wait for.

based on the mask given.

task/affinity.

thread in a core.

it can bind cyclically across sockets.

Previous logic assumed cpus_per_task=1, so ntasks_per_core option
could spread the job across more cores than desired

cpus-per-task support: Try to pack all CPUs of each tasks onto one socket.
Previous logic could spread the tasks CPUs across multiple sockets.

Previous logic was counting CPUs, but assuming each task would
only use one CPU.

This reverts commit b22268d8.

Conflicts:
	src/slurmctld/job_mgr.c

Properly enforce job --requeue and --norequeue options. Previous
logic was in three places not doing so (either ignoring the value,
ANDing it with the JobRequeue configuration option or using the
JobRequeue configuration option by itself).
bug 821

There should be no change in behavior with the production code,
but this will improve the robustness of the code if someone makes
changes to the logic.

Add SelectTypeParameters option of CR_PACK_NODES to pack a job's tasks
tightly on its allocated nodes rather than distributing them evenly across
the allocated nodes.
bug 819

something you also get a signal which would produce deadlock.

Fix Bug 601.

Run EpilogSlurmctld for a job is killed during slurmctld reconfiguration.
bug 806

Only if ALL of their partitions are hidden will a job be hidden by default.
bug 812

If a batch job launch request can not be built (the script file
is missing, a credential can not be created, or the user does
not exist on the selected compute node), then cancel the job
in a graceful fashion. Previously, the bad RPC would be sent to
the compute node and that node DRAINED.
see bug 807

Correct SelectTypeParameters=CR_LLN with job selecition of specific nodes.
Previous logic would in most instances allocate resources on all nodes
to the job.

Correct squeue's job node and CPU counts for requeued jobs.
Previously, when a job was requeued, its CPU count reported
was that of the previous execution. When combined with the
--ntasks-per-node option, squeue would compute the expected
node count. If the --exclusive option is also used, the node
count reported by squeue could be off by a large margin (e.g.
"sbatch --exclusive --ntasks-per-node=1 -N1 .." on requeue
would use the number of CPUs on the allocated node to recompute
the expected node count).
bug 756

jobacct_gather/cgroup.

Support SLURM_CONF path which does not have "slurm.conf" as the file name.
bug 803

If a job has non-responding node, retry job step create rather than
returning with DOWN node error.
bug 734

Also correct related documentation

Add force option to all file removals ("rm ..." to "rm -f ...").
bug 673

Completing nodes is removed when calling _try_sched() for a job, which
is the case in select_nodes(). If _try_sched() thinks the job can run
now but select_nodes() returns ESLURM_NODES_BUSY, the backfill loop will
be ended.