mac80211: hardware scan rework (V2)

PATCH In review Zhu Yi 6 participants 9 messages Nov 20, 2007 – Nov 21, 2007 mac80211

lore

Share & tools

lore

Zhu Yi

author

2007-11-20 05:21 UTC #0 lore raw

PATCH 1/1 mac80211: hardware scan rework (V2)

Hi,

This is the second version of the hardware scan patch. Comparing with the first version, it does the following:

1. fix a hw scan bug in ieee80211_rx_bss_info() for setting beacon supported rates
2. let control frames pass in ieee80211_sta_rx_scan() during hw scan
3. set local->sta_{hw|sw}_scanning type to bool
4. avoid channel setting during hw scan
5. rework ieee80211_scan_completed() to make it symmetric for hw scan

The scan code in mac80211 makes the software scan assumption in various places. For example, we stop the Tx queue during a software scan so that all the Tx packets will be queued by the stack. We also drop frames not related to scan in the software scan process. But these are not true for hardware scan.

Some wireless hardwares (for example iwl3945/4965) has the ability to perform the whole scan process by hardware and/or firmware. The hardware scan is relative powerful in that it tries to maintain normal network traffic while doing a scan in the background. Some drivers (i.e iwlwifi) do provide a way to tune the hardware scan parameters (for example if the STA is associated, what's the max time could the STA leave from the associated channel, how long the scans get suspended after returning to the service channel, etc). But basically this is transparent to the stack. mac80211 should not stop Tx queues or drop Rx packets during a hardware scan.

This patch resolves the above problem by spliting the current scan indicator local->sta_scanning into local->sta_sw_scanning and local->sta_hw_scanning. It then changes the scan related code to be aware of hardware scan or software scan in various places. With this patch, iwlwifi performs much better in the scan-while-associated condition and disable_hw_scan=1 should never be required.

Cc Johannes Berg <[email protected]>

Cc Mohamed Abbas <[email protected]>

Cc Ben Cahill <[email protected]>

Signed-off-by Zhu Yi <[email protected]>

---

net/mac80211/ieee80211.c +8 -5

index e0ee65a..d230419 100644
@@ -328,11 +328,14 @@ static int ieee80211_stop(struct net_device *dev)
 		synchronize_rcu();
 		skb_queue_purge(&sdata->u.sta.skb_queue);
 
-		if (!local->ops->hw_scan &&
-		    local->scan_dev == sdata->dev) {
-			local->sta_scanning = 0;
-			cancel_delayed_work(&local->scan_work);
+		if (local->scan_dev == sdata->dev) {
+			if (!local->ops->hw_scan) {
+				local->sta_sw_scanning = 0;
+				cancel_delayed_work(&local->scan_work);
+			} else
+				local->sta_hw_scanning = 0;
 		}
+
 		flush_workqueue(local->hw.workqueue);
 		/* fall through */
 	default:

@@ -500,7 +503,7 @@ int ieee80211_hw_config(struct ieee80211_local *local)
 	struct ieee80211_channel *chan;
 	int ret = 0;
 
-	if (local->sta_scanning) {
+	if (local->sta_sw_scanning) {
 		chan = local->scan_channel;
 		mode = local->scan_hw_mode;
 	} else {

net/mac80211/ieee80211_i.h +4 -2

index 72e1c93..35829b1 100644
@@ -470,7 +470,8 @@ struct ieee80211_local {
 
 	struct list_head interfaces;
 
-	int sta_scanning;
+	bool sta_sw_scanning;
+	bool sta_hw_scanning;
 	int scan_channel_idx;
 	enum { SCAN_SET_CHANNEL, SCAN_SEND_PROBE } scan_state;
 	unsigned long last_scan_completed;

