[PATCH] b43: RF-kill support

[PATCH] b43: RF-kill support

[Michael Buesch]

This adds full support for the RFKILL button and
the RFKILL LED trigger.

Signed-off-by: Michael Buesch <mb@bu3sch.de>

Index: wireless-2.6/drivers/net/wireless/b43/Kconfig
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/Kconfig	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/Kconfig	2007-09-27 20:56:53.000000000 +0200
@@ -64,7 +64,13 @@ config B43_PCMCIA
 # LED support
 config B43_LEDS
 	bool
-	depends on MAC80211_LEDS
+	depends on B43 && MAC80211_LEDS
+	default y
+
+# RFKILL support
+config B43_RFKILL
+	bool
+	depends on B43 && RFKILL
 	default y
 
 config B43_DEBUG
Index: wireless-2.6/drivers/net/wireless/b43/Makefile
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/Makefile	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/Makefile	2007-09-27 20:56:53.000000000 +0200
@@ -5,6 +5,8 @@ b43-y				+= phy.o
 b43-y				+= sysfs.o
 b43-y				+= xmit.o
 b43-y				+= lo.o
+# b43 RFKILL button support
+b43-$(CONFIG_B43_RFKILL)	+= rfkill.o
 # b43 LED support
 b43-$(CONFIG_B43_LEDS)		+= leds.o
 # b43 PCMCIA support
Index: wireless-2.6/drivers/net/wireless/b43/b43.h
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/b43.h	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/b43.h	2007-09-27 20:56:53.000000000 +0200
@@ -10,6 +10,7 @@
 
 #include "debugfs.h"
 #include "leds.h"
+#include "rfkill.h"
 #include "lo.h"
 #include "phy.h"
 
@@ -636,6 +637,9 @@ struct b43_wl {
 	u8 rng_initialized;
 	char rng_name[30 + 1];
 
+	/* The RF-kill button */
+	struct b43_rfkill rfkill;
+
 	/* List of all wireless devices on this chip */
 	struct list_head devlist;
 	u8 nr_devs;
@@ -711,6 +715,7 @@ struct b43_wldev {
 	struct b43_led led_tx;
 	struct b43_led led_rx;
 	struct b43_led led_assoc;
+	struct b43_led led_radio;
 
 	/* Reason code of the last interrupt. */
 	u32 irq_reason;
Index: wireless-2.6/drivers/net/wireless/b43/main.c
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/main.c	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/main.c	2007-09-27 20:56:53.000000000 +0200
@@ -2201,7 +2201,7 @@ static bool b43_is_hw_radio_enabled(stru
 /* This is the opposite of b43_chip_init() */
 static void b43_chip_exit(struct b43_wldev *dev)
 {
-	b43_radio_turn_off(dev);
+	b43_radio_turn_off(dev, 1);
 	b43_leds_exit(dev);
 	b43_gpio_cleanup(dev);
 	/* firmware is released later */
@@ -2312,7 +2312,7 @@ out:
 	return err;
 
 err_radio_off:
-	b43_radio_turn_off(dev);
+	b43_radio_turn_off(dev, 1);
 err_leds_exit:
 	b43_leds_exit(dev);
 	b43_gpio_cleanup(dev);
@@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
 	radio_hw_enable = b43_is_hw_radio_enabled(dev);
 	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
 		dev->radio_hw_enable = radio_hw_enable;
-		b43info(dev->wl, "Radio hardware status changed to %s\n",
-			radio_hw_enable ? "ENABLED" : "DISABLED");
+		b43_rfkill_toggled(dev, radio_hw_enable);
 	}
 }
 
@@ -2893,7 +2892,7 @@ static int b43_dev_config(struct ieee802
 					"Press the button to turn it on.\n");
 			}
 		} else {
-			b43_radio_turn_off(dev);
+			b43_radio_turn_off(dev, 0);
 			b43info(dev->wl, "Radio turned off by software\n");
 		}
 	}
@@ -3355,11 +3354,15 @@ static void b43_wireless_core_exit(struc
 		return;
 	b43_set_status(dev, B43_STAT_UNINIT);
 
+	mutex_unlock(&dev->wl->mutex);
+	b43_rfkill_exit(dev);
+	mutex_lock(&dev->wl->mutex);
+
 	b43_rng_exit(dev->wl);
 	b43_pio_free(dev);
 	b43_dma_free(dev);
 	b43_chip_exit(dev);
-	b43_radio_turn_off(dev);
+	b43_radio_turn_off(dev, 1);
 	b43_switch_analog(dev, 0);
 	if (phy->dyn_tssi_tbl)
 		kfree(phy->tssi2dbm);
@@ -3482,6 +3485,7 @@ static int b43_wireless_core_init(struct
 	memset(wl->bssid, 0, ETH_ALEN);
 	b43_upload_card_macaddress(dev, NULL);
 	b43_security_init(dev);
+	b43_rfkill_init(dev);
 	b43_rng_init(wl);
 
 	b43_set_status(dev, B43_STAT_INITIALIZED);
@@ -3803,7 +3807,7 @@ static int b43_wireless_core_attach(stru
 		wl->current_dev = dev;
 	INIT_WORK(&dev->restart_work, b43_chip_reset);
 
