| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
This is useful for applications such as PulseAudio to identify and
recognize sound cards by their id (serial) or their path.
|
|
|
|
|
|
| |
It is more likely that the vendor id is in the databases than the
product id. Hence, check for the vendor id to figure out whether
pci-db/usb-db still needs to be run.
|
|
|
|
| |
device
|
| |
|
|
|
|
| |
As discussed with Kay Sievers in #udev.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
libgphoto supports a lot of old cameras which don't speak PTP yet and have a
proprietary protocol. With a recent patch [1] libgphoto now
creates udev rules which tag those with ID_GPHOTO2.
Discussed on devkit-devel@ [2].
[1] https://sourceforge.net/tracker/?func=detail&aid=2801117&group_id=8874&atid=308874
[2] http://lists.freedesktop.org/archives/devkit-devel/2009-June/000191.html
|
|
|
|
|
|
|
| |
Without it, /dev/dri/card* is not accessible to the user, thus no X.org
acceleration love.
This coincides with the equivalent rule in old hal's 20-acl-management.fdi.
|
|\ |
|
| |
| |
| |
| |
| | |
Reported on hal ML as hal-info patch by Francis Russell
<francis@unchartedbackwaters.co.uk>.
|
| |
| |
| |
| |
| | |
Reported by Stefan Kriwanek <kriwanek@physik.hu-berlin.de> as hal-info
patch on hal ML.
|
| | |
|
|/ |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
This makes it much simpler to use libudev from programs already using
GObject. It also makes it possible to easily use libudev from other
programming languages, such as Javascript, because of GObject
introspection support.
|
| |
|
| |
|
|
|
|
|
|
| |
Add a KERNEL=="input*" check to the input module matches, to ensure that the
ATTRS{name} check is really done against the input device, not to any random
name attached to a subdevice. Thanks to Kay Sievers!
|
|
|
|
|
|
|
|
| |
Read [dmi/id]sys_vendor just once into a variable and only use the
variable in all the rules. Short-circuit rules file if vendor is empty.
Also fix the ThinkPad/Asus Extra Buttons rules to check parent's name as
well, for more robustness.
|
|
|
|
|
|
|
| |
Using DEVNAME=="..." is invalid, we mean KERNEL=="..." here.
Instead of adding this to individual rules, short-circuit the entire keymap
rules file for non-event devices.
|
|
|
|
|
|
| |
The name "Sony Vaio Keys" is attached to the parent input device (like
"input5"), the evdev device which we actually want (input/event5) is a child of
it. Update udev rules accordingly.
|
| |
|
| |
|
| |
|
| |
|
|\
| |
| |
| | |
git+ssh://master.kernel.org/pub/scm/linux/hotplug/udev-extras
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| | |
findkeyboards prints out the device names of all "real" keyboards (atkbd and
the separate modules as found in ThinkPads or Sonys). This makes it much easier
to use a tool like keymap -i or do other debugging.
|
| |
| |
| |
| | |
This was forgotten to be added to EXTRA_DIST for building source distributions.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Add -i/--interactive mode to keymap tool, for easy debugging of unknown and
wrong keymaps. Use like
sudo /lib/udev/keymap -i input/event4
and note down the wrong/missing keys.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This patch adds persistant device node symlinks for ALSA control
devices. This mimics the behaviour of the persistancy rules for other
subsystems, such as v4l or storage.
This patch is useful only when alsa-lib with this patch is used:
http://git.alsa-project.org/?p=alsa-lib.git;a=commit;h=33ab0b5381c87e151e87e5000964edd868484886
This will allow referring to ALSA devices with device strings such as:
hw:/dev/snd/by-path/pci-0000:00:1a.1-usb-0:1:1.0
instead of
hw:0
where the device index might be unstable.
|
| |
| |
| |
| |
| | |
Reported in https://launchpad.net/bugs/356019 by Roger E Critchlow Jr
<rec@elf.org>.
|
| |
| |
| |
| | |
Verify that we have added all key map files to Makefile.am's _DATA list.
|
| | |
|
| |
| |
| |
| | |
Verify that maps referred to in the rules files actually exist in keymaps/.
|
| |
| |
| |
| |
| |
| | |
Fix thinko in keymap/check-keymaps.sh: keys.txt is not in ${srcdir}, but
in the build dir. This broke "make check" when using a separate build
tree.
|
| |
| |
| |
| |
| | |
Automake falls over with the previous glob when building in a separate
build tree.
|
| |
| |
| |
| |
| | |
Add keymap for HP TX2, as reported on the hal ML by Stephen Sugden
<glurgle@gmail.com>. Translated from hal-info patch.
|
| |
| |
| |
| |
| |
| |
| |
| | |
Fix the "auto brightness" substitute key on HP from f21 to f23 to be
consistent with other mappings, as per recommendation on
http://people.freedesktop.org/~hughsient/quirk/quirk-keymap-try.html.
Reported on hal ML by Stephen Sugden.
|
| |
| |
| |
| | |
As reported on hal ML.
|
|/ |
|
| |
|
|
|
|
|
|
|
| |
Add keymap/check-keymaps.sh and run it in make check.
This verifies that all key names in keymap/keymaps/* are known in the
current <linux/input.h>.
|
|
|
|
|
|
| |
keymap's Makefile uses awk. Instead of hardcoding it, use AC_PROG_AWK properly
and use the detected one in the Makefiles. Thanks to Dan Nicholson for the
suggestion.
|
|
|
|
|
| |
We detect gperf presence and path in configure, so we should use it in keymap's
Makefile. Thanks to Dan Nicholson for noticing!
|
| |
|
| |
|