diff --git a/NEWS b/NEWS
index 3435733b1efc5e0829eb46c938e34a2a3b7883ef..7e323aaaed032d2e2051f3e719110878ddc6a566 100644
--- a/NEWS
+++ b/NEWS
@@ -164,6 +164,13 @@ documents those changes that are of interest to users and administrators.
to a insufficient priviledges.
-- Add warning about libcurl-devel not being installed during configure.
-- Streamline job purge by handling file deletion on a separate thread.
+ -- Always set RLIMIT_CORE to the maximum permitted for slurmd, to ensure
+ core files are created even on non-developer builds.
+ -- Fix --ntasks-per-core option/environment variable parsing to set
+ the requested value, instead of always setting one.
+ -- If trying to cancel a step that hasn't started yet for some reason return
+ a good return code.
+ -- Fix issue with sacctmgr show where user=''
* Changes in Slurm 17.02.3
==========================
diff --git a/doc/html/faq.shtml b/doc/html/faq.shtml
index 4ebf4ccbbbbdea421016ee3069e5b04216a8cdb3..c3e1fe8d67070460ac8e5f2db942baeeae672e42 100644
--- a/doc/html/faq.shtml
+++ b/doc/html/faq.shtml
@@ -62,6 +62,7 @@
<li><a href="#sbatch_srun">What is the difference between the sbatch
and srun commands?</a></li>
<li><a href="#squeue_color">Can squeue output be color coded?</a></li>
+<li><a href="#x11">Can Slurm export an X11 display on an allocated compute node?</a></li>
</ol>
<h2>For Administrators</h2>
@@ -1056,6 +1057,46 @@ regex white:default ^JOBID.*
</pre>
<img src="squeue_color.png" width=600>
+<p><a name="x11"><b>33. Can Slurm export an X11 display on an allocated compute node?</b></a><br>
+You will need to build and install an optional SPANK plugin for that functionality
+today. The functionality should be provided natively in Slurm soon. Instructions
+to build and install the plugin follow. Update the Slurm installation path as needed.</p>
+<pre>
+# Maybe obvious, but don't forget the -X on ssh
+$ ssh -X alex@testserver.com
+
+# Get the plugin
+$ mkdir git
+$ cd git
+$ git clone https://github.com/hautreux/slurm-spank-x11.git
+$ cd slurm-spank-x11
+
+# Manually edit the X11_LIBEXEC_PROG macro definition
+$ vi slurm-spank-x11.c
+$ vi slurm-spank-x11-plug.c
+$ grep "define X11_" slurm-spank-x11.c
+#define X11_LIBEXEC_PROG "/opt/slurm/17.02/libexec/slurm-spank-x11"
+$ grep "define X11_LIBEXEC_PROG" slurm-spank-x11-plug.c
+#define X11_LIBEXEC_PROG "/opt/slurm/17.02/libexec/slurm-spank-x11"
+
+
+# Compile
+$ gcc -g -o slurm-spank-x11 slurm-spank-x11.c
+$ gcc -g -I/opt/slurm/17.02/include -shared -fPIC -o x11.so slurm-spank-x11-plug.c
+
+# Install
+$ mkdir -p /opt/slurm/17.02/libexec
+$ install -m 755 slurm-spank-x11 /opt/slurm/17.02/libexec
+$ install -m 755 x11.so /opt/slurm/17.02/lib/slurm
+
+# Configure
+$ echo -e "optional\tx11.so" >> /opt/slurm/17.02/etc/plugstack.conf
+$ cd ~/tests
+
+# Run
+$ srun -n1 --pty --x11 xclock
+adam@node1's password:
+</pre>
<p class="footer"><a href="#top">top</a></p>
@@ -2194,6 +2235,6 @@ slurmctld for those changes to be recognized.
<p class="footer"><a href="#top">top</a></p>
-<p style="text-align:center;">Last modified 12 May 2017</p>
+<p style="text-align:center;">Last modified 6 June 2017</p>
<!--#include virtual="footer.txt"-->
diff --git a/src/common/slurm_protocol_defs.c b/src/common/slurm_protocol_defs.c
index 157a79feab37470f1926b9bbec1d134f75054233..fb815a3c38e011dc7f6073e104699dc4611c95dd 100644
--- a/src/common/slurm_protocol_defs.c
+++ b/src/common/slurm_protocol_defs.c
@@ -238,7 +238,7 @@ extern int slurm_char_list_copy(List dst, List src)
/* returns number of objects added to list */
extern int slurm_addto_char_list(List char_list, char *names)
{
- int i = 0, start = 0;
+ int i = 0, start = 0, cnt = 0;
char *name = NULL;
ListIterator itr = NULL;
char quote_c = '\0';
@@ -263,6 +263,7 @@ extern int slurm_addto_char_list(List char_list, char *names)
i++;
}
start = i;
+ cnt = list_count(char_list);
while (names[i]) {
//info("got %d - %d = %d", i, start, i-start);
if (quote && (names[i] == quote_c))
@@ -366,7 +367,8 @@ extern int slurm_addto_char_list(List char_list, char *names)
i++;
}
- if (i-start) {
+ /* check for empty strings user='' etc */
+ if ((cnt == list_count(char_list)) || (i - start)) {
name = xstrndup(names+start, (i-start));
/* If we get a duplicate remove the
* first one and tack this on the end.
