[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20090204225942.3590b9c1@zork.foo.com>
Date: Wed, 4 Feb 2009 22:59:42 +0100
From: Max Weninger <max.weninger@...il.com>
To: Daniel Ritz <daniel.ritz-ml@...ssonline.ch>
Cc: Andrew Morton <akpm@...ux-foundation.org>,
linux-kernel@...r.kernel.org,
Niels de Vos <niels.devos@...cor-nixdorf.com>,
linux-usb@...r.kernel.org
Subject: Re: SOLVED support of touchscreen reported as eGalax but is not
working
Hi
On Wed, 4 Feb 2009 00:44:11 +0100
Daniel Ritz <daniel.ritz-ml@...ssonline.ch> wrote:
> On Tuesday 03 February 2009 07:59:29 Andrew Morton wrote:
> > On Sat, 31 Jan 2009 23:09:01 +0100 Max Weninger
> > <max.weninger@...il.com> wrote:
> [snip]
> > > > > After comparing the usbtouchscreen.c sources between the
> > > > > different versions I got it working by commenting the
> > > > > follwoing lines
> > > > >
> > > > > ...
> > > > > #ifdef CONFIG_TOUCHSCREEN_USB_EGALAX
> > > > > /* ignore the HID capable devices, handled by usbhid
> > > > > */ {USB_DEVICE_HID_CLASS(0x0eef, 0x0001), .driver_info =
> > > > > DEVTYPE_IGNORE}, {USB_DEVICE_HID_CLASS(0x0eef,
> > > > > 0x0002), .driver_info = DEVTYPE_IGNORE}, ...
> > > > >
> > > > > Now the entry in /proc/bus/input/devices is created as in the
> > > > > older versions.
> > > > >
> > > > > So it seems that usbhid is not correctly handling my type
> > > > > of touchscreen if I understand the comment there correctly
> > > > >
> > > >
> > > > So... how do we fix this for real?
> > >
> > > Well I am really not an expert in kernel programming :-)
> > > I found out that these lines where added based on this thread
> > > from 2008
> > >
> > > http://kerneltrap.org/mailarchive/linux-kernel/2008/6/28/2286264/thread
> > >
> > > So all I can do is to provide any help in testing if needed
> > >
> >
> > (optimistically cc'ing Daniel).
>
> looks like it worked :)
>
> ermm, that ignoring code...seems like i was just a little bit stupid
> when i wrote it...
>
> Max, please try the attached patch...
I patched usbtouchscreen.c in my 2.6.28.2 kernel
But if I look in /proc/bus/usb/devices
the touchscreen is now managed from the usbhid driver
so usbtouchscreen is again ignoring my device
/proc/bus/input/devices is again empty and no
entry exists for the touchscreen :-(
This is the same behaviour that I had before
copying back "my patched" version and everything is
working again.
Just to make sure I understand how this should work
the entry must be there also if usbhid is managing
the touchscreen right?
BTW: dmesg contains a new line with the patched version
generic-usb: probe of 0003:0EEF:0001.0002 failed with error -32
when initializing the touchscreen
So sorry no success here
But I have to say that I got this touchscreen very cheap
second hand and maybe this is just some stupid problem
only for this device.
If your code works for all the other devices you tested
I am fine doing my change for now and you should not
waste more time on this :-)
Thanks
Max
> rgds
> -daniel
>
> --------
>
> [PATCH] usbtouchscreen: fix eGalax/ETTI HID device ignoring
>
> Fix the eGalax/ETTI HID device ignoring by adding a match for the
> protocol too. This is what the marco USB_DEVICE_HID_CLASS() was
> supposed to do anyways.
>
> Devices confirmed to be real HID have the class set to HID and the
> protocol set to 'mouse'. Some have HID class but protocol set to
> 'none'. Those are not HID and should be driven by usbtouchscreen.
>
> Signed-off-by: Daniel Ritz <daniel.ritz@....ch>
>
> diff --git a/drivers/input/touchscreen/usbtouchscreen.c
> b/drivers/input/touchscreen/usbtouchscreen.c index 5080b26..ba6237a
> 100644 --- a/drivers/input/touchscreen/usbtouchscreen.c
> +++ b/drivers/input/touchscreen/usbtouchscreen.c
> @@ -118,7 +118,8 @@ enum {
>
> #define USB_DEVICE_HID_CLASS(vend, prod) \
> .match_flags = USB_DEVICE_ID_MATCH_INT_CLASS \
> - | USB_DEVICE_ID_MATCH_DEVICE, \
> + | USB_DEVICE_ID_MATCH_DEVICE \
> + | USB_DEVICE_ID_MATCH_DEV_PROTOCOL, \
> .idVendor = (vend), \
> .idProduct = (prod), \
> .bInterfaceClass = USB_INTERFACE_CLASS_HID, \
>
>
Download attachment "signature.asc" of type "application/pgp-signature" (198 bytes)
Powered by blists - more mailing lists