This also makes use of the -j (join) flag as a normal qsub does.

Fix for uninitialized variable in task binding logic, could leave tasks
    with fewer CPUs than intended.
bug 2766

Fix GRES task layout bug that could cause slurmctld to abort with
invalid memory reference.
bug 2766

Coverity 44891.

sizeof(optarg) is incorrect, that's the size of the pointer
not the length of the character string that must be parsed.

Coverity 53128.

Coverity 44992.

bug 2760

Needs to unlock here, not re-lock the lock.

Some tests fail if the partition is configured with Shared=exclusive.
This change disables the same tests as if select/linear is configured.

Prevent '--preserve' from being inadvertanly enabled by '-j'.

Logic introduced in v16.05.0-rc2 could attempt state change for
nid00000, even if no such node exists
bug 2758

Still testing 16.05 on my NetBSD/amd64 workstation ...
Just encountered a crash with scancel(1).
njoly@lanfeust [~]> sbatch --wrap "sleep 3600"
Submitted batch job 4680
njoly@lanfeust [~]> scancel 4680
scancel: Error detected by libpthread: Invalid condition variable.
Detected by file "/local/src/NetBSD/src/lib/libpthread/pthread_cond.c", line 140, function "pthread_cond_timedwait".
See pthread(3) for information.
zsh: abort (core dumped)  scancel 4680
Checking the code show indeed that pthread_cond_wait() call from scancel.c:_signal_job_by_str() use an uninitialised condition variable "num_active_threads_cond"
The attached patch, which add the missing pthread_cond_init() seems to fix it.
bug 2753

Change how Slurm determines the NUMA count of a node. Ignore KNL NUMA
    that only include memory.
bug 2745

Fix test for large CPU count nodes

Need thread_id to distinguish between multiple threads with the same
name.

Already installed by 'make install-contrib'. Bug 2747.

# Conflicts:
#	NEWS

and the slurmctld doesn't wait long enough for the response it would give
up leaving the connection open and create a situation where the next message
sent could receive the response of the first one.

Bug 2739

Correct logic that calculates a step's cpus_per_task allocation
on a heterogenous job allocation. Mixing a KNL with a Xeon resulted
in a count that was between the CPU count on the two node types
and invalid on the node with smaller CPU count (e.g. 272 CPUs on
KNL, 8 on Xeon, and 2 tasks, cpus_per_task = 140).

JOB_COMPLETING is a flag and not an enum.

accounting may show job or step in COMPLETING state for a while

Search for step state happened before the step started.

Correction to environment variable name in
commit a09a0795