Merge tag 'nfsd-4.8' of git://linux-nfs.org/~bfields/linux

Pull nfsd updates from Bruce Fields:
 "Highlights:

   - Trond made a change to the server's tcp logic that allows a fast
     client to better take advantage of high bandwidth networks, but may
     increase the risk that a single client could starve other clients;
     a new sunrpc.svc_rpc_per_connection_limit parameter should help
     mitigate this in the (hopefully unlikely) event this becomes a
     problem in practice.

   - Tom Haynes added a minimal flex-layout pnfs server, which is of no
     use in production for now--don't build it unless you're doing
     client testing or further server development"

* tag 'nfsd-4.8' of git://linux-nfs.org/~bfields/linux: (32 commits)
  nfsd: remove some dead code in nfsd_create_locked()
  nfsd: drop unnecessary MAY_EXEC check from create
  nfsd: clean up bad-type check in nfsd_create_locked
  nfsd: remove unnecessary positive-dentry check
  nfsd: reorganize nfsd_create
  nfsd: check d_can_lookup in fh_verify of directories
  nfsd: remove redundant zero-length check from create
  nfsd: Make creates return EEXIST instead of EACCES
  SUNRPC: Detect immediate closure of accepted sockets
  SUNRPC: accept() may return sockets that are still in SYN_RECV
  nfsd: allow nfsd to advertise multiple layout types
  nfsd: Close race between nfsd4_release_lockowner and nfsd4_lock
  nfsd/blocklayout: Make sure calculate signature/designator length aligned
  xfs: abstract block export operations from nfsd layouts
  SUNRPC: Remove unused callback xpo_adjust_wspace()
  SUNRPC: Change TCP socket space reservation
  SUNRPC: Add a server side per-connection limit
  SUNRPC: Micro optimisation for svc_data_ready
  SUNRPC: Call the default socket callbacks instead of open coding
  SUNRPC: lock the socket while detaching it
  ...

4 files changed, 86 insertions, 122 deletions

diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c
index e085f5ae1548..1d281816f2bf 100644
--- a/net/sunrpc/auth_gss/svcauth_gss.c
+++ b/net/sunrpc/auth_gss/svcauth_gss.c
@@ -1230,8 +1230,9 @@ static int svcauth_gss_proxy_init(struct svc_rqst *rqstp,
 	if (status)
 		goto out;
 
-	dprintk("RPC:       svcauth_gss: gss major status = %d\n",
-			ud.major_status);
+	dprintk("RPC:       svcauth_gss: gss major status = %d "
+			"minor status = %d\n",
+			ud.major_status, ud.minor_status);
 
 	switch (ud.major_status) {
 	case GSS_S_CONTINUE_NEEDED:
diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c
index 553bf95f7003..4d8e11f94a35 100644
--- a/net/sunrpc/cache.c
+++ b/net/sunrpc/cache.c
@@ -362,7 +362,7 @@ void sunrpc_destroy_cache_detail(struct cache_detail *cd)
 	cache_purge(cd);
 	spin_lock(&cache_list_lock);
 	write_lock(&cd->hash_lock);
-	if (cd->entries || atomic_read(&cd->inuse)) {
+	if (cd->entries) {
 		write_unlock(&cd->hash_lock);
 		spin_unlock(&cache_list_lock);
 		goto out;
diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index 4f01f63102ee..c3f652395a80 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -21,6 +21,10 @@
 
 #define RPCDBG_FACILITY	RPCDBG_SVCXPRT
 
+static unsigned int svc_rpc_per_connection_limit __read_mostly;
+module_param(svc_rpc_per_connection_limit, uint, 0644);
+
+
 static struct svc_deferred_req *svc_deferred_dequeue(struct svc_xprt *xprt);
 static int svc_deferred_recv(struct svc_rqst *rqstp);
 static struct cache_deferred_req *svc_defer(struct cache_req *req);
@@ -329,12 +333,45 @@ char *svc_print_addr(struct svc_rqst *rqstp, char *buf, size_t len)
 }
 EXPORT_SYMBOL_GPL(svc_print_addr);
 
