2012-06-20 10:10:24

by Eyal Shapira

Subject: [PATCH] mac80211: fix cleanup if driver suspend callback fails

In case the driver suspend callback fails, mac80211 is left
with stopped queues which prevents any further traffic as well
as all STAs are left marked with WLAN_STA_BLOCK_BA which will
cause any further ADDBA requests to be declined. Fix it by
undoing both before returning from __iee80211_suspend.

Reported-by: Vitaly Wool <[email protected]>
Signed-off-by: Eyal Shapira <[email protected]>
---

This was discussed in this thread:
http://comments.gmane.org/gmane.linux.kernel.wireless.general/89571
but no v2 was submitted. We've had this patch applied
on our internal tree for a while so here it is.

net/mac80211/pm.c | 10 ++++++++++
1 files changed, 10 insertions(+), 0 deletions(-)

diff --git a/net/mac80211/pm.c b/net/mac80211/pm.c
index 98c128b..5c572e7 100644
--- a/net/mac80211/pm.c
+++ b/net/mac80211/pm.c
@@ -78,6 +78,16 @@ int __ieee80211_suspend(struct ieee80211_hw *hw, struct cfg80211_wowlan *wowlan)
if (err < 0) {
local->quiescing = false;
local->wowlan = false;
+ if (hw->flags & IEEE80211_HW_AMPDU_AGGREGATION) {
+ mutex_lock(&local->sta_mtx);
+ list_for_each_entry(sta,
+ &local->sta_list, list) {
+ clear_sta_flag(sta, WLAN_STA_BLOCK_BA);
+ }
+ mutex_unlock(&local->sta_mtx);
+ }
+ ieee80211_wake_queues_by_reason(hw,
+ IEEE80211_QUEUE_STOP_REASON_SUSPEND);
return err;
} else if (err > 0) {
WARN_ON(err != 1);
--
1.7.4.1

2012-06-20 15:40:17

by Johannes Berg

Subject: Re: [PATCH] mac80211: fix cleanup if driver suspend callback fails

On Wed, 2012-06-20 at 13:10 +0300, Eyal Shapira wrote:
> In case the driver suspend callback fails, mac80211 is left
> with stopped queues which prevents any further traffic as well
> as all STAs are left marked with WLAN_STA_BLOCK_BA which will
> cause any further ADDBA requests to be declined. Fix it by
> undoing both before returning from __iee80211_suspend.
>
> Reported-by: Vitaly Wool <[email protected]>
> Signed-off-by: Eyal Shapira <[email protected]>
> ---
>
> This was discussed in this thread:
> http://comments.gmane.org/gmane.linux.kernel.wireless.general/89571
> but no v2 was submitted. We've had this patch applied
> on our internal tree for a while so here it is.

Applied.

johannes