[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAM0EoMnzsSspPockj3kA93ZUD5ibroA8by93cwC11uCTgLEP=g@mail.gmail.com>
Date: Thu, 15 Feb 2024 12:33:00 -0500
From: Jamal Hadi Salim <jhs@...atatu.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: davem@...emloft.net, netdev@...r.kernel.org, edumazet@...gle.com,
pabeni@...hat.com, Marcelo Ricardo Leitner <marcelo.leitner@...il.com>,
Davide Caratti <dcaratti@...hat.com>, xiyou.wangcong@...il.com, jiri@...nulli.us,
shmulik.ladkani@...il.com
Subject: Re: [PATCH net v3 1/2] net/sched: act_mirred: use the backlog for
mirred ingress
On Thu, Feb 15, 2024 at 9:33 AM Jakub Kicinski <kuba@...nel.org> wrote:
>
> The test Davide added in commit ca22da2fbd69 ("act_mirred: use the backlog
> for nested calls to mirred ingress") hangs our testing VMs every 10 or so
> runs, with the familiar tcp_v4_rcv -> tcp_v4_rcv deadlock reported by
> lockdep.
>
> The problem as previously described by Davide (see Link) is that
> if we reverse flow of traffic with the redirect (egress -> ingress)
> we may reach the same socket which generated the packet. And we may
> still be holding its socket lock. The common solution to such deadlocks
> is to put the packet in the Rx backlog, rather than run the Rx path
> inline. Do that for all egress -> ingress reversals, not just once
> we started to nest mirred calls.
>
> In the past there was a concern that the backlog indirection will
> lead to loss of error reporting / less accurate stats. But the current
> workaround does not seem to address the issue.
>
> Fixes: 53592b364001 ("net/sched: act_mirred: Implement ingress actions")
> Cc: Marcelo Ricardo Leitner <marcelo.leitner@...il.com>
> Suggested-by: Davide Caratti <dcaratti@...hat.com>
> Link: https://lore.kernel.org/netdev/33dc43f587ec1388ba456b4915c75f02a8aae226.1663945716.git.dcaratti@redhat.com/
> Signed-off-by: Jakub Kicinski <kuba@...nel.org>
Acked-by: Jamal Hadi Salim <jhs@...atatu.com>
cheers,
jamal
> ---
> CC: jhs@...atatu.com
> CC: xiyou.wangcong@...il.com
> CC: jiri@...nulli.us
> CC: shmulik.ladkani@...il.com
> ---
> net/sched/act_mirred.c | 14 +++++---------
> .../testing/selftests/net/forwarding/tc_actions.sh | 3 ---
> 2 files changed, 5 insertions(+), 12 deletions(-)
>
> diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c
> index 0a1a9e40f237..291d47c9eb69 100644
> --- a/net/sched/act_mirred.c
> +++ b/net/sched/act_mirred.c
> @@ -232,18 +232,14 @@ static int tcf_mirred_init(struct net *net, struct nlattr *nla,
> return err;
> }
>
> -static bool is_mirred_nested(void)
> -{
> - return unlikely(__this_cpu_read(mirred_nest_level) > 1);
> -}
> -
> -static int tcf_mirred_forward(bool want_ingress, struct sk_buff *skb)
> +static int
> +tcf_mirred_forward(bool at_ingress, bool want_ingress, struct sk_buff *skb)
> {
> int err;
>
> if (!want_ingress)
> err = tcf_dev_queue_xmit(skb, dev_queue_xmit);
> - else if (is_mirred_nested())
> + else if (!at_ingress)
> err = netif_rx(skb);
> else
> err = netif_receive_skb(skb);
> @@ -319,9 +315,9 @@ static int tcf_mirred_to_dev(struct sk_buff *skb, struct tcf_mirred *m,
>
> skb_set_redirected(skb_to_send, skb_to_send->tc_at_ingress);
>
> - err = tcf_mirred_forward(want_ingress, skb_to_send);
> + err = tcf_mirred_forward(at_ingress, want_ingress, skb_to_send);
> } else {
> - err = tcf_mirred_forward(want_ingress, skb_to_send);
> + err = tcf_mirred_forward(at_ingress, want_ingress, skb_to_send);
> }
>
> if (err) {
> diff --git a/tools/testing/selftests/net/forwarding/tc_actions.sh b/tools/testing/selftests/net/forwarding/tc_actions.sh
> index b0f5e55d2d0b..589629636502 100755
> --- a/tools/testing/selftests/net/forwarding/tc_actions.sh
> +++ b/tools/testing/selftests/net/forwarding/tc_actions.sh
> @@ -235,9 +235,6 @@ mirred_egress_to_ingress_tcp_test()
> check_err $? "didn't mirred redirect ICMP"
> tc_check_packets "dev $h1 ingress" 102 10
> check_err $? "didn't drop mirred ICMP"
> - local overlimits=$(tc_rule_stats_get ${h1} 101 egress .overlimits)
> - test ${overlimits} = 10
> - check_err $? "wrong overlimits, expected 10 got ${overlimits}"
>
> tc filter del dev $h1 egress protocol ip pref 100 handle 100 flower
> tc filter del dev $h1 egress protocol ip pref 101 handle 101 flower
> --
> 2.43.0
>
Powered by blists - more mailing lists