This happens when a job has multiple partitions and
priority/multifactor is NOT in use

fix of the following problem:
if a node is excised from a job and a reconfiguration(e.g., update
partition) is done when the job is still running, the node will be left
in state idle but not available any more until the next
reconfiguration/restart of slurmctld after the job finished.

to be edited more.

to be running in the calling program.

the node tab and we didn't notice.

which reads the array boundary.

the daemon to core dump.

Response to bug 274

If a user requests multiple counts of a specific GRES type (e.g.
"--gres=gpu:2") and those GRES are associated with specific CPUs,
the job submit could fail without this change.