Re: [git pull 02/14] Add Wistron driver

[Vojtech Pavlik <vojtech@suse.cz>]

On Sun, Dec 11, 2005 at 01:10:13PM +0000, Bernhard Rosenkraenzer wrote:
> On Tuesday 06 December 2005 15:21, Dmitry Torokhov wrote:
> 
> Hi, sorry for the delay, just returning from Spain...
> 
> > > There are acpi daemon for any evetnts that needs user space attention.
> > > I'm not sure if these events should be routed to input layer.
> >
> > How do you suggest handle buttons such as "Mail", "WWW", etc? Through
> > acpid? And then tunnel them to X somewhow?
> 
> I think routing them to the input layer makes most sense because they are keys 
> like everything else -- of course hacking acpid to pass on ACPI key events to 
> Xorg via the XTest extension is not exactly hard, but that would break the 
> keys in text mode (who knows, maybe someone wants to map his mail key to 
> "mutt[RETURN]"?), and of course launching an application from acpid is a bit 
> hard (acpid runs as root --> need to figure out which user is pressed the 
> button, switch user IDs, find the correct X display if any, .....) if it's an 
> input event, solutions for the expected functionality already exist - e.g. 
> khotkeys.

You also can hack acpid to use uinput to feed the events back to the
input subsystem, but I agree with you that going there directly is
probably the best way to go.

> > > With acpi enabled - wistron module, bluetooth works.
> > > From these test cases, do you still think wistron driver can help my
> > > laptop?
> >
> > No, you have proven that the driver will not help to your laptop. Now,
> > as it is, it won't even load on your laptop either, because of
> > different DMI signature. So why are you complaining? I am pretty sure
> > Bernhard (who added bluetooth handling) has his working with ACPI.
> > Bernhard, any input on this?
> 
> I have ACPI + wistron module. Can't tell if bluetooth actually works because I 
> don't have any bluetooth hardware, but I can tell the bluetooth LED can be 
> turned on and off.

You should see the bluetooth USB device appearing and disappearing in
'lsusb'.

> ACPI works on this box, but not the ACPI buttons stuff (guess they aren't 
> following standards...). I'm running a modified 2.6.15-rc5-mm1 here (none of 
> the modifications touch ACPI though).

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR