Re: oomkillers gone wild.

[David Rientjes <rientjes@google.com>]

On Sun, 10 Jun 2012, Dave Jones wrote:

> To double check, here it is in rc2 (which has that patch)..
> 
> $ uname -r
> 3.5.0-rc2+
> $ cat /proc/$(pidof dbus-daemon)/oom_score{_adj,}
> -900
> 7441500919753
> $ grep RSS /proc/$(pidof dbus-daemon)/status
> VmRSS:	    1604 kB

Eek, yes, that's definitely wrong.  The following should fix it.
---
 mm/oom_kill.c |   15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)

diff --git a/mm/oom_kill.c b/mm/oom_kill.c
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -184,6 +184,7 @@ unsigned long oom_badness(struct task_struct *p, struct mem_cgroup *memcg,
 			  const nodemask_t *nodemask, unsigned long totalpages)
 {
 	long points;
+	long adj;
 
 	if (oom_unkillable_task(p, memcg, nodemask))
 		return 0;
@@ -192,7 +193,8 @@ unsigned long oom_badness(struct task_struct *p, struct mem_cgroup *memcg,
 	if (!p)
 		return 0;
 
-	if (p->signal->oom_score_adj == OOM_SCORE_ADJ_MIN) {
+	adj = p->signal->oom_score_adj;
+	if (adj == OOM_SCORE_ADJ_MIN) {
 		task_unlock(p);
 		return 0;
 	}
@@ -210,14 +212,11 @@ unsigned long oom_badness(struct task_struct *p, struct mem_cgroup *memcg,
 	 * implementation used by LSMs.
 	 */
 	if (has_capability_noaudit(p, CAP_SYS_ADMIN))
-		points -= 30 * totalpages / 1000;
+		adj -= 30;
 
-	/*
-	 * /proc/pid/oom_score_adj ranges from -1000 to +1000 such that it may
-	 * either completely disable oom killing or always prefer a certain
-	 * task.
-	 */
-	points += p->signal->oom_score_adj * totalpages / 1000;
+	/* Normalize to oom_score_adj units */
+	adj *= totalpages / 1000;
+	points += adj;
 
 	/*
 	 * Never return 0 for an eligible task regardless of the root bonus and

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>