Received-SPF: pass (google.com: domain of linux-kernel+bounces-10907-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20231222102255.56993-1-ryncsn@gmail.com> <20231222102255.56993-4-ryncsn@gmail.com>
In-Reply-To: <20231222102255.56993-4-ryncsn@gmail.com>
From: Yu Zhao <yuzhao@google.com>
Date: Sun, 24 Dec 2023 23:41:31 -0700
Message-ID: <CAOUHufYrzw7dDbchNkrrXnSqpeYasnfPh6qFoMNgAmbK9GXNaQ@mail.gmail.com>
Subject: Re: [PATCH 3/3] mm, lru_gen: try to prefetch next page when canning LRU
To: Kairui Song <kasong@tencent.com>, Matthew Wilcox <willy@infradead.org>
Cc: linux-mm@kvack.org, Andrew Morton <akpm@linux-foundation.org>, 
	linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Fri, Dec 22, 2023 at 3:24=E2=80=AFAM Kairui Song <ryncsn@gmail.com> wrot=
e:
>
> From: Kairui Song <kasong@tencent.com>
>
> Prefetch for inactive/active LRU have been long exiting, apply the same
> optimization for MGLRU.

I seriously doubt that prefetch helps in this case.

Willy, any thoughts on this? Thanks.

> Tested in a 4G memcg on a EPYC 7K62 with:
>
>   memcached -u nobody -m 16384 -s /tmp/memcached.socket \
>     -a 0766 -t 16 -B binary &
>
>   memtier_benchmark -S /tmp/memcached.socket \
>     -P memcache_binary -n allkeys \
>     --key-minimum=3D1 --key-maximum=3D16000000 -d 1024 \
>     --ratio=3D1:0 --key-pattern=3DP:P -c 2 -t 16 --pipeline 8 -x 6
>
> Average result of 18 test runs:
>
> Before:           44017.78 Ops/sec
> After patch 1-3:  44890.50 Ops/sec (+1.8%)
>
> Signed-off-by: Kairui Song <kasong@tencent.com>
> ---
>  mm/vmscan.c | 12 ++++++++----
>  1 file changed, 8 insertions(+), 4 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index af1266129c1b..1e9d69e18443 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -3764,10 +3764,12 @@ static bool inc_min_seq(struct lruvec *lruvec, in=
t type, bool can_swap)
>                         VM_WARN_ON_ONCE_FOLIO(folio_is_file_lru(folio) !=
=3D type, folio);
>                         VM_WARN_ON_ONCE_FOLIO(folio_zonenum(folio) !=3D z=
one, folio);
>
> -                       if (unlikely(list_is_first(&folio->lru, head)))
> +                       if (unlikely(list_is_first(&folio->lru, head))) {
>                                 prev =3D NULL;
> -                       else
> +                       } else {
>                                 prev =3D lru_to_folio(&folio->lru);
> +                               prefetchw(&prev->flags);
> +                       }
>
>                         new_gen =3D folio_inc_gen(lruvec, folio, false, &=
batch);
>                         lru_gen_try_inc_bulk(lrugen, folio, bulk_gen, new=
_gen, type, zone, &batch);
> @@ -4434,10 +4436,12 @@ static int scan_folios(struct lruvec *lruvec, str=
uct scan_control *sc,
>                         VM_WARN_ON_ONCE_FOLIO(folio_zonenum(folio) !=3D z=
one, folio);
>
>                         scanned +=3D delta;
> -                       if (unlikely(list_is_first(&folio->lru, head)))
> +                       if (unlikely(list_is_first(&folio->lru, head))) {
>                                 prev =3D NULL;
> -                       else
> +                       } else {
>                                 prev =3D lru_to_folio(&folio->lru);
> +                               prefetchw(&prev->flags);
> +                       }
>
>                         if (sort_folio(lruvec, folio, sc, tier, bulk_gen,=
 &batch))
>                                 sorted +=3D delta;
> --
> 2.43.0
>