-	b43_radio_turn_off(dev);
+	b43_radio_turn_off(dev, 1);
 	b43_switch_analog(dev, 0);
 	ssb_device_disable(dev->dev, 0);
 	ssb_bus_may_powerdown(bus);
Index: wireless-2.6/drivers/net/wireless/b43/rfkill.c
===================================================================
--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ wireless-2.6/drivers/net/wireless/b43/rfkill.c	2007-09-27 21:04:07.000000000 +0200
@@ -0,0 +1,155 @@
+/*
+
+  Broadcom B43 wireless driver
+  RFKILL support
+
+  Copyright (c) 2007 Michael Buesch <mb@bu3sch.de>
+
+  This program is free software; you can redistribute it and/or modify
+  it under the terms of the GNU General Public License as published by
+  the Free Software Foundation; either version 2 of the License, or
+  (at your option) any later version.
+
+  This program is distributed in the hope that it will be useful,
+  but WITHOUT ANY WARRANTY; without even the implied warranty of
+  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+  GNU General Public License for more details.
+
+  You should have received a copy of the GNU General Public License
+  along with this program; see the file COPYING.  If not, write to
+  the Free Software Foundation, Inc., 51 Franklin Steet, Fifth Floor,
+  Boston, MA 02110-1301, USA.
+
+*/
+
+#include "rfkill.h"
+#include "b43.h"
+
+
+static void b43_notify_rfkill_press(struct work_struct *work)
+{
+	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
+					      notify_work);
+	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
+	struct b43_wldev *dev;
+	enum rfkill_state state;
+
+	mutex_lock(&wl->mutex);
+	dev = wl->current_dev;
+	if (b43_status(dev) < B43_STAT_INITIALIZED) {
+		mutex_unlock(&wl->mutex);
+		return;
+	}
+	if (dev->radio_hw_enable)
+		state = RFKILL_STATE_ON;
+	else
+		state = RFKILL_STATE_OFF;
+	b43info(wl, "Radio hardware status changed to %s\n",
+		dev->radio_hw_enable ? "ENABLED" : "DISABLED");
+	mutex_unlock(&wl->mutex);
+
+	if (rfk->rfkill) {
+		/* Be careful. This calls back into the software toggle routines.
+		 * So we must unlock before calling. */
+		rfkill_switch_all(rfk->rfkill->type, state);
+	}
+}
+
+/* Called when the RFKILL toggled in hardware.
+ * This is called with the mutex locked. */
+void b43_rfkill_toggled(struct b43_wldev *dev, bool on)
+{
+	struct b43_wl *wl = dev->wl;
+
+	B43_WARN_ON(b43_status(dev) < B43_STAT_INITIALIZED);
+	/* Update the RF status asynchronously, as rfkill will
+	 * call back into the software toggle handler.
+	 * This would deadlock if done synchronously. */
+	queue_work(wl->hw->workqueue, &wl->rfkill.notify_work);
+}
+
+/* Called when the RFKILL toggled in software.
+ * This is called without locking. */
+static int b43_rfkill_soft_toggle(void *data, enum rfkill_state state)
+{
+	struct b43_wldev *dev = data;
+	struct b43_wl *wl = dev->wl;
+	int err = 0;
+
+	mutex_lock(&wl->mutex);
+	if (b43_status(dev) < B43_STAT_INITIALIZED)
+		goto out_unlock;
+
+	switch (state) {
+	case RFKILL_STATE_ON:
+		if (!dev->radio_hw_enable) {
+			/* No luck. We can't toggle the hardware RF-kill
+			 * button from software. */
+			err = -EBUSY;
+			goto out_unlock;
+		}
+		if (!dev->phy.radio_on)
+			b43_radio_turn_on(dev);
+		break;
+	case RFKILL_STATE_OFF:
+		if (dev->phy.radio_on)
+			b43_radio_turn_off(dev, 0);
+		break;
+	}
+
+out_unlock:
+	mutex_unlock(&wl->mutex);
+
+	return err;
+}
+
+char * b43_rfkill_led_name(struct b43_wldev *dev)
+{
+	struct b43_wl *wl = dev->wl;
+
+	if (!wl->rfkill.rfkill)
+		return NULL;
+	return rfkill_get_led_name(wl->rfkill.rfkill);
+}
+
+void b43_rfkill_init(struct b43_wldev *dev)
+{
+	struct b43_wl *wl = dev->wl;
+	struct b43_rfkill *rfk = &(wl->rfkill);
+	int err;
+
+	snprintf(rfk->name, sizeof(rfk->name),
+		 "b43-%s", wiphy_name(wl->hw->wiphy));
+	rfk->rfkill = rfkill_allocate(dev->dev->dev, RFKILL_TYPE_WLAN);
+	if (!rfk->rfkill)
+		goto error;
+	rfk->rfkill->name = rfk->name;
+	rfk->rfkill->state = RFKILL_STATE_ON;
+	rfk->rfkill->data = dev;
+	rfk->rfkill->toggle_radio = b43_rfkill_soft_toggle;
+	rfk->rfkill->user_claim_unsupported = 1;
+
+	INIT_WORK(&rfk->notify_work, b43_notify_rfkill_press);
+
+	err = rfkill_register(rfk->rfkill);
+	if (err)
+		goto error;
+
+	return;
+error:
+	b43warn(dev->wl, "Failed to initialize the RF-kill button\n");
+	rfkill_free(rfk->rfkill);
+	rfk->rfkill = NULL;
+}
+
+void b43_rfkill_exit(struct b43_wldev *dev)
+{
+	struct b43_rfkill *rfk = &(dev->wl->rfkill);
+
+	if (!rfk->rfkill)
+		return;
+	cancel_work_sync(&rfk->notify_work);
+	rfkill_unregister(rfk->rfkill);
+	rfkill_free(rfk->rfkill);
+	rfk->rfkill = NULL;
+}
Index: wireless-2.6/drivers/net/wireless/b43/rfkill.h
===================================================================
--- /dev/null	1970-01-01 00:00:00.000000000 +0000
+++ wireless-2.6/drivers/net/wireless/b43/rfkill.h	2007-09-27 20:56:53.000000000 +0200
@@ -0,0 +1,49 @@
+#ifndef B43_RFKILL_H_
+#define B43_RFKILL_H_
+
+struct b43_wldev;
+
+
+#ifdef CONFIG_B43_RFKILL
+
+#include <linux/rfkill.h>
+
+struct b43_rfkill {
+	/* The RFKILL subsystem data structure */
+	struct rfkill *rfkill;
+	/* The unique name of this rfkill switch */
+	char name[32];
+	/* Workqueue for asynchronous notification. */
+	struct work_struct notify_work;
+};
+
+void b43_rfkill_init(struct b43_wldev *dev);
+void b43_rfkill_exit(struct b43_wldev *dev);
+void b43_rfkill_toggled(struct b43_wldev *dev, bool on);
+char * b43_rfkill_led_name(struct b43_wldev *dev);
+
+
+#else /* CONFIG_B43_RFKILL */
+/* No RFKILL support. */
+
+struct b43_rfkill {
+	/* empty */
+};
+
+static inline void b43_rfkill_init(struct b43_wldev *dev)
+{
+}
+static inline void b43_rfkill_exit(struct b43_wldev *dev)
+{
+}
+static inline void b43_rfkill_toggled(struct b43_wldev *dev, bool on)
+{
+}
+static inline char * b43_rfkill_led_name(struct b43_wldev *dev)
+{
+	return NULL;
+}
+
+#endif /* CONFIG_B43_RFKILL */
+
+#endif /* B43_RFKILL_H_ */
Index: wireless-2.6/drivers/net/wireless/b43/leds.c
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/leds.c	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/leds.c	2007-09-27 20:56:53.000000000 +0200
@@ -154,12 +154,16 @@ static void b43_map_led(struct b43_wldev
 				 ieee80211_get_rx_led_name(hw),
 				 led_index, activelow);
 		break;
-	/*FIXME: We need another trigger for the "radio-on" LEDs below.
-	 *       Wiggle that somehow into the rfkill subsystem. */
 	case B43_LED_RADIO_ALL:
 	case B43_LED_RADIO_A:
 	case B43_LED_RADIO_B:
 	case B43_LED_MODE_BG:
+		snprintf(name, sizeof(name),
+			 "b43-%s:radio", wiphy_name(hw->wiphy));
+		b43_register_led(dev, &dev->led_radio, name,
+				 b43_rfkill_led_name(dev),
+				 led_index, activelow);
+		break;
 	case B43_LED_WEIRD:
 	case B43_LED_ASSOC:
 		snprintf(name, sizeof(name),
Index: wireless-2.6/drivers/net/wireless/b43/phy.c
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/phy.c	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/phy.c	2007-09-27 20:56:53.000000000 +0200
@@ -4349,10 +4349,13 @@ void b43_radio_turn_on(struct b43_wldev 
 	phy->radio_on = 1;
 }
 
