cray/burst_buffer - Preserve job ID and don't translate to job array ID
  after slurmctld restart. Prior logic would not set array_task_id to
  NO_VAL, so all job-buffer IDs would be reported in the form
  "JobID=0_0(123)" rather than "JobID=123"

cray/busrt_buffer - Internally track both allocated and unusable space.
    The reported UsedSpace in a pool is now the allocated space (previously was
    unusable space). Base available space on whichever value leaves least free
    space.
bug 3222

Previously disconnected from build system, and most code removed by
commit 0b14a3a7 back on 15.08-pre1.

We don't build on Tru-64, and there are a lot more platform-dependent
pieces of code within Slurm than we've indicated here.

Remove stray Bull logo as well.

Remove stray Bull logo as well.

They're full of dead links, and the plugins are deprecated
(announced at SLUG16).

OSX is not currently supported, and the build is likely broken
due to differences in dynamic library loading.

and request --ntasks-per-core=1 and only 1 task on the node
the slurmd would abort on an infinite loop fatal.

Regression is from commit 5265420d.

Without this fix you can get into an infinite loop in the task/affinity
plugin.  The loop is handled by producing a fatal.

Bug 3118

cray/busrt_buffer - Fix for double counting of used_space at slurmctld
    startup.
bug 3222

Add the POLLRDHUP hack used elsewhere to work around non-standard flag use.

Bug 3227.

cray/busrt_buffer - If total_space in a pool decreases, reset used_space
    rather than trying to account for buffer allocations in progress.
bug 3222

more time than should be allowed would be accounted for.

This only happened on jobs in the completing state when the slurmctld
was shutdown.

This will also be enhanced in 17.02 as the job's end_time_exp is not
stored which is needed to determine if the job has already been through
the decay_thread at end of job.

Bug 3162

issue with gang scheduling.

Bug 3211

This option specifies minimum characteristics of the compute nodes
  which should be considered for use, not the resource allocation
  size.
bug 3118

Create separate check_hosts_contiguous procedure in globals and use it
for both test1.83 and test15.21.

Bug 3006.

Fix bug that was clearing MAINT mode on nodes scheduled for reboot (bug
    introduced in version 16.05.5 to address bug in overlapping reservations,
    commit 5eee1d28). Note that a node's
    MAINT flag is used for both a requested reboot and maintenance reservation.
    What I'd like to do is add a new node state flag to differenciate
    between these two cases, but that involves some significant changes
    that could introduce instability, so it will be defered to version
    17.02
bug 3210

Correct/expand description of NODE_STATE_FLAG

requested with -n tasks < hosts from -w hostlist.

Bug 3171.

bug 2149

In addition to "show <ENTITY> <ID>" slurm also accepts
  "show <ENTITY>=<ID>"
bug 2162

Bug 3194

Document that node Weight takes precedence over load with LLN scheduling.
bug 3204

task/cray's _get_numa_nodes() function needs to run before task/cgroup
cleans up the cgroup hierarchies, otherwise ALPS memory compaction will
never run.

Also move task_p_add_pid() outside the #ifdef HAVE_NATIVE_CRAY
block so that the plugin will load (albeit without any functionality)
on non-Cray systems for testing purposes.

Revise documentation and provided slurm.conf templates as well.

Bug 3154.

There is a problem with gathering batch step statistics for jobs which are allocated on more than one node.

Sstat asks wrong node for batch step stats. It requests info from last node from hostlist while it should ask first host from hostlist (i.e. BatchHost), because only on the first node the batch step actually executes.

For example, when you have a job allocated on nodes n000[1-2] with BatchHost=p0001. You should be able to check its statistics by running sstat [ with -vv switch for more verbose output] (e.g. sstat -j 1234.batch -vv). Then you can see lines:

sstat: debug:  slurm_job_step_stat: getting pid information of job 1234.4294967294 on nodes n0002
sstat: debug:  job step 1234.4294967294 has already completed

The problem lays in sstat source code. For batch step a hostlist variable is taken from the hostlist_pop function, which returns last host from given hostlist. This should be replaced with the hostlist_shift function, which returns first host from the given hostlist. Patch attached.

bug 2975

This commit fixes a bug in the multi-prog handling. When running
salloc -N 2 srun -O --multi-prog mp.conf where mp.conf reads

0-192 true

srun crashes can be observed. valgrind reports:

==6857== Invalid read of size 4
==6857==    at 0x45938D: bit_realloc (bitstring.c:189)
==6857==    by 0x5977A9: _update_task_mask (multi_prog.c:335)
==6857==    by 0x597A5E: _validate_ranks (multi_prog.c:403)
==6857==    by 0x597D1E: verify_multi_name (multi_prog.c:469)
==6857==    by 0x6E7B4BE: launch_p_handle_multi_prog_verify (launch_slurm.c:453)
==6857==    by 0x58A25D: launch_g_handle_multi_prog_verify (launch.c:493)
==6857==    by 0x58E556: _opt_args (opt.c:1927)
==6857==    by 0x58A3B9: initialize_and_process_args (opt.c:270)
==6857==    by 0x591F82: init_srun (srun_job.c:459)
==6857==    by 0x427E70: srun (srun.c:193)
==6857==    by 0x428E23: main (srun.wrapper.c:17)
==6857==  Address 0x5ace440 is 16 bytes inside a block of size 28 free'd
==6857==    at 0x4C2BB4A: realloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==6857==    by 0x446886: slurm_xrealloc (xmalloc.c:139)
==6857==    by 0x45944C: bit_realloc (bitstring.c:191)
==6857==    by 0x5977A9: _update_task_mask (multi_prog.c:335)
==6857==    by 0x597A5E: _validate_ranks (multi_prog.c:403)
==6857==    by 0x597D1E: verify_multi_name (multi_prog.c:469)
==6857==    by 0x6E7B4BE: launch_p_handle_multi_prog_verify (launch_slurm.c:453)
==6857==    by 0x58A25D: launch_g_handle_multi_prog_verify (launch.c:493)
==6857==    by 0x58E556: _opt_args (opt.c:1927)
==6857==    by 0x58A3B9: initialize_and_process_args (opt.c:270)
==6857==    by 0x591F82: init_srun (srun_job.c:459)
==6857==    by 0x427E70: srun (srun.c:193)

_select_nodes_parts() was resetting state_reason to an admin hold
without regard to admin vs user hold state. state_reason is the only
place that user vs. admin is distinguished, so this prevented users
from releasing these jobs.

Bug introduced by commit fb46c84b in 16.05.5.

Bug 3197.