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

John W. Linville says:

====================
Please pull this batch of fixes intended for the 3.14 stream...

For the mac80211 bits, Johannes says:

"This time I have a fix to get out of an 'infinite error state' in case
regulatory domain updates failed and two fixes for VHT associations: one
to not disconnect immediately when the AP uses more bandwidth than the
new regdomain would allow after a change due to association country
information getting used, and one for an issue in the code where
mac80211 doesn't correctly ignore a reserved field and then uses an HT
instead of VHT association."

For the iwlwifi bits, Emmanuel says:

"Johannes fixes a long standing bug in the AMPDU status reporting.
Max fixes the listen time which was way too long and causes trouble
to several APs."

Along with those, Bing Zhao marks the mwifiex_usb driver as _not_
supporting USB autosuspend after a number of problems with that have
been reported.
====================

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

2 files changed, 32 insertions, 1 deletions

diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c
index 57d5482b10f..245dce969b3 100644
--- a/net/mac80211/mlme.c
+++ b/net/mac80211/mlme.c
@@ -222,6 +222,7 @@ ieee80211_determine_chantype(struct ieee80211_sub_if_data *sdata,
 	switch (vht_oper->chan_width) {
 	case IEEE80211_VHT_CHANWIDTH_USE_HT:
 		vht_chandef.width = chandef->width;
+		vht_chandef.center_freq1 = chandef->center_freq1;
 		break;
 	case IEEE80211_VHT_CHANWIDTH_80MHZ:
 		vht_chandef.width = NL80211_CHAN_WIDTH_80;
@@ -271,6 +272,28 @@ ieee80211_determine_chantype(struct ieee80211_sub_if_data *sdata,
 	ret = 0;
 
 out:
+	/*
+	 * When tracking the current AP, don't do any further checks if the
+	 * new chandef is identical to the one we're currently using for the
+	 * connection. This keeps us from playing ping-pong with regulatory,
+	 * without it the following can happen (for example):
+	 *  - connect to an AP with 80 MHz, world regdom allows 80 MHz
+	 *  - AP advertises regdom US
+	 *  - CRDA loads regdom US with 80 MHz prohibited (old database)
+	 *  - the code below detects an unsupported channel, downgrades, and
+	 *    we disconnect from the AP in the caller
+	 *  - disconnect causes CRDA to reload world regdomain and the game
+	 *    starts anew.
+	 * (see https://bugzilla.kernel.org/show_bug.cgi?id=70881)
+	 *
+	 * It seems possible that there are still scenarios with CSA or real
+	 * bandwidth changes where a this could happen, but those cases are
+	 * less common and wouldn't completely prevent using the AP.
+	 */
+	if (tracking &&
+	    cfg80211_chandef_identical(chandef, &sdata->vif.bss_conf.chandef))
+		return ret;
+
 	/* don't print the message below for VHT mismatch if VHT is disabled */
 	if (ret & IEEE80211_STA_DISABLE_VHT)
 		vht_chandef = *chandef;
diff --git a/net/wireless/reg.c b/net/wireless/reg.c
index 4c50c21d6f5..f0541370e68 100644
--- a/net/wireless/reg.c
+++ b/net/wireless/reg.c
@@ -2373,6 +2373,7 @@ static int reg_set_rd_country_ie(const struct ieee80211_regdomain *rd,
 int set_regdom(const struct ieee80211_regdomain *rd)
 {
 	struct regulatory_request *lr;
+	bool user_reset = false;
 	int r;
 
 	if (!reg_is_valid_request(rd->alpha2)) {
@@ -2389,6 +2390,7 @@ int set_regdom(const struct ieee80211_regdomain *rd)
 		break;
 	case NL80211_REGDOM_SET_BY_USER:
 		r = reg_set_rd_user(rd, lr);
+		user_reset = true;
 		break;
 	case NL80211_REGDOM_SET_BY_DRIVER:
 		r = reg_set_rd_driver(rd, lr);
@@ -2402,8 +2404,14 @@ int set_regdom(const struct ieee80211_regdomain *rd)
 	}
 
 	if (r) {
-		if (r == -EALREADY)
+		switch (r) {
+		case -EALREADY:
 			reg_set_request_processed();
+			break;
+		default:
+			/* Back to world regulatory in case of errors */
+			restore_regulatory_settings(user_reset);
+		}
 
 		kfree(rd);
 		return r;