-void b43_radio_turn_off(struct b43_wldev *dev)
+void b43_radio_turn_off(struct b43_wldev *dev, bool force)
 {
 	struct b43_phy *phy = &dev->phy;
 
+	if (!phy->radio_on && !force)
+		return;
+
 	if (phy->type == B43_PHYTYPE_A) {
 		b43_radio_write16(dev, 0x0004, 0x00FF);
 		b43_radio_write16(dev, 0x0005, 0x00FB);
@@ -4364,9 +4367,11 @@ void b43_radio_turn_off(struct b43_wldev
 
 		rfover = b43_phy_read(dev, B43_PHY_RFOVER);
 		rfoverval = b43_phy_read(dev, B43_PHY_RFOVERVAL);
-		phy->radio_off_context.rfover = rfover;
-		phy->radio_off_context.rfoverval = rfoverval;
-		phy->radio_off_context.valid = 1;
+		if (!force) {
+			phy->radio_off_context.rfover = rfover;
+			phy->radio_off_context.rfoverval = rfoverval;
+			phy->radio_off_context.valid = 1;
+		}
 		b43_phy_write(dev, B43_PHY_RFOVER, rfover | 0x008C);
 		b43_phy_write(dev, B43_PHY_RFOVERVAL, rfoverval & 0xFF73);
 	} else
Index: wireless-2.6/drivers/net/wireless/b43/phy.h
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/b43/phy.h	2007-09-27 20:53:39.000000000 +0200
+++ wireless-2.6/drivers/net/wireless/b43/phy.h	2007-09-27 20:56:53.000000000 +0200
@@ -267,7 +267,7 @@ u16 b43_radio_init2050(struct b43_wldev 
 void b43_radio_init2060(struct b43_wldev *dev);
 
 void b43_radio_turn_on(struct b43_wldev *dev);
-void b43_radio_turn_off(struct b43_wldev *dev);
+void b43_radio_turn_off(struct b43_wldev *dev, bool force);
 
 int b43_radio_selectchannel(struct b43_wldev *dev, u8 channel,
 			    int synthetic_pu_workaround);

Re: [PATCH] b43: RF-kill support

[Ivo van Doorn]

Hi,

> @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
>  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
>  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
>  		dev->radio_hw_enable = radio_hw_enable;
> -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> -			radio_hw_enable ? "ENABLED" : "DISABLED");
> +		b43_rfkill_toggled(dev, radio_hw_enable);

Isn't it better to use the input_polldev for scheduled input device checking?

> +static void b43_notify_rfkill_press(struct work_struct *work)
> +{
> +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> +					      notify_work);
> +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> +	struct b43_wldev *dev;
> +	enum rfkill_state state;

Same here, input_polldev was created especially for hardware
devices that don't trigger interrupts when the button was pressed.

> +	mutex_lock(&wl->mutex);
> +	dev = wl->current_dev;
> +	if (b43_status(dev) < B43_STAT_INITIALIZED) {
> +		mutex_unlock(&wl->mutex);
> +		return;
> +	}
> +	if (dev->radio_hw_enable)
> +		state = RFKILL_STATE_ON;
> +	else
> +		state = RFKILL_STATE_OFF;
> +	b43info(wl, "Radio hardware status changed to %s\n",
> +		dev->radio_hw_enable ? "ENABLED" : "DISABLED");
> +	mutex_unlock(&wl->mutex);
> +
> +	if (rfk->rfkill) {
> +		/* Be careful. This calls back into the software toggle routines.
> +		 * So we must unlock before calling. */
> +		rfkill_switch_all(rfk->rfkill->type, state);
> +	}
> +}

This is incorrect. Instead of rfkill_switch_all the driver must send
a signal over the input device it has registered. This will assure
the event is send to userspace or handled by the kernel when
the user has chosen that option.

I have recently send a patch to the list which added documentation
for rfkill and drivers to the Documentation/ folder.

It basically came down to this:
	- driver allocated rfkill
	- driver allocates input device or input_polldev
	- driver registers rfkill & input device
	- When button is pressed, driver should send a signal to the input device
	using the input_report_key(input_dev, KEY_WLAN, 1) event

rt2x00 did it wrong for some time, but this has been fixed now, so you
could use that as a reference.

