Downing the node will kill all jobs allocated to the node, very bad
on something like a BlueGene system

multiple memory operations which are not necessary.

If slurm.conf defines fewer of a gres than gres.conf AND gres.conf
defines specific CPUs to be associated with the gres, this patch
prevents the bit_overlap call being made with bitmaps of different
sizes, which would cause an abort.

Also modify the epilog script to use this new output field

Timing changes due to faster job step launch

This can be especially useful to schedule GPUs. For example a node can be
associated with two Slurm partitions (e.g. "cpu" and "gpu") and the
partition/queue "cpu" could be limited to only a subset of the node's CPUs,
insuring that one or more CPUs would be available to jobs in the "gpu"
partition/queue.

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.