+static bool svc_xprt_slots_in_range(struct svc_xprt *xprt)
+{
+	unsigned int limit = svc_rpc_per_connection_limit;
+	int nrqsts = atomic_read(&xprt->xpt_nr_rqsts);
+
+	return limit == 0 || (nrqsts >= 0 && nrqsts < limit);
+}
+
+static bool svc_xprt_reserve_slot(struct svc_rqst *rqstp, struct svc_xprt *xprt)
+{
+	if (!test_bit(RQ_DATA, &rqstp->rq_flags)) {
+		if (!svc_xprt_slots_in_range(xprt))
+			return false;
+		atomic_inc(&xprt->xpt_nr_rqsts);
+		set_bit(RQ_DATA, &rqstp->rq_flags);
+	}
+	return true;
+}
+
+static void svc_xprt_release_slot(struct svc_rqst *rqstp)
+{
+	struct svc_xprt	*xprt = rqstp->rq_xprt;
+	if (test_and_clear_bit(RQ_DATA, &rqstp->rq_flags)) {
+		atomic_dec(&xprt->xpt_nr_rqsts);
+		svc_xprt_enqueue(xprt);
+	}
+}
+
 static bool svc_xprt_has_something_to_do(struct svc_xprt *xprt)
 {
 	if (xprt->xpt_flags & ((1<<XPT_CONN)|(1<<XPT_CLOSE)))
 		return true;
-	if (xprt->xpt_flags & ((1<<XPT_DATA)|(1<<XPT_DEFERRED)))
-		return xprt->xpt_ops->xpo_has_wspace(xprt);
+	if (xprt->xpt_flags & ((1<<XPT_DATA)|(1<<XPT_DEFERRED))) {
+		if (xprt->xpt_ops->xpo_has_wspace(xprt) &&
+		    svc_xprt_slots_in_range(xprt))
+			return true;
+		trace_svc_xprt_no_write_space(xprt);
+		return false;
+	}
 	return false;
 }
 
@@ -480,8 +517,6 @@ void svc_reserve(struct svc_rqst *rqstp, int space)
 		atomic_sub((rqstp->rq_reserved - space), &xprt->xpt_reserved);
 		rqstp->rq_reserved = space;
 
-		if (xprt->xpt_ops->xpo_adjust_wspace)
-			xprt->xpt_ops->xpo_adjust_wspace(xprt);
 		svc_xprt_enqueue(xprt);
 	}
 }
@@ -512,8 +547,8 @@ static void svc_xprt_release(struct svc_rqst *rqstp)
 
 	rqstp->rq_res.head[0].iov_len = 0;
 	svc_reserve(rqstp, 0);
+	svc_xprt_release_slot(rqstp);
 	rqstp->rq_xprt = NULL;
-
 	svc_xprt_put(xprt);
 }
 
@@ -781,7 +816,7 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt)
 			svc_add_new_temp_xprt(serv, newxpt);
 		else
 			module_put(xprt->xpt_class->xcl_owner);
-	} else {
+	} else if (svc_xprt_reserve_slot(rqstp, xprt)) {
 		/* XPT_DATA|XPT_DEFERRED case: */
 		dprintk("svc: server %p, pool %u, transport %p, inuse=%d\n",
 			rqstp, rqstp->rq_pool->sp_id, xprt,
@@ -871,6 +906,7 @@ EXPORT_SYMBOL_GPL(svc_recv);
  */
 void svc_drop(struct svc_rqst *rqstp)
 {
+	trace_svc_drop(rqstp);
 	dprintk("svc: xprt %p dropped request\n", rqstp->rq_xprt);
 	svc_xprt_release(rqstp);
 }
@@ -1148,6 +1184,7 @@ static void svc_revisit(struct cache_deferred_req *dreq, int too_many)
 		spin_unlock(&xprt->xpt_lock);
 		dprintk("revisit canceled\n");
 		svc_xprt_put(xprt);
+		trace_svc_drop_deferred(dr);
 		kfree(dr);
 		return;
 	}
@@ -1205,6 +1242,7 @@ static struct cache_deferred_req *svc_defer(struct cache_req *req)
 	set_bit(RQ_DROPME, &rqstp->rq_flags);
 
 	dr->handle.revisit = svc_revisit;
+	trace_svc_defer(rqstp);
 	return &dr->handle;
 }
 
