tcp: Fix a data-race around sysctl_tcp_comp_sack_delay_ns.

[ Upstream commit 4866b2b ]

While reading sysctl_tcp_comp_sack_delay_ns, it can be changed
concurrently.  Thus, we need to add READ_ONCE() to its reader.

Fixes: 6d82aa2 ("tcp: add tcp_comp_sack_delay_ns sysctl")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/ipv4/tcp_input.c

@@ -5546,7 +5546,8 @@ static void __tcp_ack_snd_check(struct sock *sk, int ofo_possible)
 	if (tp->srtt_us && tp->srtt_us < rtt)
 		rtt = tp->srtt_us;
 
-	delay = min_t(unsigned long, sock_net(sk)->ipv4.sysctl_tcp_comp_sack_delay_ns,
+	delay = min_t(unsigned long,
+		      READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_comp_sack_delay_ns),
 		      rtt * (NSEC_PER_USEC >> 3)/20);
 	sock_hold(sk);
 	hrtimer_start_range_ns(&tp->compressed_ack_timer, ns_to_ktime(delay),