Skip to contentSkip to conversation

ath9k's regulatory domain code changes (for ar9170)

RFC In review Christian Lamparter 5 participants 26 messages Mar 24, 2009 – Mar 30, 2009 Networking drivers (wireless)

lore
Share & tools
lore
#0 lore raw

RFC 1/1 ath9k's regulatory domain code changes (for ar9170)

Hi Luis,

This is more or less what I need from ath9k's regulatory domain code. What's your opinion? Would you accept the changes, or do you see a potential conflict/problem with the design?

Regards,
Chr
---

drivers/net/wireless/ath9k/hw.c +3 -3
index 9818945..334a20f 100644
@@ -1366,7 +1366,7 @@ static int ath9k_hw_process_ini(struct ath_hw *ah,
 		ath9k_olc_init(ah);
 
 	status = ah->eep_ops->set_txpower(ah, chan,
-				  ath9k_regd_get_ctl(ah, chan),
+				  ath9k_regd_get_ctl(&ah->regulatory, chan),
 				  channel->max_antenna_gain * 2,
 				  channel->max_power * 2,
 				  min((u32) MAX_RATE_POWER,
@@ -1706,7 +1706,7 @@ static bool ath9k_hw_channel_change(struct ath_hw *ah,
 	}
 
 	if (ah->eep_ops->set_txpower(ah, chan,
-			     ath9k_regd_get_ctl(ah, chan),
+			     ath9k_regd_get_ctl(&ah->regulatory, chan),
 			     channel->max_antenna_gain * 2,
 			     channel->max_power * 2,
 			     min((u32) MAX_RATE_POWER,
@@ -3768,7 +3768,7 @@ bool ath9k_hw_set_txpowerlimit(struct ath_hw *ah, u32 limit)
 	ah->regulatory.power_limit = min(limit, (u32) MAX_RATE_POWER);
 
 	if (ah->eep_ops->set_txpower(ah, chan,
-			     ath9k_regd_get_ctl(ah, chan),
+			     ath9k_regd_get_ctl(&ah->regulatory, chan),
 			     channel->max_antenna_gain * 2,
 			     channel->max_power * 2,
 			     min((u32) MAX_RATE_POWER,
drivers/net/wireless/ath9k/main.c +15 -4
index c13e4e5..8db019f 100644
@@ -1362,6 +1362,16 @@ void ath_detach(struct ath_softc *sc)
 	ath9k_ps_restore(sc);
 }
 
+struct ath9k_regulatory *ath9k_reg_get_from_wiphy(struct wiphy *wiphy)
+{
+	struct ieee80211_hw *hw = wiphy_to_ieee80211_hw(wiphy);
+	struct ath_wiphy *aphy = hw->priv;
+	struct ath_softc *sc = aphy->sc;
+	struct ath_hw *ah = sc->sc_ah;
+
+	return &ah->regulatory;
+}
+
 static int ath_init(u16 devid, struct ath_softc *sc)
 {
 	struct ath_hw *ah = NULL;
@@ -1416,7 +1426,8 @@ static int ath_init(u16 devid, struct ath_softc *sc)
 	for (i = 0; i < sc->keymax; i++)
 		ath9k_hw_keyreset(ah, (u16) i);
 
-	if (ath9k_regd_init(sc->sc_ah))
+	sc->sc_ah->regulatory.debug = sc;
+	if (ath9k_regd_init(&sc->sc_ah->regulatory))
 		goto bad;
 
 	/* default to MONITOR mode */
@@ -1666,10 +1677,10 @@ int ath_attach(u16 devid, struct ath_softc *sc)
 		goto error_attach;
 #endif
 
-	if (ath9k_is_world_regd(sc->sc_ah)) {
+	if (ath9k_is_world_regd(&sc->sc_ah->regulatory)) {
 		/* Anything applied here (prior to wiphy registration) gets
 		 * saved on the wiphy orig_* parameters */
-		regd = ath9k_world_regdomain(sc->sc_ah);
+		regd = ath9k_world_regdomain(&sc->sc_ah->regulatory);
 		hw->wiphy->custom_regulatory = true;
 		hw->wiphy->strict_regulatory = false;
 	} else {
@@ -1688,7 +1699,7 @@ int ath_attach(u16 devid, struct ath_softc *sc)
 
 	error = ieee80211_register_hw(hw);
 
-	if (!ath9k_is_world_regd(sc->sc_ah)) {
+	if (!ath9k_is_world_regd(&sc->sc_ah->regulatory)) {
 		error = regulatory_hint(hw->wiphy,
 			sc->sc_ah->regulatory.alpha2);
 		if (error)
drivers/net/wireless/ath9k/regd.c +43 -48

193 lines not on this page · raw message

drivers/net/wireless/ath9k/regd.h +9 -6
index 9f5fbd4..dc23883 100644
@@ -49,6 +49,7 @@ struct ath9k_regulatory {
 	u16 current_rd_ext;
 	int16_t power_limit;
 	struct reg_dmn_pair_mapping *regpair;
+	void *debug;
 };
 
 enum CountryCode {
@@ -233,15 +234,17 @@ enum CountryCode {
 	CTRY_BELGIUM2 = 5002
 };
 
