This prevents premature re-sending of job kill RPC
(e.g. "Resending TERMINATE_JOB request JobId=#")

If the default partition has shared=force, then each job is allocated
whole nodes and core reservations tests are not valid

This fixes another error in job priority calculations

The pending jobs will have their reservation info removed
bug 455

Increase maximum number of hostlist ranges from 12k to 64k and
use malloc to allocate memory rather than using the stack
bug 458

Initiate jobs pending to run in a reservation as soon as the reservation
becomes active.
Partial fix for bug 455

Revert commit 626be3ea
It was causing stack overflow and memory corruption

Previous logic only reported un-reserved node map.
New logging adds information about each job testing and where/when
it is scheduled resources.

slurm.conf when using the DBD.

Induced by bf_continue option and deleting a partition.

to NULL.

At least if you try the compiler gives this warning

warning: left-hand operand of comma expression has no effect [-Wunused-value]

In 13.12 this is treated as an error

to reflect only the latest supported format.

if bf_continue option is configured and slurm is reconfigured during
one of the sleep cycles, then the backfill scheduler will reference
an invalid partition pointer.

Result of bad merge

Previous logic would place more tasks on each node than specified
by --ntasks-per-node, using fewer nodes than desired. This only
happens with exclusive node allocations (e.g. in partition
configuration Shared=Exclusive).

EpilogSlurmctld pthread is passed required arguments rather than a pointer
to the job record, which under some conditions could be purged and result
in an invalid memory reference.