@@ -749,7 +750,8 @@ int ieee80211_sta_req_scan(struct net_device *dev, u8 *ssid, size_t ssid_len);
 void ieee80211_sta_req_auth(struct net_device *dev,
 			    struct ieee80211_if_sta *ifsta);
 int ieee80211_sta_scan_results(struct net_device *dev, char *buf, size_t len);
-void ieee80211_sta_rx_scan(struct net_device *dev, struct sk_buff *skb,
+ieee80211_txrx_result ieee80211_sta_rx_scan(struct net_device *dev,
+					    struct sk_buff *skb,
 			   struct ieee80211_rx_status *rx_status);
 void ieee80211_rx_bss_list_init(struct net_device *dev);
 void ieee80211_rx_bss_list_deinit(struct net_device *dev);

net/mac80211/ieee80211_ioctl.c +4 -2

index 942b9cc..741515a 100644
@@ -315,7 +315,7 @@ int ieee80211_set_channel(struct ieee80211_local *local, int channel, int freq)
 	}
 
 	if (set) {
-		if (local->sta_scanning)
+		if (local->sta_sw_scanning || local->sta_hw_scanning)
 			ret = 0;
 		else
 			ret = ieee80211_hw_config(local);

@@ -558,8 +558,10 @@ static int ieee80211_ioctl_giwscan(struct net_device *dev,
 {
 	int res;
 	struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr);
-	if (local->sta_scanning)
+
+	if (local->sta_sw_scanning || local->sta_hw_scanning)
 		return -EAGAIN;
+
 	res = ieee80211_sta_scan_results(dev, extra, data->length);
 	if (res >= 0) {
 		data->length = res;

net/mac80211/ieee80211_sta.c +44 ~~-22~~

157 lines not on this page · raw message

net/mac80211/rx.c +9 -3

index 428a9fc..596ff71 100644
@@ -338,8 +338,14 @@ ieee80211_rx_h_passive_scan(struct ieee80211_txrx_data *rx)
 	struct ieee80211_local *local = rx->local;
 	struct sk_buff *skb = rx->skb;
 
-	if (unlikely(local->sta_scanning != 0)) {
-		ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+	if (unlikely(local->sta_hw_scanning))
+		return ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+
+	if (unlikely(local->sta_sw_scanning)) {
+		/* drop all the other packets during a software scan anyway */
+		if (ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status)
+		    != TXRX_QUEUED)
+			dev_kfree_skb(skb);
 		return TXRX_QUEUED;
 	}

@@ -1486,7 +1492,7 @@ void __ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb,
 		goto end;
 	}
 
-	if (unlikely(local->sta_scanning))
+	if (unlikely(local->sta_sw_scanning || local->sta_hw_scanning))
 		rx.flags |= IEEE80211_TXRXD_RXIN_SCAN;
 
 	if (__ieee80211_invoke_rx_handlers(local, local->rx_pre_handlers, &rx,

net/mac80211/tx.c +1 -1

index 1a53154..a6a657f 100644
@@ -225,7 +225,7 @@ ieee80211_tx_h_check_assoc(struct ieee80211_txrx_data *tx)
 	if (unlikely(tx->flags & IEEE80211_TXRXD_TX_INJECTED))
 		return TXRX_CONTINUE;
 
-	if (unlikely(tx->local->sta_scanning != 0) &&
+	if (unlikely(tx->local->sta_sw_scanning) &&
 	    ((tx->fc & IEEE80211_FCTL_FTYPE) != IEEE80211_FTYPE_MGMT ||
 	     (tx->fc & IEEE80211_FCTL_STYPE) != IEEE80211_STYPE_PROBE_REQ))
 		return TXRX_DROP;

Johannes Berg

maintainer

2007-11-20 13:50 UTC #1 lore raw

Hi,

Thanks!

Quoted 5 lines

1. fix a hw scan bug in ieee80211_rx_bss_info() for setting beacon supported rates
2. let control frames pass in ieee80211_sta_rx_scan() during hw scan
3. set local->sta_{hw|sw}_scanning type to bool
4. avoid channel setting during hw scan

Why this? I was confused at first and I'm sorry if I confused you, but since we're now fully unaware of hardware scan thanks to your item 5, I think we need to allow setting channel during hardware scan so the firmware will change to the right channel once it finished scanning.

