Conflicts:
	src/plugins/select/bluegene/bg_read_config.c
	src/slurmctld/job_mgr.c

This reverts https://github.com/hautreux/slurm/commit/35075c10995f4e83d0104662f147cd7b413d25f4
My version of RPM doesn't seem to understand the "replace"
parameter for %config, so rpm builds bomb out with
    Invalid %config token: replace
    Invalid %config token: replace

I think the default for %config is to not replace config
files, so I'm not sure this commit is needed. Since the
release agents are not really config files, but need to
be tightly coupled to the release of task/cgroup and
proctrack/cgroup, I would say the release agents should
be moved under /usr/libexec/slurm

At the very least, this commit needs to be reverted before slurm-2.3.1
in order to allow rpms to be built.

Thanks,
Mark Grondona

Report correct job "Reason" if needed nodes are DOWN, DRAINED, or
NOT_RESPONDING, "Resources" rather than "PartitionNodeLimit".

This insures that the job association pointer is correct for the job
in case the QOS permits partition limits to be over-ridden.

plugins in the slurmd.

    value for jobs.

Cgroup plugins update

For a long time configure has modified the SLURM Release number
as set in META by stripping off everything before the last '.'
when building the SLURM_VERSION_STRING. This was done so that a
release number of 0.pre1 would become just 'pre1' in the version
string printed by SLURM commands. (e.g. slurm-2.3.0-0.pre1 becomes
slurm-2.3.0-pre1 in sinfo --version).

In attempting to create a new version 2.3.0-2.x of SLURM (branched
from 2.3.0-2), it was found that this method is overzealous, and
results in a version string of just "2.3.0-1" instead of the expected
"2.3.0-2.1". Since the intent of the sed command is only to remove
'0.' from prereleases, this patch makes that explicit, so that
non-prerelease versions branched of tagged SLURM releases keep the
original Release number in the version string.

Preserve the address information about nodes in the cloud after a reconfigure

if any message comes through, and since we have states of cnodes in
the status threads we don't need to keep retrying to send the message of
cnodes in error if the slurmctld is down.

parameter to the bluegene.conf file, MaxBlockInError, which is a percentage
of a block that can be in an error state before future jobs are disallowed
to run on the block.

in the system.  cnode_err_bitmap is used to tell if there are cnodes in
an error state in the usable portion of the midplane.  cnode_usable_bitmap
is used to tell what part of the midplane is usable, typically used for
small blocks which only use a portion of a midplane.