Skip to contentSkip to conversation

libertas: Remove carrier signaling from the scan code

PATCH In review Samuel Ortiz 3 participants 4 messages Dec 18, 2009 – Jan 7, 2010 Networking drivers (wireless)

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 libertas: Remove carrier signaling from the scan code

There is no reason to signal a carrier off when doing a 802.11 scan.

Cc Dan Williams <[email protected]>
Cc Holger Schurig <[email protected]>
---
 drivers/net/wireless/libertas/scan.c |   22 ++++++++--------------
 1 files changed, 8 insertions(+), 14 deletions(-)
drivers/net/wireless/libertas/scan.c +8 -14
index c6a6c04..b0b1c78 100644
@@ -567,11 +567,8 @@ int lbs_scan_networks(struct lbs_private *priv, int full_scan)
 	chan_count = lbs_scan_create_channel_list(priv, chan_list);
 
 	netif_stop_queue(priv->dev);
-	netif_carrier_off(priv->dev);
-	if (priv->mesh_dev) {
+	if (priv->mesh_dev)
 		netif_stop_queue(priv->mesh_dev);
-		netif_carrier_off(priv->mesh_dev);
-	}
 
 	/* Prepare to continue an interrupted scan */
 	lbs_deb_scan("chan_count %d, scan_channel %d\n",
@@ -635,16 +632,13 @@ out2:
 	priv->scan_channel = 0;
 
 out:
-	if (priv->connect_status == LBS_CONNECTED) {
-		netif_carrier_on(priv->dev);
-		if (!priv->tx_pending_len)
-			netif_wake_queue(priv->dev);
-	}
-	if (priv->mesh_dev && (priv->mesh_connect_status == LBS_CONNECTED)) {
-		netif_carrier_on(priv->mesh_dev);
-		if (!priv->tx_pending_len)
-			netif_wake_queue(priv->mesh_dev);
-	}
+	if (priv->connect_status == LBS_CONNECTED && !priv->tx_pending_len)
+		netif_wake_queue(priv->dev);
+
+	if (priv->mesh_dev && (priv->mesh_connect_status == LBS_CONNECTED) &&
+	    !priv->tx_pending_len)
+		netif_wake_queue(priv->mesh_dev);
+
 	kfree(chan_list);
 
 	lbs_deb_leave_args(LBS_DEB_SCAN, "ret %d", ret);
Signature · 4 lines
1.6.3.3

Intel Open Source Technology Centre
http://oss.intel.com/
Acked-by on v1 1/1
#1 lore raw
Quoted 5 lines · Samuel Ortiz

There is no reason to signal a carrier off when doing a 802.11 scan.

Cc Dan Williams <[email protected]>
Cc Holger Schurig <[email protected]>
Quoted 45 lines
---
 drivers/net/wireless/libertas/scan.c |   22 ++++++++--------------
 1 files changed, 8 insertions(+), 14 deletions(-)
drivers/net/wireless/libertas/scan.c +8 -14
index c6a6c04..b0b1c78 100644
@@ -567,11 +567,8 @@ int lbs_scan_networks(struct lbs_private *priv, int full_scan)
 	chan_count = lbs_scan_create_channel_list(priv, chan_list);
 
 	netif_stop_queue(priv->dev);
-	netif_carrier_off(priv->dev);
-	if (priv->mesh_dev) {
+	if (priv->mesh_dev)
 		netif_stop_queue(priv->mesh_dev);
-		netif_carrier_off(priv->mesh_dev);
-	}
 
 	/* Prepare to continue an interrupted scan */
 	lbs_deb_scan("chan_count %d, scan_channel %d\n",
@@ -635,16 +632,13 @@ out2:
 	priv->scan_channel = 0;
 
 out:
-	if (priv->connect_status == LBS_CONNECTED) {
-		netif_carrier_on(priv->dev);
-		if (!priv->tx_pending_len)
-			netif_wake_queue(priv->dev);
-	}
-	if (priv->mesh_dev && (priv->mesh_connect_status == LBS_CONNECTED)) {
-		netif_carrier_on(priv->mesh_dev);
-		if (!priv->tx_pending_len)
-			netif_wake_queue(priv->mesh_dev);
-	}
+	if (priv->connect_status == LBS_CONNECTED && !priv->tx_pending_len)
+		netif_wake_queue(priv->dev);
+
+	if (priv->mesh_dev && (priv->mesh_connect_status == LBS_CONNECTED) &&
+	    !priv->tx_pending_len)
+		netif_wake_queue(priv->mesh_dev);
+
 	kfree(chan_list);
 
 	lbs_deb_leave_args(LBS_DEB_SCAN, "ret %d", ret);
#2 lore raw

Hi John,

Quoted 7 lines
Quoted 5 lines

There is no reason to signal a carrier off when doing a 802.11 scan.

Cc Dan Williams <[email protected]>
Cc Holger Schurig <[email protected]>

can we get this driver merged since no wireless driver should play with carrier on/off settings during scanning.

I would actually go that far that we should fix it in 2.6.33 and also backport it to -stable.

Regards

Marcel

#3 lore raw
Quoted 8 lines · Marcel Holtmann

Hi John,

Quoted 6 lines
Quoted 5 lines

There is no reason to signal a carrier off when doing a 802.11 scan.

Cc Dan Williams <[email protected]>
Cc Holger Schurig <[email protected]>
Quoted 7 lines
Quoted 1 line

can we get this driver merged since no wireless driver should play with carrier on/off settings during scanning.

I would actually go that far that we should fix it in 2.6.33 and also backport it to -stable.

Well likewise even if it *did* play with carrier, nothing should be listening to the carrier signal for wifi devices anyway. So it shouldn't really be a problem.

Dan