Quoted 14 lines

5. rework ieee80211_scan_completed() to make it symmetric for hw scan

net/mac80211/ieee80211_sta.c +1 -1

index 015b3f8..26f404a 100644
@@ -1487,8 +1487,18 @@ static void ieee80211_rx_bss_info(struct net_device *dev,
 		u32 supp_rates, prev_rates;
 		int i, j;
 
-		mode = local->sta_scanning ?
+		mode = local->sta_sw_scanning ?
 		       local->scan_hw_mode : local->oper_hw_mode;

net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev, · line 1501

Quoted 10 lines

net/mac80211/ieee80211_sta.c +10 -0

+
+		if (local->sta_hw_scanning) {
+			/* search for the correct mode matches the beacon */
+			list_for_each_entry(mode, &local->modes_list, list)
+				if (mode->mode == rx_status->phymode)
+					break;
+
+			if (mode == NULL)
+				mode = local->oper_hw_mode;
+		}

Good catch.

net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct *work) · line 2007

Quoted 7 lines

net/mac80211/ieee80211_sta.c +1 -1

@@ -1985,7 +2003,7 @@ void ieee80211_sta_work(struct work_struct *work)
 	if (!netif_running(dev))
 		return;
 
-	if (local->sta_scanning)
+	if (local->sta_sw_scanning || local->sta_hw_scanning)
 		return;

Shouldn't the sta work be able to run normally while hw scan is in progress?

net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct *work)

Quoted 10 lines

net/mac80211/ieee80211_sta.c +8 -2

-	if (unlikely(local->sta_scanning != 0)) {
-		ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+	if (unlikely(local->sta_hw_scanning))
+		return ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+
+	if (unlikely(local->sta_sw_scanning)) {
+		/* drop all the other packets during a software scan anyway */
+		if (ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status)
+		    != TXRX_QUEUED)
+			dev_kfree_skb(skb);

Not entirely sure why we do this, but nothing we should change with this patch.

Other than the ioctl thing it looks good to me.

johannes

Cahill, Ben M

2007-11-20 16:23 UTC #2 lore raw

Quoted 18 lines

-----Original Message-----
From: Johannes Berg [mailto:[email protected]]
Sent: Tuesday, November 20, 2007 8:51 AM
To: Zhu, Yi
Cc: [email protected]; [email protected]; Abbas, Mohamed; Cahill, Ben M
Subject: Re: [PATCH] mac80211: hardware scan rework (V2)
Hi,
Thanks!
Quoted 3 lines
1. fix a hw scan bug in ieee80211_rx_bss_info() for setting beacon supported rates
2. let control frames pass in ieee80211_sta_rx_scan()
during hw scan

Quoted 8 lines

Quoted 2 lines
3. set local->sta_{hw|sw}_scanning type to bool 4. avoid channel setting during hw scan
Why this? I was confused at first and I'm sorry if I confused you, but since we're now fully unaware of hardware scan thanks to your item 5, I think we need to allow setting channel during hardware scan so the firmware will change to the right channel once it finished scanning.

The firmware-driven scan on 4965 (3945 also) returns to the original channel by itself. It might even do this several times while scanning a given list of channels, depending on how the timing is set up. Within the uCode scan command, the driver sets "max_out_time" to restrict the scan engine from being away from the service (network) channel longer than a certain amount of time. The scan engine will work its way through the channel list until it doesn't have enough time to complete the next channel's scan, then it returns to the service channel.

Conversely, the "suspend_time" controls how long the scan engine stays back on the service channel (after returning from a scan channel) before resuming the scan with the next channel on the list.

If the "max_out_time" is set short, then the scan engine will take several iterations before it completes the list of scan channels. This is the mechanism that keeps traffic "flowing" during the course of a scan.

Once done with the list, the scan engine returns to the service channel automatically. While it *might* work okay to change or reset that channel, it is not necessary, and might be asking for trouble.

