Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
On Thursday 25 January 2007 04:21, Larry Finger wrote: Michael Buesch wrote: On Thursday 25 January 2007 00:16, Steve Brown wrote: My understanding of the proposed behavior is: Radio on: led on Radio off: led off Activity: flash on-off-on ... This is the same behavior as my Buffalo WHR-HP-G54 and seems reasonable. That's not what is implemented in larry's patch. In this patch the activity led does only indicate activity and not radio status. (Well, it also indicates a the-card-is-up state, but that's not necessarily indicating that the radio is switched on). Actually, it is what is in my patch. The LED is on if the hardware-enable switch is on and the interface is up. The activity LED switches off, if you turn off the rf through rfkill button? I don't think it does. I'd say it would always stay on, if you switch rf off. -- Greetings Michael. ___ Bcm43xx-dev mailing list Bcm43xx-dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
The current bcm43xx-d80211 driver ignores any wireless-enable switches on
mini-PCI
and mini-PCI-E cards. This patch implements a new routine to interrogate the
radio hardware enabled bit in the interface, logs the initial state and any
changes in the switch (if debugging enabled), activates the LED to show the
state, and changes the periodic work handler to provide 1 second response
to switch changes. The changes in the periodic work specs have not been
implemented.
This should be applied to http://bu3sch.de/git/wireless-dev.git.
Signed-off-by: Larry Finger [EMAIL PROTECTED]
---
Index: wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx.h
===
--- wireless-dev.orig/drivers/net/wireless/d80211/bcm43xx/bcm43xx.h
+++ wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx.h
@@ -238,6 +238,9 @@ enum {
#define BCM43xx_UCODEFLAG_UNKPACTRL0x0040
#define BCM43xx_UCODEFLAG_JAPAN0x0080
+/* Hardware Radio Enable masks */
+#define BCM43xx_MMIO_RADIO_HWENABLED_HI_MASK (1 16)
+#define BCM43xx_MMIO_RADIO_HWENABLED_LO_MASK (1 4)
/* HostFlags. See bcm43xx_hf_read/write() */
#define BCM43xx_HF_ANTDIVHELP 0x0001 /* ucode antenna div helper
*/
@@ -735,7 +738,8 @@ struct bcm43xx_wldev {
reg124_set_0x4:1, /* Some variable to keep track of IRQ
stuff. */
short_preamble:1, /* TRUE, if short preamble is enabled.
*/
short_slot:1, /* TRUE, if short slot timing is
enabled. */
- firmware_norelease:1; /* Do not release the firmware. Used on
suspend. */
+ firmware_norelease:1, /* Do not release the firmware. Used on
suspend. */
+ radio_hw_enable:1; /* saved state of radio hardware
enabled state */
/* PHY/Radio device. */
struct bcm43xx_phy phy;
Index: wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx_leds.c
===
--- wireless-dev.orig/drivers/net/wireless/d80211/bcm43xx/bcm43xx_leds.c
+++ wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx_leds.c
@@ -27,7 +27,7 @@
#include bcm43xx_leds.h
#include bcm43xx.h
-
+#include bcm43xx_main.h
static void bcm43xx_led_changestate(struct bcm43xx_led *led)
{
@@ -108,6 +108,7 @@ static void bcm43xx_led_init_hardcoded(s
switch (led_index) {
case 0:
led-behaviour = BCM43xx_LED_ACTIVITY;
+ led-activelow = 1;
if (bus-board_vendor == PCI_VENDOR_ID_COMPAQ)
led-behaviour = BCM43xx_LED_RADIO_ALL;
break;
@@ -196,18 +197,19 @@ void bcm43xx_leds_update(struct bcm43xx_
turn_on = activity;
break;
case BCM43xx_LED_RADIO_ALL:
- turn_on = phy-radio_on;
+ turn_on = phy-radio_on
bcm43xx_is_hw_radio_enabled(dev);
break;
case BCM43xx_LED_RADIO_A:
- turn_on = (phy-radio_on phy-type ==
BCM43xx_PHYTYPE_A);
+ turn_on = (phy-radio_on
bcm43xx_is_hw_radio_enabled(dev)
+ phy-type == BCM43xx_PHYTYPE_A);
break;
case BCM43xx_LED_RADIO_B:
- turn_on = (phy-radio_on
+ turn_on = (phy-radio_on
bcm43xx_is_hw_radio_enabled(dev)
(phy-type == BCM43xx_PHYTYPE_B ||
phy-type == BCM43xx_PHYTYPE_G));
break;
case BCM43xx_LED_MODE_BG:
- if (phy-type == BCM43xx_PHYTYPE_G
+ if (phy-type == BCM43xx_PHYTYPE_G
bcm43xx_is_hw_radio_enabled(dev)
1/*FIXME: using G rates.*/)
turn_on = 1;
break;
Index: wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx_main.c
===
--- wireless-dev.orig/drivers/net/wireless/d80211/bcm43xx/bcm43xx_main.c
+++ wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx_main.c
@@ -2080,6 +2080,9 @@ static int bcm43xx_chip_init(struct bcm4
if (err)
goto err_gpio_cleanup;
bcm43xx_radio_turn_on(dev);
+ dev-radio_hw_enable = bcm43xx_is_hw_radio_enabled(dev);
+ dprintk(KERN_INFO PFX Radio %s by hardware\n,
+ (dev-radio_hw_enable == 0) ? disabled : enabled);
bcm43xx_write16(dev, 0x03E6, 0x);
err = bcm43xx_phy_init(dev);
@@ -2226,22 +2229,38 @@ static void bcm43xx_periodic_every15sec(
//TODO for APHY (temperature?)
