Merge tag 'nfs-for-5.16-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client updates from Trond Myklebust:
 "Highlights include:

  Features:
   - NFSv4.1 can always retrieve and cache the ACCESS mode on OPEN
   - Optimisations for READDIR and the 'ls -l' style workload
   - Further replacements of dprintk() with tracepoints and other
     tracing improvements
   - Ensure we re-probe NFSv4 server capabilities when the user does a
     "mount -o remount"

  Bugfixes:
   - Fix an Oops in pnfs_mark_request_commit()
   - Fix up deadlocks in the commit code
   - Fix regressions in NFSv2/v3 attribute revalidation due to the
     change_attr_type optimisations
   - Fix some dentry verifier races
   - Fix some missing dentry verifier settings
   - Fix a performance regression in nfs_set_open_stateid_locked()
   - SUNRPC was sending multiple SYN calls when re-establishing a TCP
     connection.
   - Fix multiple NFSv4 issues due to missing sanity checking of server
     return values
   - Fix a potential Oops when FREE_STATEID races with an unmount

  Cleanups:
   - Clean up the labelled NFS code
   - Remove unused header <linux/pnfs_osd_xdr.h>"

* tag 'nfs-for-5.16-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (84 commits)
  NFSv4: Sanity check the parameters in nfs41_update_target_slotid()
  NFS: Remove the nfs4_label argument from decode_getattr_*() functions
  NFS: Remove the nfs4_label argument from nfs_setsecurity
  NFS: Remove the nfs4_label argument from nfs_fhget()
  NFS: Remove the nfs4_label argument from nfs_add_or_obtain()
  NFS: Remove the nfs4_label argument from nfs_instantiate()
  NFS: Remove the nfs4_label from the nfs_setattrres
  NFS: Remove the nfs4_label from the nfs4_getattr_res
  NFS: Remove the f_label from the nfs4_opendata and nfs_openres
  NFS: Remove the nfs4_label from the nfs4_lookupp_res struct
  NFS: Remove the label from the nfs4_lookup_res struct
  NFS: Remove the nfs4_label from the nfs4_link_res struct
  NFS: Remove the nfs4_label from the nfs4_create_res struct
  NFS: Remove the nfs4_label from the nfs_entry struct
  NFS: Create a new nfs_alloc_fattr_with_label() function
  NFS: Always initialise fattr->label in nfs_fattr_alloc()
  NFSv4.2: alloc_file_pseudo() takes an open flag, not an f_mode
  NFS: Don't allocate nfs_fattr on the stack in __nfs42_ssc_open()
  NFSv4: Remove unnecessary 'minor version' check
  NFSv4: Fix potential Oops in decode_op_map()
  ...

1 files changed, 41 insertions, 68 deletions

diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 04f1b78bcbca..ae48c9c84ee1 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -1134,6 +1134,7 @@ static void xs_run_error_worker(struct sock_xprt *transport, unsigned int nr)
 
 static void xs_sock_reset_connection_flags(struct rpc_xprt *xprt)
 {
+	xprt->connect_cookie++;
 	smp_mb__before_atomic();
 	clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
 	clear_bit(XPRT_CLOSING, &xprt->state);
@@ -1153,14 +1154,13 @@ static void xs_error_report(struct sock *sk)
 	struct sock_xprt *transport;
 	struct rpc_xprt *xprt;
 
-	read_lock_bh(&sk->sk_callback_lock);
 	if (!(xprt = xprt_from_sock(sk)))
-		goto out;
+		return;
 
 	transport = container_of(xprt, struct sock_xprt, xprt);
 	transport->xprt_err = -sk->sk_err;
 	if (transport->xprt_err == 0)
-		goto out;
+		return;
 	dprintk("RPC:       xs_error_report client %p, error=%d...\n",
 			xprt, -transport->xprt_err);
 	trace_rpc_socket_error(xprt, sk->sk_socket, transport->xprt_err);
@@ -1168,8 +1168,6 @@ static void xs_error_report(struct sock *sk)
 	/* barrier ensures xprt_err is set before XPRT_SOCK_WAKE_ERROR */
 	smp_mb__before_atomic();
 	xs_run_error_worker(transport, XPRT_SOCK_WAKE_ERROR);
- out:
-	read_unlock_bh(&sk->sk_callback_lock);
 }
 
 static void xs_reset_transport(struct sock_xprt *transport)
@@ -1188,7 +1186,7 @@ static void xs_reset_transport(struct sock_xprt *transport)
 	kernel_sock_shutdown(sock, SHUT_RDWR);
 
