Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.36 as permitted sender) client-ip=23.128.96.36;
MIME-Version: 1.0
References: <20231201083926.1817394-1-judyhsiao@chromium.org>
In-Reply-To: <20231201083926.1817394-1-judyhsiao@chromium.org>
From:   Eric Dumazet <edumazet@google.com>
Date:   Fri, 1 Dec 2023 10:10:03 +0100
Message-ID: <CANn89iJMbMZdnJRP0CUVfEi20whhShBfO+DAmdaerhiXfiTx5A@mail.gmail.com>
Subject: Re: [PATCH v1] neighbour: Don't let neigh_forced_gc() disable
 preemption for long
To:     Judy Hsiao <judyhsiao@chromium.org>
Cc:     David Ahern <dsahern@kernel.org>, Simon Horman <horms@kernel.org>,
        Douglas Anderson <dianders@chromium.org>,
        Brian Haley <haleyb.dev@gmail.com>,
        "David S. Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>,
        Joel Granados <joel.granados@gmail.com>,
        Julian Anastasov <ja@ssi.bg>,
        Leon Romanovsky <leon@kernel.org>,
        Luis Chamberlain <mcgrof@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, linux-kernel@vger.kernel.org,
        netdev@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Fri, Dec 1, 2023 at 9:39=E2=80=AFAM Judy Hsiao <judyhsiao@chromium.org> =
wrote:
>
> We are seeing cases where neigh_cleanup_and_release() is called by
> neigh_forced_gc() many times in a row with preemption turned off.
> When running on a low powered CPU at a low CPU frequency, this has
> been measured to keep preemption off for ~10 ms. That's not great on a
> system with HZ=3D1000 which expects tasks to be able to schedule in
> with ~1ms latency.

This will not work in general, because this code runs with BH blocked.

jiffies will stay untouched for many more ms on systems with only one CPU.

I would rather not rely on jiffies here but ktime_get_ns() [1]

Also if we break the loop based on time, we might be unable to purge
the last elements in gc_list.
We might need to use a second list to make sure to cycle over all
elements eventually.


[1]
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index df81c1f0a57047e176b7c7e4809d2dae59ba6be5..e2340e6b07735db8cf6e75d23ef=
09bb4b0db53b4
100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -253,9 +253,11 @@ static int neigh_forced_gc(struct neigh_table *tbl)
 {
        int max_clean =3D atomic_read(&tbl->gc_entries) -
                        READ_ONCE(tbl->gc_thresh2);
+       u64 tmax =3D ktime_get_ns() + NSEC_PER_MSEC;
        unsigned long tref =3D jiffies - 5 * HZ;
        struct neighbour *n, *tmp;
        int shrunk =3D 0;
+       int loop =3D 0;

        NEIGH_CACHE_STAT_INC(tbl, forced_gc_runs);

@@ -279,10 +281,16 @@ static int neigh_forced_gc(struct neigh_table *tbl)
                        if (shrunk >=3D max_clean)
                                break;
                }
+               if (++loop =3D=3D 16) {
+                       if (ktime_get_ns() > tmax)
+                               goto unlock;
+                       loop =3D 0;
+               }
        }

        WRITE_ONCE(tbl->last_flush, jiffies);

+unlock:
        write_unlock_bh(&tbl->lock);

        return shrunk;