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"

Use PowerParameters instead for greater flexibility

an error in powercap logic
a bug in reservation
a bug in licenses
forgotten declaration of functions
forgotten include
error in format
fix incompatible plugin version for power layout
correct target for wrong protocol version on pack/unpack functions

The previous power reservation check, replicated from job_test_lic_resv
is not well protected against consecutives reservations for which
reserved watts (or licenses count) can not be added directly.
Thus, if the job overlaps multiple non-overlapping reservations,
it will be prevented to use more watts (or licenses) than necessary.

Correct this bug building a planning of the evolution of the
constraints and then extracting the max amount of watts reserved
by potentially overlapping reservations.

Also correct a bug resulting in valid power check when more watts
than the defined powercap were reserved.

(Adapted for 15.08 by Yiannis Georgiou <yiannis.georgiou@bull.netwq>)

(Adapted for 15.08 from Yiannis Georgiou <yiannis.georgiou@bull.net>)

(Adapted for 15.08 by Yiannis Georgiou <yiannis.georgiou@bull.net>)

modify select_nodes() logic in order to compute the amount of power
requested by the system if the returned list of nodes are really used to
run the job. If this value is greater than the current power cap, then
return ESLURM_POWER_NOT_AVAIL, a new error code equivalent to
ESLURM_NODES_BUSY for the scheduling logic.

When the backfilling is active, this error code will prevent the main
schedule() function of slurm to attempt scheduling other jobs on the
same resources/partition and will leave the job to the backfill logic
(as it is the case with ESLURM_NODES_BUSY).

Note that no eligible start time is currently computed for jobs blocked
because of ESLURM_POWER_NOT_AVAIL, thus the backfilling logic will only
act as a FIFO concerning the power resource. Thus a job with a lower
priority could steal resources to run if the available power is
sufficient for it, creating starvation of large power rejected jobs.
This should be modified in the future.

Based upon the layouts power framework.

(Adapted for layouts and 15.08 by Yiannis Georgiou <yiannis.georgiou@bull.net>)

(Adapted for 15.08 by Yiannis Georgiou <yiannis.georgiou@bull.net>)