}
+static void bcm43xx_periodic_every1sec(struct bcm43xx_wldev *dev)
+{
+ int radio_hw_enable;
+
+ /* check if radio hardware enabled status changed */
+
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
On Wednesday 24 January 2007 19:39, Larry Finger wrote:
The current bcm43xx-d80211 driver ignores any wireless-enable switches on
mini-PCI
and mini-PCI-E cards. This patch implements a new routine to interrogate the
radio hardware enabled bit in the interface, logs the initial state and any
changes in the switch (if debugging enabled), activates the LED to show the
state, and changes the periodic work handler to provide 1 second response
to switch changes. The changes in the periodic work specs have not been
implemented.
This should be applied to http://bu3sch.de/git/wireless-dev.git.
Signed-off-by: Larry Finger [EMAIL PROTECTED]
---
Index: wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx.h
===
--- wireless-dev.orig/drivers/net/wireless/d80211/bcm43xx/bcm43xx.h
+++ wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx.h
@@ -238,6 +238,9 @@ enum {
#define BCM43xx_UCODEFLAG_UNKPACTRL 0x0040
#define BCM43xx_UCODEFLAG_JAPAN 0x0080
+/* Hardware Radio Enable masks */
+#define BCM43xx_MMIO_RADIO_HWENABLED_HI_MASK (1 16)
+#define BCM43xx_MMIO_RADIO_HWENABLED_LO_MASK (1 4)
/* HostFlags. See bcm43xx_hf_read/write() */
#define BCM43xx_HF_ANTDIVHELP0x0001 /* ucode antenna div
helper */
@@ -735,7 +738,8 @@ struct bcm43xx_wldev {
reg124_set_0x4:1, /* Some variable to keep track of IRQ
stuff. */
short_preamble:1, /* TRUE, if short preamble is enabled.
*/
short_slot:1, /* TRUE, if short slot timing is
enabled. */
- firmware_norelease:1; /* Do not release the firmware. Used on
suspend. */
+ firmware_norelease:1, /* Do not release the firmware. Used on
suspend. */
+ radio_hw_enable:1; /* saved state of radio hardware
enabled state */
/* PHY/Radio device. */
struct bcm43xx_phy phy;
Index: wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx_leds.c
===
--- wireless-dev.orig/drivers/net/wireless/d80211/bcm43xx/bcm43xx_leds.c
+++ wireless-dev/drivers/net/wireless/d80211/bcm43xx/bcm43xx_leds.c
@@ -27,7 +27,7 @@
#include bcm43xx_leds.h
#include bcm43xx.h
-
+#include bcm43xx_main.h
static void bcm43xx_led_changestate(struct bcm43xx_led *led)
{
@@ -108,6 +108,7 @@ static void bcm43xx_led_init_hardcoded(s
switch (led_index) {
case 0:
led-behaviour = BCM43xx_LED_ACTIVITY;
+ led-activelow = 1;
Why activelow?
--
Greetings Michael.
___
Bcm43xx-dev mailing list
Bcm43xx-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
Michael Buesch wrote:
On Wednesday 24 January 2007 19:39, Larry Finger wrote:
switch (led_index) {
case 0:
led-behaviour = BCM43xx_LED_ACTIVITY;
+led-activelow = 1;
Why activelow?
It makes the light be on when there is no activity and blink off when there is.
With that behavior,
it is a lot easier to see that the switch is on.
Larry
___
Bcm43xx-dev mailing list
Bcm43xx-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
Michael Buesch wrote:
On Wednesday 24 January 2007 20:08, Larry Finger wrote:
Michael Buesch wrote:
On Wednesday 24 January 2007 19:39, Larry Finger wrote:
switch (led_index) {
case 0:
led-behaviour = BCM43xx_LED_ACTIVITY;
+ led-activelow = 1;
Why activelow?
It makes the light be on when there is no activity and blink off when there
is. With that behavior,
it is a lot easier to see that the switch is on.
