DEVICE_ORDINAL environment variable.

Splitting the job's shared field into whole_node and share_res
in order to support gang scheduling of jobs with exclusive node
use broke the logic used to support advanced reservations containing
specific cores.

Conflicts:
	META
	NEWS
	contribs/perlapi/libslurm/perl/Slurm.xs

Previous logic would try to pick a specific node count and on a
heterogeneous system, this would cause a problem. This change
largely reverts commit a270417b

beside the numerical values.

the unit test program.

Added whole_node field to job_resources structure
Enable gang scheduling for jobs with core specialization and other
jobs allocated whole nodes.

If a batch job is submitted that can not be started due to its
core specialization count being equal to or larger than the
node's core count, then reject it at submit time rather than
queuing the job.

linking)

If the Cray MPMD commands have different arguments then they should
appear in separate lines in the MPMD data structure.

starts and then a minor typo fix

i.e. salloc -n32 doesn't request the number of nodes and with the previous
code if this request used 4 nodes and only 1 was left in GrpNodes it
would just run with no issue since we were checking things before we
selected how many nodes it ran on.

Now we check this afterwards so we always check the limits on how many
nodes, cpus and how much memory is to be used.