Add slurmd "-b" option to report node rebooted at daemon start time. Used
    for testing purposes.

Remove obsolete packaging format. Project is dead, website no
longer resolves, and if the package depends on slurm-switch-elan
that's a good sign it hasn't worked for a long time.

Remove obsolete SVN $Id$ tags from top level and src/{api,common}.

reduce reliance on fixed-sized buffers for output, helps reduce
warnings from coverity et al.

split up key/value pairs in preparation for JSON output work.
xstrfmtcat exists and is cleaner than snprintf followed by xstrcat.

use a consistent line ending rather than repeat conditional block.

output format should be unchanged, and has been tested to match
on common cases and passes all relevant regression tests.

Warning flag -Wempty-body in clang-3.7 now errors on this, split
macro definition up over multiple lines to silence.

Both the original logic and modified logic failed to lock the job
data structure prior to decrementing "prolog_running" counter.

in the forward message logic.

Previously under some conditions that boot completion was ignored
and the job kept pending.

silence clang warning about unsupported flag. if we want
performance we have other ways to chase after it, the reason
for this specific flag has long since passed. if you're still
using gcc 3.4 there are plenty of other issues in modern slurm.

run autogen.sh to account for removal of auxdir/x_ac_cflags.m4

Superfluous xstrdup call to passed argument causes
a small memory leak in sreport, caught by Coverity.

Update some node feature logic to support KNL nodes.

If a user submits a job with duplcate partition names (e.g.
  "sbatch --partition=debug,debug,debug", slurm was creating
  multiple (duplicate) records in the job's partition list.
  This change prevents duplicate entries in that list.

Bug 2364

Commit fa331e30 fixes this.  The logic was bad to begin with...

uint32_t new_cpus = detail_ptr->num_tasks
	/ detail_ptr->cpus_per_task;

The / should had been * this whole time.  This was the reason we found
this in the first place.

bug 2369

# Conflicts:
#	src/slurmctld/reservation.c

If scancel is operating on large number of jobs and RPC responses from
    slurmctld daemon are slow then introduce a delay in sending the cancel job
    requests from scancel in order to reduce load on slurmctld.
bug 2256

If a job launch is delayed, the test was failing due to bad parsing.
These lines were being interpretted as a counter folloed by node
names of "queued" and "has":
  srun: job 1332712 queued and waiting for resources
  srun: job 1332712 has been allocated resources

Expand documentation for advanced reservation Purge_Comp flag in
  scontrol man page, add to advanced reservation web page, increase
  minimum start time from 1 to 5 minutes

bug 2366

Conflicts:
	META
	NEWS

Backfill scheduling properly synchronized with Cray Node Health Check.
    Prior logic could result in highest priority job getting improperly
    postponed.
bug 2350