Add a new GRES function to return the count of elements of a specific
  GRES name
Add a new GRES funtion to reset environment variables based upon the
  GRES which are usable from the CPUs to which a task is bound

Add srun --accel-bind option to control how tasks are bound to GPUs and NIC
    Generic RESources (GRES).
Based in part upon work by Matthieu Ospici (ATOS).
gres/nic plugin modified to set OMPI_MCA_btl_openib_if_include environment
    variable based upon allocated devices (usable with OpenMPI and Melanox).
Reset GRES env vars after task affinity set

Conflicts:
	src/squeue/print.c

If an salloc or srun command is executed on a "front-end" configuration,
    that job will be assigned a slurmd shepherd daemon on the same host as used
    to execute the command when possible rather than an slurmd daemon on an
    arbitrary front-end node.

Disable a test for select/linear and only one node

Disable test with select/linear and only one node

Bug 1495

This implements the same logic changes to the SGI hypercube and
  dragonfly network logic was was done to the regular and topology
  portions of the code in version 14.11, commit
  58623ec7

Conflicts:
	src/sreport/job_reports.c

It was deplicated years ago. The option will still work though.

Correct count of CPUs allocated to job on system with hyperthreads.
The bug was introduced in commit a6d3074d
On a system with hyperthreads:
srun -n1 --ntasks-per-core=1 hostname
you would get:
slurmctld: error: job_update_cpu_cnt: cpu_cnt underflow on job_id 67072

preempt/job_prio plugin: Implement the concept of Warm-up Time here. Use
the QoS GraceTime as the amount of time to wait before preempting.
Basically, skip preemption if your time is not up.

Conflicts:
	src/plugins/accounting_storage/mysql/as_mysql_rollup.c
	src/slurmctld/acct_policy.c

-W/--wait is only supported by srun and should not show
up in the usage string of sbatch or salloc.

These fields are not currently in use, but have been added for
  future use.

a job runs past it's time limit.

Permit the powercap to be changed when NOT using the layouts/power framework
Fix memory leak
Cosmetic changes

have one.

Bug 1705

Don't clear a reservations power cap in update reqeust comes from older
  version of scontrol
Fix two memory leaks due to missing list_iterator_destroy() functions
Some minor format changes

In "if (use_cluster_list != arch_cond->job_cond->cluster_list)"
"arch_cond->job_cond" could be NULL

Change error message contents from "PowerCap" to "PowerParameters"
Fix scontrol parsing for update argument so "partition" requires at
  least two characters ("pa") as does "powercapping" ("po")
Prevent use of unitialized variable "rc" in scontrol_update_layout()
Remove dead (unused) variable store in "tag_len"