Merge branch 'master' into upstream

11 files changed, 90 insertions, 67 deletions

diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
index 06abb6634f5b..53086fb75089 100644
--- a/net/bridge/br_netlink.c
+++ b/net/bridge/br_netlink.c
@@ -85,7 +85,7 @@ void br_ifinfo_notify(int event, struct net_bridge_port *port)
 		goto err_out;
 
 	err = br_fill_ifinfo(skb, port, current->pid, 0, event, 0);
-	if (err)
+	if (err < 0)
 		goto err_kfree;
 
 	NETLINK_CB(skb).dst_group = RTNLGRP_LINK;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 738dad9f7d49..104af5d5bcbc 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3541,7 +3541,8 @@ void tcp_cwnd_application_limited(struct sock *sk)
 	if (inet_csk(sk)->icsk_ca_state == TCP_CA_Open &&
 	    sk->sk_socket && !test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
 		/* Limited by application or receiver window. */
-		u32 win_used = max(tp->snd_cwnd_used, 2U);
+		u32 init_win = tcp_init_cwnd(tp, __sk_dst_get(sk));
+		u32 win_used = max(tp->snd_cwnd_used, init_win);
 		if (win_used < tp->snd_cwnd) {
 			tp->snd_ssthresh = tcp_current_ssthresh(sk);
 			tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1;
diff --git a/net/lapb/lapb_iface.c b/net/lapb/lapb_iface.c
index d504eed416f6..7e6bc41eeb21 100644
--- a/net/lapb/lapb_iface.c
+++ b/net/lapb/lapb_iface.c
@@ -238,11 +238,13 @@ int lapb_setparms(struct net_device *dev, struct lapb_parms_struct *parms)
 		goto out_put;
 
 	if (lapb->state == LAPB_STATE_0) {
-		if (((parms->mode & LAPB_EXTENDED) &&
-		     (parms->window < 1 || parms->window > 127)) ||
-		    (parms->window < 1 || parms->window > 7))
-			goto out_put;
-
+		if (parms->mode & LAPB_EXTENDED) {
+			if (parms->window < 1 || parms->window > 127)
+				goto out_put;
+		} else {
+			if (parms->window < 1 || parms->window > 7)
+				goto out_put;
+		}
 		lapb->mode    = parms->mode;
 		lapb->window  = parms->window;
 	}
diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c
index d6cfe84d521b..2652ead96c64 100644
--- a/net/llc/af_llc.c
+++ b/net/llc/af_llc.c
@@ -784,24 +784,20 @@ static int llc_ui_recvmsg(struct kiocb *iocb, struct socket *sock,
 		copied += used;
 		len -= used;
 
-		if (used + offset < skb->len)
-			continue;
-
 		if (!(flags & MSG_PEEK)) {
 			sk_eat_skb(sk, skb, 0);
 			*seq = 0;
 		}
+
+		/* For non stream protcols we get one packet per recvmsg call */
+		if (sk->sk_type != SOCK_STREAM)
+			goto copy_uaddr;
+
+		/* Partial read */
+		if (used + offset < skb->len)
+			continue;
 	} while (len > 0);
 
-	/* 
-	 * According to UNIX98, msg_name/msg_namelen are ignored
-	 * on connected socket. -ANK
-	 * But... af_llc still doesn't have separate sets of methods for
-	 * SOCK_DGRAM and SOCK_STREAM :-( So we have to do this test, will
-	 * eventually fix this tho :-) -acme
-	 */
-	if (sk->sk_type == SOCK_DGRAM)
-		goto copy_uaddr;
 out:
 	release_sock(sk);
 	return copied;
diff --git a/net/llc/llc_sap.c b/net/llc/llc_sap.c
index 20c4eb5c1ac6..42eb0c3a9780 100644
--- a/net/llc/llc_sap.c
+++ b/net/llc/llc_sap.c
@@ -51,10 +51,10 @@ void llc_save_primitive(struct sock *sk, struct sk_buff* skb, u8 prim)
 {
 	struct sockaddr_llc *addr;
 