@@ -1245,6 +1283,7 @@ static struct svc_deferred_req *svc_deferred_dequeue(struct svc_xprt *xprt)
 				struct svc_deferred_req,
 				handle.recent);
 		list_del_init(&dr->handle.recent);
+		trace_svc_revisit_deferred(dr);
 	} else
 		clear_bit(XPT_DEFERRED, &xprt->xpt_flags);
 	spin_unlock(&xprt->xpt_lock);
diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c
index dadfec66dbd8..57625f64efd5 100644
--- a/net/sunrpc/svcsock.c
+++ b/net/sunrpc/svcsock.c
@@ -60,7 +60,6 @@
 
 static struct svc_sock *svc_setup_socket(struct svc_serv *, struct socket *,
 					 int flags);
-static void		svc_udp_data_ready(struct sock *);
 static int		svc_udp_recvfrom(struct svc_rqst *);
 static int		svc_udp_sendto(struct svc_rqst *);
 static void		svc_sock_detach(struct svc_xprt *);
@@ -398,48 +397,21 @@ static int svc_sock_secure_port(struct svc_rqst *rqstp)
 	return svc_port_is_privileged(svc_addr(rqstp));
 }
 
-static bool sunrpc_waitqueue_active(wait_queue_head_t *wq)
-{
-	if (!wq)
-		return false;
-	/*
-	 * There should normally be a memory * barrier here--see
-	 * wq_has_sleeper().
-	 *
-	 * It appears that isn't currently necessary, though, basically
-	 * because callers all appear to have sufficient memory barriers
-	 * between the time the relevant change is made and the
-	 * time they call these callbacks.
-	 *
-	 * The nfsd code itself doesn't actually explicitly wait on
-	 * these waitqueues, but it may wait on them for example in
-	 * sendpage() or sendmsg() calls.  (And those may be the only
-	 * places, since it it uses nonblocking reads.)
-	 *
-	 * Maybe we should add the memory barriers anyway, but these are
-	 * hot paths so we'd need to be convinced there's no sigificant
-	 * penalty.
-	 */
-	return waitqueue_active(wq);
-}
-
 /*
  * INET callback when data has been received on the socket.
  */
-static void svc_udp_data_ready(struct sock *sk)
+static void svc_data_ready(struct sock *sk)
 {
 	struct svc_sock	*svsk = (struct svc_sock *)sk->sk_user_data;
-	wait_queue_head_t *wq = sk_sleep(sk);
 
 	if (svsk) {
 		dprintk("svc: socket %p(inet %p), busy=%d\n",
 			svsk, sk,
 			test_bit(XPT_BUSY, &svsk->sk_xprt.xpt_flags));
-		set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags);
-		svc_xprt_enqueue(&svsk->sk_xprt);
+		svsk->sk_odata(sk);
+		if (!test_and_set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags))
+			svc_xprt_enqueue(&svsk->sk_xprt);
 	}
-	if (sunrpc_waitqueue_active(wq))
-		wake_up_interruptible(wq);
 }
 
 /*
@@ -448,56 +420,22 @@ static void svc_udp_data_ready(struct sock *sk)
 static void svc_write_space(struct sock *sk)
 {
 	struct svc_sock	*svsk = (struct svc_sock *)(sk->sk_user_data);
-	wait_queue_head_t *wq = sk_sleep(sk);
 
 	if (svsk) {
 		dprintk("svc: socket %p(inet %p), write_space busy=%d\n",
 			svsk, sk, test_bit(XPT_BUSY, &svsk->sk_xprt.xpt_flags));
+		svsk->sk_owspace(sk);
 		svc_xprt_enqueue(&svsk->sk_xprt);
 	}
-
-	if (sunrpc_waitqueue_active(wq)) {
-		dprintk("RPC svc_write_space: someone sleeping on %p\n",
-		       svsk);
-		wake_up_interruptible(wq);
-	}
 }
 
 static int svc_tcp_has_wspace(struct svc_xprt *xprt)
 {
-	struct svc_sock *svsk =	container_of(xprt, struct svc_sock, sk_xprt);
-	struct svc_serv *serv = svsk->sk_xprt.xpt_server;
-	int required;
+	struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt);
 
