Merge tag 'master-2014-11-20' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next

John W. Linville says:

====================
pull request: wireless-next 2014-11-21

Please pull this batch of updates intended for the 3.19 stream...

For the mac80211 bits, Johannes says:

"It has been a while since my last pull request, so we accumulated
another relatively large set of changes:
 * TDLS off-channel support set from Arik/Liad, with some support
   patches I did
 * custom regulatory fixes from Arik
 * minstrel VHT fix (and a small optimisation) from Felix
 * add back radiotap vendor namespace support (myself)
 * random MAC address scanning for cfg80211/mac80211/hwsim (myself)
 * CSA improvements (Luca)
 * WoWLAN Net Detect (wake on network found) support (Luca)
 * and lots of other smaller changes from many people"

For the Bluetooth bits, Johan says:

"Here's another set of patches for 3.19. Most of it is again fixes and
cleanups to ieee802154 related code from Alexander Aring. We've also got
better handling of hardware error events along with a proper API for HCI
drivers to notify the HCI core of such situations. There's also a minor
fix for mgmt events as well as a sparse warning fix. The code for
sending HCI commands synchronously also gets a fix where we might loose
the completion event in the case of very fast HW (particularly easily
reproducible with an emulated HCI device)."

And...

"Here's another bluetooth-next pull request for 3.19. We've got:

 - Various fixes, cleanups and improvements to ieee802154/mac802154
 - Support for a Broadcom BCM20702A1 variant
 - Lots of lockdep fixes
 - Fixed handling of LE CoC errors that should trigger SMP"

For the Atheros bits, Kalle says:

"One ath6kl patch and rest for ath10k, but nothing really major which
stands out. Most notable:

o fix resume (Bartosz)

o firmware restart is now faster and more reliable (Michal)

o it's now possible to test hardware restart functionality without
  crashing the firmware using hw-restart parameter with
  simulate_fw_crash debugfs file (Michal)"

On top of that...both ath9k and mwifiex get their usual level of
updates.  Of note is the ath9k spectral scan work from Oleksij Rempel.

I also pulled from the wireless tree in order to avoid some merge issues.

Please let me know if there are problems!
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 25 insertions, 4 deletions

diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c
index 31f106e61ca2..b0efb7202957 100644
--- a/net/bluetooth/l2cap_sock.c
+++ b/net/bluetooth/l2cap_sock.c
@@ -285,6 +285,12 @@ static int l2cap_sock_listen(struct socket *sock, int backlog)
 	sk->sk_max_ack_backlog = backlog;
 	sk->sk_ack_backlog = 0;
 
+	/* Listening channels need to use nested locking in order not to
+	 * cause lockdep warnings when the created child channels end up
+	 * being locked in the same thread as the parent channel.
+	 */
+	atomic_set(&chan->nesting, L2CAP_NESTING_PARENT);
+
 	chan->state = BT_LISTEN;
 	sk->sk_state = BT_LISTEN;
 
@@ -301,7 +307,7 @@ static int l2cap_sock_accept(struct socket *sock, struct socket *newsock,
 	long timeo;
 	int err = 0;
 
-	lock_sock_nested(sk, SINGLE_DEPTH_NESTING);
+	lock_sock_nested(sk, L2CAP_NESTING_PARENT);
 
 	timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK);
 
@@ -333,7 +339,7 @@ static int l2cap_sock_accept(struct socket *sock, struct socket *newsock,
 
 		release_sock(sk);
 		timeo = schedule_timeout(timeo);
-		lock_sock_nested(sk, SINGLE_DEPTH_NESTING);
+		lock_sock_nested(sk, L2CAP_NESTING_PARENT);
 	}
 	__set_current_state(TASK_RUNNING);
 	remove_wait_queue(sk_sleep(sk), &wait);
@@ -1096,6 +1102,8 @@ static int l2cap_sock_shutdown(struct socket *sock, int how)
 	chan = l2cap_pi(sk)->chan;
 	conn = chan->conn;
 
+	BT_DBG("chan %p state %s", chan, state_to_string(chan->state));
+
 	if (conn)
 		mutex_lock(&conn->chan_lock);
 
@@ -1153,12 +1161,16 @@ static void l2cap_sock_cleanup_listen(struct sock *parent)
 {
 	struct sock *sk;
 
-	BT_DBG("parent %p", parent);
+	BT_DBG("parent %p state %s", parent,
+	       state_to_string(parent->sk_state));
 
 	/* Close not yet accepted channels */
 	while ((sk = bt_accept_dequeue(parent, NULL))) {
 		struct l2cap_chan *chan = l2cap_pi(sk)->chan;
 
+		BT_DBG("child chan %p state %s", chan,
+		       state_to_string(chan->state));
+
 		l2cap_chan_lock(chan);
 		__clear_chan_timer(chan);
 		l2cap_chan_close(chan, ECONNRESET);
@@ -1246,7 +1258,16 @@ static void l2cap_sock_teardown_cb(struct l2cap_chan *chan, int err)
 	struct sock *sk = chan->data;
 	struct sock *parent;
 
-	lock_sock(sk);
+	BT_DBG("chan %p state %s", chan, state_to_string(chan->state));
+
+	/* This callback can be called both for server (BT_LISTEN)
+	 * sockets as well as "normal" ones. To avoid lockdep warnings
+	 * with child socket locking (through l2cap_sock_cleanup_listen)
+	 * we need separation into separate nesting levels. The simplest
+	 * way to accomplish this is to inherit the nesting level used
+	 * for the channel.
+	 */
+	lock_sock_nested(sk, atomic_read(&chan->nesting));
 
 	parent = bt_sk(sk)->parent;