cherry-pick tcp reset bug fix

[pve-kernel.git] / 0001-tcp-reset-sk_rx_dst-in-tcp_disconnect.patch

From d747a7a51b00984127a88113cdbbc26f91e9d815 Mon Sep 17 00:00:00 2001
From: WANG Cong <xiyou.wangcong@gmail.com>
Date: Sat, 24 Jun 2017 23:50:30 -0700
Subject: [PATCH] tcp: reset sk_rx_dst in tcp_disconnect()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We have to reset the sk->sk_rx_dst when we disconnect a TCP
connection, because otherwise when we re-connect it this
dst reference is simply overridden in tcp_finish_connect().

This fixes a dst leak which leads to a loopback dev refcnt
leak. It is a long-standing bug, Kevin reported a very similar
(if not same) bug before. Thanks to Andrei for providing such
a reliable reproducer which greatly narrows down the problem.

Fixes: 41063e9dd119 ("ipv4: Early TCP socket demux.")
Reported-by: Andrei Vagin <avagin@gmail.com>
Reported-by: Kevin Xu <kaiwen.xu@hulu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
 net/ipv4/tcp.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index b5ea036ca781..40aca7803cf2 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2330,6 +2330,8 @@ int tcp_disconnect(struct sock *sk, int flags)
 	tcp_init_send_head(sk);
 	memset(&tp->rx_opt, 0, sizeof(tp->rx_opt));
 	__sk_dst_reset(sk);
+	dst_release(sk->sk_rx_dst);
+	sk->sk_rx_dst = NULL;
 	tcp_saved_syn_free(tp);
 
 	/* Clean up fastopen related fields */
-- 
2.11.0