> +/* Called when the RFKILL toggled in hardware.
> + * This is called with the mutex locked. */
> +void b43_rfkill_toggled(struct b43_wldev *dev, bool on)
> +{
> +	struct b43_wl *wl = dev->wl;
> +
> +	B43_WARN_ON(b43_status(dev) < B43_STAT_INITIALIZED);
> +	/* Update the RF status asynchronously, as rfkill will
> +	 * call back into the software toggle handler.
> +	 * This would deadlock if done synchronously. */
> +	queue_work(wl->hw->workqueue, &wl->rfkill.notify_work);
> +}
> +
> +/* Called when the RFKILL toggled in software.
> + * This is called without locking. */
> +static int b43_rfkill_soft_toggle(void *data, enum rfkill_state state)
> +{
> +	struct b43_wldev *dev = data;
> +	struct b43_wl *wl = dev->wl;
> +	int err = 0;
> +
> +	mutex_lock(&wl->mutex);
> +	if (b43_status(dev) < B43_STAT_INITIALIZED)
> +		goto out_unlock;
> +
> +	switch (state) {
> +	case RFKILL_STATE_ON:
> +		if (!dev->radio_hw_enable) {
> +			/* No luck. We can't toggle the hardware RF-kill
> +			 * button from software. */
> +			err = -EBUSY;
> +			goto out_unlock;
> +		}
> +		if (!dev->phy.radio_on)
> +			b43_radio_turn_on(dev);
> +		break;
> +	case RFKILL_STATE_OFF:
> +		if (dev->phy.radio_on)
> +			b43_radio_turn_off(dev, 0);
> +		break;
> +	}
> +
> +out_unlock:
> +	mutex_unlock(&wl->mutex);
> +
> +	return err;
> +}
> +
> +char * b43_rfkill_led_name(struct b43_wldev *dev)
> +{
> +	struct b43_wl *wl = dev->wl;
> +
> +	if (!wl->rfkill.rfkill)
> +		return NULL;
> +	return rfkill_get_led_name(wl->rfkill.rfkill);
> +}
> +
> +void b43_rfkill_init(struct b43_wldev *dev)
> +{
> +	struct b43_wl *wl = dev->wl;
> +	struct b43_rfkill *rfk = &(wl->rfkill);
> +	int err;
> +
> +	snprintf(rfk->name, sizeof(rfk->name),
> +		 "b43-%s", wiphy_name(wl->hw->wiphy));
> +	rfk->rfkill = rfkill_allocate(dev->dev->dev, RFKILL_TYPE_WLAN);
> +	if (!rfk->rfkill)
> +		goto error;
> +	rfk->rfkill->name = rfk->name;
> +	rfk->rfkill->state = RFKILL_STATE_ON;
> +	rfk->rfkill->data = dev;
> +	rfk->rfkill->toggle_radio = b43_rfkill_soft_toggle;
> +	rfk->rfkill->user_claim_unsupported = 1;
> +
> +	INIT_WORK(&rfk->notify_work, b43_notify_rfkill_press);
> +
> +	err = rfkill_register(rfk->rfkill);
> +	if (err)
> +		goto error;
> +
> +	return;
> +error:
> +	b43warn(dev->wl, "Failed to initialize the RF-kill button\n");
> +	rfkill_free(rfk->rfkill);
> +	rfk->rfkill = NULL;
> +}
> +
> +void b43_rfkill_exit(struct b43_wldev *dev)
> +{
> +	struct b43_rfkill *rfk = &(dev->wl->rfkill);
> +
> +	if (!rfk->rfkill)
> +		return;
> +	cancel_work_sync(&rfk->notify_work);
> +	rfkill_unregister(rfk->rfkill);
> +	rfkill_free(rfk->rfkill);
> +	rfk->rfkill = NULL;
> +}
> Index: wireless-2.6/drivers/net/wireless/b43/rfkill.h
> ===================================================================
> --- /dev/null	1970-01-01 00:00:00.000000000 +0000
> +++ wireless-2.6/drivers/net/wireless/b43/rfkill.h	2007-09-27 20:56:53.000000000 +0200
> @@ -0,0 +1,49 @@
> +#ifndef B43_RFKILL_H_
> +#define B43_RFKILL_H_
> +
> +struct b43_wldev;
> +
> +
> +#ifdef CONFIG_B43_RFKILL
> +
> +#include <linux/rfkill.h>
> +
> +struct b43_rfkill {
> +	/* The RFKILL subsystem data structure */
> +	struct rfkill *rfkill;
> +	/* The unique name of this rfkill switch */
> +	char name[32];
> +	/* Workqueue for asynchronous notification. */
> +	struct work_struct notify_work;
> +};
> +
> +void b43_rfkill_init(struct b43_wldev *dev);
> +void b43_rfkill_exit(struct b43_wldev *dev);
> +void b43_rfkill_toggled(struct b43_wldev *dev, bool on);
> +char * b43_rfkill_led_name(struct b43_wldev *dev);
> +
> +
> +#else /* CONFIG_B43_RFKILL */