-bool ath9k_is_world_regd(struct ath_hw *ah);
-const struct ieee80211_regdomain *ath9k_world_regdomain(struct ath_hw *ah);
+struct ath9k_channel;
+
+bool ath9k_is_world_regd(struct ath9k_regulatory *reg);
+const struct ieee80211_regdomain *ath9k_world_regdomain(struct ath9k_regulatory *reg);
 const struct ieee80211_regdomain *ath9k_default_world_regdomain(void);
 void ath9k_reg_apply_world_flags(struct wiphy *wiphy,
 				 enum nl80211_reg_initiator initiator);
 void ath9k_reg_apply_radar_flags(struct wiphy *wiphy);
-int ath9k_regd_init(struct ath_hw *ah);
-bool ath9k_regd_is_eeprom_valid(struct ath_hw *ah);
-u32 ath9k_regd_get_ctl(struct ath_hw *ah, struct ath9k_channel *chan);
+int ath9k_regd_init(struct ath9k_regulatory *reg);
+bool ath9k_regd_is_eeprom_valid(struct ath9k_regulatory *reg);
+u32 ath9k_regd_get_ctl(struct ath9k_regulatory *reg, struct ath9k_channel *chan);
 int ath9k_reg_notifier(struct wiphy *wiphy, struct regulatory_request *request);
-
+struct ath9k_regulatory *ath9k_reg_get_from_wiphy(struct wiphy *wiphy);
 #endif

---

of course, here are the ar9170 changes:
---

drivers/net/wireless/ar9170/Makefile +1 -1
index 8d91c7e..66e89ce 100644
@@ -1,3 +1,3 @@
-ar9170usb-objs := usb.o main.o cmd.o mac.o phy.o led.o
+ar9170usb-objs := usb.o ../ath9k/regd.o main.o cmd.o mac.o phy.o led.o
 
 obj-$(CONFIG_AR9170_USB) += ar9170usb.o
drivers/net/wireless/ar9170/ar9170.h +6 -0
index f4fb2e9..49d4995 100644
@@ -48,6 +48,9 @@
 #include "eeprom.h"
 #include "hw.h"
 
+/* ath9k definitions */
+#include "../ath9k/regd.h"
+
 #define PAYLOAD_MAX	(AR9170_MAX_CMD_LEN/4 - 1)
 
 enum ar9170_bw {
@@ -156,6 +159,9 @@ struct ar9170 {
 	struct sk_buff_head global_tx_status;
 	struct sk_buff_head global_tx_status_waste;
 	struct delayed_work tx_status_janitor;
+
+	/* regulatory domain */
+	struct ath9k_regulatory regulatory;
 };
 
 struct ar9170_sta_info {
drivers/net/wireless/ar9170/main.c +69 -6

115 lines not on this page · raw message

---

#1 lore raw
Quoted 5 lines · Christian Lamparter

Hi Luis,

This is more or less what I need from ath9k's regulatory domain code. What's your opinion? Would you accept the changes, or do you see a potential conflict/problem with the design?

FWIW I had done a small bit of rework for sharing with ath5k, and Luis has those patches now in somewhat better shape. The last patch set I made is here:

http://bobcopeland.com/kernel/wl/20090309/

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#2 lore raw
Quoted 7 lines · Bob Copeland
Quoted 5 lines · Christian Lamparter

Hi Luis,

This is more or less what I need from ath9k's regulatory domain code. What's your opinion? Would you accept the changes, or do you see a potential conflict/problem with the design?

Quoted 5 lines

FWIW I had done a small bit of rework for sharing with ath5k, and Luis has those patches now in somewhat better shape. The last patch set I made is here:

http://bobcopeland.com/kernel/wl/20090309/

I rebased these on top of the latest wl, and slightly modified them to apply cleanly with checkpatch and I think I had a few very minor changes. You can find the latest here:

http://bombadil.infradead.org/~mcgrof/patches/ath5k/ath5k-reg.patch.txt

I thought I was going to be able to work on this last week but that wasn't the case and it seems I won't get to it this week either. Do any of you guys have time to pick it up? Only change I'd recommend is to first make ath5k use a small set of channels as ath9k and ar9170 has, then you can actually debug regulatory stuff using iw by calling 'iw list' and checking the channel info. I was going to start with that but saw its a little pain. Essentially the way I'd do it is to modify ath5k_setup_bands() to remove the channel hw specific settings and move that stuff to a routine which can be called dynamically upon channel change (IEEE80211_CONF_CHANGE_CHANNEL) through the mac80211 config callback. In ath9k we do this with the ath9k_update_ichannel(). The idea was to eventually condense that down to a ath9k_update_ht_mode() as I think that's all that would be required. But do do that we'd need to slowly completley remove ath9k_channel. That's a bit of a challenge, provided you want to make sure everything else works. But anyway that should give you an idea of where I was heading if you want to try something similar for ath5k.

Technically speaking it'd be even nice to not have to compute this upon channel change so that way we save time during a channel change, but that again is a nice challenge. For example it'd be hot of we wouldn't have to compute the hw value for the frequency on ath9k every single time as its a static value.

Anyway that's my feedback.

Luis

#3 lore raw
Quoted 1 line · Luis R. Rodriguez

Do any of you guys have time to pick it up?

Well my patch #2 is pretty similar to what Christian has, and I'd like to upstream the first few patches soon (especially the ath.ko creation, it is a beast to rebase as you saw). That would bring up the first contentious point of the merge: is "ath/ath.ko" ok with everyone? ath5k/9k specific bits can eventually go in ath/{5k,9k} subdirs... not sure where 9170 lives then :)

I probably can't get to it until the weekend, but if you have more time, Christian, feel free to adopt whatever you want from mine. Otherwise I'll pick up where Luis left off on the weekend.

I didn't push it yet because ath5k was having problems once I turned it on for ath5k, and I have no other hardware to test with so wanted to figure out what was broken first. Channel list is a good hint, maybe it's time to fix iw/nl80211 to send back all the channels :)

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#4 lore raw

On Tue, Mar 24, 2009 at 1:03 PM, Bob Copeland <[email protected]> wrot= e:

Quoted 5 lines
Quoted 1 line · Luis R. Rodriguez

Do any of you guys have time to pick it up?

Well my patch #2 is pretty similar to what Christian has, and I'd lik=

e to

Quoted 1 line

upstream the first few patches soon (especially the ath.ko creation, =

it is a

Quoted 1 line

beast to rebase as you saw). =C2=A0That would bring up the first cont=

entious

Quoted 1 line

point of the merge: is "ath/ath.ko" ok with everyone?

