Previously the SIGCONT would be generated easily as part of gang
scheduling and cause the job step initiaition to abort.
This is being back-ported from SLURM v2.5/master.

One of our testers discovered an inconsistency between trigger names that
are displayed with 'strigger --get' and the trigger names passed to trigger
programs.

accounting as well.

We've discovered that our implementation of proctrack/cgroups includes
internal threads as well as the process pids so that "scontrol list pids"
shows them as well. The attached patch (courtesy of Matthieu Hautreux)
to 2.4.2 fixes this problem.

running on it are preemptable by scheduling job.

reported some suspended time.

static blocks.

correctly when dealing with systems larger than 1 midplane :).

This code was obsoleted by patch
f365e8c5

and going to create.

when none are requested.  The Block allocator sets it up temporarily so
this isn't needed.

and a job comes through backfill and can fit on the block without ending
jobs don't set an end_time for the running jobs since they don't need to
end to start the job.

confused with the functions jobinfo.

Return ESLURM_NODES_BUSY rather than ESLURM_NODE_NOT_AVAIL error on job
submit when required nodes are up, but completing a job or in exclusive
job allocation.

previous 20 minute time limit.  The previous behavior would fail for
large files 20 minutes into the transfer.

If a user tries to to clear message that he does not own, return
access denied rather than not found error