 	mutex_lock(&transport->recv_mutex);
-	write_lock_bh(&sk->sk_callback_lock);
+	lock_sock(sk);
 	transport->inet = NULL;
 	transport->sock = NULL;
 	transport->file = NULL;
@@ -1197,10 +1195,10 @@ static void xs_reset_transport(struct sock_xprt *transport)
 
 	xs_restore_old_callbacks(transport, sk);
 	xprt_clear_connected(xprt);
-	write_unlock_bh(&sk->sk_callback_lock);
 	xs_sock_reset_connection_flags(xprt);
 	/* Reset stream record info */
 	xs_stream_reset_connect(transport);
+	release_sock(sk);
 	mutex_unlock(&transport->recv_mutex);
 
 	trace_rpc_socket_close(xprt, sock);
@@ -1364,7 +1362,6 @@ static void xs_data_ready(struct sock *sk)
 {
 	struct rpc_xprt *xprt;
 
-	read_lock_bh(&sk->sk_callback_lock);
 	dprintk("RPC:       xs_data_ready...\n");
 	xprt = xprt_from_sock(sk);
 	if (xprt != NULL) {
@@ -1379,7 +1376,6 @@ static void xs_data_ready(struct sock *sk)
 		if (!test_and_set_bit(XPRT_SOCK_DATA_READY, &transport->sock_state))
 			queue_work(xprtiod_workqueue, &transport->recv_worker);
 	}
-	read_unlock_bh(&sk->sk_callback_lock);
 }
 
 /*
@@ -1408,9 +1404,8 @@ static void xs_tcp_state_change(struct sock *sk)
 	struct rpc_xprt *xprt;
 	struct sock_xprt *transport;
 
-	read_lock_bh(&sk->sk_callback_lock);
 	if (!(xprt = xprt_from_sock(sk)))
-		goto out;
+		return;
 	dprintk("RPC:       xs_tcp_state_change client %p...\n", xprt);
 	dprintk("RPC:       state %x conn %d dead %d zapped %d sk_shutdown %d\n",
 			sk->sk_state, xprt_connected(xprt),
@@ -1471,8 +1466,6 @@ static void xs_tcp_state_change(struct sock *sk)
 		/* Trigger the socket release */
 		xs_run_error_worker(transport, XPRT_SOCK_WAKE_DISCONNECT);
 	}
- out:
-	read_unlock_bh(&sk->sk_callback_lock);
 }
 
 static void xs_write_space(struct sock *sk)
@@ -1511,13 +1504,9 @@ out:
  */
 static void xs_udp_write_space(struct sock *sk)
 {
-	read_lock_bh(&sk->sk_callback_lock);
-
 	/* from net/core/sock.c:sock_def_write_space */
 	if (sock_writeable(sk))
 		xs_write_space(sk);
-
-	read_unlock_bh(&sk->sk_callback_lock);
 }
 
 /**
@@ -1532,13 +1521,9 @@ static void xs_udp_write_space(struct sock *sk)
  */
 static void xs_tcp_write_space(struct sock *sk)
 {
-	read_lock_bh(&sk->sk_callback_lock);
-
 	/* from net/core/stream.c:sk_stream_write_space */
 	if (sk_stream_is_writeable(sk))
 		xs_write_space(sk);
-
-	read_unlock_bh(&sk->sk_callback_lock);
 }
 
 static void xs_udp_do_set_buffer_size(struct rpc_xprt *xprt)
@@ -1833,7 +1818,7 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt,
 	if (!transport->inet) {
 		struct sock *sk = sock->sk;
 
-		write_lock_bh(&sk->sk_callback_lock);
+		lock_sock(sk);
 
 		xs_save_old_callbacks(transport, sk);
 
@@ -1849,7 +1834,7 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt,
 		transport->sock = sock;
 		transport->inet = sk;
 
-		write_unlock_bh(&sk->sk_callback_lock);
+		release_sock(sk);
 	}
 
 	xs_stream_start_connect(transport);
@@ -2031,7 +2016,7 @@ static void xs_udp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
 	if (!transport->inet) {
 		struct sock *sk = sock->sk;
 
-		write_lock_bh(&sk->sk_callback_lock);
+		lock_sock(sk);
 
 		xs_save_old_callbacks(transport, sk);
 
@@ -2048,7 +2033,7 @@ static void xs_udp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
 
 		xs_set_memalloc(xprt);
 
-		write_unlock_bh(&sk->sk_callback_lock);
+		release_sock(sk);
 	}
 	xs_udp_do_set_buffer_size(xprt);
 
@@ -2174,7 +2159,6 @@ static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt,
 static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
 {
 	struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
-	int ret = -ENOTCONN;
 
 	if (!transport->inet) {
 		struct sock *sk = sock->sk;
@@ -2194,7 +2178,7 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
 		xs_tcp_set_socket_timeouts(xprt, sock);
 		tcp_sock_set_nodelay(sk);
 
