diff --git a/[refs] b/[refs]
index 2b9187ed9157..7e36e50813ef 100644
--- a/[refs]
+++ b/[refs]
@@ -1,2 +1,2 @@
 ---
-refs/heads/master: 501e89d3aef10f9a5b79137fbb0bcd5bf9cac2ac
+refs/heads/master: 0ad92ad03aa444b312bd318b0341011a8be09d13
diff --git a/trunk/net/ipv4/udp.c b/trunk/net/ipv4/udp.c
index 131d8a720086..ab0966df1e2a 100644
--- a/trunk/net/ipv4/udp.c
+++ b/trunk/net/ipv4/udp.c
@@ -1397,6 +1397,8 @@ int udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 	nf_reset(skb);
 
 	if (up->encap_type) {
+		int (*encap_rcv)(struct sock *sk, struct sk_buff *skb);
+
 		/*
 		 * This is an encapsulation socket so pass the skb to
 		 * the socket's udp_encap_rcv() hook. Otherwise, just
@@ -1409,11 +1411,11 @@ int udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 		 */
 
 		/* if we're overly short, let UDP handle it */
-		if (skb->len > sizeof(struct udphdr) &&
-		    up->encap_rcv != NULL) {
+		encap_rcv = ACCESS_ONCE(up->encap_rcv);
+		if (skb->len > sizeof(struct udphdr) && encap_rcv != NULL) {
 			int ret;
 
-			ret = (*up->encap_rcv)(sk, skb);
+			ret = encap_rcv(sk, skb);
 			if (ret <= 0) {
 				UDP_INC_STATS_BH(sock_net(sk),
 						 UDP_MIB_INDATAGRAMS,