net: speedup udp receive path

Since commit 95766ff ([UDP]: Add memory accounting.), 
each received packet needs one extra sock_lock()/sock_release() pair.

This added latency because of possible backlog handling. Then later,
ticket spinlocks added yet another latency source in case of DDOS.

This patch introduces lock_sock_bh() and unlock_sock_bh()
synchronization primitives, avoiding one atomic operation and backlog
processing.

skb_free_datagram_locked() uses them instead of full blown
lock_sock()/release_sock(). skb is orphaned inside locked section for
proper socket memory reclaim, and finally freed outside of it.

UDP receive path now take the socket spinlock only once.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/sock.h

@@ -1021,6 +1021,16 @@ extern void release_sock(struct sock *sk);
 				SINGLE_DEPTH_NESTING)
 #define bh_unlock_sock(__sk)	spin_unlock(&((__sk)->sk_lock.slock))
 
+static inline void lock_sock_bh(struct sock *sk)
+{
+	spin_lock_bh(&sk->sk_lock.slock);
+}
+
+static inline void unlock_sock_bh(struct sock *sk)
+{
+	spin_unlock_bh(&sk->sk_lock.slock);
+}
+
 extern struct sock		*sk_alloc(struct net *net, int family,
 					  gfp_t priority,
 					  struct proto *prot);

net/core/datagram.c

@@ -229,9 +229,13 @@ EXPORT_SYMBOL(skb_free_datagram);
 
 void skb_free_datagram_locked(struct sock *sk, struct sk_buff *skb)
 {
-	lock_sock(sk);
-	skb_free_datagram(sk, skb);
-	release_sock(sk);
+	lock_sock_bh(sk);
+	skb_orphan(skb);
+	sk_mem_reclaim_partial(sk);
+	unlock_sock_bh(sk);
+
+	/* skb is now orphaned, might be freed outside of locked section */
+	consume_skb(skb);
 }
 EXPORT_SYMBOL(skb_free_datagram_locked);
 

net/ipv4/udp.c

@@ -1062,10 +1062,10 @@ static unsigned int first_packet_length(struct sock *sk)
 	spin_unlock_bh(&rcvq->lock);
 
 	if (!skb_queue_empty(&list_kill)) {
-		lock_sock(sk);
+		lock_sock_bh(sk);
 		__skb_queue_purge(&list_kill);
 		sk_mem_reclaim_partial(sk);
-		release_sock(sk);
+		unlock_sock_bh(sk);
 	}
 	return res;
 }
@@ -1196,10 +1196,10 @@ int udp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 	return err;
 
 csum_copy_err:
-	lock_sock(sk);
+	lock_sock_bh(sk);
 	if (!skb_kill_datagram(sk, skb, flags))
 		UDP_INC_STATS_USER(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
-	release_sock(sk);
+	unlock_sock_bh(sk);
 
 	if (noblock)
 		return -EAGAIN;
@@ -1624,9 +1624,9 @@ int udp_rcv(struct sk_buff *skb)
 
 void udp_destroy_sock(struct sock *sk)
 {
-	lock_sock(sk);
+	lock_sock_bh(sk);
 	udp_flush_pending_frames(sk);
-	release_sock(sk);
+	unlock_sock_bh(sk);
 }
 
 /*

net/ipv6/udp.c

@@ -424,7 +424,7 @@ int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
 	return err;
 
 csum_copy_err:
-	lock_sock(sk);
+	lock_sock_bh(sk);
 	if (!skb_kill_datagram(sk, skb, flags)) {
 		if (is_udp4)
 			UDP_INC_STATS_USER(sock_net(sk),
@@ -433,7 +433,7 @@ int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
 			UDP6_INC_STATS_USER(sock_net(sk),
 					UDP_MIB_INERRORS, is_udplite);
 	}
-	release_sock(sk);
+	unlock_sock_bh(sk);
 
 	if (flags & MSG_DONTWAIT)
 		return -EAGAIN;