Johannes,
What is the reason that rfkill_get_global_sw_state() is not exported?
A call to this routine seems to be the ideal replacement for the
deprecated radio_enabled.
Larry
Johannes,
> No need, radio_enabled is intended to be removed, because if you rfkill
> now cfg80211 will force the interfaces down. Some time later way may
> need to add a way to disable just the radio again, but that will be
> different and for different purposes.
In b43/b43legacy, the driver can determine when the hw has required an
rfkill, but what is the "approved" method for determining if userspace
has killed it?
Larry
On Sat, 2009-06-20 at 07:59 -0500, Larry Finger wrote:
> Johannes,
>
> > No need, radio_enabled is intended to be removed, because if you rfkill
> > now cfg80211 will force the interfaces down. Some time later way may
> > need to add a way to disable just the radio again, but that will be
> > different and for different purposes.
>
> In b43/b43legacy, the driver can determine when the hw has required an
> rfkill, but what is the "approved" method for determining if userspace
> has killed it?
That isn't necessary. In that case, cfg80211 notices that it was killed,
and (via mac80211) ends up calling the ->stop() callback.
johannes
Larry,
> What is the reason that rfkill_get_global_sw_state() is not exported?
> A call to this routine seems to be the ideal replacement for the
> deprecated radio_enabled.
No need, radio_enabled is intended to be removed, because if you rfkill
now cfg80211 will force the interfaces down. Some time later way may
need to add a way to disable just the radio again, but that will be
different and for different purposes.
johannes