af_unix: split 'u->readlock' into two: 'iolock' and 'bindlock'

[ Upstream commit 6e1ce3c3 ]

Right now we use the 'readlock' both for protecting some of the af_unix
IO path and for making the bind be single-threaded.

The two are independent, but using the same lock makes for a nasty
deadlock due to ordering with regards to filesystem locking.  The bind
locking would want to nest outside the VSF pathname locking, but the IO
locking wants to nest inside some of those same locks.

We tried to fix this earlier with commit c845acb3 ("af_unix: Fix
splice-bind deadlock") which moved the readlock inside the vfs locks,
but that caused problems with overlayfs that will then call back into
filesystem routines that take the lock in the wrong order anyway.

Splitting the locks means that we can go back to having the bind lock be
the outermost lock, and we don't have any deadlocks with lock ordering.
Acked-by: Rainer Weikusat <rweikusat@cyberadapt.com>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

include/net/af_unix.h

@@ -52,7 +52,7 @@ struct unix_sock {
 	struct sock		sk;
 	struct unix_address     *addr;
 	struct path		path;
-	struct mutex		readlock;
+	struct mutex		iolock, bindlock;
 	struct sock		*peer;
 	struct list_head	link;
 	atomic_long_t		inflight;

net/unix/af_unix.c

@@ -661,11 +661,11 @@ static int unix_set_peek_off(struct sock *sk, int val)
 {
 	struct unix_sock *u = unix_sk(sk);
 
-	if (mutex_lock_interruptible(&u->readlock))
+	if (mutex_lock_interruptible(&u->iolock))
 		return -EINTR;
 
 	sk->sk_peek_off = val;
-	mutex_unlock(&u->readlock);
+	mutex_unlock(&u->iolock);
 
 	return 0;
 }
@@ -778,7 +778,8 @@ static struct sock *unix_create1(struct net *net, struct socket *sock, int kern)
 	spin_lock_init(&u->lock);
 	atomic_long_set(&u->inflight, 0);
 	INIT_LIST_HEAD(&u->link);
-	mutex_init(&u->readlock); /* single task reading lock */
+	mutex_init(&u->iolock); /* single task reading lock */
+	mutex_init(&u->bindlock); /* single task binding lock */
 	init_waitqueue_head(&u->peer_wait);
 	init_waitqueue_func_entry(&u->peer_wake, unix_dgram_peer_wake_relay);
 	unix_insert_socket(unix_sockets_unbound(sk), sk);
@@ -847,7 +848,7 @@ static int unix_autobind(struct socket *sock)
 	int err;
 	unsigned int retries = 0;
 
-	err = mutex_lock_interruptible(&u->readlock);
+	err = mutex_lock_interruptible(&u->bindlock);
 	if (err)
 		return err;
 
@@ -894,7 +895,7 @@ static int unix_autobind(struct socket *sock)
 	spin_unlock(&unix_table_lock);
 	err = 0;
 
-out:	mutex_unlock(&u->readlock);
+out:	mutex_unlock(&u->bindlock);
 	return err;
 }
 
@@ -1008,7 +1009,7 @@ static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
 		goto out;
 	addr_len = err;
 
-	err = mutex_lock_interruptible(&u->readlock);
+	err = mutex_lock_interruptible(&u->bindlock);
 	if (err)
 		goto out;
 
@@ -1062,7 +1063,7 @@ static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
 out_unlock:
 	spin_unlock(&unix_table_lock);
 out_up:
-	mutex_unlock(&u->readlock);
+	mutex_unlock(&u->bindlock);
 out:
 	return err;
 }
@@ -1954,17 +1955,17 @@ static ssize_t unix_stream_sendpage(struct socket *socket, struct page *page,
 	if (false) {
 alloc_skb:
 		unix_state_unlock(other);
-		mutex_unlock(&unix_sk(other)->readlock);
+		mutex_unlock(&unix_sk(other)->iolock);
 		newskb = sock_alloc_send_pskb(sk, 0, 0, flags & MSG_DONTWAIT,
 					      &err, 0);
 		if (!newskb)
 			goto err;
 	}
 