Yes.

Quoted 3 lines

ath5k/9k specific
bits can eventually go in ath/{5k,9k} subdirs... not sure where 9170 lives then :)

Wherever Christian wants it. If he finds re-usable bits then great.

Quoted 1 line

I probably can't get to it until the weekend, but if you have more ti=

me,

Quoted 1 line

Christian, feel free to adopt whatever you want from mine. =C2=A0Othe=

rwise I'll

Quoted 3 lines

pick up where Luis left off on the weekend.

I didn't push it yet because ath5k was having problems once I turned =

it

Quoted 1 line

on for ath5k,

So this is why I recommend to trim the channel list down of ath5k. The other reason is you won't have users spending what may be 3/4 of their time scanning on channels that they won't ever find APs in. What can be done here is add flag or command as we had reviewed a while back to allow the user to enable these channels if he knows what he is doing so that way we know we don't mind bothering scanning on the gazillion channels. But that is just my advice. My concern first is to get users working happily and quickly associated to an AP and I think that parallels those goals and puts as secondary the bells and whistles of adding every single possible channel.

Quoted 4 lines

and I have no other hardware to test with so wanted to figure out what was broken first.
Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

So that is the other option. I tried looking at that yesterday but wasn't able to find a way to do this. I went through the netlink documenation and even tried modifying the skb on the kernel side as its a dump. Not sure what to do here to fix this.

So yes, both are possible options. Just keep in mind that without being able to see the channel list with 'iw list' debugging regulatory is a real royal pain in the ass. So I'd try to fix that first either through channel reduction (you get another bonus enhancement for scanning time for users) or through fixing this through iw/nl80211.

Johannes probably can advise best on the second path.

Luis
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

maintainer
#5 lore raw
Quoted 7 lines · Luis R. Rodriguez
Quoted 2 lines

Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

So that is the other option. I tried looking at that yesterday but wasn't able to find a way to do this. I went through the netlink documenation and even tried modifying the skb on the kernel side as its a dump. Not sure what to do here to fix this.

I looked into yesterday actually, and it's completely non-trivial and requires userspace interface changes.

johannes

#6 lore raw
Quoted 5 lines · Johannes Berg
Quoted 3 lines · Luis R. Rodriguez
Quoted 2 lines

Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

Quoted 7 lines
Quoted 4 lines

So that is the other option. I tried looking at that yesterday but wasn't able to find a way to do this. I went through the netlink documenation and even tried modifying the skb on the kernel side as its a dump. Not sure what to do here to fix this.

I looked into yesterday actually, and it's completely non-trivial and requires userspace interface changes.

Heh, there goes that idea then. Then in that case I'd advocate even more trimming the channels down.

Luis

#7 lore raw
Quoted 8 lines · Luis R. Rodriguez
Quoted 5 lines
Quoted 2 lines
Quoted 2 lines

Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

I looked into yesterday actually, and it's completely non-trivial and requires userspace interface changes.

Heh, there goes that idea then. Then in that case I'd advocate even more trimming the channels down.

Nick, what do you think of this? I think it's the best compromise for now - people can still get all the channels, normal users get a nice speedup in the default case.

--------------------

Quoted 1 line

From 7289f60a3acf3bb1cfcc2a7ef7c4d88ffd0e13f1 Mon Sep 17 00:00:00 2001

From: Bob Copeland <[email protected]>
Date: Sat, 28 Mar 2009 12:27:31 -0400
Subject: [PATCH] ath5k: reduce exported channel list

Claiming every available 5 ghz channel has a couple of negative
side-effects: scanning takes a long time, and the channel list overflows the available buffer space for netlink commands, resulting in:

    $ iw phy phy0 info
    command failed: No buffer space available (-105)

This patch adds a modparam so people who want to see all the channels can do so by passing all_channels=1. By default users will see a smaller list of channels that works with iw. This also halves scan time, from 10 seconds down to less than 5.

Changes-licensed-under: 3-Clause-BSD

---
 drivers/net/wireless/ath5k/base.c |   21 +++++++++++++++++++++
 1 files changed, 21 insertions(+), 0 deletions(-)
drivers/net/wireless/ath5k/base.c +21 -0
index 5d57d77..a018106 100644
@@ -64,6 +64,10 @@ static int modparam_nohwcrypt;
 module_param_named(nohwcrypt, modparam_nohwcrypt, int, 0444);
 MODULE_PARM_DESC(nohwcrypt, "Disable hardware encryption.");
 
+static int modparam_all_channels;
+module_param_named(all_channels, modparam_all_channels, int, 0444);
+MODULE_PARM_DESC(all_channels, "Expose all channels the device can use.");
+
 
 /******************\
 * Internal defines *
@@ -862,6 +866,20 @@ ath5k_ieee2mhz(short chan)
 		return 2212 + chan * 20;
 }
 
+/*
+ * Returns true for the channel numbers user without all_channels modparam.
+ */
+static bool ath5k_is_standard_channel(short chan)
+{
+	return ((chan <= 14) ||
+		/* UNII 1,2 */
+		((chan & 3) == 0 && chan >= 36 && chan <= 64) ||
+		/* midband */
+		((chan & 3) == 0 && chan >= 100 && chan <= 140) ||
+		/* UNII-3 */
+		((chan & 3) == 1 && chan >= 149 && chan <= 165));
+}
+
 static unsigned int
 ath5k_copy_channels(struct ath5k_hw *ah,
 		struct ieee80211_channel *channels,
@@ -899,6 +917,9 @@ ath5k_copy_channels(struct ath5k_hw *ah,
 		if (!ath5k_channel_ok(ah, freq, chfreq))
 			continue;
 
+		if (!modparam_all_channels && !ath5k_is_standard_channel(ch))
+			continue;
+
 		/* Write channel info and increment counter */
 		channels[count].center_freq = freq;
 		channels[count].band = (chfreq == CHANNEL_2GHZ) ?
Signature · 5 lines
1.6.0.6


-- 
Bob Copeland %% www.bobcopeland.com
#8 lore raw

2009/3/28 Bob Copeland <[email protected]>:

Quoted 5 lines · Luis R. Rodriguez
Quoted 4 lines
Quoted 4 lines
Quoted 2 lines
Quoted 2 lines

Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

I looked into yesterday actually, and it's completely non-trivial =

and

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

requires userspace interface changes.

