Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 16 Dec 2022 07:36:40 +0000
In-Reply-To: <13c73151-4b28-4324-afd5-8b84b82bc25d@shopee.com>
Mime-Version: 1.0
References: <20221215091907.763801-1-haifeng.xu@shopee.com>
 <20221215181803.ome46pkh6g7qu3t4@google.com> <abec13da-c74a-0fd8-74e0-3cdc547e4e08@shopee.com>
 <20221216064210.krxtxebuwc7dijzu@google.com> <13c73151-4b28-4324-afd5-8b84b82bc25d@shopee.com>
Message-ID: <20221216073640.xjtpsyigoej77v5h@google.com>
Subject: Re: [PATCH] mm/memcontrol: Skip root memcg in memcg_memory_event_mm
From:   Shakeel Butt <shakeelb@google.com>
To:     Haifeng Xu <haifeng.xu@shopee.com>
Cc:     akpm@linux-foundation.org, roman.gushchin@linux.dev,
        songmuchun@bytedance.com, hannes@cmpxchg.org, vbabka@suse.cz,
        willy@infradead.org, vasily.averin@linux.dev,
        linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="us-ascii"
Precedence: bulk

On Fri, Dec 16, 2022 at 03:28:53PM +0800, Haifeng Xu wrote:
> 
> 
> On 2022/12/16 14:42, Shakeel Butt wrote:
> > On Fri, Dec 16, 2022 at 09:43:02AM +0800, Haifeng Xu wrote:
> >>
> >>
> >> On 2022/12/16 02:18, Shakeel Butt wrote:
> >>> On Thu, Dec 15, 2022 at 09:19:07AM +0000, Haifeng Xu wrote:
> >>>> The memory events aren't supported on root cgroup, so there is no need
> >>>> to account MEMCG_OOM_KILL on root memcg.
> >>>>
> >>>
> >>> Can you explain the scenario where this is happening and causing issue
> >>> for you?
> >>>
> >> If the victim selected by oom killer belongs to root memcg, memcg_memory_event_mm
> >> still counts the MEMCG_OOM_KILL event. This behavior is meaningless because the
> >> flag of events/events.local in memory_files is CFTYPE_NOT_ON_ROOT. The root memcg
> >> does not count any memory event.
> >>
> > 
> > What about v1's memory.oom_control?
> >
>  
> The memory.oom_control doesn't set the CFTYPE_NOT_ON_ROOT flag. But oom_kill_disable or
> under_oom actually only support non-root memcg, so the memory_events should be consistent
> with them.

Did you take a look at mem_cgroup_oom_control_read()? It is displaying
MEMCG_OOM_KILL for root memcg. Irrespective it makes sense or not, you
want to change behavior of user visible interface. If you really want to
then propose for the deprecation of that interface.