Make test more robust for compute nodes with large CPU counts.

This change fixes Slurm's ability to optimize selection of resources
for a job requesting feature counts where some of those node features
are currently inactive (require node reboot to claim).

# Conflicts:
#	src/common/slurm_acct_gather_energy.c
#	src/common/slurm_acct_gather_filesystem.c
#	src/common/slurm_acct_gather_infiniband.c
#	src/common/slurm_acct_gather_profile.c
#	src/common/slurm_jobacct_gather.c

when in use.

The problem here is the polling threads in the various acct_gather codes
were detached and could possibly still be polling after the plugin had
been unloaded making a seg fault with a backtrace like this...

#0  0x00007fe7af008c00 in ?? ()
#1  0x00007fe7b1138479 in __nptl_deallocate_tsd () at pthread_create.c:175
#2  0x00007fe7b11398b0 in __nptl_deallocate_tsd () at pthread_create.c:326
#3  start_thread (arg=0x7fe7b1f12700) at pthread_create.c:346
#4  0x00007fe7b0e6fb5d in clone ()
    at ../sysdeps/unix/sysv/linux/x86_64/clone.S:109

The fix was to make the threads non-detached and join them before calling
a dlclose.

Whenever possible, avoid allocating nodes that require a reboot.
Previous logic failed to re-sort the job set table based upon
the need for rebooting to achieve the desired features (e.g. KNL
MCDRAM or CACHE mode).
bug 2726

Put header files in alphabetic order, No change in logic

# Conflicts:
#	src/slurmctld/controller.c

trying to verify the cluster name (which may try to /create/ files or
directories) *before* dropping privs results in a fatal error as
slurmctld tries to create items which ultimately fail.  Moving
this process until after the privs and uid have changed allows
the process to succeed.

Reported by Jon Nelson <jdnelson@dyn.com>

Bug 2728

Reject invalid step at submit time rather than leaving it queued.
Bug 2722 describes one of the use cases triggering the bug.

Minor update to commit 2fad3bcf
This leaves the files locked until file write completes.

Disable a job deadline test with no input time limit if the default
partition has a default time limit.

This partially restores commit 03b2cfb5
Logic was not closing file descriptor, which left the file locked and
leaked an open file descriptor.

tasks-per-node/nodes != tasks print warning and ignore
ntasks-per-node.

Bug 2520

On a Cray, the output file isn't being created the second time.

Make test_id in more tests be just the numeric value rather than "test#.#"
for consistency with the other tests.

Make test_id in test30.1 be just the numeric value rather than "test30.1"
for consistency with the other tests.

The account still had maxnodes=1 set preventing the qos grpnodes to take
affect. This showed up on slower machines because it takes a second for
the changes to get to the controller.

Test would originally try to start more jobs than default_queue_depth
in SchedulerParameters and fail

Job was failing on Cray/kachina due to timeout. Increase job time
  limit from 1 to 2 minutes.

make it to the slurmctld when using message aggregation.

make sure we handle it correctly when the database comes back up.

Give test job an extra second to start. Test was failing by one
  second on kachina.

Get the maximum file pathname size from system include file rather
  than local #define. This was causing failures on kachina test.