Heh, there goes that idea then. Then in that case I'd advocate even more trimming the channels down.

Nick, what do you think of this? =C2=A0I think it's the best compromi=

se

Quoted 5 lines

for now - people can still get all the channels, normal users get a nice speedup in the default case.

--------------------
From 7289f60a3acf3bb1cfcc2a7ef7c4d88ffd0e13f1 Mon Sep 17 00:00:00 200=

1

Quoted 9 lines

From: Bob Copeland <[email protected]>
Date: Sat, 28 Mar 2009 12:27:31 -0400
Subject: [PATCH] ath5k: reduce exported channel list

Claiming every available 5 ghz channel has a couple of negative
side-effects: scanning takes a long time, and the channel list overflows the available buffer space for netlink commands, resulting in:

Quoted 5 lines

=C2=A0 =C2=A0$ iw phy phy0 info
=C2=A0 =C2=A0command failed: No buffer space available (-105)

This patch adds a modparam so people who want to see all the channels can do so by passing all_channels=3D1. =C2=A0By default users will se=

e a

Quoted 1 line

smaller list of channels that works with iw. =C2=A0This also halves s=

can

Quoted 7 lines

time, from 10 seconds down to less than 5.

Changes-licensed-under: 3-Clause-BSD

---
=C2=A0drivers/net/wireless/ath5k/base.c | =C2=A0 21 +++++++++++++++++=

++++

drivers/net/wireless=

Quoted 3 lines

=C2=A01 files changed, 21 insertions(+), 0 deletions(-)

drivers/net/wireless= renamed from drivers/net/wireless/ath5k/base.c +0 -0

/ath5k/base.c

drivers/net/wireless/ath5k/base.c static int modparam_nohwcrypt; · line 69

Quoted 10 lines

index 5d57d77..a018106 100644

drivers/net/wireless/ath5k/base.c +3 -0
@@ -64,6 +64,10 @@ static int modparam_nohwcrypt;
 =C2=A0module_param_named(nohwcrypt, modparam_nohwcrypt, int, 0444);
 =C2=A0MODULE_PARM_DESC(nohwcrypt, "Disable hardware encryption.");
 
+static int modparam_all_channels;
+module_param_named(all_channels, modparam_all_channels, int, 0444);
+MODULE_PARM_DESC(all_channels, "Expose all channels the device can u=

se.");

Quoted 6 lines
drivers/net/wireless/ath5k/base.c +1 -0
+
 
 =C2=A0/******************\

=C2=A0* Internal defines *

drivers/net/wireless/ath5k/base.c +0 -0
@@ -862,6 +866,20 @@ ath5k_ieee2mhz(short chan)

=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0return 2212 + =

chan * 20;

Quoted 4 lines

=C2=A0}

+/*
+ * Returns true for the channel numbers user without all_channels mo=

dparam.

drivers/net/wireless/ath5k/base.c

Quoted 6 lines

+ */

drivers/net/wireless/ath5k/base.c +5 -0
+static bool ath5k_is_standard_channel(short chan)
+{
+ =C2=A0 =C2=A0 =C2=A0 return ((chan <=3D 14) ||
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 /* UNII 1,2 */
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ((chan & 3) =3D=3D=

0 && chan >=3D 36 && chan <=3D 64) ||

Quoted 2 lines

+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 /* midband */
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ((chan & 3) =3D=3D=

0 && chan >=3D 100 && chan <=3D 140) ||

Quoted 2 lines

+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 /* UNII-3 */
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ((chan & 3) =3D=3D=

1 && chan >=3D 149 && chan <=3D 165));

Quoted 5 lines
drivers/net/wireless/ath5k/base.c +2 -0
+}
+

=C2=A0static unsigned int
=C2=A0ath5k_copy_channels(struct ath5k_hw *ah,
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0struct ieee802=

11_channel *channels,

Quoted 2 lines
drivers/net/wireless/ath5k/base.c +0 -0
@@ -899,6 +917,9 @@ ath5k_copy_channels(struct ath5k_hw *ah,

=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0if (!ath5k_cha=

nnel_ok(ah, freq, chfreq))

Quoted 1 line

=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=

=C2=A0 =C2=A0continue;

Quoted 2 lines

+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 if (!modparam_all_=

channels && !ath5k_is_standard_channel(ch))

Quoted 1 line

+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=

=C2=A0 continue;

Quoted 2 lines

+
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0/* Write chann=

el info and increment counter */

Quoted 1 line

=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0channels[count=

].center_freq =3D freq;

Quoted 1 line

=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0channels[count=

].band =3D (chfreq =3D=3D CHANNEL_2GHZ) ?

Seems fine ;-)

--=20
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-) Nick
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#9 lore raw

2009/3/30 Nick Kossifidis <[email protected]>:

Quoted 6 lines

2009/3/28 Bob Copeland <[email protected]>:

Quoted 5 lines · Luis R. Rodriguez
Quoted 4 lines
Quoted 4 lines
Quoted 2 lines
Quoted 2 lines

Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

I looked into yesterday actually, and it's completely non-trivial=

and

Quoted 6 lines
Quoted 6 lines
Quoted 4 lines
Quoted 1 line

requires userspace interface changes.