> +/* No RFKILL support. */
> +
> +struct b43_rfkill {
> +	/* empty */
> +};
> +
> +static inline void b43_rfkill_init(struct b43_wldev *dev)
> +{
> +}
> +static inline void b43_rfkill_exit(struct b43_wldev *dev)
> +{
> +}
> +static inline void b43_rfkill_toggled(struct b43_wldev *dev, bool on)
> +{
> +}
> +static inline char * b43_rfkill_led_name(struct b43_wldev *dev)
> +{
> +	return NULL;
> +}
> +
> +#endif /* CONFIG_B43_RFKILL */
> +
> +#endif /* B43_RFKILL_H_ */
> Index: wireless-2.6/drivers/net/wireless/b43/leds.c
> ===================================================================
> --- wireless-2.6.orig/drivers/net/wireless/b43/leds.c	2007-09-27 20:53:39.000000000 +0200
> +++ wireless-2.6/drivers/net/wireless/b43/leds.c	2007-09-27 20:56:53.000000000 +0200
> @@ -154,12 +154,16 @@ static void b43_map_led(struct b43_wldev
>  				 ieee80211_get_rx_led_name(hw),
>  				 led_index, activelow);
>  		break;
> -	/*FIXME: We need another trigger for the "radio-on" LEDs below.
> -	 *       Wiggle that somehow into the rfkill subsystem. */
>  	case B43_LED_RADIO_ALL:
>  	case B43_LED_RADIO_A:
>  	case B43_LED_RADIO_B:
>  	case B43_LED_MODE_BG:
> +		snprintf(name, sizeof(name),
> +			 "b43-%s:radio", wiphy_name(hw->wiphy));
> +		b43_register_led(dev, &dev->led_radio, name,
> +				 b43_rfkill_led_name(dev),
> +				 led_index, activelow);
> +		break;
>  	case B43_LED_WEIRD:
>  	case B43_LED_ASSOC:
>  		snprintf(name, sizeof(name),
> Index: wireless-2.6/drivers/net/wireless/b43/phy.c
> ===================================================================
> --- wireless-2.6.orig/drivers/net/wireless/b43/phy.c	2007-09-27 20:53:39.000000000 +0200
> +++ wireless-2.6/drivers/net/wireless/b43/phy.c	2007-09-27 20:56:53.000000000 +0200
> @@ -4349,10 +4349,13 @@ void b43_radio_turn_on(struct b43_wldev 
>  	phy->radio_on = 1;
>  }
>  
> -void b43_radio_turn_off(struct b43_wldev *dev)
> +void b43_radio_turn_off(struct b43_wldev *dev, bool force)
>  {
>  	struct b43_phy *phy = &dev->phy;
>  
> +	if (!phy->radio_on && !force)
> +		return;
> +
>  	if (phy->type == B43_PHYTYPE_A) {
>  		b43_radio_write16(dev, 0x0004, 0x00FF);
>  		b43_radio_write16(dev, 0x0005, 0x00FB);
> @@ -4364,9 +4367,11 @@ void b43_radio_turn_off(struct b43_wldev
>  
>  		rfover = b43_phy_read(dev, B43_PHY_RFOVER);
>  		rfoverval = b43_phy_read(dev, B43_PHY_RFOVERVAL);
> -		phy->radio_off_context.rfover = rfover;
> -		phy->radio_off_context.rfoverval = rfoverval;
> -		phy->radio_off_context.valid = 1;
> +		if (!force) {
> +			phy->radio_off_context.rfover = rfover;
> +			phy->radio_off_context.rfoverval = rfoverval;
> +			phy->radio_off_context.valid = 1;
> +		}
>  		b43_phy_write(dev, B43_PHY_RFOVER, rfover | 0x008C);
>  		b43_phy_write(dev, B43_PHY_RFOVERVAL, rfoverval & 0xFF73);
>  	} else
> Index: wireless-2.6/drivers/net/wireless/b43/phy.h
> ===================================================================
> --- wireless-2.6.orig/drivers/net/wireless/b43/phy.h	2007-09-27 20:53:39.000000000 +0200
> +++ wireless-2.6/drivers/net/wireless/b43/phy.h	2007-09-27 20:56:53.000000000 +0200
> @@ -267,7 +267,7 @@ u16 b43_radio_init2050(struct b43_wldev 
>  void b43_radio_init2060(struct b43_wldev *dev);
>  
>  void b43_radio_turn_on(struct b43_wldev *dev);
> -void b43_radio_turn_off(struct b43_wldev *dev);
> +void b43_radio_turn_off(struct b43_wldev *dev, bool force);
>  
>  int b43_radio_selectchannel(struct b43_wldev *dev, u8 channel,
>  			    int synthetic_pu_workaround);
> -
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: [PATCH] b43: RF-kill support

