[PATCH 1/2] mac80211: set CLEAR_PS for pspolled frames

[PATCH 1/2] mac80211: set CLEAR_PS for pspolled frames

[Christian Lamparter]

This patch sets IEEE80211_TX_CTL_CLEAR_PS_FILT for outgoing
frames for a half-wake station.

this is necessary if one wants to get ps-poll working properly with a p54 ap.

Signed-off-by: Christian Lamparter <chunkeey@web.de>
---
diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c
index 3fb04a8..4c2e19e 100644
--- a/net/mac80211/tx.c
+++ b/net/mac80211/tx.c
@@ -409,8 +409,19 @@ ieee80211_tx_h_unicast_ps_buf(struct ieee80211_tx_data *tx)
 		       sta->sta.addr);
 	}
 #endif /* CONFIG_MAC80211_VERBOSE_PS_DEBUG */
-	clear_sta_flags(sta, WLAN_STA_PSPOLL);
+	if (test_and_clear_sta_flags(sta, WLAN_STA_PSPOLL)) {
+		/*
+		 * The sleeping station with pending data is now snoozing.
+		 * It queried us for its buffered frames and will go back
+		 * to deep sleep once it got everything.
+		 *
+		 * inform the driver, in case the hardware does powersave
+		 * frame filtering and keeps a station  blacklist on its own
+		 * (e.g: p54), so that frames can be delivered unimpeded.
+		 */
 
+		info->flags |= IEEE80211_TX_CTL_CLEAR_PS_FILT;
+	}
 	return TX_CONTINUE;
 }
 

[PATCH 1/2 v2] mac80211: set CLEAR_PS for pspolled frames

[Christian Lamparter]

This patch sets IEEE80211_TX_CTL_CLEAR_PS_FILT for outgoing
frames for a half-wake station.
 
this is necessary if one wants to get ps-poll working properly with a p54 ap.
 
Signed-off-by: Christian Lamparter <chunkeey@web.de>
---
changes:
	added a explanation, why we can safely remove the filter now.
---
diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c
index 3fb04a8..f336cc7 100644
--- a/net/mac80211/tx.c
+++ b/net/mac80211/tx.c
@@ -409,8 +409,24 @@ ieee80211_tx_h_unicast_ps_buf(struct ieee80211_tx_data *tx)
 		       sta->sta.addr);
 	}
 #endif /* CONFIG_MAC80211_VERBOSE_PS_DEBUG */
-	clear_sta_flags(sta, WLAN_STA_PSPOLL);
+	if (test_and_clear_sta_flags(sta, WLAN_STA_PSPOLL)) {
+		/*
+		 * The sleeping station with pending data is now snoozing.
+		 * It queried us for its buffered frames and will go back
+		 * to deep sleep once it got everything.
+		 *
+		 * inform the driver, in case the hardware does powersave
+		 * frame filtering and keeps a station  blacklist on its own
+		 * (e.g: p54), so that frames can be delivered unimpeded.
+		 *
+		 * Note: It should be save to disable the filter now.
+		 * As, it is really unlikely that we still have any pending
+		 * frame for this station in the hw's buffers/fifos left,
+		 * that is not rejected with a unsuccessful tx_status yet.
+		 */
 
+		info->flags |= IEEE80211_TX_CTL_CLEAR_PS_FILT;
+	}
 	return TX_CONTINUE;
 }