Heh, there goes that idea then. Then in that case I'd advocate even more trimming the channels down.

Nick, what do you think of this? =C2=A0I think it's the best comprom=

ise

Quoted 3 lines
Quoted 3 lines

for now - people can still get all the channels, normal users get a nice speedup in the default case.

By the way it would be nice if you could also clean up CHANNEL_DEBUG stuff from ath5k.h etc ;-)

--=20
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-) Nick
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#10 lore raw
Quoted 7 lines · Nick Kossifidis
Quoted 4 lines
Quoted 4 lines

Nick, what do you think of this? ??I think it's the best compromise for now - people can still get all the channels, normal users get a nice speedup in the default case.

By the way it would be nice if you could also clean up CHANNEL_DEBUG stuff from ath5k.h etc ;-)

Ah, sure, will do that in a followup (also ATH_CHAN_MAX & friends).

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#11 lore raw
In reply to Luis R. Rodriguez
Quoted 11 lines · On Tuesday 24 March 2009 21:10:21 Luis R. Rodriguez wrote:
Quoted 8 lines · Bob Copeland
Quoted 1 line · Luis R. Rodriguez

Do any of you guys have time to pick it up?

Well my patch #2 is pretty similar to what Christian has, and I'd like to upstream the first few patches soon (especially the ath.ko creation, it is a beast to rebase as you saw). That would bring up the first contentious point of the merge: is "ath/ath.ko" ok with everyone?

Yes.

Yes.

Quoted 5 lines
Quoted 3 lines

ath5k/9k specific
bits can eventually go in ath/{5k,9k} subdirs... not sure where 9170 lives then :)

Wherever Christian wants it. If he finds re-usable bits then great.

maybe ath/9kusb ?

Quoted 3 lines
Quoted 3 lines

I probably can't get to it until the weekend, but if you have more time, Christian, feel free to adopt whatever you want from mine. Otherwise I'll pick up where Luis left off on the weekend.

Done! Patch attached, please add to the tree ;)

Quoted 30 lines
Quoted 2 lines

I didn't push it yet because ath5k was having problems once I turned it on for ath5k,

So this is why I recommend to trim the channel list down of ath5k. The other reason is you won't have users spending what may be 3/4 of their time scanning on channels that they won't ever find APs in. What can be done here is add flag or command as we had reviewed a while back to allow the user to enable these channels if he knows what he is doing so that way we know we don't mind bothering scanning on the gazillion channels. But that is just my advice. My concern first is to get users working happily and quickly associated to an AP and I think that parallels those goals and puts as secondary the bells and whistles of adding every single possible channel.

Quoted 4 lines

and I have no other hardware to test with so wanted to figure out what was broken first.
Channel list is a good hint, maybe
it's time to fix iw/nl80211 to send back all the channels :)

So that is the other option. I tried looking at that yesterday but wasn't able to find a way to do this. I went through the netlink documenation and even tried modifying the skb on the kernel side as its a dump. Not sure what to do here to fix this.

So yes, both are possible options. Just keep in mind that without being able to see the channel list with 'iw list' debugging regulatory is a real royal pain in the ass. So I'd try to fix that first either through channel reduction (you get another bonus enhancement for scanning time for users) or through fixing this through iw/nl80211.

Quoted 3 lines

Johannes probably can advise best on the second path.

Luis

---
From: Christian Lamparter <[email protected]>
Date: Tue, 24 Mar 2009 21:24:14 +0100
Subject: [PATCH] ar9170: use regulatory infrastructure
To: [email protected]

Make ar9170 select the ath module and add in the hooks to make the eeprom regulatory hint and reg notifier take effect.

---

drivers/net/wireless/ar9170/Kconfig +1 -0
index de4281f..b99e326 100644
@@ -2,6 +2,7 @@ config AR9170_USB
 	tristate "Atheros AR9170 802.11n USB support"
 	depends on USB && MAC80211 && WLAN_80211 && EXPERIMENTAL
 	select FW_LOADER
+	select ATH_COMMON
 	help
 	  This is a driver for the Atheros "otus" 802.11n USB devices.
 
drivers/net/wireless/ar9170/ar9170.h +3 -0
index f4fb2e9..87c1985 100644
@@ -48,6 +48,8 @@
 #include "eeprom.h"
 #include "hw.h"
 
+#include "../ath/regd.h"
+
 #define PAYLOAD_MAX	(AR9170_MAX_CMD_LEN/4 - 1)
 
 enum ar9170_bw {
@@ -151,6 +153,7 @@ struct ar9170 {
 
 	/* EEPROM */
 	struct ar9170_eeprom eeprom;
+	struct ath_regulatory regulatory;
 
 	/* global tx status for unregistered Stations. */
 	struct sk_buff_head global_tx_status;
drivers/net/wireless/ar9170/main.c +18 -0
index 5996ff9..4a251fc 100644
@@ -1619,12 +1619,24 @@ static int ar9170_read_eeprom(struct ar9170 *ar)
 	else
 		ar->hw->channel_change_time = 80 * 1000;
 
+	ar->regulatory.current_rd = le16_to_cpu(ar->eeprom.reg_domain[0]);
+	ar->regulatory.current_rd_ext = le16_to_cpu(ar->eeprom.reg_domain[1]);
+
 	/* second part of wiphy init */
 	SET_IEEE80211_PERM_ADDR(ar->hw, addr);
 
 	return bands ? 0 : -EINVAL;
 }
 
+int ar9170_reg_notifier(struct wiphy *wiphy,
+			struct regulatory_request *request)
+{
+	struct ieee80211_hw *hw = wiphy_to_ieee80211_hw(wiphy);
+	struct ar9170 *ar = hw->priv;
+
+	return ath_reg_notifier_apply(wiphy, request, &ar->regulatory);
+}
+
 int ar9170_register(struct ar9170 *ar, struct device *pdev)
 {
 	int err;
@@ -1634,10 +1646,16 @@ int ar9170_register(struct ar9170 *ar, struct device *pdev)
 	if (err)
 		goto err_out;
 