 	if (test_bit(XPT_LISTENER, &xprt->xpt_flags))
 		return 1;
-	required = atomic_read(&xprt->xpt_reserved) + serv->sv_max_mesg;
-	if (sk_stream_wspace(svsk->sk_sk) >= required ||
-	    (sk_stream_min_wspace(svsk->sk_sk) == 0 &&
-	     atomic_read(&xprt->xpt_reserved) == 0))
-		return 1;
-	set_bit(SOCK_NOSPACE, &svsk->sk_sock->flags);
-	return 0;
-}
-
-static void svc_tcp_write_space(struct sock *sk)
-{
-	struct svc_sock *svsk = (struct svc_sock *)(sk->sk_user_data);
-	struct socket *sock = sk->sk_socket;
-
-	if (!sk_stream_is_writeable(sk) || !sock)
-		return;
-	if (!svsk || svc_tcp_has_wspace(&svsk->sk_xprt))
-		clear_bit(SOCK_NOSPACE, &sock->flags);
-	svc_write_space(sk);
-}
-
-static void svc_tcp_adjust_wspace(struct svc_xprt *xprt)
-{
-	struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt);
-
-	if (svc_tcp_has_wspace(xprt))
-		clear_bit(SOCK_NOSPACE, &svsk->sk_sock->flags);
+	return !test_bit(SOCK_NOSPACE, &svsk->sk_sock->flags);
 }
 
 /*
@@ -746,7 +684,7 @@ static void svc_udp_init(struct svc_sock *svsk, struct svc_serv *serv)
 	svc_xprt_init(sock_net(svsk->sk_sock->sk), &svc_udp_class,
 		      &svsk->sk_xprt, serv);
 	clear_bit(XPT_CACHE_AUTH, &svsk->sk_xprt.xpt_flags);
-	svsk->sk_sk->sk_data_ready = svc_udp_data_ready;
+	svsk->sk_sk->sk_data_ready = svc_data_ready;
 	svsk->sk_sk->sk_write_space = svc_write_space;
 
 	/* initialise setting must have enough space to
@@ -786,11 +724,12 @@ static void svc_udp_init(struct svc_sock *svsk, struct svc_serv *serv)
 static void svc_tcp_listen_data_ready(struct sock *sk)
 {
 	struct svc_sock	*svsk = (struct svc_sock *)sk->sk_user_data;
-	wait_queue_head_t *wq;
 
 	dprintk("svc: socket %p TCP (listen) state change %d\n",
 		sk, sk->sk_state);
 
+	if (svsk)
+		svsk->sk_odata(sk);
 	/*
 	 * This callback may called twice when a new connection
 	 * is established as a child socket inherits everything
@@ -808,10 +747,6 @@ static void svc_tcp_listen_data_ready(struct sock *sk)
 		} else
 			printk("svc: socket %p: no user data\n", sk);
 	}
-
-	wq = sk_sleep(sk);
-	if (sunrpc_waitqueue_active(wq))
-		wake_up_interruptible_all(wq);
 }
 
 /*
@@ -820,7 +755,6 @@ static void svc_tcp_listen_data_ready(struct sock *sk)
 static void svc_tcp_state_change(struct sock *sk)
 {
 	struct svc_sock	*svsk = (struct svc_sock *)sk->sk_user_data;
-	wait_queue_head_t *wq = sk_sleep(sk);
 
 	dprintk("svc: socket %p TCP (connected) state change %d (svsk %p)\n",
 		sk, sk->sk_state, sk->sk_user_data);
@@ -828,26 +762,12 @@ static void svc_tcp_state_change(struct sock *sk)
 	if (!svsk)
 		printk("svc: socket %p: no user data\n", sk);
 	else {
-		set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags);
-		svc_xprt_enqueue(&svsk->sk_xprt);
-	}
-	if (sunrpc_waitqueue_active(wq))
-		wake_up_interruptible_all(wq);
-}
-
-static void svc_tcp_data_ready(struct sock *sk)
-{
-	struct svc_sock *svsk = (struct svc_sock *)sk->sk_user_data;
-	wait_queue_head_t *wq = sk_sleep(sk);
-
-	dprintk("svc: socket %p TCP data ready (svsk %p)\n",
-		sk, sk->sk_user_data);
-	if (svsk) {
-		set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags);
-		svc_xprt_enqueue(&svsk->sk_xprt);
+		svsk->sk_ostate(sk);
+		if (sk->sk_state != TCP_ESTABLISHED) {
+			set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags);
+			svc_xprt_enqueue(&svsk->sk_xprt);
+		}
 	}
-	if (sunrpc_waitqueue_active(wq))
-		wake_up_interruptible(wq);
 }
 
