Downing the node will kill all jobs allocated to the node, very bad
on something like a BlueGene system

fix of the following problem:
if a node is excised from a job and a reconfiguration(e.g., update
partition) is done when the job is still running, the node will be left
in state idle but not available any more until the next
reconfiguration/restart of slurmctld after the job finished.

the node tab and we didn't notice.

Response to bug 274

Make test work if current working directory not in the search path
Check for appropriate task rank on POE based systems
Disable the entire test on POE systems

Without this change pmdv12 was bound to one CPU and could not use
all of the resources allocated to the job step for the tasks that
it launches

This only changes behaviour when the --ntasks option is not used,
but the --cpus-per-task option is use

we used cpus erroneously but now we use tasks.  The cpus variable will
be taken out in 2.6.

also "-euidevice sn_single".

Avoid placing pending jobs in AdminHold state due to backfill scheduler
interactions with advanced reservation.
Specifically, if the backfill scheduler tests a pending job can be
scheduled after it's advanced reservation ends then the job was
assigned a priority of zero (AdminHold).

undefined variable.

requested and allocated.

allocation as taking up the entire node instead of just part of the node
allocated.  And always enforce exclusive on a step request.