+	err = ath_regd_init(&ar->regulatory, ar->hw->wiphy,
+			    ar9170_reg_notifier);
+
 	err = ieee80211_register_hw(ar->hw);
 	if (err)
 		goto err_out;
 
+	if (!ath_is_world_regd(&ar->regulatory))
+		regulatory_hint(ar->hw->wiphy, ar->regulatory.alpha2);
+
 	err = ar9170_init_leds(ar);
 	if (err)
 		goto err_unreg;
#12 lore raw
Quoted 12 lines · Christian Lamparter
Quoted 11 lines · On Tuesday 24 March 2009 21:10:21 Luis R. Rodriguez wrote:
Quoted 8 lines · Bob Copeland
Quoted 1 line · Luis R. Rodriguez

Do any of you guys have time to pick it up?

Well my patch #2 is pretty similar to what Christian has, and I'd like to upstream the first few patches soon (especially the ath.ko creation, it is a beast to rebase as you saw). That would bring up the first contentious point of the merge: is "ath/ath.ko" ok with everyone?

Yes.

Quoted 8 lines

Yes.

Quoted 5 lines
Quoted 3 lines

ath5k/9k specific
bits can eventually go in ath/{5k,9k} subdirs... not sure where 9170 lives then :)

Wherever Christian wants it. If he finds re-usable bits then great.

maybe ath/9kusb ?

In that case maybe 9170/, we do have future USB plans.

Quoted 4 lines
Quoted 3 lines
Quoted 3 lines

I probably can't get to it until the weekend, but if you have more time, Christian, feel free to adopt whatever you want from mine. Otherwise I'll pick up where Luis left off on the weekend.

Done! Patch attached, please add to the tree ;)

Nice....... so it seems all we need now is ath5k figured out and properly tested.

Luis

#13 lore raw

On Tue, 24 Mar 2009 13:58:47 -0700, Luis R. Rodriguez wrote

Quoted 5 lines
Quoted 2 lines
Quoted 1 line
Quoted 1 line

pick up where Luis left off on the weekend.

Done! Patch attached, please add to the tree ;)

Nice....... so it seems all we need now is ath5k figured out and properly tested.

BTW there's a dumb bug in my first patch, missing 'break' in the switch statements. Oops. But ath5k wasn't using any of that since it didn't do the CTL stuff yet.

I'll go ahead and rebase the stuff and repost it later in the week, as long as you guys can test ath9k/ar9170 then I can do the debugging for 5k..

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#14 lore raw
Quoted 7 lines · Bob Copeland

On Tue, 24 Mar 2009 13:58:47 -0700, Luis R. Rodriguez wrote

Quoted 5 lines
Quoted 2 lines
Quoted 1 line
Quoted 1 line

pick up where Luis left off on the weekend.

Done! Patch attached, please add to the tree ;)

Nice....... so it seems all we need now is ath5k figured out and properly tested.

Quoted 7 lines

BTW there's a dumb bug in my first patch, missing 'break' in the switch statements. Oops. But ath5k wasn't using any of that since it didn't do the CTL stuff yet.

I'll go ahead and rebase the stuff and repost it later in the week, as long as you guys can test ath9k/ar9170 then I can do the debugging for 5k..

Nice, I think I can at least commit now to reviewing and testing ath9k stuff. The good news is this is all going for a .31 so we'll have a good amount of time to test.

Luis

#15 lore raw
In reply to Luis R. Rodriguez
Quoted 8 lines · On Tuesday 24 March 2009 21:58:47 Luis R. Rodriguez wrote:
Quoted 7 lines · Christian Lamparter
Quoted 6 lines · On Tuesday 24 March 2009 21:10:21 Luis R. Rodriguez wrote:
Quoted 4 lines · Bob Copeland

On Tue, Mar 24, 2009 at 2:41 PM, Luis R. Rodriguez <[email protected]> wrote:
ath5k/9k specific
bits can eventually go in ath/{5k,9k} subdirs... not sure where 9170 lives then :)

Quoted 4 lines
Quoted 2 lines
Quoted 1 line

Wherever Christian wants it. If he finds re-usable bits then great.

maybe ath/9kusb ?

In that case maybe 9170/, we do have future USB plans.

alright, ath/9170/ it is!

Quoted 7 lines
Quoted 4 lines
Quoted 3 lines
Quoted 3 lines

I probably can't get to it until the weekend, but if you have more time, Christian, feel free to adopt whatever you want from mine. Otherwise I'll pick up where Luis left off on the weekend.

Done! Patch attached, please add to the tree ;)

Nice....... so it seems all we need now is ath5k figured out and properly tested.

uhh, looks like ath5k/regd.c got dropped by accident?!

Regards,
Chr

#16 lore raw

On Tue, 24 Mar 2009 23:24:33 +0100, Christian Lamparter wrote

Quoted 3 lines
Quoted 2 lines

Nice....... so it seems all we need now is ath5k figured out and properly tested.

uhh, looks like ath5k/regd.c got dropped by accident?!

Well, it wasn't working when I had it either :)

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#17 lore raw
Quoted 7 lines · Bob Copeland

On Tue, 24 Mar 2009 23:24:33 +0100, Christian Lamparter wrote

Quoted 3 lines
Quoted 2 lines

Nice....... so it seems all we need now is ath5k figured out and properly tested.

uhh, looks like ath5k/regd.c got dropped by accident?!

Well, it wasn't working when I had it either :)

My brain is fuzzy here but I think I had decided to not have an ath5k/regd.c. Hm but I see the file there now in my tree.

Anyway we should get a newer shinier patch which has all these changes so far + ar9170 changes and test them. Whoever gets to it first please post one.

