[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87fsbinsvq.fsf@nvidia.com>
Date: Tue, 07 Feb 2023 12:50:24 +1100
From: Alistair Popple <apopple@...dia.com>
To: Tejun Heo <tj@...nel.org>
Cc: Waiman Long <longman@...hat.com>,
Yosry Ahmed <yosryahmed@...gle.com>, linux-mm@...ck.org,
cgroups@...r.kernel.org, linux-kernel@...r.kernel.org,
jgg@...dia.com, jhubbard@...dia.com, tjmercier@...gle.com,
hannes@...xchg.org, surenb@...gle.com, mkoutny@...e.com,
daniel@...ll.ch, "Daniel P . Berrange" <berrange@...hat.com>,
Alex Williamson <alex.williamson@...hat.com>,
Zefan Li <lizefan.x@...edance.com>,
Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH 14/19] mm: Introduce a cgroup for pinned memory
Tejun Heo <tj@...nel.org> writes:
> On Mon, Feb 06, 2023 at 08:00:54PM -0500, Waiman Long wrote:
>> If it should not be part of the memcg, does it make sense to make it a
>> resource in the existing misc controller? I believe we don't want a
>> proliferation of new cgroup controllers.
>
> Yeah, if it's gonna be an independent knob, I suppose so, but I really think
> the locked accounting should be tied to the page, which mostly likely would
> mean that it'd be tied to the page ownership too making its natural place
> memcg.
Yes, I think it might be possible. I looked briefly at doing it when
initially developing the series but I would like to resolve the question
of independent knob vs. memcg before heading too far down either path.
> Thanks.
Powered by blists - more mailing lists