From 52f671db18823089a02f07efc04efdb2272ddc17 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 15 Feb 2024 06:33:45 -0800 Subject: 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 Suggested-by: Davide Caratti Link: https://lore.kernel.org/netdev/33dc43f587ec1388ba456b4915c75f02a8aae226.1663945716.git.dcaratti@redhat.com/ Signed-off-by: Jakub Kicinski Acked-by: Jamal Hadi Salim Signed-off-by: David S. Miller --- net/sched/act_mirred.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) (limited to 'net/sched/act_mirred.c') 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 @@ 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) { -- cgit v1.2.3-70-g09d2 From 166c2c8a6a4dc2e4ceba9e10cfe81c3e469e3210 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Thu, 15 Feb 2024 06:33:46 -0800 Subject: net/sched: act_mirred: don't override retval if we already lost the skb If we're redirecting the skb, and haven't called tcf_mirred_forward(), yet, we need to tell the core to drop the skb by setting the retcode to SHOT. If we have called tcf_mirred_forward(), however, the skb is out of our hands and returning SHOT will lead to UaF. Move the retval override to the error path which actually need it. Reviewed-by: Michal Swiatkowski Fixes: e5cf1baf92cb ("act_mirred: use TC_ACT_REINSERT when possible") Signed-off-by: Jakub Kicinski Acked-by: Jamal Hadi Salim Signed-off-by: David S. Miller --- net/sched/act_mirred.c | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) (limited to 'net/sched/act_mirred.c') diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c index 291d47c9eb69..6faa7d00da09 100644 --- a/net/sched/act_mirred.c +++ b/net/sched/act_mirred.c @@ -266,8 +266,7 @@ static int tcf_mirred_to_dev(struct sk_buff *skb, struct tcf_mirred *m, if (unlikely(!(dev->flags & IFF_UP)) || !netif_carrier_ok(dev)) { net_notice_ratelimited("tc mirred to Houston: device %s is down\n", dev->name); - err = -ENODEV; - goto out; + goto err_cant_do; } /* we could easily avoid the clone only if called by ingress and clsact; @@ -279,10 +278,8 @@ static int tcf_mirred_to_dev(struct sk_buff *skb, struct tcf_mirred *m, tcf_mirred_can_reinsert(retval); if (!dont_clone) { skb_to_send = skb_clone(skb, GFP_ATOMIC); - if (!skb_to_send) { - err = -ENOMEM; - goto out; - } + if (!skb_to_send) + goto err_cant_do; } want_ingress = tcf_mirred_act_wants_ingress(m_eaction); @@ -319,15 +316,16 @@ static int tcf_mirred_to_dev(struct sk_buff *skb, struct tcf_mirred *m, } else { err = tcf_mirred_forward(at_ingress, want_ingress, skb_to_send); } - - if (err) { -out: + if (err) tcf_action_inc_overlimit_qstats(&m->common); - if (is_redirect) - retval = TC_ACT_SHOT; - } return retval; + +err_cant_do: + if (is_redirect) + retval = TC_ACT_SHOT; + tcf_action_inc_overlimit_qstats(&m->common); + return retval; } static int tcf_blockcast_redir(struct sk_buff *skb, struct tcf_mirred *m, -- cgit v1.2.3-70-g09d2