PERFORCE change 123399 for review

[Hans Petter Selasky]

On Friday 13 July 2007 15:42, Rui Paulo wrote:
> Hans Petter Selasky wrote:
> > On Thursday 12 July 2007 21:18, Rui Paulo wrote:
> >> http://perforce.freebsd.org/chv.cgi?CH=123399
> >>
> >> Change 123399 by rpaulo at rpaulo_epsilon on 2007/07/12 19:18:01
> >>
> >> 	For some reason, moused confuses sysmouse after attaching to the
> >> 	HID version of the MacBook Bluetooth device.
> >> 	For now, blacklist the product id of the Bluetooth controller
> >> 	(in HID mode) inside ums(4).
> >
> > You plan to remove this Quirk at some point?
>
> Yes.
>
> > I personally don't like all these quirks.
>
> What do you mean by "all these quirks" ? This is the only quirk that can
> be overcome. The other fix I added for the Fn key in Apple USB keyboards
>  is not optional.

You should try to autodetect this.

>
> > Could you provide a "udesc_dump" from your MacBook and all its devices,
> > so that we can have a look?

Remove all USB device drivers from your kernel config file, except "device 
ugen". All your USB devices should show up like "ugenX" devices. Then 
install "/usr/ports/sysutils/udesc_dump" and run this command on 
all /dev/ugenX instances. Redirect the contents to a set of files and upload 
them somewhere. I want to have a look at the descriptors. Maybe there is a 
USB descriptor that you can detect, that makes a keyboard an Apple keyboard. 
That will be more generic than adding these quirks!

> When I switch my bluetooth controller from HID mode to HCI mode, I see:
> ukbd1: at uhub3 port 1 (addr 2) disconnected
> ukbd1: detached
> ums1: at uhub3 port 1 (addr 2) disconnected
> ums1: detached
> uhid2: at uhub3 port 1 (addr 2) disconnected
> uhid2: detached
> ubt0: <vendor 0x05ac product 0x8205, class 224/1, rev 2.00/19.65, addr
> 2> on uhub3
> ubt0: Interface 0 endpoints: interrupt=0x81, bulk-in=0x82, bulk-out=0x2
> ubt0: Interface 1 (alt.config 5) endpoints: isoc-in=0x83, isoc-out=0x3;
> wMaxPacketSize=49; nframes=6, buffer size=294
>
> If I start X after that, there's something fishy going on that prevents
> me from using the keyboard and the other usb mouses that I have attached.
>
> The workaround is to restart all moused instances.
>
> I don't really understand why this happens, because when the bluetooth
> controller is switched, the moused on ums1 dies.

Have you configured your system to detect Bluetooth Keyboard and Mouse ?

--HPS