Merge branch 'for-davem' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next

John W. Linville says:

====================
This probably is the last big pull request for wireless bits
for 3.9.  Of course, I'm sure there will be a few stragglers here
and there...surely a few bug fixes as well... :-) (In fact, I see
that Johannes has already queued-up a few more for me while I was
preparing this...)

Included are a number of pulls...

For mac80211-next, Johannes says:

"The biggest change I have is undoubtedly Marco's mesh powersave
implementation. Beyond that, I have a patch from Emmanuel to modify the
DTIM period API in mac80211, scan improvements and a removal of some
previous workaround code from Stanislaw, dynamic short slot time from
Thomas and 64-bit station byte counters from Vladimir. I also made a
number of changes myself, some related to WoWLAN, some auth/deauth
improvements and most of them BSS list cleanups."

"This time, I have relatively large number of fixes in various areas of
the code (a memory leak in regulatory, an RX race in mac80211, the new
radar checking caused a P2P device problem, some mesh issues with
stations, an older bug in tracing and for kernel-doc) as well as a
number of small new features. The biggest (in the diffstat) is my work
on hidden SSID tracking."

"Please pull to get
 * radar detection work from Simon
 * mesh improvements from Thomas
 * a connection monitoring/powersave fix from Wojciech
 * TDLS-related station management work from Jouni
 * VLAN crypto fixes from Michael Braun
 * CCK support in minstrel_ht from Felix
 * an SMPS (not SMSP, oops) related improvement in mac80211 (Emmanuel)
 * some WoWLAN work from Amitkumar Karwar: pattern match offset and a
   documentation fix
 * some WoWLAN work from myself (TCP connection wakeup feature API)
 * and a lot of VHT (and some HT) work (also from myself)

And a number of more random cleanups/fixes. I merged mac80211/master to
avoid a merge problem there."

And regarding iwlwifi-next, Johannes says:

"We continue work on our new driver, but I also have a WoWLAN and AP mode
improvement for the previous driver and a change to use threaded
interrupts to prepare us for working with non-PCIe devices."

Regarding wl12xx, Luca says:

"A few more patches intended for 3.9.  Mostly some clean-ups I've been
doing to make it easier to support device-tree.  Also including one bug
fix for wl12xx where the rates we advertise were wrong and an update in
the wlconf structure to support newer firmwares."

For the nfc-next bits, Samuel says:

"This is the second NFC pull request for 3.9.

We have:

- A few pn533 fixes on top of Waldemar refactorization of the driver, one of
  them fixes target mode.

- A new driver for Inside Secure microread chipset. It supports two
  physical layers: i2c and MEI. The MEI one depends on a patchset that's
  been sent to Greg Kroah-Hartman for inclusion into the 3.9 kernel [1]. The
  dependency is a KConfig one which means this code is not buildable as long
  as the MEI API is not usptream."

"This 3rd NFC pull request for 3.9 contains a fix for the microread MEI
physical layer support, as the MEI bus API changed.

From the MEI code, we now pass the MEI id back to the driver probe routine,
and we also pass a name and a MEI id table through the mei_bus_driver
structure. A few renames as well like e.g. mei_bus_driver to mei_driver or
mei_bus_client to mei_device in order to be closer to the driver model
practices."

For the ath6kl bits, Kalle says:

"There's not anything special here, most of the patches are just code
cleanup. The only functional changes are using the beacon interval from user
space and fixing a crash which happens when inserting and removing the
module in a loop."

Also, I pulled the wireless tree in order to resolve some pending
merge issues.  On top of that, there is a bunch of work on brcmfmac
that leads up to P2P support.  Also, mwifiex, rtlwifi, and a variety
of other drivers see some basic cleanups and minor enhancements.

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

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

1 files changed, 23 insertions, 43 deletions

diff --git a/net/mac80211/scan.c b/net/mac80211/scan.c
index 607684c47d55..43a45cf00e06 100644
--- a/net/mac80211/scan.c
+++ b/net/mac80211/scan.c
@@ -27,22 +27,15 @@
 
 #define IEEE80211_PROBE_DELAY (HZ / 33)
 #define IEEE80211_CHANNEL_TIME (HZ / 33)
-#define IEEE80211_PASSIVE_CHANNEL_TIME (HZ / 8)
-
-static void ieee80211_rx_bss_free(struct cfg80211_bss *cbss)
-{
-	struct ieee80211_bss *bss = (void *)cbss->priv;
-
-	kfree(bss_mesh_id(bss));
-	kfree(bss_mesh_cfg(bss));
-}
+#define IEEE80211_PASSIVE_CHANNEL_TIME (HZ / 9)
 
 void ieee80211_rx_bss_put(struct ieee80211_local *local,
 			  struct ieee80211_bss *bss)
 {
 	if (!bss)
 		return;
-	cfg80211_put_bss(container_of((void *)bss, struct cfg80211_bss, priv));
+	cfg80211_put_bss(local->hw.wiphy,
+			 container_of((void *)bss, struct cfg80211_bss, priv));
 }
 
 static bool is_uapsd_supported(struct ieee802_11_elems *elems)
@@ -85,10 +78,12 @@ ieee80211_bss_info_update(struct ieee80211_local *local,
 	if (!cbss)
 		return NULL;
 
-	cbss->free_priv = ieee80211_rx_bss_free;
 	bss = (void *)cbss->priv;
 
