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>)

However, --mem=0 now reflects the appropriate amount of memory in the
system, --mem-per-cpu=0 hasn't changed.  This allows all the memory to
be allocated in a cgroup but is not "consumed" and is available for
other jobs running on the same host.
Eric Martin, Washington University School of Medicine

Use sleep with value less than 1 (e.g. 0.5)

These changes are required to support NetBSD.
1) Use standard -p option for time command in test17.35 instead of
GNU specific format string to fetch the elapsed (real) time.

2) Adjust date command calls in test22.1 to use short option
(--date -> -d), and move options before arguments. NetBSD,
unlike Linux, disallow non standard order.

3) Use standard -R option instead of -r for cp command in test30.1.

Fix for bug introduced in commit 935b0cbe

Conflicts:
	testsuite/expect/test12.7

bug 1649

bug 1642

Conflicts:
	doc/man/man1/sbatch.1
	src/slurmctld/job_scheduler.c

Correct list of unavailable nodes reported in a job's "reason" field when
that job can not start.
bug 1614

which can be used to aggregate messages to the slurmctld into a single
message to reduce communication to the slurmctld.  Currently only epilog
complete messages and node registration messages use this logic.

'sacctmgr archive dump' is ran.

bug 1684

bug 1679

Do not log invalid layout update calls as errors in slurmctld, that's
  something for the user to address.
Make sure that data structure pointer are not NULL before attempting to
  free fields within them.
Initialize memory for a data structure taken from the stack.

This replaces the logic that reads until we run out of buffer
  space and setting the record count from that.
Also add an unlock if the layout dump failed to prevent a slurmctld
  deadlock.