Document limitations in burst buffer use by the salloc command (possible
    access problems from a login node).
bug 2883

task/cgroup plugin is configured with ConstrainRAMSpace=yes, then set soft
    memory limit to allocated memory limit (previously no soft limit was set).
bug 2679

of 0.

This might be the cause of run away jobs.  I couldn't see how an end_time
could be 0, but if it was it would just exit and never set time_end to
anything.  At least if it happens now we can have an idea that it is
possible and we will have an idea this is the place it happens.

This will keep from referencing the task array that might not be set up
correctly in src/common/plugstack.c _spank_handle_init().

Prevent possible incorrect counting of GRES of a given type if a node has
    the multiple "types" of a given GRES "name", which could over-subscribe
    GRES of a given type.
bug 2836

cleaning up on a restart.

sleep is killed.

for steps.

for a step.

hostlists when brackets == 0.

Fix for invalid array pointer when creating advanced reservation when job
    allocations span heterogeneous nodes (differing core or socket counts).
bug 2876

Prevent backfill scheduler from starting a second "singleton" job if another
    one started during a backfill sleep.
related to bug 2808

Previous logic required listing each task of job array individually.
For example --dependency="afterany:123_4:123_5" can not be expressed
as --dependency="afterany:123_[4-5]".
bug 2644

we have the pids added to the system correctly.

Most likely related to bug 2874

back to the slurmctld.

Add salloc/sbatch/srun --spread-job to distribute tasks over as many nodes
    as possible. This also treats the --ntasks-node-node option as a maximum
    value. For example "srun -n16 --ntasks-per-node=2 --spread-job ..." will
    run the job on at least 8 nodes (with 2 tasks each), but will attempt to
    use as many nodes as possible up to the task count (1 task per node).
bug 2732

prctld(PR_GET_NAME) >= 2.6.11

In bug #1499, customer asked to precisely identify which srun options apply
to job allocations, which to step allocations and which apply to both. So
the srun man page has been updated with this information.

Move test for job wait reason value of BurstBufferResources and
    BurstBufferStageIn later in the scheduling logic. This will
    result in job reasons related to QOS, account, and partition
    limits to be displayed instead of burst buffer (if applicable)
bug 2835

id wasn't able to be made.

Values are already time_t / printed with %ld so this should
keep Y2038 at bay.

Bug 2862.

node_features/knl_cray plugin modified to get current node MCDRAM
   mode using cnselect command rather than capmc command. This gets the
   current BIOS setting, which could differ from what is configured to be
   set on boot, at least if something other than Slurm is modifying
   KNL modes.
bug 2850

node_features/knl_cray plugin modified to get current node NUMA mode using
    cnselect command rather than capmc command. This gets the current
    BIOS setting, which could differ from what is configured to be
    set on boot, at least if something other than Slurm is modifying
    KNL modes.
bug 2850