From b6e99ea80011d1e3911e1ba2ae62143a90c5deab Mon Sep 17 00:00:00 2001
From: Moe Jette <jette1@llnl.gov>
Date: Tue, 26 Sep 2006 16:45:21 +0000
Subject: [PATCH] Report a job's exit status in "scontrol show job".
---
NEWS | 1 +
doc/man/man1/scontrol.1 | 10 ++++++++--
slurm/slurm.h.in | 1 +
src/api/job_info.c | 12 ++++++++++--
src/common/slurm_protocol_pack.c | 1 +
src/slurmctld/job_mgr.c | 1 +
src/slurmctld/slurmctld.h | 3 ++-
7 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/NEWS b/NEWS
index a1345c7b217..caec6a36367 100644
--- a/NEWS
+++ b/NEWS
@@ -17,6 +17,7 @@ documents those changes that are of interest to users and admins.
-- Removed addresses from slurm_step_layout_t
-- Added new job field, "comment". Set by srun, salloc and sbatch. See
with "scontrol show job". Used in sched/wiki2.
+ -- Report a job's exit status in "scontrol show job".
* Changes in SLURM 1.2.0-pre2
=============================
diff --git a/doc/man/man1/scontrol.1 b/doc/man/man1/scontrol.1
index ccec95e2eba..3552eb4e28a 100644
--- a/doc/man/man1/scontrol.1
+++ b/doc/man/man1/scontrol.1
@@ -1,4 +1,4 @@
-.TH SCONTROL "1" "May 2006" "scontrol 1.2" "Slurm components"
+.TH SCONTROL "1" "September 2006" "scontrol 1.2" "Slurm components"
.SH "NAME"
scontrol \- Used view and modify Slurm configuration and state.
@@ -180,7 +180,7 @@ Display the version number of scontrol being executed.
\fI!!\fP
Repeat the last command executed.
.TP
-\fBSPECIFICATIONS FOR UPDATE COMMAND, JOBS\fR
+\fBSPECIFICATIONS FOR SHOW AND UPDATE COMMANDS, JOBS\fR
.TP
\fIAccount\fP=<account>
Account name to be changed for this job's resource use.
@@ -194,6 +194,12 @@ Possible values are"YES" and "NO".
Defer job's initiation until specified job_id completes.
Cancel dependency with job_id value of "0", "Depedency=0".
.TP
+\fIExitCode\fP=<exit>:<sig>
+Exit status reported for the job by the wait() function.
+The first number is the exit code, typically as set by the exit() function.
+The second number of the signal that caused the process to terminate if
+it was terminated by a signal.
+.TP
\fIFeatures\fP=<features>
Set the job's required features on nodes specified value. Multiple values
may be comma separated if all features are required (AND operation) or
diff --git a/slurm/slurm.h.in b/slurm/slurm.h.in
index 9f6e44ed0cb..abbf1043257 100644
--- a/slurm/slurm.h.in
+++ b/slurm/slurm.h.in
@@ -470,6 +470,7 @@ typedef struct job_info {
* start_range_2, .., -1 */
char *features; /* comma separated list of required features */
uint32_t dependency; /* defer until specified job completes */
+ uint32_t exit_code; /* exit code for job (status from wait call) */
char *account; /* charge to specified account */
uint16_t wait_reason; /* reason job still pending, see
* slurm.h:enum job_wait_reason */
diff --git a/src/api/job_info.c b/src/api/job_info.c
index 92e007e51a5..5597d95d04e 100644
--- a/src/api/job_info.c
+++ b/src/api/job_info.c
@@ -47,6 +47,7 @@
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
+#include <sys/wait.h>
#include <time.h>
#include <slurm/slurm.h>
@@ -117,6 +118,7 @@ slurm_sprint_job_info ( job_info_t * job_ptr, int one_liner )
char tmp_line[128];
uint16_t quarter = (uint16_t) NO_VAL;
uint16_t nodecard = (uint16_t) NO_VAL;
+ uint16_t term_sig = 0;
char *out = NULL;
#ifdef HAVE_BG
@@ -178,9 +180,15 @@ slurm_sprint_job_info ( job_info_t * job_ptr, int one_liner )
if (job_ptr->time_limit == INFINITE)
sprintf(tmp_line, "UNLIMITED");
else if (job_ptr->time_limit == NO_VAL)
- sprintf(tmp_line, "Partition_Limit");
+ sprintf(tmp_line, "Partition_Limit ");
else
- sprintf(tmp_line, "%u", job_ptr->time_limit);
+ sprintf(tmp_line, "%u ", job_ptr->time_limit);
+ xstrcat(out, tmp_line);
+ if (WIFSIGNALED(job_ptr->exit_code))
+ term_sig = WTERMSIG(job_ptr->exit_code);
+ snprintf(tmp_line, sizeof(tmp_line),
+ "ExitCode=%u:%u",
+ WEXITSTATUS(job_ptr->exit_code), term_sig);
xstrcat(out, tmp_line);
if (one_liner)
xstrcat(out, " ");
diff --git a/src/common/slurm_protocol_pack.c b/src/common/slurm_protocol_pack.c
index d33e191dd71..941b0d7a2f4 100644
--- a/src/common/slurm_protocol_pack.c
+++ b/src/common/slurm_protocol_pack.c
@@ -1929,6 +1929,7 @@ _unpack_job_info_members(job_info_t * job, Buf buffer)
safe_unpackstr_xmalloc(&job->network, &uint16_tmp, buffer);
safe_unpackstr_xmalloc(&job->comment, &uint16_tmp, buffer);
safe_unpack32(&job->dependency, buffer);
+ safe_unpack32(&job->exit_code, buffer);
safe_unpackstr_xmalloc(&job->name, &uint16_tmp, buffer);
safe_unpackstr_xmalloc(&job->alloc_node, &uint16_tmp, buffer);
diff --git a/src/slurmctld/job_mgr.c b/src/slurmctld/job_mgr.c
index e1a37ddefb0..6e612e28ede 100644
--- a/src/slurmctld/job_mgr.c
+++ b/src/slurmctld/job_mgr.c
@@ -2797,6 +2797,7 @@ void pack_job(struct job_record *dump_job_ptr, Buf buffer)
packstr(dump_job_ptr->network, buffer);
packstr(dump_job_ptr->comment, buffer);
pack32((uint32_t)dump_job_ptr->dependency, buffer);
+ pack32((uint32_t)dump_job_ptr->exit_code, buffer);
packstr(dump_job_ptr->name, buffer);
packstr(dump_job_ptr->alloc_node, buffer);
diff --git a/src/slurmctld/slurmctld.h b/src/slurmctld/slurmctld.h
index 67ecc367373..6566915f976 100644
--- a/src/slurmctld/slurmctld.h
+++ b/src/slurmctld/slurmctld.h
@@ -356,7 +356,8 @@ struct job_record {
uint16_t mail_type; /* see MAIL_JOB_* in slurm.h */
char *mail_user; /* user to get e-mail notification */
uint32_t requid; /* requester user ID */
- uint32_t exit_code; /* exit code for the job */
+ uint32_t exit_code; /* exit code for job (status from
+ * wait call) */
};
struct step_record {
--
GitLab