[Michael Buesch]

On Thursday 27 September 2007 22:54:44 Ivo van Doorn wrote:
> Hi,
> 
> > @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
> >  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
> >  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
> >  		dev->radio_hw_enable = radio_hw_enable;
> > -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> > -			radio_hw_enable ? "ENABLED" : "DISABLED");
> > +		b43_rfkill_toggled(dev, radio_hw_enable);
> 
> Isn't it better to use the input_polldev for scheduled input device checking?
> 
> > +static void b43_notify_rfkill_press(struct work_struct *work)
> > +{
> > +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> > +					      notify_work);
> > +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> > +	struct b43_wldev *dev;
> > +	enum rfkill_state state;
> 
> Same here, input_polldev was created especially for hardware
> devices that don't trigger interrupts when the button was pressed.

Ah, well. Will change that later. This works for now.

-- 
Greetings Michael.

Re: [PATCH] b43: RF-kill support

[Ivo van Doorn]

On Thursday 27 September 2007, Michael Buesch wrote:
> On Thursday 27 September 2007 22:54:44 Ivo van Doorn wrote:
> > Hi,
> > 
> > > @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
> > >  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
> > >  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
> > >  		dev->radio_hw_enable = radio_hw_enable;
> > > -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> > > -			radio_hw_enable ? "ENABLED" : "DISABLED");
> > > +		b43_rfkill_toggled(dev, radio_hw_enable);
> > 
> > Isn't it better to use the input_polldev for scheduled input device checking?
> > 
> > > +static void b43_notify_rfkill_press(struct work_struct *work)
> > > +{
> > > +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> > > +					      notify_work);
> > > +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> > > +	struct b43_wldev *dev;
> > > +	enum rfkill_state state;
> > 
> > Same here, input_polldev was created especially for hardware
> > devices that don't trigger interrupts when the button was pressed.
> 
> Ah, well. Will change that later. This works for now.