 /*
@@ -901,6 +821,11 @@ static struct svc_xprt *svc_tcp_accept(struct svc_xprt *xprt)
 	dprintk("%s: connect from %s\n", serv->sv_name,
 		__svc_print_addr(sin, buf, sizeof(buf)));
 
+	/* Reset the inherited callbacks before calling svc_setup_socket */
+	newsock->sk->sk_state_change = svsk->sk_ostate;
+	newsock->sk->sk_data_ready = svsk->sk_odata;
+	newsock->sk->sk_write_space = svsk->sk_owspace;
+
 	/* make sure that a write doesn't block forever when
 	 * low on memory
 	 */
@@ -1317,7 +1242,6 @@ static struct svc_xprt_ops svc_tcp_ops = {
 	.xpo_has_wspace = svc_tcp_has_wspace,
 	.xpo_accept = svc_tcp_accept,
 	.xpo_secure_port = svc_sock_secure_port,
-	.xpo_adjust_wspace = svc_tcp_adjust_wspace,
 };
 
 static struct svc_xprt_class svc_tcp_class = {
@@ -1357,8 +1281,8 @@ static void svc_tcp_init(struct svc_sock *svsk, struct svc_serv *serv)
 	} else {
 		dprintk("setting up TCP socket for reading\n");
 		sk->sk_state_change = svc_tcp_state_change;
-		sk->sk_data_ready = svc_tcp_data_ready;
-		sk->sk_write_space = svc_tcp_write_space;
+		sk->sk_data_ready = svc_data_ready;
+		sk->sk_write_space = svc_write_space;
 
 		svsk->sk_reclen = 0;
 		svsk->sk_tcplen = 0;
@@ -1368,8 +1292,13 @@ static void svc_tcp_init(struct svc_sock *svsk, struct svc_serv *serv)
 		tcp_sk(sk)->nonagle |= TCP_NAGLE_OFF;
 
 		set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags);
-		if (sk->sk_state != TCP_ESTABLISHED)
+		switch (sk->sk_state) {
+		case TCP_SYN_RECV:
+		case TCP_ESTABLISHED:
+			break;
+		default:
 			set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags);
+		}
 	}
 }
 
@@ -1428,17 +1357,14 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv,
 	/* Initialize the socket */
 	if (sock->type == SOCK_DGRAM)
 		svc_udp_init(svsk, serv);
-	else {
-		/* initialise setting must have enough space to
-		 * receive and respond to one request.
-		 */
-		svc_sock_setbufsize(svsk->sk_sock, 4 * serv->sv_max_mesg,
-					4 * serv->sv_max_mesg);
+	else
 		svc_tcp_init(svsk, serv);
-	}
 
-	dprintk("svc: svc_setup_socket created %p (inet %p)\n",
-				svsk, svsk->sk_sk);
+	dprintk("svc: svc_setup_socket created %p (inet %p), "
+			"listen %d close %d\n",
+			svsk, svsk->sk_sk,
+			test_bit(XPT_LISTENER, &svsk->sk_xprt.xpt_flags),
+			test_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags));
 
 	return svsk;
 }
@@ -1606,18 +1532,16 @@ static void svc_sock_detach(struct svc_xprt *xprt)
 {
 	struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt);
 	struct sock *sk = svsk->sk_sk;
-	wait_queue_head_t *wq;
 
 	dprintk("svc: svc_sock_detach(%p)\n", svsk);
 
 	/* put back the old socket callbacks */
+	lock_sock(sk);
 	sk->sk_state_change = svsk->sk_ostate;
 	sk->sk_data_ready = svsk->sk_odata;
 	sk->sk_write_space = svsk->sk_owspace;
-
-	wq = sk_sleep(sk);
-	if (sunrpc_waitqueue_active(wq))
-		wake_up_interruptible(wq);
+	sk->sk_user_data = NULL;
+	release_sock(sk);
 }
 
 /*