tcp: clear xmit timers in tcp_v4_syn_recv_sock()

Simon Kirby reported divides by zero errors in __tcp_select_window()

This happens when inet_csk_route_child_sock() returns a NULL pointer :

We free new socket while we eventually armed keepalive timer in
tcp_create_openreq_child()

Fix this by a call to tcp_clear_xmit_timers()

[ This is a followup to commit 918eb39 (net: add missing
bh_unlock_sock() calls) ]

Reported-by: Simon Kirby <sim@hostway.ca>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Tested-by: Simon Kirby <sim@hostway.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/ipv4/tcp_ipv4.c

@@ -1510,6 +1510,7 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
 	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENDROPS);
 	return NULL;
 put_and_exit:
+	tcp_clear_xmit_timers(newsk);
 	bh_unlock_sock(newsk);
 	sock_put(newsk);
 	goto exit;