Messages in this thread | | | Date | Wed, 16 Jun 2010 23:41:27 +0900 | From | Minchan Kim <> | Subject | Re: [PATCH 1/9] oom: don't try to kill oom_unkillable child |
| |
On Wed, Jun 16, 2010 at 08:29:13PM +0900, KOSAKI Motohiro wrote: > Now, badness() doesn't care neigher CPUSET nor mempolicy. Then > if the victim child process have disjoint nodemask, __out_of_memory() > can makes kernel hang eventually. > > This patch fixes it. > > Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
This patch inclues two things.
1. consider cpuset and mempolicy in oom_kill_process 2. Simplify mempolicy oom check with nodemask != NULL in select_bad_process.
1) change behavior but 2) is just cleanup. It should have been in another patch to reivew easily. :)
-- Kind regards, Minchan Kim
| |