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;
MIME-Version: 1.0
References: <20220225012819.1807147-1-surenb@google.com> <YiY7K4ftAI3t6km8@dhcp22.suse.cz>
 <CAJuCfpEt6V+v_FcLsYWpGLA1vwCK01vv0PqNernfKM2GTzNqKg@mail.gmail.com>
In-Reply-To: <CAJuCfpEt6V+v_FcLsYWpGLA1vwCK01vv0PqNernfKM2GTzNqKg@mail.gmail.com>
From:   Suren Baghdasaryan <surenb@google.com>
Date:   Thu, 17 Mar 2022 16:04:11 -0700
Message-ID: <CAJuCfpHgTb5Uvd7yfwSHnPx7gf2wXDfRa0U+XOwr=qHm+JJTNA@mail.gmail.com>
Subject: Re: [RFC 1/1] mm: page_alloc: replace mm_percpu_wq with kthreads in drain_all_pages
To:     Michal Hocko <mhocko@suse.com>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Petr Mladek <pmladek@suse.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Roman Gushchin <guro@fb.com>,
        Shakeel Butt <shakeelb@google.com>,
        Minchan Kim <minchan@kernel.org>,
        Tim Murray <timmurray@google.com>,
        linux-mm <linux-mm@kvack.org>,
        LKML <linux-kernel@vger.kernel.org>,
        kernel-team <kernel-team@android.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Mar 7, 2022 at 9:24 AM Suren Baghdasaryan <surenb@google.com> wrote:
>
> On Mon, Mar 7, 2022 at 9:04 AM 'Michal Hocko' via kernel-team
> <kernel-team@android.com> wrote:
> >
> > On Thu 24-02-22 17:28:19, Suren Baghdasaryan wrote:
> > > Sending as an RFC to confirm if this is the right direction and to
> > > clarify if other tasks currently executed on mm_percpu_wq should be
> > > also moved to kthreads. The patch seems stable in testing but I want
> > > to collect more performance data before submitting a non-RFC version.
> > >
> > >
> > > Currently drain_all_pages uses mm_percpu_wq to drain pages from pcp
> > > list during direct reclaim. The tasks on a workqueue can be delayed
> > > by other tasks in the workqueues using the same per-cpu worker pool.
> > > This results in sizable delays in drain_all_pages when cpus are highly
> > > contended.
> >
> > This is not about cpus being highly contended. It is about too much work
> > on the WQ context.
>
> Ack.
>
> >
> > > Memory management operations designed to relieve memory pressure should
> > > not be allowed to block by other tasks, especially if the task in direct
> > > reclaim has higher priority than the blocking tasks.
> >
> > Agreed here.
> >
> > > Replace the usage of mm_percpu_wq with per-cpu low priority FIFO
> > > kthreads to execute draining tasks.
> >
> > This looks like a natural thing to do when WQ context is not suitable
> > but I am not sure the additional resources is really justified. Large
> > machines with a lot of cpus would create a lot of kernel threads. Can we
> > do better than that?
> >
> > Would it be possible to have fewer workers (e.g. 1 or one per numa node)
> > and it would perform the work on a dedicated cpu by changing its
> > affinity? Or would that introduce an unacceptable overhead?
>
> Not sure but I can try implementing per-node kthreads and measure the
> performance of the reclaim path, comparing with the current and with
> per-cpu approach.

Just to update on this RFC. In my testing I don't see a meaningful
improvement from using the kthreads yet. This might be due to my test
setup, so I'll keep exploring. Will post the next version only if I
get demonstrable improvements.
Thanks!

>
> >
> > Or would it be possible to update the existing WQ code to use rescuer
> > well before the WQ is completely clogged?
> > --
> > Michal Hocko
> > SUSE Labs
> >
> > --
> > To unsubscribe from this group and stop receiving emails from it, send an email to kernel-team+unsubscribe@android.com.
> >