[PATCH 01/10] oom: don't try to kill oom_unkillable child

[KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>]

Now, badness() doesn't care neigher CPUSET nor mempolicy. Then
if the victim child process is oom_unkillable()==1, __out_of_memory()
can makes kernel hang eventually.

This patch fixes it.


[remark: this is needed to fold "oom: sacrifice child with highest
badness score for parent"]
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
---
 mm/oom_kill.c |    5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index d49d542..0d7397b 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -387,9 +387,6 @@ static void dump_header(struct task_struct *p, gfp_t gfp_mask, int order,
 static int __oom_kill_process(struct task_struct *p, struct mem_cgroup *mem,
 			      int verbose)
 {
-	if (oom_unkillable(p, mem))
-		return 1;
-
 	p = find_lock_task_mm(p);
 	if (!p)
 		return 1;
@@ -440,6 +437,8 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order,
 
 			if (c->mm == p->mm)
 				continue;
+			if (oom_unkillable(c, mem, nodemask))
+				continue;
 
 			/* badness() returns 0 if the thread is unkillable */
 			cpoints = badness(c, uptime.tv_sec);
-- 
1.6.5.2