ipv4: tcp: dont cache output dst for syncookies

Don't cache output dst for syncookies, as this adds pressure on IP route
cache and rcu subsystem for no gain.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Hans Schillstrom <hans.schillstrom@ericsson.com>
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 21e22a0..b52934f 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -825,7 +825,8 @@
 static int tcp_v4_send_synack(struct sock *sk, struct dst_entry *dst,
 			      struct request_sock *req,
 			      struct request_values *rvp,
-			      u16 queue_mapping)
+			      u16 queue_mapping,
+			      bool nocache)
 {
 	const struct inet_request_sock *ireq = inet_rsk(req);
 	struct flowi4 fl4;
@@ -833,7 +834,7 @@
 	struct sk_buff * skb;
 
 	/* First, grab a route. */
-	if (!dst && (dst = inet_csk_route_req(sk, &fl4, req)) == NULL)
+	if (!dst && (dst = inet_csk_route_req(sk, &fl4, req, nocache)) == NULL)
 		return -1;
 
 	skb = tcp_make_synack(sk, dst, req, rvp);
@@ -855,7 +856,7 @@
 			      struct request_values *rvp)
 {
 	TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_RETRANSSEGS);
-	return tcp_v4_send_synack(sk, NULL, req, rvp, 0);
+	return tcp_v4_send_synack(sk, NULL, req, rvp, 0, false);
 }
 
 /*
@@ -1388,7 +1389,7 @@
 		 */
 		if (tmp_opt.saw_tstamp &&
 		    tcp_death_row.sysctl_tw_recycle &&
-		    (dst = inet_csk_route_req(sk, &fl4, req)) != NULL &&
+		    (dst = inet_csk_route_req(sk, &fl4, req, want_cookie)) != NULL &&
 		    fl4.daddr == saddr &&
 		    (peer = rt_get_peer((struct rtable *)dst, fl4.daddr)) != NULL) {
 			inet_peer_refcheck(peer);
@@ -1424,7 +1425,8 @@
 
 	if (tcp_v4_send_synack(sk, dst, req,
 			       (struct request_values *)&tmp_ext,
-			       skb_get_queue_mapping(skb)) ||
+			       skb_get_queue_mapping(skb),
+			       want_cookie) ||
 	    want_cookie)
 		goto drop_and_free;