Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
References: <20201014190749.24607-1-rpalethorpe@suse.com>
 <20201016094702.GA95052@blackbook> <20201016145308.GA312010@cmpxchg.org>
 <20201016171502.GA102311@blackbook>
User-agent: mu4e 1.4.13; emacs 27.1
From:   Richard Palethorpe <rpalethorpe@suse.de>
To:     Michal =?utf-8?Q?Koutn=C3=BD?= <mkoutny@suse.com>
Cc:     Johannes Weiner <hannes@cmpxchg.org>, Roman Gushchin <guro@fb.com>,
        ltp@lists.linux.it, Andrew Morton <akpm@linux-foundation.org>,
        Shakeel Butt <shakeelb@google.com>,
        Christoph Lameter <cl@linux.com>,
        Michal Hocko <mhocko@kernel.org>, Tejun Heo <tj@kernel.org>,
        Vlastimil Babka <vbabka@suse.cz>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH] mm: memcg/slab: Stop reparented obj_cgroups from
 charging root
Reply-To: rpalethorpe@suse.de
In-reply-to: <20201016171502.GA102311@blackbook>
Date:   Mon, 19 Oct 2020 09:45:32 +0100
Message-ID: <87lfg2ob83.fsf@suse.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8BIT
Precedence: bulk

Hello,

Michal Koutný <mkoutny@suse.com> writes:

> On Fri, Oct 16, 2020 at 10:53:08AM -0400, Johannes Weiner <hannes@cmpxchg.org> wrote:
>> The central try_charge() function charges recursively all the way up
>> to and including the root.
> Except for use_hiearchy=0 (which is the case here as Richard
> wrote). The reparenting is hence somewhat incompatible with
> new_parent.use_hiearchy=0 :-/
>

Yes and it also seems

new_parent.use_hierarch=0 -> new_child.use_hierarchy=0

and

new_parent.use_hierarch=0 -> new_child.use_hierarchy=1

are considered valid on cgroupsV1. The kernel will also allow more
descendants on new_child.use_hierarchy=0, but sets
broken_hierarchy=1. However this will not stop the stack trace occuring
(AFAICT) when the reparenting happens between two descendants.

>> We should clean this up one way or another: either charge the root or
>> don't, but do it consistently.
> I agree this'd be good to unify. One upside of excluding root memcg from
> charging is that users are spared from the charging overhead when memcg
> tree is not created.  (Actually, I thought that was the reason for this
> exception.)
>
> Michal


-- 
Thank you,
Richard.