This fixes a bug introduced in commit 52fe3de1
in the event the fork() call fails in slurmstepd.

Insure that a job is completely launched before trying to suspend it.
Previous logic would start suspend logic early in the life of the
slurmstepd process, after it's listening socket was open but before
the tasks were launched. This defers the suspend logic until after
all prologs and setup completes and the tasks are launched. This is
important in the case of gang scheduling, in which newly launched
jobs can be immediately suspended.
bug 2494

Bug 1976

Add note to slurm.conf man page about setting "--cpu_bind=no" as part
of SallocDefaultCommand if a TaskPlugin is in use.

Test 14.10 in the test suite (of slurm 15.08.8, at least) uses

  $sinfo -tidle -h -o%n

to find idle nodes.  This only works if NodeHostname == NodeName on the
nodes.  The following should work regardless of this:

  $scontrol show hostnames \$($sinfo -tidle -h -o%N)

Since the function is inlined the single definition let GCC build everything
properly, but debug builds (which disable inline) resulted in:
slurmstepd: [465.0]: symbol lookup error:
(trimmed path)/task_cgroup.so: undefined symbol: val_to_char
when running srun --cpu_bind=v.

task/affinity had this definition already, task/cgroup didn't.

Reported by valgrind running test7.2, but shouldn't cause any real problem

was also given.

a partition.

This also reverts most of commit fa331e30 as well as commit bd9fa830
which would try to set the pn_min_cpus every time a job was updated.
If a job didn't request node counts then they were hosed.

This commit takes away the magic which was screwing things up.  Now the
person gets what they asked for without magic changing things.

Bug 2302
Bug 2742
Bug 2478

erroneously.

Failure has never been observed, but initialize the used variable
  before calling the function so we don't re-use old data if the
  function returns an error.

Rename an improperly named variable in the logic scontrol uses to
  print node information ("total_used" was really "idle_cpus"), so
  the logic looks the same as that used in sinfo to determine node
  state.

Include warning for Cray simulation as reminder for developers to
change code as needed.

This whole process could probably be done better by keeping track of
old values and new values and only calling one function instead of a
pre and post function, but that can probably wait for future generations
of the code as it works now and is probably adequate for the time being.

Bug 2352

Consistantly use American English for
existant -> existent
assocation -> association

Correct some typos, and one grammatical mistake.

BurstBuffer/cray - Defer job cancellation or time limit while "pre-run"
    operation in progress to avoid inconsistent state due to multiple calls
    to job termination functions.
bug 2454

Otherwise call fclose(NULL) iff the ClusterName is not set and the
clustername file does not exist. Should not happen in production.

Coverity #67041.

a new account and making it a default all at once.

Bug 2428

Match acct_gather_energy/rapl plugin. Bug 2397.

Control whether the scheduler will continue to try to run jobs in
a partition if a higher priority job is stuck due to an association
limit.

Can cause starvation for larger jobs, but will improve throughput and
utilization for systems that have extensively divvyed up their resources
through association/QOS limits.

Bug 2388 and 2452.

Bug 2453

This should have no effect, but is a belt-and-suspenders approach
  to checking node state.

have been selected set the time limit appropriately if the job didn't
request one.

If the partition has no DefaultTime setting, and no time_limit was given
for the job, job_ptr->time_limit == NO_VAL.

With AccountingStorageEnforce=safe this will prevent jobs from ever
starting if the association has any limit set for CPUMins.
(NO_VAL * cpus is a very large number, but if no time_limit is given
anywhere that is what they get :))

Bug 2388.