Problems with hardware radio kill switch on Sony Vaio notebook

Problems with hardware radio kill switch on Sony Vaio notebook

[Ralph Benzinger]

Hello,

My Sony Vaio VPC-Y11S1E notebook has an Atheros AR9285 chip, which is
supported just fine by ath9k.  The notebook also has a physical radio
kill switch at the front that can be slid into one of two positions.
Alas, the rfkill tool shows wlan0 as "hard blocked: yes" for either
switch position.  Consequently, txpower is off, and wireless
networking cannot be activated.

Could you kindly indicate which part of the kernel would be
responsible for handling the radio kill switches -- cfg80211, ath9k,
sony-laptop, ...?  Are there any known fixes or workarounds?  I'm
using Kubuntu 9.10 with their stock kernel 2.6.31 and the wireless
backports modules right now, but tests with Kubuntu 10.4 Beta and
2.6.32 failed as well.

(Note that I cannot get ndiswrapper to work either, although it seems
to power on the wireless card.)

Thanks
Ralph

Re: Problems with hardware radio kill switch on Sony Vaio notebook

[John W. Linville]

On Mon, Mar 29, 2010 at 05:27:21PM +0200, Ralph Benzinger wrote:

> My Sony Vaio VPC-Y11S1E notebook has an Atheros AR9285 chip, which is
> supported just fine by ath9k.  The notebook also has a physical radio
> kill switch at the front that can be slid into one of two positions.
> Alas, the rfkill tool shows wlan0 as "hard blocked: yes" for either
> switch position.  Consequently, txpower is off, and wireless
> networking cannot be activated.
> 
> Could you kindly indicate which part of the kernel would be
> responsible for handling the radio kill switches -- cfg80211, ath9k,
> sony-laptop, ...?  Are there any known fixes or workarounds?  I'm
> using Kubuntu 9.10 with their stock kernel 2.6.31 and the wireless
> backports modules right now, but tests with Kubuntu 10.4 Beta and
> 2.6.32 failed as well.

Generally, sony-laptop would be responsible for correctly reporting
the rfkill switch state.

Hth!

John
-- 
John W. Linville		Someday the world will need a hero, and you
linville@tuxdriver.com			might be all we have.  Be ready.

Re: Problems with hardware radio kill switch on Sony Vaio notebook

[Ralph Benzinger]

On March 29, you wrote:
> Generally, sony-laptop would be responsible for correctly reporting
> the rfkill switch state.

John,

Thanks for the info, I appreciate it!

Just a naive follow-up question then: Wouldn't it be possible for the
ath9k driver to ignore the rfkill switch settings?  (Assuming, of
course, that there is no actual physical power disconnect, which I
doubt.)

Is there a simple way to override the rfkill effect on ath9k via
/proc or /sys?  Or alternatively, if I were to recompile the ath9k
module, what would I need to change/disable?  (On the other hand,
hacking sony-laptop seems to be the easier way.)

Any pointers would be greatly appreciated.

Regards
Ralph

Re: Problems with hardware radio kill switch on Sony Vaio notebook

[Mattia Dongili]

Hi Raplh,

On Mon, Mar 29, 2010 at 01:06:05PM -0400, John W. Linville wrote:
> On Mon, Mar 29, 2010 at 05:27:21PM +0200, Ralph Benzinger wrote:
> 
> > My Sony Vaio VPC-Y11S1E notebook has an Atheros AR9285 chip, which is
> > supported just fine by ath9k.  The notebook also has a physical radio
> > kill switch at the front that can be slid into one of two positions.
> > Alas, the rfkill tool shows wlan0 as "hard blocked: yes" for either
> > switch position.  Consequently, txpower is off, and wireless
> > networking cannot be activated.
> > 
> > Could you kindly indicate which part of the kernel would be
> > responsible for handling the radio kill switches -- cfg80211, ath9k,
> > sony-laptop, ...?  Are there any known fixes or workarounds?  I'm
> > using Kubuntu 9.10 with their stock kernel 2.6.31 and the wireless
> > backports modules right now, but tests with Kubuntu 10.4 Beta and
> > 2.6.32 failed as well.
> 
> Generally, sony-laptop would be responsible for correctly reporting
> the rfkill switch state.

Yes, it's most likely that sony-laptop would be the one module to point
fingers at.

Could you send the DSDT from your laptop and the output of dmesg after
loading sony-laptop with debug=1 and switching the rfkill trigger back
and forth?

also the output from:
  $ grep . /sys/class/rfkill/rfkill*/*

Thanks
-- 
mattia
:wq!