allocate tasks in blocks of the plane size.

    dependency that can not be satisfied.

    just setting the priority to 1.

Improve handling of job updates when using limits in accounting, and updating jobs as a non-admin user.

Clear pending job's start_time between executions of sched/backfill.

Fix for sview started on a non-bluegene system to pick colors correctly when talking to a real bluegene system.

    a full midplane on BlueGene computers. Treat cnodes as liceses which can
    be reserved and are consumed by jobs.

    Perl can not build if both are set.

    step names, and account names with an underscore to insure proper parsing.

BLUEGENE - Fix for handling extremely overloaded system on Dynamic system dealing with starting jobs on overlapping blocks.  Previous fallout was job would be requeued.  (happens very rarely)

    "%j" in the job's output or error file specification.

BLUEGENE - If using OVERLAP mode fixed issue with jobs getting denied at submit if the only option for their job was overlapping a block in error state.

    the slurmd daemon(s). This insures that tasks in the process of getting
    spawned are killed.

    having a job start that runs into the reservation when later advanced.

and explain the change in job allocation man pages

   sockets option" behavior.

   affinity with plane distribution" bug.

Gerrit Renker:  I found an easier test: after making sure that stdin is not
 redirected, check whether there is a non-negative terminal process group (it
 is -1 if there is no controlling terminal).

 Also, I think that the heuristic "only interactive if from login shell" is
 needlessly complicated/obscure. Most (if not all) shells that support job
 control place processes in their own process group (pid == pgrp): it now uses
 this as indicator of interactive mode.

The following tests now pass:

  salloc &			# from login shell and from nested shell
  salloc --no-shell&		# ok
  salloc /bin/date&		# wants to be placed into the foreground
  salloc /bin/date </dev/null &	# ok

  setsid salloc	 		# complains there is no controlling terminal
  setsid salloc --no-shell&	# ok

    or resumed when appropriate.

    ("setsid salloc ..." is disabled). Patch from Gerrit Renker, CSCS.

    foreground. Patch from Gerrit Renker, CSCS.
    a) salloc is not run in allocation-only (--no-shell) mode,
    b) stdin is from a terminal (tcgetattr does not return ENOTTY),
    c) salloc has been invoked from a login shell (not a nested one),
    d) salloc has been configured at compile-time to support background
       execution and is not currently in the background process group.

Fix for handling job names with a "'" in the name within MySQL accounting. Patch from Gerrit Renker, CSCS.

    associated with a given NodeHostName when running multiple slurmd daemons
    per compute node (typically used for testing purposes). Patch from
    Matthieu Hautreux, CEA.

    MessageTimeout configuration parameter rather than always 3 seconds.
    Patch from Matthieu Hautreux, CEA.

    Bjorn-Helge Mevik, University of Oslo.

   block second distribution and cpus-per-task > 1" bug.