It might work, but this means that userspace is not made aware of
the key toggling and would thus make rfkill useless in this case.

Ivo

Re: [PATCH] b43: RF-kill support

[Michael Buesch]

On Thursday 27 September 2007 23:12:43 Ivo van Doorn wrote:
> On Thursday 27 September 2007, Michael Buesch wrote:
> > On Thursday 27 September 2007 22:54:44 Ivo van Doorn wrote:
> > > Hi,
> > > 
> > > > @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
> > > >  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
> > > >  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
> > > >  		dev->radio_hw_enable = radio_hw_enable;
> > > > -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> > > > -			radio_hw_enable ? "ENABLED" : "DISABLED");
> > > > +		b43_rfkill_toggled(dev, radio_hw_enable);
> > > 
> > > Isn't it better to use the input_polldev for scheduled input device checking?
> > > 
> > > > +static void b43_notify_rfkill_press(struct work_struct *work)
> > > > +{
> > > > +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> > > > +					      notify_work);
> > > > +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> > > > +	struct b43_wldev *dev;
> > > > +	enum rfkill_state state;
> > > 
> > > Same here, input_polldev was created especially for hardware
> > > devices that don't trigger interrupts when the button was pressed.
> > 
> > Ah, well. Will change that later. This works for now.
> 
> It might work, but this means that userspace is not made aware of
> the key toggling and would thus make rfkill useless in this case.

Yeah, sure. But it's not useless. The other functionality is still available.
I'll fix that. But in a seperate patch. No need to delay this one because
of this (IMO) minor issue.

-- 
Greetings Michael.

Re: [PATCH] b43: RF-kill support

[Ivo van Doorn]

On Thursday 27 September 2007, Michael Buesch wrote:
> On Thursday 27 September 2007 23:12:43 Ivo van Doorn wrote:
> > On Thursday 27 September 2007, Michael Buesch wrote:
> > > On Thursday 27 September 2007 22:54:44 Ivo van Doorn wrote:
> > > > Hi,
> > > > 
> > > > > @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
> > > > >  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
> > > > >  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
> > > > >  		dev->radio_hw_enable = radio_hw_enable;
> > > > > -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> > > > > -			radio_hw_enable ? "ENABLED" : "DISABLED");
> > > > > +		b43_rfkill_toggled(dev, radio_hw_enable);
> > > > 
> > > > Isn't it better to use the input_polldev for scheduled input device checking?
> > > > 
> > > > > +static void b43_notify_rfkill_press(struct work_struct *work)
> > > > > +{
> > > > > +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> > > > > +					      notify_work);
> > > > > +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> > > > > +	struct b43_wldev *dev;
> > > > > +	enum rfkill_state state;
> > > > 
> > > > Same here, input_polldev was created especially for hardware
> > > > devices that don't trigger interrupts when the button was pressed.
> > > 
> > > Ah, well. Will change that later. This works for now.
> > 
> > It might work, but this means that userspace is not made aware of
> > the key toggling and would thus make rfkill useless in this case.
> 
> Yeah, sure. But it's not useless. The other functionality is still available.
> I'll fix that. But in a seperate patch. No need to delay this one because
> of this (IMO) minor issue.

I agree, no need to delay this patch.

But since I never added proper documentation from the start of rfkill,
and rt2x00 did it wrong for a while as well, I just wanted to make sure you
were aware of the correct approach. :)

Ivo

Re: [PATCH] b43: RF-kill support

[Michael Buesch]

On Thursday 27 September 2007 23:23:45 Ivo van Doorn wrote:
> On Thursday 27 September 2007, Michael Buesch wrote:
> > On Thursday 27 September 2007 23:12:43 Ivo van Doorn wrote:
> > > On Thursday 27 September 2007, Michael Buesch wrote:
> > > > On Thursday 27 September 2007 22:54:44 Ivo van Doorn wrote:
> > > > > Hi,
> > > > > 
> > > > > > @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
> > > > > >  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
> > > > > >  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
> > > > > >  		dev->radio_hw_enable = radio_hw_enable;
> > > > > > -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> > > > > > -			radio_hw_enable ? "ENABLED" : "DISABLED");
> > > > > > +		b43_rfkill_toggled(dev, radio_hw_enable);
> > > > > 
> > > > > Isn't it better to use the input_polldev for scheduled input device checking?
> > > > > 
> > > > > > +static void b43_notify_rfkill_press(struct work_struct *work)
> > > > > > +{
> > > > > > +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> > > > > > +					      notify_work);
> > > > > > +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> > > > > > +	struct b43_wldev *dev;
> > > > > > +	enum rfkill_state state;
> > > > > 
> > > > > Same here, input_polldev was created especially for hardware
> > > > > devices that don't trigger interrupts when the button was pressed.
> > > > 
> > > > Ah, well. Will change that later. This works for now.
> > > 
> > > It might work, but this means that userspace is not made aware of
> > > the key toggling and would thus make rfkill useless in this case.
> > 
> > Yeah, sure. But it's not useless. The other functionality is still available.
> > I'll fix that. But in a seperate patch. No need to delay this one because
> > of this (IMO) minor issue.
> 
> I agree, no need to delay this patch.
> 
> But since I never added proper documentation from the start of rfkill,
> and rt2x00 did it wrong for a while as well, I just wanted to make sure you
> were aware of the correct approach. :)

Yeah, thanks. I thought this was correct.
So if it's not correct to use the "switch" API function from outside, please
remove it from the public header and put it into some private inside of
the net/rfkill directory.