-- Ben --

Quoted 64 lines

Quoted 2 lines
5. rework ieee80211_scan_completed() to make it symmetric
for hw scan
net/mac80211/ieee80211_sta.c static void
Quoted 5 lines
(quoted diff) +0 -0
b/net/mac80211/ieee80211_sta.c index 015b3f8..26f404a 100644
net/mac80211/ieee80211_sta.c +0 -0
@@ -1487,8 +1487,18 @@ static void 
ieee80211_rx_bss_info(struct net_device *dev,
net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev,
Quoted 9 lines
net/mac80211/ieee80211_sta.c +4 -1
 		u32 supp_rates, prev_rates;
 		int i, j;
 
-		mode = local->sta_scanning ?
+		mode = local->sta_sw_scanning ?
 		       local->scan_hw_mode : local->oper_hw_mode;
+
+		if (local->sta_hw_scanning) {
+			/* search for the correct mode matches 
the beacon */
net/mac80211/ieee80211_sta.c static void · line 1495
Quoted 1 line
net/mac80211/ieee80211_sta.c +1 -0
+			list_for_each_entry(mode, 
&local->modes_list, list)
net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev, · line 1501
Quoted 6 lines
net/mac80211/ieee80211_sta.c +6 -0
+				if (mode->mode == rx_status->phymode)
+					break;
+
+			if (mode == NULL)
+				mode = local->oper_hw_mode;
+		}
Good catch.
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct
Quoted 1 line
net/mac80211/ieee80211_sta.c +0 -0
@@ -1985,7 +2003,7 @@ void ieee80211_sta_work(struct 
work_struct *work)
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct *work) · line 2007
Quoted 6 lines
net/mac80211/ieee80211_sta.c +1 -1
 	if (!netif_running(dev))
 		return;
 
-	if (local->sta_scanning)
+	if (local->sta_sw_scanning || local->sta_hw_scanning)
 		return;
Shouldn't the sta work be able to run normally while hw scan is in progress?
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct · line 2009
Quoted 4 lines
net/mac80211/ieee80211_sta.c +2 -2
-	if (unlikely(local->sta_scanning != 0)) {
-		ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+	if (unlikely(local->sta_hw_scanning))
+		return ieee80211_sta_rx_scan(rx->dev, skb, 
rx->u.rx.status);
net/mac80211/ieee80211_sta.c
Quoted 3 lines
net/mac80211/ieee80211_sta.c +3 -0
+
+	if (unlikely(local->sta_sw_scanning)) {
+		/* drop all the other packets during a software 
scan anyway */
net/mac80211/ieee80211_sta.c
Quoted 3 lines
net/mac80211/ieee80211_sta.c +3 -0
+		if (ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status)
+		    != TXRX_QUEUED)
+			dev_kfree_skb(skb);
Not entirely sure why we do this, but nothing we should change with this patch.
Other than the ioctl thing it looks good to me.
johannes

Johannes Berg

maintainer

2007-11-20 16:28 UTC #3 lore raw

Quoted 3 lines

The firmware-driven scan on 4965 (3945 also) returns to the original channel by itself. [...]

Right. Thanks for the description.

Quoted 3 lines

Once done with the list, the scan engine returns to the service channel automatically. While it *might* work okay to change or reset that channel, it is not necessary, and might be asking for trouble.

Well, in the patch right now the channel the user sets will be completely ignored because the scan_finished thing doesn't reconfigure the hardware as it should.

