Re: [Linuxwacom-discuss] Bamboo Fun multi touch

[Chris B. <ch...@cn...>]

On Wed, Apr 21, 2010 at 2:48 PM, Edward Johns <ejd...@gm...> wrote:
> On 21 April 2010 20:31, Edward Johns <ejd...@gm...> wrote:
>> On 21 April 2010 16:26, Chris Bagwell <ch...@cn...> wrote:
>>> On Wed, Apr 21, 2010 at 3:41 AM, Edward Johns <ejd...@gm...> wrote:
>>>>
>>>> Hi,
>>>> I'm not currently using an fdi file, I'm using xorg.conf.
>>>>
>>>> When I tried just using an fdi (with no mention of wacom in xorg.conf)
>>>> xinput --list and xsetwacom list dev both showed nothing. The pad
>>>> worked but the resolution didn't match the screen. The cursor also
>>>> jumped around when the stylus/finger went in and out of range of the
>>>> pad.
>>>>
>>>> When using xorg.conf the output of both xinput and xsetwacom show the
>>>> correct list of devices, the pad resolution matches the screen
>>>> resolution and the cursor behaves as expected.
>>>> Using xorg.conf but removing wacom_drv gives the same behaviour as
>>>> trying to use an fdi file.
>>>>
>>>> I was starting to think that the wacom_drv is being used but so is
>>>> something else at the sametime?
>>>>
>>>> Should I be seeing mouse3 and mouse4 or is that a red herring? I had a
>>>> look in /dev/input/by-path and it showed that they are used by the
>>>> wacom pad (which is confirmed by them disappearing when the pad is not
>>>> connected).
>>>>
>>>> Ed.
>>>
>>> Can you restore all your input drivers and send that Xorg.0.log file?
>>> Also, can you send your xorg.conf at same time?  Probably you've
>>> already sent these but I've lost it.
>>>
>>> What looks like is happening to me from your last Xorg.0.log is that
>>> you do not have a core pointer defined in xorg.conf and X is trying to
>>> find one and using /dev/input/mice which is somehow your wacom tablet.
>>>
>>> I've not used anything prior to xorg 1.7 it quite some time and forget
>>> all the behavior and its options.  I think others on the list are
>>> better to describe how to specificy a Device "/dev/input/mice", Driver
>>> "wacom", and possibly some Core options in your xorg.conf to have it
>>> stop mis-assigning the wrong driver to /dev/input/* devices.
>>>
>>> Chris
>>>
>> Ok. Using the xorg.conf Ping sent me in the other thread. My mouse
>> doesn't work, the tablet does and guess what... 2 fingers crashes.
>> I've attached the log from booting with that config.
>> I've also attached my current xorg.conf and the log from that.
>> In /dev/input/by-id I have this:
>>
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18 usb-062a_0000-mouse -> ../mouse0
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18
>> usb-Logitech_USB_Receiver-mouse -> ../mouse1
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18 usb-062a_0000-event-mouse -> ../event0
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18
>> usb-Logitech_USB_Receiver-event-kbd -> ../event1
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18
>> usb-Logitech_USB_Receiver-event-mouse -> ../event2
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18
>> usb-Wacom_Co._Ltd._CTH-461-event-mouse -> ../event5
>> lrwxrwxrwx 1 root root 9 2010-04-21 20:18
>> usb-Wacom_Co._Ltd._CTH-461-mouse -> ../mouse2
>>
>> usb-062a is my actual mouse. The logitec is from a wireless
>> keyboard/mouse combo for which I only use the keyboard.
>> wacom and wacom-touch are linked to event5 and event6 respectively in
>> /dev/input.
>> To my untrained eye that looks like wacom is indeed being treated as a mouse?
>>
>> Ed.
>>
>>
> Sorry, I was so keen to get you the logs I didn't check them first!
> Just noticed why Pings xorg.conf didn't let my mouse work, I needed
> /dev/input/mice rather than /dev/mice. With that change my mouse works
> but I still get the crash.
>
> Ed.
>

There is some udev strangeness going on with you distribution.
Hopefully, someone understands it and can suggestion something.  I've
no idea why they've chosen to bind devices to both /dev/input/event*'s
and then duplicate them and bind to /dev/input/mouse* or mice.

Your Xorg.0.log isn't advertising that its grabbing those duplicate
/dev/input devices and assigning them to mouse_drv but it seems like
it must be based on your testing.

Chris