diff --git a/src/plugins/sched/backfill/backfill.c b/src/plugins/sched/backfill/backfill.c
index 9e653b9718ef23f2a5f3ac331f5cc91d624f4764..b88fc90e67101cd7b141c0a92653f6f1345b987d 100644
--- a/src/plugins/sched/backfill/backfill.c
+++ b/src/plugins/sched/backfill/backfill.c
@@ -1109,7 +1109,8 @@ static int _attempt_backfill(void)
if (assoc_limit_stop) {
assoc_mgr_lock(&qos_read_lock);
- list_for_each(part_list, _clear_qos_blocked_times, NULL);
+ list_for_each(assoc_mgr_qos_list,
+ _clear_qos_blocked_times, NULL);
assoc_mgr_unlock(&qos_read_lock);
}
diff --git a/src/sbatch/opt.c b/src/sbatch/opt.c
index 61d3f81c66459f2c1a4c60af96f6f6057c1f60ec..ad7856e72ce7c53a65feef24a4e3d5f2d8523d12 100644
--- a/src/sbatch/opt.c
+++ b/src/sbatch/opt.c
@@ -1856,7 +1856,7 @@ static void _set_options(int argc, char **argv)
optarg, true);
setenvf(NULL, "SLURM_NTASKS_PER_CORE", "%d",
opt.ntasks_per_core);
- opt.ntasks_per_core = true;
+ opt.ntasks_per_core_set = true;
break;
case LONG_OPT_HINT:
if (!optarg)
@@ -1871,7 +1871,7 @@ static void _set_options(int argc, char **argv)
exit(error_exit);
}
opt.hint_set = true;
- opt.ntasks_per_core = true;
+ opt.ntasks_per_core_set = true;
opt.threads_per_core_set = true;
break;
case LONG_OPT_BLRTS_IMAGE:
diff --git a/src/slurmctld/proc_req.c b/src/slurmctld/proc_req.c
index c918852138d275460c9ce840b044f7162146fdb1..2026b6db4c9cfde1a125867707040144c155f8eb 100644
--- a/src/slurmctld/proc_req.c
+++ b/src/slurmctld/proc_req.c
@@ -2090,7 +2090,8 @@ static void _slurm_rpc_complete_batch_script(slurm_msg_t *msg,
/* do RPC call */
/* First set node DOWN if fatal error */
- if ((comp_msg->slurm_rc == ESLURM_ALREADY_DONE) ||
+ if ((comp_msg->slurm_rc == ESLURMD_JOB_NOTRUNNING) ||
+ (comp_msg->slurm_rc == ESLURM_ALREADY_DONE) ||
(comp_msg->slurm_rc == ESLURMD_CREDENTIAL_REVOKED)) {
/* race condition on job termination, not a real error */
info("slurmd error running JobId=%u from %s=%s: %s",
diff --git a/src/slurmd/slurmd/slurmd.c b/src/slurmd/slurmd/slurmd.c
index 95ff786f49718d1919729793990c3c8709ead69d..eda00ad142863b337ffa3713d672069e596f1de0 100644
--- a/src/slurmd/slurmd/slurmd.c
+++ b/src/slurmd/slurmd/slurmd.c
@@ -1559,12 +1559,10 @@ _slurmd_init(void)
rlim.rlim_cur = rlim.rlim_max;
setrlimit(RLIMIT_NOFILE, &rlim);
}
-#ifndef NDEBUG
if (getrlimit(RLIMIT_CORE, &rlim) == 0) {
rlim.rlim_cur = rlim.rlim_max;
setrlimit(RLIMIT_CORE, &rlim);
}
-#endif /* !NDEBUG */
/*
* Create a context for verifying slurm job credentials
diff --git a/src/slurmd/slurmstepd/step_terminate_monitor.c b/src/slurmd/slurmstepd/step_terminate_monitor.c
index b043421a0ee9f5f38519b1089c3dc0350ff18bd8..464af54e18dfc976562e4ddc7e8b55aa1a8dbc7f 100644
--- a/src/slurmd/slurmstepd/step_terminate_monitor.c
+++ b/src/slurmd/slurmstepd/step_terminate_monitor.c
@@ -130,6 +130,8 @@ static void *_monitor(void *arg)
if (rc == ETIMEDOUT) {
char entity[24], time_str[24];
time_t now = time(NULL);
+ int rc;
+
_call_external_program(job);
if (job->stepid == SLURM_BATCH_SCRIPT) {
@@ -144,16 +146,17 @@ static void *_monitor(void *arg)
}
slurm_make_time_str(&now, time_str, sizeof(time_str));
- if (job->state != SLURMSTEPD_STEP_RUNNING) {
+ if (job->state < SLURMSTEPD_STEP_RUNNING) {
error("*** %s STEPD TERMINATED ON %s AT %s DUE TO JOB NOT RUNNING ***",
entity, job->node_name, time_str);
+ rc = ESLURMD_JOB_NOTRUNNING;
} else {
error("*** %s STEPD TERMINATED ON %s AT %s DUE TO JOB NOT ENDING WITH SIGNALS ***",
entity, job->node_name, time_str);
+ rc = ESLURMD_KILL_TASK_FAILED;
}
- stepd_cleanup(NULL, job, NULL, NULL, SLURM_ERROR, 0);
- abort();
+ exit(stepd_cleanup(NULL, job, NULL, NULL, rc, 0));
} else if (rc != 0) {
error("Error waiting on condition in _monitor: %m");
}