Luis

#18 lore raw
Quoted 11 lines · Luis R. Rodriguez
Quoted 7 lines · Bob Copeland

On Tue, 24 Mar 2009 23:24:33 +0100, Christian Lamparter wrote

Quoted 3 lines
Quoted 2 lines

Nice....... so it seems all we need now is ath5k figured out and properly tested.

uhh, looks like ath5k/regd.c got dropped by accident?!

Well, it wasn't working when I had it either :)

My brain is fuzzy here but I think I had decided to not have an ath5k/regd.c. Hm but I see the file there now in my tree.

Ah yes, brain_works_still++, the file was not in the commit log and I did move things into the driver elsewhere.

Luis

#19 lore raw
Quoted 9 lines · Luis R. Rodriguez
Quoted 7 lines · Luis R. Rodriguez
Quoted 6 lines · Bob Copeland

On Tue, 24 Mar 2009 23:24:33 +0100, Christian Lamparter wrote

Quoted 3 lines
Quoted 2 lines

Nice....... so it seems all we need now is ath5k figured out and properly tested.

uhh, looks like ath5k/regd.c got dropped by accident?!

Quoted 7 lines
Quoted 4 lines
Quoted 1 line

Well, it wasn't working when I had it either :)

My brain is fuzzy here but I think I had decided to not have an ath5k/regd.c. Hm but I see the file there now in my tree.

Ah yes, brain_works_still++, the file was not in the commit log and I did move things into the driver elsewhere.

Alright here's a quick concat of all these patches (also one small sparse fix on Christian's last patch). It also now throws ath5k, ath9k and ar9170 into ath/. We can rename the drivers after, was lazy to do that. Think its easier for review to separate that too.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v3.patch.txt

Bob, where was that switch break error?

Luis

#20 lore raw
Quoted 8 lines · Luis R. Rodriguez

Alright here's a quick concat of all these patches (also one small sparse fix on Christian's last patch). It also now throws ath5k, ath9k and ar9170 into ath/. We can rename the drivers after, was lazy to do that. Think its easier for review to separate that too.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v3.patch.txt

Bob, where was that switch break error?

1st patch:

Review

Quoted 10 lines
(quoted diff) +10 -0
+	switch (band) {
+	case IEEE80211_BAND_2GHZ:
+		ctl = ah->regulatory.regpair->reg_2ghz_ctl;
+	case IEEE80211_BAND_5GHZ:
+		ctl = ah->regulatory.regpair->reg_5ghz_ctl;
+	default:
+		ctl = NO_CTL;
+	}
+	return ctl;
+}

I think I had it like the following and changed it -- actually with the returns it's probably better...

switch (band) {
case XXX:

        return ah->...;

case YYY:
...
}

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#21 lore raw

On Tue, Mar 24, 2009 at 6:06 PM, Bob Copeland <[email protected]> wrot= e:

Quoted 3 lines · Luis R. Rodriguez
Quoted 2 lines

Alright here's a quick concat of all these patches (also one small sparse fix on Christian's last patch). It also now throws ath5k, ath=

9k

Quoted 1 line
Quoted 1 line

and ar9170 into ath/. We can rename the drivers after, was lazy to d=

o

Quoted 3 lines
Quoted 3 lines

that. Think its easier for review to separate that too.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v3.patc=

h.txt

Quoted 8 lines
Quoted 2 lines

Bob, where was that switch break error?

1st patch:

Quoted 3 lines

+ =C2=A0 =C2=A0 switch (band) {
+ =C2=A0 =C2=A0 case IEEE80211_BAND_2GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulatory.r=

egpair->reg_2ghz_ctl;

Quoted 2 lines
Quoted 2 lines

+ =C2=A0 =C2=A0 case IEEE80211_BAND_5GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulatory.r=

egpair->reg_5ghz_ctl;

Quoted 9 lines
Quoted 5 lines

+ =C2=A0 =C2=A0 default:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D NO_CTL;
+ =C2=A0 =C2=A0 }
+ =C2=A0 =C2=A0 return ctl;
+}

I think I had it like the following and changed it -- actually with the returns it's probably better...

Quoted 6 lines

switch (band) {
case XXX:
=C2=A0 =C2=A0 =C2=A0 =C2=A0return ah->...;
case YYY:
...
}

Alright thanks, I've fixed that here, it also now does not introduce any new checkpatch or sparse complaints.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v4.patch.t=
xt

My shiny new wl rebased on 29 kernel is compiling, will let you know in the morning how it goes.

Luis
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#22 lore raw
Quoted 1 line · Luis R. Rodriguez

On Tue, Mar 24, 2009 at 6:06 PM, Bob Copeland <[email protected]> wr=

ote:

Quoted 3 lines
Quoted 3 lines · Luis R. Rodriguez
Quoted 2 lines

Alright here's a quick concat of all these patches (also one small sparse fix on Christian's last patch). It also now throws ath5k, at=

h9k

Quoted 1 line
Quoted 1 line
Quoted 1 line

and ar9170 into ath/. We can rename the drivers after, was lazy to =

do

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

that. Think its easier for review to separate that too.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v3.pat=

ch.txt

Quoted 8 lines
Quoted 8 lines
Quoted 2 lines

Bob, where was that switch break error?

1st patch:

Quoted 3 lines

+ =C2=A0 =C2=A0 switch (band) {
+ =C2=A0 =C2=A0 case IEEE80211_BAND_2GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulatory.=

regpair->reg_2ghz_ctl;

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

+ =C2=A0 =C2=A0 case IEEE80211_BAND_5GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulatory.=

regpair->reg_5ghz_ctl;

Quoted 16 lines
Quoted 9 lines
Quoted 5 lines

+ =C2=A0 =C2=A0 default:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D NO_CTL;
+ =C2=A0 =C2=A0 }
+ =C2=A0 =C2=A0 return ctl;
+}

I think I had it like the following and changed it -- actually with the returns it's probably better...

