MIME-Version: 1.0
References: <20230306154138.3775-1-findns94@gmail.com> <20230306154138.3775-4-findns94@gmail.com>
 <ZAYoi8ZwwbXT9j7f@dhcp22.suse.cz>
In-Reply-To: <ZAYoi8ZwwbXT9j7f@dhcp22.suse.cz>
From:   Martin Zhao <findns94@gmail.com>
Date:   Wed, 8 Mar 2023 22:51:54 +0800
Message-ID: <CADfL_jA3fa6HrGLRp25avQm+yamxqUZhK6BLjnjANbaJF7tsBw@mail.gmail.com>
Subject: Re: [PATCH v2, 3/4] mm, memcg: Prevent memory.oom_control load/store tearing
To:     Michal Hocko <mhocko@suse.com>
Cc:     akpm@linux-foundation.org, roman.gushchin@linux.dev,
        hannes@cmpxchg.org, shakeelb@google.com, muchun.song@linux.dev,
        willy@infradead.org, linux-mm@kvack.org, cgroups@vger.kernel.org,
        linux-kernel@vger.kernel.org, tangyeechou@gmail.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Mar 7, 2023 at 1:53=E2=80=AFAM Michal Hocko <mhocko@suse.com> wrote=
:
>
> On Mon 06-03-23 23:41:37, Yue Zhao wrote:
> > The knob for cgroup v1 memory controller: memory.oom_control
> > is not protected by any locking so it can be modified while it is used.
> > This is not an actual problem because races are unlikely.
> > But it is better to use READ_ONCE/WRITE_ONCE to prevent compiler from
> > doing anything funky.
> >
> > The access of memcg->oom_kill_disable is lockless,
> > so it can be concurrently set at the same time as we are
> > trying to read it.
> >
> > Signed-off-by: Yue Zhao <findns94@gmail.com>
> > ---
> >  mm/memcontrol.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > index dca895c66a9b..26605b2f51b1 100644
> > --- a/mm/memcontrol.c
> > +++ b/mm/memcontrol.c
> > @@ -4515,7 +4515,7 @@ static int mem_cgroup_oom_control_read(struct seq=
_file *sf, void *v)
> >  {
> >       struct mem_cgroup *memcg =3D mem_cgroup_from_seq(sf);
> >
> > -     seq_printf(sf, "oom_kill_disable %d\n", memcg->oom_kill_disable);
> > +     seq_printf(sf, "oom_kill_disable %d\n", READ_ONCE(memcg->oom_kill=
_disable));
> >       seq_printf(sf, "under_oom %d\n", (bool)memcg->under_oom);
> >       seq_printf(sf, "oom_kill %lu\n",
> >                  atomic_long_read(&memcg->memory_events[MEMCG_OOM_KILL]=
));
> > @@ -4531,7 +4531,7 @@ static int mem_cgroup_oom_control_write(struct cg=
roup_subsys_state *css,
> >       if (mem_cgroup_is_root(memcg) || !((val =3D=3D 0) || (val =3D=3D =
1)))
> >               return -EINVAL;
> >
> > -     memcg->oom_kill_disable =3D val;
> > +     WRITE_ONCE(memcg->oom_kill_disable, val);
> >       if (!val)
> >               memcg_oom_recover(memcg);
>
> Any specific reasons you haven't covered other accesses
> (mem_cgroup_css_alloc, mem_cgroup_oom, mem_cgroup_oom_synchronize)?

Thanks for point this out, you are right, we should add
[READ|WRITE]_ONCE for all used places.
Let me create PATCH v3 later.
Also for the memcg->soft_limit, I will update as well.

> >
> > --
> > 2.17.1
>
> --
> Michal Hocko
> SUSE Labs