net/sched: act_mirred: use the backlog for mirred ingress

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@gmail.com>
Suggested-by: Davide Caratti <dcaratti@redhat.com>
Link: https://lore.kernel.org/netdev/33dc43f587ec1388ba456b4915c75f02a8aae226.1663945716.git.dcaratti@redhat.com/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c
index 0a1a9e40f237012ecaa561bd563162bbc1802f9b..291d47c9eb69dedb30faca00c3c9145470ae12f6 100644 (file)
--- a/net/sched/act_mirred.c
+++ b/net/sched/act_mirred.c
@@ -232,18 +232,14 @@ release_idr:
        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 b0f5e55d2d0b2584aefacc135ffe6b2d2cab34fc..58962963650227bcc942354a052d8bf2bd95aa13 100755 (executable)
--- 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