-- 
Greetings Michael.

Re: [PATCH] b43: RF-kill support

[Ivo van Doorn]

On Thursday 27 September 2007, Michael Buesch wrote:
> On Thursday 27 September 2007 23:23:45 Ivo van Doorn wrote:
> > On Thursday 27 September 2007, Michael Buesch wrote:
> > > On Thursday 27 September 2007 23:12:43 Ivo van Doorn wrote:
> > > > On Thursday 27 September 2007, Michael Buesch wrote:
> > > > > On Thursday 27 September 2007 22:54:44 Ivo van Doorn wrote:
> > > > > > Hi,
> > > > > > 
> > > > > > > @@ -2401,8 +2401,7 @@ static void b43_periodic_every1sec(struc
> > > > > > >  	radio_hw_enable = b43_is_hw_radio_enabled(dev);
> > > > > > >  	if (unlikely(dev->radio_hw_enable != radio_hw_enable)) {
> > > > > > >  		dev->radio_hw_enable = radio_hw_enable;
> > > > > > > -		b43info(dev->wl, "Radio hardware status changed to %s\n",
> > > > > > > -			radio_hw_enable ? "ENABLED" : "DISABLED");
> > > > > > > +		b43_rfkill_toggled(dev, radio_hw_enable);
> > > > > > 
> > > > > > Isn't it better to use the input_polldev for scheduled input device checking?
> > > > > > 
> > > > > > > +static void b43_notify_rfkill_press(struct work_struct *work)
> > > > > > > +{
> > > > > > > +	struct b43_rfkill *rfk = container_of(work, struct b43_rfkill,
> > > > > > > +					      notify_work);
> > > > > > > +	struct b43_wl *wl = container_of(rfk, struct b43_wl, rfkill);
> > > > > > > +	struct b43_wldev *dev;
> > > > > > > +	enum rfkill_state state;
> > > > > > 
> > > > > > Same here, input_polldev was created especially for hardware
> > > > > > devices that don't trigger interrupts when the button was pressed.
> > > > > 
> > > > > Ah, well. Will change that later. This works for now.
> > > > 
> > > > It might work, but this means that userspace is not made aware of
> > > > the key toggling and would thus make rfkill useless in this case.
> > > 
> > > Yeah, sure. But it's not useless. The other functionality is still available.
> > > I'll fix that. But in a seperate patch. No need to delay this one because
> > > of this (IMO) minor issue.
> > 
> > I agree, no need to delay this patch.
> > 
> > But since I never added proper documentation from the start of rfkill,
> > and rt2x00 did it wrong for a while as well, I just wanted to make sure you
> > were aware of the correct approach. :)
> 
> Yeah, thanks. I thought this was correct.
> So if it's not correct to use the "switch" API function from outside, please
> remove it from the public header and put it into some private inside of
> the net/rfkill directory.

Will do. :)

Ivo

Re: [PATCH] b43: RF-kill support

[Larry Finger]

Michael Buesch wrote:
> 
> Yeah, sure. But it's not useless. The other functionality is still available.
> I'll fix that. But in a seperate patch. No need to delay this one because
> of this (IMO) minor issue.
> 

How long will the delay be? If relatively short, I'll wait for the final one before submitting the
port to b43legacy.

Larry

Re: [PATCH] b43: RF-kill support

[Michael Buesch]

On Friday 28 September 2007 00:17:13 Larry Finger wrote:
> Michael Buesch wrote:
> > 
> > Yeah, sure. But it's not useless. The other functionality is still available.
> > I'll fix that. But in a seperate patch. No need to delay this one because
> > of this (IMO) minor issue.
> > 
> 
> How long will the delay be? If relatively short, I'll wait for the final one before submitting the
> port to b43legacy.

One or two days, maybe.
I'm currently debugging another important bug.
And I guess I'll have to sleep in an hour or two, too. :)

The rfkill fix will be incremental based on this one, then.

-- 
Greetings Michael.

Re: [PATCH] b43: RF-kill support

[Larry Finger]

Michael Buesch wrote:
> One or two days, maybe.
> I'm currently debugging another important bug.
> And I guess I'll have to sleep in an hour or two, too. :)
> 
> The rfkill fix will be incremental based on this one, then.
> 

That I expected. The way I should have stated the question was "Should I submit the port of this
patch to b43legacy ASAP, or should I wait until the rfkill fix is ready to have only 1 rather than 2
patches for John to handle?" I think the correct answer is to wait. It will probably take me that
long to verify my port of the new LED handler anyway.

Larry