Ah, ok.
It's debatable if that behaviour is desireable.
I don't like it, but if people want to have it, we can implement it.
I do not think this is what people want. If the radio is off the light
should just be off, when radio is on turn the light on, is a much more
desirable.
Thanks,
-Jory
___
Bcm43xx-dev mailing list
Bcm43xx-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
Jory A. Pratt wrote:
Michael Buesch wrote:
On Wednesday 24 January 2007 20:08, Larry Finger wrote:
Michael Buesch wrote:
On Wednesday 24 January 2007 19:39, Larry Finger wrote:
switch (led_index) {
case 0:
led-behaviour = BCM43xx_LED_ACTIVITY;
+ led-activelow = 1;
Why activelow?
It makes the light be on when there is no activity and blink off when there
is. With that behavior,
it is a lot easier to see that the switch is on.
Ah, ok.
It's debatable if that behaviour is desireable.
I don't like it, but if people want to have it, we can implement it.
I do not think this is what people want. If the radio is off the light
should just be off, when radio is on turn the light on, is a much more
desirable.
Thanks,
-Jory
___
Bcm43xx-dev mailing list
Bcm43xx-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
!DSPAM:6251,45b7e6db63428126216758!
My understanding of the proposed behavior is:
Radio on: led on
Radio off: led off
Activity: flash on-off-on ...
This is the same behavior as my Buffalo WHR-HP-G54 and seems reasonable.
Steve
___
Bcm43xx-dev mailing list
Bcm43xx-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
On Thursday 25 January 2007 00:16, Steve Brown wrote: My understanding of the proposed behavior is: Radio on: led on Radio off: led off Activity: flash on-off-on ... This is the same behavior as my Buffalo WHR-HP-G54 and seems reasonable. That's not what is implemented in larry's patch. In this patch the activity led does only indicate activity and not radio status. (Well, it also indicates a the-card-is-up state, but that's not necessarily indicating that the radio is switched on). -- Greetings Michael. ___ Bcm43xx-dev mailing list Bcm43xx-dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
Michael Buesch wrote:
On Wednesday 24 January 2007 23:59, Jory A. Pratt wrote:
Michael Buesch wrote:
On Wednesday 24 January 2007 20:08, Larry Finger wrote:
Michael Buesch wrote:
On Wednesday 24 January 2007 19:39, Larry Finger wrote:
switch (led_index) {
case 0:
led-behaviour = BCM43xx_LED_ACTIVITY;
+led-activelow = 1;
Why activelow?
It makes the light be on when there is no activity and blink off when
there is. With that behavior,
it is a lot easier to see that the switch is on.
Ah, ok.
It's debatable if that behaviour is desireable.
I don't like it, but if people want to have it, we can implement it.
I do not think this is what people want. If the radio is off the light
should just be off, when radio is on turn the light on, is a much more
desirable.
Well, we are talking about the activity LED, not about the radio on
LED. There are many cards out there (including ethernet cards) which
always have the LED on and flash it for activity. So it's not _that_
wrong to do it.
Personally, I think it's more intuitive to have a LED off, when there
is no activity, but hey... .
It could also save a little bit of power on notebooks, if we leave
it activehigh. But on the other side, if someone is concerned about
the LEDs drawing too much current, he can as well completely disable
the LEDs with the module parameter.
If the LED that shows the radio switch state were different than the one used
to show activity, then
I would agree with you; however, it is the same one. If we want the light on
when the switch is on
(and the interface is up) and I think we do, then we need the behavior I coded.
If there were a way to determine if the hardware-enabled bit were determined by
a switch, then we
could make activelow conditional, but that is not the case. It doesn't matter
to me as my switch is
a slide, and I always know if the radio is enabled; however, a lot of them are
momentary switches
that seem to toggle a bit in the hardware. Those people need positive feedback
that the radio is
enabled. Without the activelow, their light would be on if the radio were off,
and vice versa. That
would be too much confusion!
Larry
___
Bcm43xx-dev mailing list
Bcm43xx-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
Re: RFC: bcm43xx-d80211: Add radio hardware switch code to mb tree
Michael Buesch wrote: On Thursday 25 January 2007 00:16, Steve Brown wrote: My understanding of the proposed behavior is: Radio on: led on Radio off: led off Activity: flash on-off-on ... This is the same behavior as my Buffalo WHR-HP-G54 and seems reasonable. That's not what is implemented in larry's patch. In this patch the activity led does only indicate activity and not radio status. (Well, it also indicates a the-card-is-up state, but that's not necessarily indicating that the radio is switched on). Actually, it is what is in my patch. The LED is on if the hardware-enable switch is on and the interface is up. Larry ___ Bcm43xx-dev mailing list Bcm43xx-dev@lists.berlios.de https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