-	if (skb->sk->sk_type == SOCK_STREAM) /* See UNIX98 */
-		return;
        /* save primitive for use by the user. */
 	addr		  = llc_ui_skb_cb(skb);
+
+	memset(addr, 0, sizeof(*addr));
 	addr->sllc_family = sk->sk_family;
 	addr->sllc_arphrd = skb->dev->type;
 	addr->sllc_test   = prim == LLC_TEST_PRIM;
diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c
index c7844bacbbcb..a19eff12cf78 100644
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -430,7 +430,7 @@ qdisc_create(struct net_device *dev, u32 handle, struct rtattr **tca, int *errp)
 	}
 #endif
 
-	err = -EINVAL;
+	err = -ENOENT;
 	if (ops == NULL)
 		goto err_out;
 
diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c
index 7026b0866b7b..00cb388ece03 100644
--- a/net/sunrpc/cache.c
+++ b/net/sunrpc/cache.c
@@ -71,7 +71,12 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
 	new = detail->alloc();
 	if (!new)
 		return NULL;
+	/* must fully initialise 'new', else
+	 * we might get lose if we need to
+	 * cache_put it soon.
+	 */
 	cache_init(new);
+	detail->init(new, key);
 
 	write_lock(&detail->hash_lock);
 
@@ -85,7 +90,6 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
 			return tmp;
 		}
 	}
-	detail->init(new, key);
 	new->next = *head;
 	*head = new;
 	detail->entries++;
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 4ba271f892c8..d6409e757219 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -921,26 +921,43 @@ call_transmit(struct rpc_task *task)
 	task->tk_status = xprt_prepare_transmit(task);
 	if (task->tk_status != 0)
 		return;
+	task->tk_action = call_transmit_status;
 	/* Encode here so that rpcsec_gss can use correct sequence number. */
 	if (rpc_task_need_encode(task)) {
-		task->tk_rqstp->rq_bytes_sent = 0;
+		BUG_ON(task->tk_rqstp->rq_bytes_sent != 0);
 		call_encode(task);
 		/* Did the encode result in an error condition? */
 		if (task->tk_status != 0)
-			goto out_nosend;
+			return;
 	}
-	task->tk_action = call_transmit_status;
 	xprt_transmit(task);
 	if (task->tk_status < 0)
 		return;
-	if (!task->tk_msg.rpc_proc->p_decode) {
-		task->tk_action = rpc_exit_task;
-		rpc_wake_up_task(task);
-	}
-	return;
-out_nosend:
-	/* release socket write lock before attempting to handle error */
-	xprt_abort_transmit(task);
+	/*
+	 * On success, ensure that we call xprt_end_transmit() before sleeping
+	 * in order to allow access to the socket to other RPC requests.
+	 */
+	call_transmit_status(task);
+	if (task->tk_msg.rpc_proc->p_decode != NULL)
+		return;
+	task->tk_action = rpc_exit_task;
+	rpc_wake_up_task(task);
+}
+
+/*
+ * 5a.	Handle cleanup after a transmission
+ */
+static void
+call_transmit_status(struct rpc_task *task)
+{
+	task->tk_action = call_status;
+	/*
+	 * Special case: if we've been waiting on the socket's write_space()
+	 * callback, then don't call xprt_end_transmit().
+	 */
+	if (task->tk_status == -EAGAIN)
+		return;
+	xprt_end_transmit(task);
 	rpc_task_force_reencode(task);
 }
 