Quoted 6 lines

switch (band) {
case XXX:
=C2=A0 =C2=A0 =C2=A0 =C2=A0return ah->...;
case YYY:
...
}

Quoted 4 lines

Alright thanks, I've fixed that here, it also now does not introduce any new checkpatch or sparse complaints.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v4.patch=

=2Etxt

Quoted 3 lines

My shiny new wl rebased on 29 kernel is compiling, will let you know in the morning how it goes.

Ah poo, ath5k_reg_notifier() got lost.

Luis
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#23 lore raw
Quoted 3 lines · Luis R. Rodriguez
Quoted 1 line · Luis R. Rodriguez

On Tue, Mar 24, 2009 at 6:06 PM, Bob Copeland <[email protected]> w=

rote:

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines · Luis R. Rodriguez
Quoted 2 lines

Alright here's a quick concat of all these patches (also one small sparse fix on Christian's last patch). It also now throws ath5k, a=

th9k

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

and ar9170 into ath/. We can rename the drivers after, was lazy to=

do

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

that. Think its easier for review to separate that too.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v3.pa=

tch.txt

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 2 lines

Bob, where was that switch break error?

1st patch:

Quoted 3 lines

+ =C2=A0 =C2=A0 switch (band) {
+ =C2=A0 =C2=A0 case IEEE80211_BAND_2GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulatory=

=2Eregpair->reg_2ghz_ctl;

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

+ =C2=A0 =C2=A0 case IEEE80211_BAND_5GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulatory=

=2Eregpair->reg_5ghz_ctl;

Quoted 16 lines
Quoted 16 lines
Quoted 15 lines
Quoted 5 lines

+ =C2=A0 =C2=A0 default:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D NO_CTL;
+ =C2=A0 =C2=A0 }
+ =C2=A0 =C2=A0 return ctl;
+}

I think I had it like the following and changed it -- actually with the returns it's probably better...

switch (band) {
case XXX:
=C2=A0 =C2=A0 =C2=A0 =C2=A0return ah->...;
case YYY:
...
}

Quoted 4 lines
Quoted 4 lines

Alright thanks, I've fixed that here, it also now does not introduce any new checkpatch or sparse complaints.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v4.patc=

h.txt

Quoted 5 lines
Quoted 3 lines

My shiny new wl rebased on 29 kernel is compiling, will let you know in the morning how it goes.

Ah poo, ath5k_reg_notifier() got lost.

=46ixed..

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v5.patch.t=
xt

Luis
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#24 lore raw
Quoted 5 lines · Luis R. Rodriguez
Quoted 3 lines · Luis R. Rodriguez
Quoted 1 line · Luis R. Rodriguez

On Tue, Mar 24, 2009 at 6:06 PM, Bob Copeland <[email protected]> =

Quoted 2 lines · wrote:
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines · Luis R. Rodriguez
Quoted 1 line

Alright here's a quick concat of all these patches (also one smal=

l

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

sparse fix on Christian's last patch). It also now throws ath5k, =

ath9k

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

and ar9170 into ath/. We can rename the drivers after, was lazy t=

o do

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

that. Think its easier for review to separate that too.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v3.p=

atch.txt

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 2 lines

Bob, where was that switch break error?

1st patch:

Quoted 3 lines

+ =C2=A0 =C2=A0 switch (band) {
+ =C2=A0 =C2=A0 case IEEE80211_BAND_2GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulator=

y.regpair->reg_2ghz_ctl;

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

+ =C2=A0 =C2=A0 case IEEE80211_BAND_5GHZ:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D ah->regulator=

y.regpair->reg_5ghz_ctl;

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 5 lines

+ =C2=A0 =C2=A0 default:
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ctl =3D NO_CTL;
+ =C2=A0 =C2=A0 }
+ =C2=A0 =C2=A0 return ctl;
+}

I think I had it like the following and changed it -- actually wit=

h

Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 8 lines

the returns it's probably better...

switch (band) {
case XXX:
=C2=A0 =C2=A0 =C2=A0 =C2=A0return ah->...;
case YYY:
...
}

Alright thanks, I've fixed that here, it also now does not introduc=

e

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

any new checkpatch or sparse complaints.

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v4.pat=

ch.txt

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

My shiny new wl rebased on 29 kernel is compiling, will let you kno=

w

Quoted 7 lines
Quoted 3 lines
Quoted 1 line

in the morning how it goes.

Ah poo, ath5k_reg_notifier() got lost.

Fixed..

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v5.patch=

=2Etxt

Nope.. no luck, something got messed up along the way... needs a closer=
look.

Luis
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#25 lore raw
Quoted 5 lines · Luis R. Rodriguez
Quoted 3 lines

Fixed..

http://bombadil.infradead.org/~mcgrof/patches/ath/ath-common-v5.patch.txt

Nope.. no luck, something got messed up along the way... needs a closer look.

So I tested this series and it works fine, with a caveat:

Currently, ath5k exposes tons of 5 GHz channels. In addition to the other problems I mentioned in http://marc.info/?l=linux-wireless&m=123825852320853,
it also essentially breaks with regulatory hints from the driver. This is because because the Atheros world regdomain allows most channels in the 5 GHz range for passive scanning, whereas the default world regdomain disables many of these. Consequently, scanning takes long enough that NM/wpa_supplicant bails.

So, I think the way forward is to include the patch linked above with the rest of the series. Although perhaps the last patch moving ath5k/ath9k etc under the ath/ directory can be put off a while in case people have patches sitting around against the old dir structure.

Anyone who still wants to use all the channels with ath5k can use a custom regd or configure the supplicant to scan the specific channels they are interested in.

Signature · 1 line
Bob Copeland %% www.bobcopeland.com