Re: [PATCH 1/2] mm,oom: Move last second allocation to inside the OOM killer.

2017-11-02 Thread Tetsuo Handa
Michal Hocko wrote: > I would really suggest you to stick with the changelog I have suggested. > Well, I think that this patch needs to clarify why using ALLOC_WMARK_HIGH. > On Wed 01-11-17 20:54:27, Tetsuo Handa wrote: > [...] > > diff --git a/mm/oom_kill.c b/mm/oom_kill.c > > index 26add8a..118

Re: [PATCH 1/2] mm,oom: Move last second allocation to inside the OOM killer.

2017-11-01 Thread Michal Hocko
I would really suggest you to stick with the changelog I have suggested. On Wed 01-11-17 20:54:27, Tetsuo Handa wrote: [...] > diff --git a/mm/oom_kill.c b/mm/oom_kill.c > index 26add8a..118ecdb 100644 > --- a/mm/oom_kill.c > +++ b/mm/oom_kill.c > @@ -870,6 +870,19 @@ static void oom_kill_process(

[PATCH 1/2] mm,oom: Move last second allocation to inside the OOM killer.

2017-11-01 Thread Tetsuo Handa
__alloc_pages_may_oom() is doing last second allocation attempt using ALLOC_WMARK_HIGH before calling out_of_memory(). This has two motivations. The first one is explained by the comment that it aims to catch potential parallel OOM killing and the second one was explained by Andrea Arcangeli as fol