@@ -992,18 +1009,7 @@ call_status(struct rpc_task *task)
 }
 
 /*
- * 6a.	Handle transmission errors.
- */
-static void
-call_transmit_status(struct rpc_task *task)
-{
-	if (task->tk_status != -EAGAIN)
-		rpc_task_force_reencode(task);
-	call_status(task);
-}
-
-/*
- * 6b.	Handle RPC timeout
+ * 6a.	Handle RPC timeout
  * 	We do not release the request slot, so we keep using the
  *	same XID for all retransmits.
  */
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index dc6cb93c8830..a3bd2db2e024 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -667,10 +667,11 @@ rpc_mkdir(char *path, struct rpc_clnt *rpc_client)
 			RPCAUTH_info, RPCAUTH_EOF);
 	if (error)
 		goto err_depopulate;
+	dget(dentry);
 out:
 	mutex_unlock(&dir->i_mutex);
 	rpc_release_path(&nd);
-	return dget(dentry);
+	return dentry;
 err_depopulate:
 	rpc_depopulate(dentry);
 	__rpc_rmdir(dir, dentry);
@@ -731,10 +732,11 @@ rpc_mkpipe(char *path, void *private, struct rpc_pipe_ops *ops, int flags)
 	rpci->flags = flags;
 	rpci->ops = ops;
 	inode_dir_notify(dir, DN_CREATE);
+	dget(dentry);
 out:
 	mutex_unlock(&dir->i_mutex);
 	rpc_release_path(&nd);
-	return dget(dentry);
+	return dentry;
 err_dput:
 	dput(dentry);
 	dentry = ERR_PTR(-ENOMEM);
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 313b68d892c6..e8c2bc4977f3 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -707,12 +707,9 @@ out_unlock:
 	return err;
 }
 
-void
-xprt_abort_transmit(struct rpc_task *task)
+void xprt_end_transmit(struct rpc_task *task)
 {
-	struct rpc_xprt	*xprt = task->tk_xprt;
-
-	xprt_release_write(xprt, task);
+	xprt_release_write(task->tk_xprt, task);
 }
 
 /**
@@ -761,8 +758,6 @@ void xprt_transmit(struct rpc_task *task)
 			task->tk_status = -ENOTCONN;
 		else if (!req->rq_received)
 			rpc_sleep_on(&xprt->pending, task, NULL, xprt_timer);
-
-		xprt->ops->release_xprt(xprt, task);
 		spin_unlock_bh(&xprt->transport_lock);
 		return;
 	}
@@ -772,18 +767,8 @@ void xprt_transmit(struct rpc_task *task)
 	 *	 schedq, and being picked up by a parallel run of rpciod().
 	 */
 	task->tk_status = status;
-
-	switch (status) {
-	case -ECONNREFUSED:
+	if (status == -ECONNREFUSED)
 		rpc_sleep_on(&xprt->sending, task, NULL, NULL);
-	case -EAGAIN:
-	case -ENOTCONN:
-		return;
-	default:
-		break;
-	}
-	xprt_release_write(xprt, task);
-	return;
 }
 
 static inline void do_xprt_reserve(struct rpc_task *task)
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index ee678ed13b6f..441bd53f5eca 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -414,6 +414,33 @@ static int xs_tcp_send_request(struct rpc_task *task)
 }
 
 /**
+ * xs_tcp_release_xprt - clean up after a tcp transmission
+ * @xprt: transport
+ * @task: rpc task
+ *
+ * This cleans up if an error causes us to abort the transmission of a request.
+ * In this case, the socket may need to be reset in order to avoid confusing
+ * the server.
+ */
+static void xs_tcp_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
+{
+	struct rpc_rqst *req;
+
+	if (task != xprt->snd_task)
+		return;
+	if (task == NULL)
+		goto out_release;
+	req = task->tk_rqstp;
+	if (req->rq_bytes_sent == 0)
+		goto out_release;
+	if (req->rq_bytes_sent == req->rq_snd_buf.len)
+		goto out_release;
+	set_bit(XPRT_CLOSE_WAIT, &task->tk_xprt->state);
+out_release:
+	xprt_release_xprt(xprt, task);
+}
+
+/**
  * xs_close - close a socket
  * @xprt: transport
  *
@@ -1250,7 +1277,7 @@ static struct rpc_xprt_ops xs_udp_ops = {
 
 static struct rpc_xprt_ops xs_tcp_ops = {
 	.reserve_xprt		= xprt_reserve_xprt,
-	.release_xprt		= xprt_release_xprt,
+	.release_xprt		= xs_tcp_release_xprt,
 	.set_port		= xs_set_port,
 	.connect		= xs_connect,
 	.buf_alloc		= rpc_malloc,