Re: [RFC] b43: A patch for control of the radio LED using rfkill

[Ivo van Doorn <ivdoorn@gmail.com>]

On Saturday 20 September 2008, Henrique de Moraes Holschuh wrote:
> On Fri, 19 Sep 2008, Ivo van Doorn wrote:
> > > Even if that means a need to 
> > > get EV_SW SW_WLAN out of Dmitry...  I don't know if rt2x000 wants SW_WLAN or
> > > SW_RFKILL_ALL.
> > 
> > Not sure either, there is no way to determine which of the 2 should be used.
> 
> Should be an module parameter or runtime (sysfs?) parameter, then...

Module parameter sounds best, I doubt people would want to change the value
at runtime (when it is just easier to do it at loadtime).

> > Oh an to make sure I get it completely right this time:
> > When I implement the input polldev, I no longer have to use rfkill_force_state() right?
> 
> Correct.  It has nothing to do with the rfkill class part of your driver at
> all, since there is no effect on the hardware.  It is just an input device
> that issues EV_SW events.

Ok, thanks. :)

I'll remove rfkill from rt2x00, implement input_polldev as seperate module,
move rfkill support into mac80211 to make mac80211 drivers automatically
detect the rfkill notifications.

Ivo