From 512a1dcba425ae3f394771b2a274ef336be4e6df Mon Sep 17 00:00:00 2001
From: Morris Jette <jette@schedmd.com>
Date: Fri, 7 Jul 2017 14:50:20 -0600
Subject: [PATCH] Disable multiple steps per pack job component with single
srun command
---
doc/html/heterogeneous_jobs.shtml | 4 ++++
src/srun/libsrun/srun_job.c | 12 ++++++++++++
2 files changed, 16 insertions(+)
diff --git a/doc/html/heterogeneous_jobs.shtml b/doc/html/heterogeneous_jobs.shtml
index c8ab10f47e8..ec6dcc4deb8 100644
--- a/doc/html/heterogeneous_jobs.shtml
+++ b/doc/html/heterogeneous_jobs.shtml
@@ -319,6 +319,10 @@ the job execute on different clusters in the federation.</p>
<p>The srun command's --no-allocate option is not supported
for heterogeneous jobs.</p>
+<p>Only one job step per heterogeneous job component can be launched by a
+single srun command (e.g.
+"srun --pack-group=0 alpha : --pack-group=0 beta" is not supported).</p>
+
<p>The sattach command can only be used to attach to a single component of
a heterogeneous job at a time.</p>
diff --git a/src/srun/libsrun/srun_job.c b/src/srun/libsrun/srun_job.c
index 4026bd66728..da54fbac8db 100644
--- a/src/srun/libsrun/srun_job.c
+++ b/src/srun/libsrun/srun_job.c
@@ -458,6 +458,7 @@ static void _pack_grp_test(List opt_list)
ListIterator iter;
opt_t *opt_local;
int pack_offset;
+ bitstr_t *master_map = NULL;
if (opt_list) {
iter = list_iterator_create(opt_list);
@@ -469,7 +470,18 @@ static void _pack_grp_test(List opt_list)
xstrfmtcat(opt_local->pack_group, "%d",
pack_offset);
}
+ if (!master_map) {
+ master_map = bit_copy(opt_local->pack_grp_bits);
+ } else {
+ if (bit_overlap(master_map,
+ opt_local->pack_grp_bits)) {
+ error("Duplicate pack groups in single srun not supported");
+ exit(error_exit);
+ }
+ bit_or(master_map, opt_local->pack_grp_bits);
+ }
}
+ FREE_NULL_BITMAP(master_map);
list_iterator_destroy(iter);
} else if (!opt.pack_group && !getenv("SLURM_PACK_SIZE")) {
FREE_NULL_BITMAP(opt.pack_grp_bits);
--
GitLab