-	/* we must acquire readlock as we modify already present
+	/* we must acquire iolock as we modify already present
 	 * skbs in the sk_receive_queue and mess with skb->len
 	 */
-	err = mutex_lock_interruptible(&unix_sk(other)->readlock);
+	err = mutex_lock_interruptible(&unix_sk(other)->iolock);
 	if (err) {
 		err = flags & MSG_DONTWAIT ? -EAGAIN : -ERESTARTSYS;
 		goto err;
@@ -2031,7 +2032,7 @@ static ssize_t unix_stream_sendpage(struct socket *socket, struct page *page,
 	}
 
 	unix_state_unlock(other);
-	mutex_unlock(&unix_sk(other)->readlock);
+	mutex_unlock(&unix_sk(other)->iolock);
 
 	other->sk_data_ready(other);
 	scm_destroy(&scm);
@@ -2040,7 +2041,7 @@ static ssize_t unix_stream_sendpage(struct socket *socket, struct page *page,
 err_state_unlock:
 	unix_state_unlock(other);
 err_unlock:
-	mutex_unlock(&unix_sk(other)->readlock);
+	mutex_unlock(&unix_sk(other)->iolock);
 err:
 	kfree_skb(newskb);
 	if (send_sigpipe && !(flags & MSG_NOSIGNAL))
@@ -2108,7 +2109,7 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
 	timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
 
 	do {
-		mutex_lock(&u->readlock);
+		mutex_lock(&u->iolock);
 
 		skip = sk_peek_offset(sk, flags);
 		skb = __skb_try_recv_datagram(sk, flags, &peeked, &skip, &err,
@@ -2116,14 +2117,14 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
 		if (skb)
 			break;
 
-		mutex_unlock(&u->readlock);
+		mutex_unlock(&u->iolock);
 
 		if (err != -EAGAIN)
 			break;
 	} while (timeo &&
 		 !__skb_wait_for_more_packets(sk, &err, &timeo, last));
 
-	if (!skb) { /* implies readlock unlocked */
+	if (!skb) { /* implies iolock unlocked */
 		unix_state_lock(sk);
 		/* Signal EOF on disconnected non-blocking SEQPACKET socket. */
 		if (sk->sk_type == SOCK_SEQPACKET && err == -EAGAIN &&
@@ -2188,7 +2189,7 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
 
 out_free:
 	skb_free_datagram(sk, skb);
-	mutex_unlock(&u->readlock);
+	mutex_unlock(&u->iolock);
 out:
 	return err;
 }
@@ -2283,7 +2284,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state)
 	/* Lock the socket to prevent queue disordering
 	 * while sleeps in memcpy_tomsg
 	 */
-	mutex_lock(&u->readlock);
+	mutex_lock(&u->iolock);
 
 	if (flags & MSG_PEEK)
 		skip = sk_peek_offset(sk, flags);
@@ -2325,7 +2326,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state)
 				break;
 			}
 
-			mutex_unlock(&u->readlock);
+			mutex_unlock(&u->iolock);
 
 			timeo = unix_stream_data_wait(sk, timeo, last,
 						      last_len);
@@ -2336,7 +2337,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state)
 				goto out;
 			}
 
-			mutex_lock(&u->readlock);
+			mutex_lock(&u->iolock);
 			goto redo;
 unlock:
 			unix_state_unlock(sk);
@@ -2439,7 +2440,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state)
 		}
 	} while (size);
 
-	mutex_unlock(&u->readlock);
+	mutex_unlock(&u->iolock);
 	if (state->msg)
 		scm_recv(sock, state->msg, &scm, flags);
 	else
@@ -2480,9 +2481,9 @@ static ssize_t skb_unix_socket_splice(struct sock *sk,
 	int ret;
 	struct unix_sock *u = unix_sk(sk);
 
-	mutex_unlock(&u->readlock);
+	mutex_unlock(&u->iolock);
 	ret = splice_to_pipe(pipe, spd);
-	mutex_lock(&u->readlock);
+	mutex_lock(&u->iolock);
 
 	return ret;
 }