-		write_lock_bh(&sk->sk_callback_lock);
+		lock_sock(sk);
 
 		xs_save_old_callbacks(transport, sk);
 
@@ -2214,11 +2198,11 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
 		transport->sock = sock;
 		transport->inet = sk;
 
-		write_unlock_bh(&sk->sk_callback_lock);
+		release_sock(sk);
 	}
 
 	if (!xprt_bound(xprt))
-		goto out;
+		return -ENOTCONN;
 
 	xs_set_memalloc(xprt);
 
@@ -2226,22 +2210,7 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
 
 	/* Tell the socket layer to start connecting... */
 	set_bit(XPRT_SOCK_CONNECTING, &transport->sock_state);
-	ret = kernel_connect(sock, xs_addr(xprt), xprt->addrlen, O_NONBLOCK);
-	switch (ret) {
-	case 0:
-		xs_set_srcport(transport, sock);
-		fallthrough;
-	case -EINPROGRESS:
-		/* SYN_SENT! */
-		if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
-			xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
-		break;
-	case -EADDRNOTAVAIL:
-		/* Source port number is unavailable. Try a new one! */
-		transport->srcport = 0;
-	}
-out:
-	return ret;
+	return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, O_NONBLOCK);
 }
 
 /**
@@ -2256,14 +2225,14 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 		container_of(work, struct sock_xprt, connect_worker.work);
 	struct socket *sock = transport->sock;
 	struct rpc_xprt *xprt = &transport->xprt;
-	int status = -EIO;
+	int status;
 
 	if (!sock) {
 		sock = xs_create_sock(xprt, transport,
 				xs_addr(xprt)->sa_family, SOCK_STREAM,
 				IPPROTO_TCP, true);
 		if (IS_ERR(sock)) {
-			status = PTR_ERR(sock);
+			xprt_wake_pending_tasks(xprt, PTR_ERR(sock));
 			goto out;
 		}
 	}
@@ -2280,21 +2249,21 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 			xprt, -status, xprt_connected(xprt),
 			sock->sk->sk_state);
 	switch (status) {
-	default:
-		printk("%s: connect returned unhandled error %d\n",
-			__func__, status);
-		fallthrough;
-	case -EADDRNOTAVAIL:
-		/* We're probably in TIME_WAIT. Get rid of existing socket,
-		 * and retry
-		 */
-		xs_tcp_force_close(xprt);
-		break;
 	case 0:
+		xs_set_srcport(transport, sock);
+		fallthrough;
 	case -EINPROGRESS:
+		/* SYN_SENT! */
+		if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
+			xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
+		fallthrough;
 	case -EALREADY:
-		xprt_unlock_connect(xprt, transport);
-		return;
+		goto out_unlock;
+	case -EADDRNOTAVAIL:
+		/* Source port number is unavailable. Try a new one! */
+		transport->srcport = 0;
+		status = -EAGAIN;
+		break;
 	case -EINVAL:
 		/* Happens, for instance, if the user specified a link
 		 * local IPv6 address without a scope-id.
@@ -2306,18 +2275,22 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 	case -EHOSTUNREACH:
 	case -EADDRINUSE:
 	case -ENOBUFS:
-		/* xs_tcp_force_close() wakes tasks with a fixed error code.
-		 * We need to wake them first to ensure the correct error code.
-		 */
-		xprt_wake_pending_tasks(xprt, status);
-		xs_tcp_force_close(xprt);
-		goto out;
+		break;
+	default:
+		printk("%s: connect returned unhandled error %d\n",
+			__func__, status);
+		status = -EAGAIN;
 	}
-	status = -EAGAIN;
+
+	/* xs_tcp_force_close() wakes tasks with a fixed error code.
+	 * We need to wake them first to ensure the correct error code.
+	 */
+	xprt_wake_pending_tasks(xprt, status);
+	xs_tcp_force_close(xprt);
 out:
 	xprt_clear_connecting(xprt);
+out_unlock:
 	xprt_unlock_connect(xprt, transport);
-	xprt_wake_pending_tasks(xprt, status);
 }
 
 /**
@@ -2341,7 +2314,7 @@ static void xs_connect(struct rpc_xprt *xprt, struct rpc_task *task)
 
 	WARN_ON_ONCE(!xprt_lock_connect(xprt, task, transport));
 
-	if (transport->sock != NULL) {
+	if (transport->sock != NULL && !xprt_connecting(xprt)) {
 		dprintk("RPC:       xs_connect delayed xprt %p for %lu "
 				"seconds\n",
 				xprt, xprt->reestablish_timeout / HZ);