Re: [PATCH v4] mm: vmpressure: don't count proactive reclaim in vmpressure

From: Michal Hocko
Date: Thu Jul 21 2022 - 07:42:35 EST


On Wed 20-07-22 10:49:53, Shakeel Butt wrote:
> On Wed, Jul 20, 2022 at 2:24 AM Michal Hocko <mhocko@xxxxxxxx> wrote:
> >
> [...]
> >
> > I think what we are missing here is
> > - explain that this doesn't have any effect on existing users of
> > vmpressure user interface because that is cgroup v1 and memory.reclaim
> > is v2 feature. This is a trivial statement but quite useful for future
> > readers of this commit
> > - explain the effect on the networking layer and typical usecases
> > memory.reclaim is used for currently and ideally document that.
>
> I agree with the above two points (Yosry, please address those) but
> the following third point is orthogonal and we don't really need to
> have an answer for this patch to be accepted.
>
> > - how are we going to deal with users who would really want to use
> > memory.reclaim interface as a replacement for existing hard/high
> > memory reclaim? Is that even something that the interface is intended
> > for?
>
> I do agree that this question is important. Nowadays I am looking at
> this from a different perspective and use-case. More concretely how
> (and why) to replace vmpressure based network throttling for cgroup
> v2. I will start a separate thread for that discussion.

I think we should be good to document this side effect for now. If you
have a plan to change to vmpressure based throttling then only better.
But one way or the other impact of the memory.reclaim interface on
netwroking should be documented properly.

--
Michal Hocko
SUSE Labs