Since the user has an expectation that the working channel is changed after setting it with the ioctl, the code should make sure. Since hw scanning is mostly transparent to mac80211, the ioctl should imho be passed to the driver right away. If there's a problem with setting it to the hardware directly, the driver can cache it for until the firmware scan finished, but I don't think we should do that in mac80211. It has to be done somewhere though and just ignoring the setting is wrong.

johannes

Cahill, Ben M

2007-11-20 16:44 UTC #4 lore raw

Quoted 26 lines

-----Original Message-----
From: Johannes Berg [mailto:[email protected]]
Sent: Tuesday, November 20, 2007 11:30 AM
To: Cahill, Ben M
Cc: Zhu, Yi; [email protected]; [email protected]; Abbas, Mohamed
Subject: RE: [PATCH] mac80211: hardware scan rework (V2)
Quoted 1 line
The firmware-driven scan on 4965 (3945 also) returns to the
original
Quoted 1 line
channel by itself. [...]
Right. Thanks for the description.
Quoted 2 lines
Once done with the list, the scan engine returns to the service channel automatically. While it *might* work okay to
change or reset
Quoted 1 line
that channel, it is not necessary, and might be asking for trouble.
Well, in the patch right now the channel the user sets will be completely ignored because the scan_finished thing doesn't reconfigure the hardware as it should.

Quoted 8 lines

Since the user has an expectation that the working channel is changed after setting it with the ioctl, the code should make sure. Since hw scanning is mostly transparent to mac80211, the ioctl should imho be passed to the driver right away. If there's a problem with setting it to the hardware directly, the driver can cache it for until the firmware scan finished, but I don't think we should do that in mac80211. It has to be done somewhere though and just ignoring the setting is wrong.

I see ... thanks, I didn't quite understand the context. Agreed, if the user says to do something, we should not throw it on the floor.

-- Ben --

Quoted 3 lines

johannes

Abbas, Mohamed

2007-11-20 17:19 UTC #5 lore raw

In reply to Johannes Berg

Johannes
The firmware will go back to the old setting, the ones before scan, once done scanning by it self. The calls by mac80211 after scan complete to set channel will be ignored by the iwl driver if we still tuned to the same channel, iwl wont call firmware on the same setting, it has to be a change to call down to firmware. I need to check what happen if user change channel during scan, this call might be necessary in this case. Mohamed

-----Original Message-----
From: Johannes Berg [mailto:[email protected]]
Sent: Tuesday, November 20, 2007 5:51 AM
To: Zhu, Yi
Cc: [email protected]; [email protected]; Abbas, Mohamed; Cahill, Ben M
Subject: Re: [PATCH] mac80211: hardware scan rework (V2)

Hi,

Thanks!

Quoted 5 lines

1. fix a hw scan bug in ieee80211_rx_bss_info() for setting beacon supported rates
2. let control frames pass in ieee80211_sta_rx_scan() during hw scan
3. set local->sta_{hw|sw}_scanning type to bool
4. avoid channel setting during hw scan

Review

Quoted 4 lines

5. rework ieee80211_scan_completed() to make it symmetric for hw scan
(quoted diff) +0 -0

b/net/mac80211/ieee80211_sta.c

net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct

Quoted 4 lines

index 015b3f8..26f404a 100644
net/mac80211/ieee80211_sta.c +0 -0
@@ -1487,8 +1487,18 @@ static void ieee80211_rx_bss_info(struct

net_device *dev,

net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev,

Quoted 9 lines

net/mac80211/ieee80211_sta.c +4 -1

 		u32 supp_rates, prev_rates;
 		int i, j;
 
-		mode = local->sta_scanning ?
+		mode = local->sta_sw_scanning ?
 		       local->scan_hw_mode : local->oper_hw_mode;
+
+		if (local->sta_hw_scanning) {
+			/* search for the correct mode matches the

beacon */

net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct · line 1495

Quoted 1 line

net/mac80211/ieee80211_sta.c +1 -0
+			list_for_each_entry(mode, &local->modes_list,

list)

net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev, · line 1501

Quoted 6 lines

net/mac80211/ieee80211_sta.c +6 -0

+				if (mode->mode == rx_status->phymode)
+					break;
+
+			if (mode == NULL)
+				mode = local->oper_hw_mode;
+		}

Good catch.

net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct

Quoted 1 line

net/mac80211/ieee80211_sta.c +0 -0
@@ -1985,7 +2003,7 @@ void ieee80211_sta_work(struct work_struct

