On Tue, Jan 28, 2020 at 11:44 AM Cong Wang <xiyou.wangcong@xxxxxxxxx> wrote:
>
> On Tue, Jan 28, 2020 at 3:39 AM Michal Hocko <mhocko@xxxxxxxxxx> wrote:
> >
> > On Tue 28-01-20 02:48:57, Matthew Wilcox wrote:
> > > Doesn't the stack trace above indicate that we're doing migration as
> > > the result of an allocation in add_to_page_cache_lru()?
> >
> > Which stack trace do you refer to? Because the one above doesn't show
> > much more beyond mem_cgroup_iter and likewise others in this email
> > thread. I do not really remember any stack with lock_page on the trace.
>
> I think the page is locked in add_to_page_cache_lru() by
> __SetPageLocked(), as the stack trace shows __add_to_page_cache_locked().
> It is not yet unlocked, as it is still looping inside try_charge().
>
> I will write a script to see if I can find the longest time spent in reclaim
> as you suggested.
After digging the changelog, I believe the following commit could fix
the problem:
commit f9c645621a28e37813a1de96d9cbd89cde94a1e4
Author: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
Date: Mon Sep 23 15:37:08 2019 -0700
memcg, oom: don't require __GFP_FS when invoking memcg OOM killer
which is not yet in our 4.19 branch yet. We will sync with 4.19 stable soon.
Thanks.
