Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190817004726.2530670-1-guro@fb.com> <CALOAHbBsMNLN6jZn83zx6EWM_092s87zvDQ7p-MZpY+HStk-1Q@mail.gmail.com>
 <20190817191419.GA11125@castle> <CALOAHbA-Z-1QDSgQ6H6QhPaPwAGyqfpd3Gbq-KLnoO=ZZxWnrw@mail.gmail.com>
 <20190819212034.GB24956@tower.dhcp.thefacebook.com>
In-Reply-To: <20190819212034.GB24956@tower.dhcp.thefacebook.com>
From:   Yafang Shao <laoar.shao@gmail.com>
Date:   Tue, 20 Aug 2019 09:29:14 +0800
Message-ID: <CALOAHbCwWHirJjmByeAVZdDoHpCMabq20tzMdhr_25Ddic9TYw@mail.gmail.com>
Subject: Re: [PATCH] Partially revert "mm/memcontrol.c: keep local VM counters
 in sync with the hierarchical ones"
To:     Roman Gushchin <guro@fb.com>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        Linux MM <linux-mm@kvack.org>,
        Michal Hocko <mhocko@kernel.org>,
        Johannes Weiner <hannes@cmpxchg.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Kernel Team <Kernel-team@fb.com>,
        "stable@vger.kernel.org" <stable@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Aug 20, 2019 at 5:20 AM Roman Gushchin <guro@fb.com> wrote:
>
> On Sun, Aug 18, 2019 at 08:30:15AM +0800, Yafang Shao wrote:
> > On Sun, Aug 18, 2019 at 3:14 AM Roman Gushchin <guro@fb.com> wrote:
> > >
> > > On Sat, Aug 17, 2019 at 11:33:57AM +0800, Yafang Shao wrote:
> > > > On Sat, Aug 17, 2019 at 8:47 AM Roman Gushchin <guro@fb.com> wrote:
> > > > >
> > > > > Commit 766a4c19d880 ("mm/memcontrol.c: keep local VM counters in sync
> > > > > with the hierarchical ones") effectively decreased the precision of
> > > > > per-memcg vmstats_local and per-memcg-per-node lruvec percpu counters.
> > > > >
> > > > > That's good for displaying in memory.stat, but brings a serious regression
> > > > > into the reclaim process.
> > > > >
> > > > > One issue I've discovered and debugged is the following:
> > > > > lruvec_lru_size() can return 0 instead of the actual number of pages
> > > > > in the lru list, preventing the kernel to reclaim last remaining
> > > > > pages. Result is yet another dying memory cgroups flooding.
> > > > > The opposite is also happening: scanning an empty lru list
> > > > > is the waste of cpu time.
> > > > >
> > > > > Also, inactive_list_is_low() can return incorrect values, preventing
> > > > > the active lru from being scanned and freed. It can fail both because
> > > > > the size of active and inactive lists are inaccurate, and because
> > > > > the number of workingset refaults isn't precise. In other words,
> > > > > the result is pretty random.
> > > > >
> > > > > I'm not sure, if using the approximate number of slab pages in
> > > > > count_shadow_number() is acceptable, but issues described above
> > > > > are enough to partially revert the patch.
> > > > >
> > > > > Let's keep per-memcg vmstat_local batched (they are only used for
> > > > > displaying stats to the userspace), but keep lruvec stats precise.
> > > > > This change fixes the dead memcg flooding on my setup.
> > > > >
> > > >
> > > > That will make some misunderstanding if the local counters are not in
> > > > sync with the hierarchical ones
> > > > (someone may doubt whether there're something leaked.).
> > >
> > > Sure, but the actual leakage is a much more serious issue.
> > >
> > > > If we have to do it like this, I think we should better document this behavior.
> > >
> > > Lru size calculations can be done using per-zone counters, which is
> > > actually cheaper, because the number of zones is usually smaller than
> > > the number of cpus. I'll send a corresponding patch on Monday.
> > >
> >
> > Looks like a good idea.
> >
> > > Maybe other use cases can also be converted?
> >
> > We'd better keep the behavior the same across counters. I think you
> > can have a try.
>
> As I said, consistency of counters is important, but not nearly as important
> as the real behavior of the system. Especially because we talk about
> per-node memcg statistics, which I believe is mostly used for debugging.
>
> So for now I think the right thing to do is to revert the change to fix
> the memory reclaim process. And then we can discuss how to get counters
> right.
>

Sure.

Thanks
Yafang