*work)

net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct *work) · line 2007

Quoted 6 lines

net/mac80211/ieee80211_sta.c +1 -1

 	if (!netif_running(dev))
 		return;
 
-	if (local->sta_scanning)
+	if (local->sta_sw_scanning || local->sta_hw_scanning)
 		return;

Shouldn't the sta work be able to run normally while hw scan is in progress?

net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct · line 2009

Quoted 4 lines

net/mac80211/ieee80211_sta.c +2 -2

-	if (unlikely(local->sta_scanning != 0)) {
-		ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+	if (unlikely(local->sta_hw_scanning))
+		return ieee80211_sta_rx_scan(rx->dev, skb,

rx->u.rx.status);

net/mac80211/ieee80211_sta.c

Quoted 3 lines

net/mac80211/ieee80211_sta.c +3 -0

+
+	if (unlikely(local->sta_sw_scanning)) {
+		/* drop all the other packets during a software scan

anyway */

net/mac80211/ieee80211_sta.c

Quoted 3 lines

net/mac80211/ieee80211_sta.c +3 -0

+		if (ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status)
+		    != TXRX_QUEUED)
+			dev_kfree_skb(skb);

Not entirely sure why we do this, but nothing we should change with this patch.

Other than the ioctl thing it looks good to me.

johannes

Dan Williams

2007-11-20 17:25 UTC #6 lore raw

Quoted 7 lines · Mohamed

Johannes
The firmware will go back to the old setting, the ones before scan, once done scanning by it self. The calls by mac80211 after scan complete to set channel will be ignored by the iwl driver if we still tuned to the same channel, iwl wont call firmware on the same setting, it has to be a change to call down to firmware. I need to check what happen if user change channel during scan, this call might be necessary in this case.

I'd expect that either (a) the scan gets canceled and then the channel change happens, or (b) the channel change gets queued and happens immediately after the scan completes. (b) is probably the easier thing to do here unless the ucode has a cancel-scan command, but then you have to wait until the scan is actually canceled anyway, which sort of implies (b) too. So maybe just do (b) and be done with it :)

Dan

Quoted 91 lines

Mohamed
-----Original Message-----
From: Johannes Berg [mailto:[email protected]]
Sent: Tuesday, November 20, 2007 5:51 AM
To: Zhu, Yi
Cc: [email protected]; [email protected]; Abbas, Mohamed; Cahill, Ben M
Subject: Re: [PATCH] mac80211: hardware scan rework (V2)
Hi,
Thanks!
Quoted 5 lines
1. fix a hw scan bug in ieee80211_rx_bss_info() for setting beacon supported rates
2. let control frames pass in ieee80211_sta_rx_scan() during hw scan
3. set local->sta_{hw|sw}_scanning type to bool
4. avoid channel setting during hw scan
Why this? I was confused at first and I'm sorry if I confused you, but since we're now fully unaware of hardware scan thanks to your item 5, I think we need to allow setting channel during hardware scan so the firmware will change to the right channel once it finished scanning.
Review
Quoted 4 lines
5. rework ieee80211_scan_completed() to make it symmetric for hw scan
(quoted diff) +0 -0
b/net/mac80211/ieee80211_sta.c
net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct
Quoted 4 lines
index 015b3f8..26f404a 100644
net/mac80211/ieee80211_sta.c +0 -0
@@ -1487,8 +1487,18 @@ static void ieee80211_rx_bss_info(struct
net_device *dev,
net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev,
Quoted 9 lines
net/mac80211/ieee80211_sta.c +4 -1
 		u32 supp_rates, prev_rates;
 		int i, j;
 
