netfilter: let reset rules clean out conntrack entries

iptables/nftables support responding to tcp packets with tcp resets.

The generated tcp reset packet passes through both output and postrouting
netfilter hooks, but conntrack will never see them because the generated
skb has its ->nfct pointer copied over from the packet that triggered the
reset rule.

If the reset rule is used for established connections, this
may result in the conntrack entry to be around for a very long
time (default timeout is 5 days).

One way to avoid this would be to not copy the nf_conn pointer
so that the rest packet passes through conntrack too.

Problem is that output rules might not have the same conntrack
zone setup as the prerouting ones, so its possible that the
reset skb won't find the correct entry.  Generating a template
entry for the skb seems error prone as well.

Add an explicit "closing" function that switches a confirmed
conntrack entry to closed state and wire this up for tcp.

If the entry isn't confirmed, no action is needed because
the conntrack entry will never be committed to the table.

Reported-by: Russel King <linux@armlinux.org.uk>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

3 files changed, 63 insertions, 0 deletions

diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index 5a6705a0e4ec..b2fdbbed2b4b 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -702,6 +702,22 @@ void nf_conntrack_destroy(struct nf_conntrack *nfct)
 }
 EXPORT_SYMBOL(nf_conntrack_destroy);
 
+void nf_ct_set_closing(struct nf_conntrack *nfct)
+{
+	const struct nf_ct_hook *ct_hook;
+
+	if (!nfct)
+		return;
+
+	rcu_read_lock();
+	ct_hook = rcu_dereference(nf_ct_hook);
+	if (ct_hook)
+		ct_hook->set_closing(nfct);
+
+	rcu_read_unlock();
+}
+EXPORT_SYMBOL_GPL(nf_ct_set_closing);
+
 bool nf_ct_get_tuple_skb(struct nf_conntrack_tuple *dst_tuple,
 			 const struct sk_buff *skb)
 {
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index c00858344f02..430bb52b6454 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -2747,11 +2747,23 @@ err_cachep:
 	return ret;
 }
 
+static void nf_conntrack_set_closing(struct nf_conntrack *nfct)
+{
+	struct nf_conn *ct = nf_ct_to_nf_conn(nfct);
+
+	switch (nf_ct_protonum(ct)) {
+	case IPPROTO_TCP:
+		nf_conntrack_tcp_set_closing(ct);
+		break;
+	}
+}
+
 static const struct nf_ct_hook nf_conntrack_hook = {
 	.update		= nf_conntrack_update,
 	.destroy	= nf_ct_destroy,
 	.get_tuple_skb  = nf_conntrack_get_tuple_skb,
 	.attach		= nf_conntrack_attach,
+	.set_closing	= nf_conntrack_set_closing,
 };
 
 void nf_conntrack_init_end(void)
diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index 16ee5ebe1ce1..4018acb1d674 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -911,6 +911,41 @@ static bool tcp_can_early_drop(const struct nf_conn *ct)
 	return false;
 }
 
+void nf_conntrack_tcp_set_closing(struct nf_conn *ct)
+{
+	enum tcp_conntrack old_state;
+	const unsigned int *timeouts;
+	u32 timeout;
+
+	if (!nf_ct_is_confirmed(ct))
+		return;
+
+	spin_lock_bh(&ct->lock);
+	old_state = ct->proto.tcp.state;
+	ct->proto.tcp.state = TCP_CONNTRACK_CLOSE;
+
+	if (old_state == TCP_CONNTRACK_CLOSE ||
+	    test_bit(IPS_FIXED_TIMEOUT_BIT, &ct->status)) {
+		spin_unlock_bh(&ct->lock);
+		return;
+	}
+
+	timeouts = nf_ct_timeout_lookup(ct);
+	if (!timeouts) {
+		const struct nf_tcp_net *tn;
+
+		tn = nf_tcp_pernet(nf_ct_net(ct));
+		timeouts = tn->timeouts;
+	}
+
+	timeout = timeouts[TCP_CONNTRACK_CLOSE];
+	WRITE_ONCE(ct->timeout, timeout + nfct_time_stamp);
+
+	spin_unlock_bh(&ct->lock);
+
+	nf_conntrack_event_cache(IPCT_PROTOINFO, ct);
+}
+
 static void nf_ct_tcp_state_reset(struct ip_ct_tcp_state *state)
 {
 	state->td_end		= 0;