They were being referenced prior to initialization

This better handles job submit failures and reduces the time spend
waiting for output files by overlapping the wait for all files.

without this the total GRES counter can keep getting incremented,
the _get_job_cnt() calculation can overflow and extra GRES be
allocated to the job

This reverts debugging accidentally committed in 48ec9695

There are a few anomalies with respect to front-end mode, which
test test changes address

Since we don't have any bluegene code this is pointless to send
NULL's over.

Bug 5698

this is related to commit 845e134b
when a step ends, flag that it is done on ALL nodes so that it's
memory can be released. Note that test1.19 would fail in front-end
mode when there was a default memory allocation. That test would
spawn 2 steps, but the first would not release its memory when
that step ended, so the second step would not start.

this is useful for testing. it is unclear why this was disabled

Bugs 5699, 5700, 5701

Bug 5698

Bug 5701

Bug 5701

Bug 5700

Bug 5700

Bug 5699

Bug 5699

Bug 5699

Bug 5699

Bug 5699

Bug 5699

Bug 5699

It doesn't do anything yet.

Two tests were failing when using --enable-front-end configuration

# Conflicts:
#	NEWS
#	src/common/hostlist.c

hosts in each list.

Bug 5741

by integers.

We understand this reverts bc7b91cd, but in all our testing we have
been unsuccessful finding what that commit actually does.

Bug 5711
Bug 5746

This fixes a problem if a job has a gres_per_job value set and
multiple required nodes (by name). Previous logic could block
use of later nodes if the gres_per_job value was reached by
earlier selected nodes.