-	bss->device_ts = rx_status->device_timestamp;
+	if (beacon)
+		bss->device_ts_beacon = rx_status->device_timestamp;
+	else
+		bss->device_ts_presp = rx_status->device_timestamp;
 
 	if (elems->parse_error) {
 		if (beacon)
@@ -146,9 +141,6 @@ ieee80211_bss_info_update(struct ieee80211_local *local,
 			bss->valid_data |= IEEE80211_BSS_VALID_WMM;
 	}
 
-	if (!beacon)
-		bss->last_probe_resp = jiffies;
-
 	return bss;
 }
 
@@ -342,6 +334,9 @@ static int ieee80211_start_sw_scan(struct ieee80211_local *local)
 
 	ieee80211_offchannel_stop_vifs(local);
 
+	/* ensure nullfunc is transmitted before leaving operating channel */
+	drv_flush(local, false);
+
 	ieee80211_configure_filter(local);
 
 	/* We need to set power level at maximum rate for scanning. */
@@ -356,6 +351,9 @@ static int ieee80211_start_sw_scan(struct ieee80211_local *local)
 static bool ieee80211_can_scan(struct ieee80211_local *local,
 			       struct ieee80211_sub_if_data *sdata)
 {
+	if (local->radar_detect_enabled)
+		return false;
+
 	if (!list_empty(&local->roc_list))
 		return false;
 
@@ -390,6 +388,11 @@ static void ieee80211_scan_state_send_probe(struct ieee80211_local *local,
 	int i;
 	struct ieee80211_sub_if_data *sdata;
 	enum ieee80211_band band = local->hw.conf.channel->band;
+	u32 tx_flags;
+
+	tx_flags = IEEE80211_TX_INTFL_OFFCHAN_TX_OK;
+	if (local->scan_req->no_cck)
+		tx_flags |= IEEE80211_TX_CTL_NO_CCK_RATE;
 
 	sdata = rcu_dereference_protected(local->scan_sdata,
 					  lockdep_is_held(&local->mtx));
@@ -401,8 +404,7 @@ static void ieee80211_scan_state_send_probe(struct ieee80211_local *local,
 			local->scan_req->ssids[i].ssid_len,
 			local->scan_req->ie, local->scan_req->ie_len,
 			local->scan_req->rates[band], false,
-			local->scan_req->no_cck,
-			local->hw.conf.channel, true);
+			tx_flags, local->hw.conf.channel, true);
 
 	/*
 	 * After sending probe requests, wait for probe responses
@@ -546,8 +548,6 @@ static void ieee80211_scan_state_decision(struct ieee80211_local *local,
 	bool associated = false;
 	bool tx_empty = true;
 	bool bad_latency;
-	bool listen_int_exceeded;
-	unsigned long min_beacon_int = 0;
 	struct ieee80211_sub_if_data *sdata;
 	struct ieee80211_channel *next_chan;
 	enum mac80211_scan_state next_scan_state;
@@ -566,11 +566,6 @@ static void ieee80211_scan_state_decision(struct ieee80211_local *local,
 			if (sdata->u.mgd.associated) {
 				associated = true;
 
-				if (sdata->vif.bss_conf.beacon_int <
-				    min_beacon_int || min_beacon_int == 0)
-					min_beacon_int =
-						sdata->vif.bss_conf.beacon_int;
-
 				if (!qdisc_all_tx_empty(sdata->dev)) {
 					tx_empty = false;
 					break;
@@ -587,34 +582,19 @@ static void ieee80211_scan_state_decision(struct ieee80211_local *local,
 	 * see if we can scan another channel without interfering
 	 * with the current traffic situation.
 	 *
-	 * Since we don't know if the AP has pending frames for us
-	 * we can only check for our tx queues and use the current
-	 * pm_qos requirements for rx. Hence, if no tx traffic occurs
-	 * at all we will scan as many channels in a row as the pm_qos
-	 * latency allows us to. Additionally we also check for the
-	 * currently negotiated listen interval to prevent losing
-	 * frames unnecessarily.
-	 *
-	 * Otherwise switch back to the operating channel.
+	 * Keep good latency, do not stay off-channel more than 125 ms.
 	 */
 
 	bad_latency = time_after(jiffies +
-			ieee80211_scan_get_channel_time(next_chan),
-			local->leave_oper_channel_time +
-			usecs_to_jiffies(pm_qos_request(PM_QOS_NETWORK_LATENCY)));
-
-	listen_int_exceeded = time_after(jiffies +
-			ieee80211_scan_get_channel_time(next_chan),
-			local->leave_oper_channel_time +
-			usecs_to_jiffies(min_beacon_int * 1024) *
-			local->hw.conf.listen_interval);
+				 ieee80211_scan_get_channel_time(next_chan),
+				 local->leave_oper_channel_time + HZ / 8);
 
 	if (associated && !tx_empty) {
 		if (local->scan_req->flags & NL80211_SCAN_FLAG_LOW_PRIORITY)
 			next_scan_state = SCAN_ABORT;
 		else
 			next_scan_state = SCAN_SUSPEND;
-	} else if (associated && (bad_latency || listen_int_exceeded)) {
+	} else if (associated && bad_latency) {
 		next_scan_state = SCAN_SUSPEND;
 	} else {
 		next_scan_state = SCAN_SET_CHANNEL;