This patch itself is againt mmotm-Feb10 but can be applied to 2.6.32.8
without problem.
==
From: KAMEZAWA Hiroyuki <[email protected]>
Now, oom-killer is memcg aware and it finds the worst process from
processes under memcg(s) in oom. Then, it kills victim's child at first.
It may kill a child in other cgroup and may not be any help for recovery.
And it will break the assumption users have...
This patch fixes it.
CC: [email protected]
CC: Minchan Kim <[email protected]>
CC: Balbir Singh <[email protected]>
CC: Daisuke Nishimura <[email protected]>
Acked-by: David Rientjes <[email protected]>
Signed-off-by: KAMEZAWA Hiroyuki <[email protected]>
---
mm/oom_kill.c | 2 ++
1 file changed, 2 insertions(+)
Index: mmotm-2.6.33-Feb10/mm/oom_kill.c
===================================================================
--- mmotm-2.6.33-Feb10.orig/mm/oom_kill.c
+++ mmotm-2.6.33-Feb10/mm/oom_kill.c
@@ -459,6 +459,8 @@ static int oom_kill_process(struct task_
list_for_each_entry(c, &p->children, sibling) {
if (c->mm == p->mm)
continue;
+ if (mem && !task_in_mem_cgroup(c, mem))
+ continue;
if (!oom_kill_task(c))
return 0;
}
On Fri, 12 Feb 2010 10:53:18 +0900, KAMEZAWA Hiroyuki <[email protected]> wrote:
> This patch itself is againt mmotm-Feb10 but can be applied to 2.6.32.8
> without problem.
>
> ==
> From: KAMEZAWA Hiroyuki <[email protected]>
>
> Now, oom-killer is memcg aware and it finds the worst process from
> processes under memcg(s) in oom. Then, it kills victim's child at first.
> It may kill a child in other cgroup and may not be any help for recovery.
> And it will break the assumption users have...
>
> This patch fixes it.
>
This bug should definitely be fixed. Thank you for finding and fixing it.
Reviewed-by: Daisuke Nishimura <[email protected]>
> CC: [email protected]
> CC: Minchan Kim <[email protected]>
> CC: Balbir Singh <[email protected]>
> CC: Daisuke Nishimura <[email protected]>
> Acked-by: David Rientjes <[email protected]>
> Signed-off-by: KAMEZAWA Hiroyuki <[email protected]>
>
> ---
> mm/oom_kill.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> Index: mmotm-2.6.33-Feb10/mm/oom_kill.c
> ===================================================================
> --- mmotm-2.6.33-Feb10.orig/mm/oom_kill.c
> +++ mmotm-2.6.33-Feb10/mm/oom_kill.c
> @@ -459,6 +459,8 @@ static int oom_kill_process(struct task_
> list_for_each_entry(c, &p->children, sibling) {
> if (c->mm == p->mm)
> continue;
> + if (mem && !task_in_mem_cgroup(c, mem))
> + continue;
> if (!oom_kill_task(c))
> return 0;
> }
>
On Fri, Feb 12, 2010 at 10:53 AM, KAMEZAWA Hiroyuki
<[email protected]> wrote:
> This patch itself is againt mmotm-Feb10 but can be applied to 2.6.32.8
> without problem.
>
> ==
> From: KAMEZAWA Hiroyuki <[email protected]>
>
> Now, oom-killer is memcg aware and it finds the worst process from
> processes under memcg(s) in oom. Then, it kills victim's child at first.
> It may kill a child in other cgroup and may not be any help for recovery.
> And it will break the assumption users have...
>
> This patch fixes it.
>
> CC: [email protected]
> CC: Minchan Kim <[email protected]>
> CC: Balbir Singh <[email protected]>
> CC: Daisuke Nishimura <[email protected]>
> Acked-by: David Rientjes <[email protected]>
> Signed-off-by: KAMEZAWA Hiroyuki <[email protected]>
Reviewed-by: Minchan Kim <[email protected]>
Sorry for noise, Kame.
--
Kind regards,
Minchan Kim
On Fri, 12 Feb 2010 11:39:35 +0900
Minchan Kim <[email protected]> wrote:
> On Fri, Feb 12, 2010 at 10:53 AM, KAMEZAWA Hiroyuki
> <[email protected]> wrote:
> > This patch itself is againt mmotm-Feb10 but can be applied to 2.6.32.8
> > without problem.
> >
> > ==
> > From: KAMEZAWA Hiroyuki <[email protected]>
> >
> > Now, oom-killer is memcg aware and it finds the worst process from
> > processes under memcg(s) in oom. Then, it kills victim's child at first.
> > It may kill a child in other cgroup and may not be any help for recovery.
> > And it will break the assumption users have...
> >
> > This patch fixes it.
> >
> > CC: [email protected]
> > CC: Minchan Kim <[email protected]>
> > CC: Balbir Singh <[email protected]>
> > CC: Daisuke Nishimura <[email protected]>
> > Acked-by: David Rientjes <[email protected]>
> > Signed-off-by: KAMEZAWA Hiroyuki <[email protected]>
> Reviewed-by: Minchan Kim <[email protected]>
>
> Sorry for noise, Kame.
>
No problem. You give me a chance to consider other problems/dirtiness of codes.
I continue review to make memcg cleaer.
Thanks,
-Kame