File /lib/udev/check_driver (referenced by /etc/udev/rules.d/60-wacom.rules) is not installed by ebuild. This script will check if the device has bound to the wacom kernel driver, and if it has not, will try to remedy that. In some cases (with USB mouses) the incorrect driver is bounded to device. Reproducible: Always Contents of file /lib/udev/check_driver: ******************** CUT *************************************** #!/bin/sh # wanted=$1 devpath=$2 bustype=$3 device=$(basename $(readlink /sys/$devpath/device)) driver=$(basename $(readlink /sys/$devpath/device/driver)) if [ x"$driver" != x"$wanted" ]; then echo -n "$device" > /sys/$devpath/device/driver/unbind echo -n "$device" > /sys/bus/$bustype/drivers/$wanted/bind fi *********************** CUT *****************************************
I added the check_driver script in linuxwacom-0.7.8_p3-r1. Please test and report back!
Thanks! Works very well. New Output from /var/log/messages: Feb 9 08:40:19 LinuxQUAD usb 4-2: new low speed USB device using uhci_hcd and address 3 Feb 9 08:40:19 LinuxQUAD usb 4-2: configuration #1 chosen from 1 choice Feb 9 08:40:19 LinuxQUAD input: WACOM CTE-640-U V4.0-3 as /class/input/input8 Feb 9 08:40:20 LinuxQUAD logger: device 4-2:1.0 is bound to the usbmouse driver Feb 9 08:40:20 LinuxQUAD logger: must rebind Feb 9 08:40:20 LinuxQUAD input: Wacom Graphire4 6x8 as /class/input/input9 Feb 9 08:40:20 LinuxQUAD logger: device 4-2:1.0 is bound to the wacom driver Feb 9 08:40:20 LinuxQUAD logger: no need to rebind
I'm confused about why this is in FILESDIR. It's not a file provided by upstream?
I think it was provided by Debian, as I believe that's where the udev rules came from. I don't think it's specific to linuxwacom, but it seems a little much for a whole ebuild to separate it out. Another solution would be to remove the catchall line at the end of the udev file I guess?
Is there some effort to get both udev rules and this script upstream, then?
(In reply to comment #5) > Is there some effort to get both udev rules and this script upstream, then? I mailed the Debian maintainer about why or why not, let's see.