This function can now be called to test for processes which are
dumping in order to avoid sending them a SIGKILL until dump
completes. Change in logic required for job_container/cray.

Handle invalid process ID better (probably never happens)
Re-arrange logic to eliminate duplicate file close on error

plugins more maintainable.

went from the old single table per enterprise style to that of separate
tables per clusters.  (2.0 -> 2.*).  If people are still running <2.2 they
really need to upgrade (long before this), and can get the translations
by upgrading to >=2.1.0 before they install this version.

the reverse of this comment to return the code to the mix.

Include a prefix of "NRT" to distinguish switch/nrt state files
from other possible switch state files.

no change in logic

Visible only using "scontrol show job" today and these fields are
only relevant for batch jobs.

Remove documentation about "*cpu" portion of gres specification
in the man pages for salloc, sbatch, and srun. Support for this
specification was never implemented nor does the GRES data structure
include a field for it.

Add mechanism for job_submit plugin to generate error message for srun
salloc or sbatch to log. New argument added to job_submit function in
the plugin.
bug 278

with a vertical bar separating each field.

Previously only the spawned job steps would be signalled rather
than the batch shell

as the job completes.

of at multifactor poll.

Used to extract plugin specific fields from the job's or step's
GRES data structure.

Conflicts:
	testsuite/expect/test1.33

Need higher memory limits due to pmdv12 size
pmdv12 fails to recognize immediate application exit, hangs with defunct process