and smaller block when using sub block allocations

to boot.  This should happen very rarely.

starts in accounting.

Add an srun shepard process to cancel a job and/or step of the srun process
is killed abnormally (e.g. SIGKILL).

starts in accounting.

There was an oversight in the GracTime preemption logic that resulted
in repeatedly reading the PreemptType configuration parameter and
doing a string comparison. This change will improve efficiency of
that logic.

One of our testers discovered a regression in version 2.3.1.  If a job is
pending due to PartitionNodeLimit and the limit is relieved with a
'sacctmgr modify qos name=<qos name> set flags=partitionmaxnodes' new jobs
exceeding the partition limit (but not the QOS limit) are allowed to run.
However, the pending job is never allowed to run.  Attached is a patch to
address this problem.  FYI, this problem doesn't exist in version 2.4.
Patch from Bill Brophy, Bull.