Bug 2071

Log the minimum and maximum exit code (or signal) and show the
  state as "Mixed" if there are multiple termination states
  (e.g. cancelled, exit 0, exit 1, SIGBUS, etc.)
bug 2058

it needed to check for INFINITE64 instead of zero.

This just removes xfrees where the xfree had already happened.

Correct scheduling logic for job features option with XOR construct that
could delay a job's initiation. Some configurations and states with job
feature XOR construct could prevent the job allocation from being made.
bug 1865

Improve sched/backfill support for job features, especially XOR construct.
bug 1865

Bug 1973

in the plugins.

when a user doesn't exist in the database.

manager.

Squeue recombines pending job array records only if their name and partition
    are identical.
bug 2057

Job array termination status email ExitCode based upon highest exit code
    from any task in the job array rather than the last task.
bug 2058

Based upon patch from Hongjia Cao
bug 2054

as it is the only TRES that can be given to another job while suspended.

sbatch --ntasks option to take precedence over --ntasks-per-node plus node
    count, as documented. Set SLURM_NTASKS/SLURM_NPROCS environment variables
    accordingly.
bug 2015

Avoid reporting more allocated CPUs than exist on a node. This can be
    triggered by resuming a previosly suspended job, resulting in
    oversubscription of CPUs.
bug 2021

Add scancel -f/--full option to signal all steps including batch script and
    all of its child processes.
bug 2031

Bug 1888

out.

Remove unneeded code that commit 8274ea54 fixed.

This code would 0 out all GRES/TRES on a reconfig which isn't what we want.

8274ea54 does the right thing by itself.

bug 2032

Needed to change a couple of variables from 32- to 64-bit.

Add new burst_buffer.conf parameters: ValidateTimeout and OtherTimeout.
See man page for details.

previously take 2 restarts of the slurmdbd to make it stick correctly.

single-threaded cores.  A regression caused only 1 socket to be used on
this kind of node instead of all that were available.

Fix case where if the backup slurmdbd has existing connections when it gives up control that the it would be killed.

If the backup had existing connections when giving up control, it would try to
signal the existing threads by using pthread_kill to send SIGKILL to the
threads. The problem is that SIGKILL doesn't go the thread but the main process
and the backup dbd would be killed.

when a cold-start (-c) happens to the slurmctld.

from a user.

This would cause the slurmctld to cache the old default which wasn't valid
and cause the user to have to request the association always.

byg 2013