wifi: mac80211: fix monitor channel with chanctx emulation

Commit Message

Johannes Berg June 12, 2024, 10:23 a.m. UTC

From: Johannes Berg <[e-mail]>

After the channel context emulation, there were reports that
changing the monitor channel no longer works. This is because
those drivers don't have WANT_MONITOR_VIF, so the setting the
channel always exits out quickly.

Fix this by always allocating the virtual monitor sdata, and
simply not telling the driver about it unless it wanted to.
This way, we have an interface/sdata to bind the chanctx to,
and the emulation can work correctly.

Cc: [e-mail]
Fixes: 0a44dfc07074 ("wifi: mac80211: simplify non-chanctx drivers")
Reported-by: Savyasaachi Vanga <[e-mail]>
Closes: https://lore.kernel.org/r/chwoymvpzwtbmzryrlitpwmta5j6mtndocxsyqvdyikqu63lon@gfds653hkknl
Signed-off-by: Johannes Berg <[e-mail]>
---
 net/mac80211/driver-ops.c | 17 +++++++++++++++++
 net/mac80211/iface.c      | 21 +++++++++------------
 net/mac80211/util.c       |  2 +-
 3 files changed, 27 insertions(+), 13 deletions(-)

Comments

Christian Heusel June 16, 2024, 10:02 p.m. UTC | #1

On 24/06/12 12:23PM, Johannes Berg wrote:
> From: Johannes Berg <[e-mail]>
> 
> After the channel context emulation, there were reports that
> changing the monitor channel no longer works. This is because
> those drivers don't have WANT_MONITOR_VIF, so the setting the
> channel always exits out quickly.
> 
> Fix this by always allocating the virtual monitor sdata, and
> simply not telling the driver about it unless it wanted to.
> This way, we have an interface/sdata to bind the chanctx to,
> and the emulation can work correctly.
> 
> Cc: [e-mail]
> Fixes: 0a44dfc07074 ("wifi: mac80211: simplify non-chanctx drivers")
> Reported-by: Savyasaachi Vanga <[e-mail]>
> Closes: https://lore.kernel.org/r/chwoymvpzwtbmzryrlitpwmta5j6mtndocxsyqvdyikqu63lon@gfds653hkknl
> Signed-off-by: Johannes Berg <[e-mail]>

Hello Johannes,

I saw that you have dropped have dropped the "Bisected-by:" tag I added
from[0], if its still not too late it would be cool if this could either
be re-added or replaced by a "Reported-By:" tag if that is what's rather
used in this subsystem.

Cheers,
chris

[0]: https://lore.kernel.org/regressions/chwoymvpzwtbmzryrlitpwmta5j6mtndocxsyqvdyikqu63lon@gfds653hkknl/

Patch

diff --git a/net/mac80211/driver-ops.c b/net/mac80211/driver-ops.c
index dce37ba8ebe3..254d745832cb 100644
--- a/net/mac80211/driver-ops.c
+++ b/net/mac80211/driver-ops.c
@@ -311,6 +311,18 @@  int drv_assign_vif_chanctx(struct ieee80211_local *local,
 	might_sleep();
 	lockdep_assert_wiphy(local->hw.wiphy);
 
+	/*
+	 * We should perhaps push emulate chanctx down and only
+	 * make it call ->config() when the chanctx is actually
+	 * assigned here (and unassigned below), but that's yet
+	 * another change to all drivers to add assign/unassign
+	 * emulation callbacks. Maybe later.
+	 */
+	if (sdata->vif.type == NL80211_IFTYPE_MONITOR &&
+	    local->emulate_chanctx &&
+	    !ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF))
+		return 0;
+
 	if (!check_sdata_in_driver(sdata))
 		return -EIO;
 
@@ -338,6 +350,11 @@  void drv_unassign_vif_chanctx(struct ieee80211_local *local,
 	might_sleep();
 	lockdep_assert_wiphy(local->hw.wiphy);
 
+	if (sdata->vif.type == NL80211_IFTYPE_MONITOR &&
+	    local->emulate_chanctx &&
+	    !ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF))
+		return;
+
 	if (!check_sdata_in_driver(sdata))
 		return;
 
diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c
index dc42902e2693..4a49e834a9f5 100644
--- a/net/mac80211/iface.c
+++ b/net/mac80211/iface.c
@@ -1121,9 +1121,6 @@  int ieee80211_add_virtual_monitor(struct ieee80211_local *local)
 	struct ieee80211_sub_if_data *sdata;
 	int ret;
 
-	if (!ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF))
-		return 0;
-
 	ASSERT_RTNL();
 	lockdep_assert_wiphy(local->hw.wiphy);
 
@@ -1145,11 +1142,13 @@  int ieee80211_add_virtual_monitor(struct ieee80211_local *local)
 
 	ieee80211_set_default_queues(sdata);
 
-	ret = drv_add_interface(local, sdata);
-	if (WARN_ON(ret)) {
-		/* ok .. stupid driver, it asked for this! */
-		kfree(sdata);
-		return ret;
+	if (ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF)) {
+		ret = drv_add_interface(local, sdata);
+		if (WARN_ON(ret)) {
+			/* ok .. stupid driver, it asked for this! */
+			kfree(sdata);
+			return ret;
+		}
 	}
 
 	set_bit(SDATA_STATE_RUNNING, &sdata->state);
@@ -1187,9 +1186,6 @@  void ieee80211_del_virtual_monitor(struct ieee80211_local *local)
 {
 	struct ieee80211_sub_if_data *sdata;
 
-	if (!ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF))
-		return;
-
 	ASSERT_RTNL();
 	lockdep_assert_wiphy(local->hw.wiphy);
 
@@ -1209,7 +1205,8 @@  void ieee80211_del_virtual_monitor(struct ieee80211_local *local)
 
 	ieee80211_link_release_channel(&sdata->deflink);
 
-	drv_remove_interface(local, sdata);
+	if (ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF))
+		drv_remove_interface(local, sdata);
 
 	kfree(sdata);
 }
diff --git a/net/mac80211/util.c b/net/mac80211/util.c
index 283bfc99417e..963ed75deb76 100644
--- a/net/mac80211/util.c
+++ b/net/mac80211/util.c
@@ -1843,7 +1843,7 @@  int ieee80211_reconfig(struct ieee80211_local *local)
 
 	/* add interfaces */
 	sdata = wiphy_dereference(local->hw.wiphy, local->monitor_sdata);
-	if (sdata) {
+	if (sdata && ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF)) {
 		/* in HW restart it exists already */
 		WARN_ON(local->resuming);
 		res = drv_add_interface(local, sdata);