-		mode = local->sta_scanning ?
+		mode = local->sta_sw_scanning ?
 		       local->scan_hw_mode : local->oper_hw_mode;
+
+		if (local->sta_hw_scanning) {
+			/* search for the correct mode matches the
beacon */
net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct · line 1495
Quoted 1 line
net/mac80211/ieee80211_sta.c +1 -0
+			list_for_each_entry(mode, &local->modes_list,
list)
net/mac80211/ieee80211_sta.c static void ieee80211_rx_bss_info(struct net_device *dev, · line 1501
Quoted 6 lines
net/mac80211/ieee80211_sta.c +6 -0
+				if (mode->mode == rx_status->phymode)
+					break;
+
+			if (mode == NULL)
+				mode = local->oper_hw_mode;
+		}
Good catch.
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct
Quoted 1 line
net/mac80211/ieee80211_sta.c +0 -0
@@ -1985,7 +2003,7 @@ void ieee80211_sta_work(struct work_struct
*work)
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct *work) · line 2007
Quoted 6 lines
net/mac80211/ieee80211_sta.c +1 -1
 	if (!netif_running(dev))
 		return;
 
-	if (local->sta_scanning)
+	if (local->sta_sw_scanning || local->sta_hw_scanning)
 		return;
Shouldn't the sta work be able to run normally while hw scan is in progress?
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct · line 2009
Quoted 4 lines
net/mac80211/ieee80211_sta.c +2 -2
-	if (unlikely(local->sta_scanning != 0)) {
-		ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status);
+	if (unlikely(local->sta_hw_scanning))
+		return ieee80211_sta_rx_scan(rx->dev, skb,
rx->u.rx.status);
net/mac80211/ieee80211_sta.c
Quoted 3 lines
net/mac80211/ieee80211_sta.c +3 -0
+
+	if (unlikely(local->sta_sw_scanning)) {
+		/* drop all the other packets during a software scan
anyway */
net/mac80211/ieee80211_sta.c
Quoted 3 lines
net/mac80211/ieee80211_sta.c +3 -0
+		if (ieee80211_sta_rx_scan(rx->dev, skb, rx->u.rx.status)
+		    != TXRX_QUEUED)
+			dev_kfree_skb(skb);
Not entirely sure why we do this, but nothing we should change with this patch.
Other than the ioctl thing it looks good to me.
johannes
-
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

mohamed salim abbas

2007-11-21 06:19 UTC #7 lore raw

Acually it is better to bock any managemnt frame task while scanning, in this function we might end up calling the driver to tune to a channel, which will cause to stop scan. I think it is better to keep the check and return if we are doing any scan. I guess this condition is added to stabilize the mac80211 and it will do the same for iwl driver with HW scan.

Quoted 13 lines

Quoted 13 lines
Review void ieee80211_sta_work(struct work_struct
Quoted 2 lines
(quoted diff) +0 -0
@@ -1985,7 +2003,7 @@ void ieee80211_sta_work(struct work_struct
*work)
net/mac80211/ieee80211_sta.c void ieee80211_sta_work(struct work_struct *work) · line 2007
Quoted 6 lines
net/mac80211/ieee80211_sta.c +1 -1
    if (!netif_running(dev))
            return;
 
-   if (local->sta_scanning)
+   if (local->sta_sw_scanning || local->sta_hw_scanning)
            return;
Shouldn't the sta work be able to run normally while hw scan is in progress?

Johannes Berg

maintainer

2007-11-21 15:09 UTC #8 lore raw

In reply to Dan Williams

Quoted 8 lines · Dan Williams

Quoted 7 lines
Johannes
The firmware will go back to the old setting, the ones before scan, once done scanning by it self. The calls by mac80211 after scan complete to set channel will be ignored by the iwl driver if we still tuned to the same channel, iwl wont call firmware on the same setting, it has to be a change to call down to firmware. I need to check what happen if user change channel during scan, this call might be necessary in this case.

Quoted 6 lines

I'd expect that either (a) the scan gets canceled and then the channel change happens, or (b) the channel change gets queued and happens immediately after the scan completes. (b) is probably the easier thing to do here unless the ucode has a cancel-scan command, but then you have to wait until the scan is actually canceled anyway, which sort of implies (b) too. So maybe just do (b) and be done with it :)

mac80211 software scanning will do (b) so I expect the driver to handle the configuration callback and queue the channel switch to after the ucode finishes scanning.

johannes