Fix MemSpecLimit to not cause deadlock.

Bug 5966

# Conflicts:
#	src/plugins/launch/aprun/Makefile.in
#	src/plugins/proctrack/sgi_job/Makefile.in
#	src/plugins/select/alps/Makefile.in
#	src/plugins/select/alps/libalps/Makefile.in
#	src/plugins/select/alps/libemulate/Makefile.in
#	src/plugins/select/cons_tres/Makefile.in
#	src/sprio/Makefile.in

Bug 5865

to not require -Wno-format-truncation CFLAG.

Bug 4310
Bug 5655
Bug 3482

This reverts commit dfd4c229.

# Conflicts:
#	NEWS

Removal of unneeded FREE_NULL_BITMAP

Bug 5774

Bug 5774

# Conflicts:
#	doc/html/heterogeneous_jobs.shtml
#	slurm/slurm.h.in

Coverity CID 188961

Fix memory leak by freeing iterator.
Simplify logic to be more intuitive.
Separate out _check_exec() into two logically different portions.
Get rid of unnecessary goto's.

Bug 5848

Bug 5848

and distributed.

Bug 5774

previously accrued for priority.

Bug 5881

Brian approved.

have the same id (for example, if a job was requeued) by also checking the
submit time.

Bug 5197

Bug 5197

Bug 5197

when not runaway.

Bug 5940

CID 189455.

Fixes an off-by-one mistake in SLURM_TOPO_LEN length handling.

Remove an undocumented max line length of 512 characters, as well as a stray
trailing space character.

Bug 5804

Bug 5755

performance hit was about 20%. now similar to cons_res

Bug 5443

This enables changing from cons_res to/fron cons_tres while
preserving all jobs EXCEPT jobs using new GPU options when
changing from cons_tres to cons_res. Those jobs will be killed
if active (pending or running).

Was accidentally reset to 0 by commit 625ec36d.

This reverts commit 0581585c.

Do not change permissions on files the testsuite does not "own".

This is a duplicate of the previous 3547869b commit, which
was previously discarded by merge commit cf98dfc0.

Bug 4118.

On second thought, since we enable debug mode automatically this
won't be useful for what I plan to use it for, and since nothing
uses it let's just remove it entirely.

Doesn't really need to be directly in configure.ac, so move it.

Note that this has been subtly broken for quite a while:
- The correct variable to test is x_ac_debug, not ac_debug.
- The correct value when debugging is enabled is "yes", not "true".

Fix regression caused by commit bb8b4214.

Commit bb8b4214 introduced this in scontrol_hold():

-		job_msg.job_id_str = _next_job_id();
-		while (job_msg.job_id_str) {
+		while ((job_msg.job_id_str = _next_job_id())) {

But didn't take out _next_job_id() at the bottom of the while loop -
so _next_job_id() is called twice in a row, skipping every other job
in the list.

Bug 5902