The main purpose of the edge zone is to detect palms in the area where we
cannot assume a full finger size and thus cannot use any other palm detection
mechanism. 8mm should be large enough that a finger should be detected based
on other properties (size, pressure, ...).
https://bugs.freedesktop.org/show_bug.cgi?id=103330
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit ed52002c8d)
If we're adding an element that's not null or not a freshly initialized list,
chances are we haven't removed it from a previous list.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jonas Ådahl <jadahl@gmail.com>
(cherry picked from commit 28ef7456d3)
Our own reference may be the last one that's still alive if the context is
currently suspended (litest_suspend()). If we unref before removing it from
the path interface, we access already freed memory.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 40144321b234f8cee0059cf5bada8c0845065851)
This was accidentally made 755 as part of e9fc59efc8
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit c7b2fcd2c947b795039326583f15d29cc42a5fc5)
The ...create() method returned the wrong device, so this one was never
actually used. Once we start using, we get test case failures related to the
device having BTN_foo events as well. For now, just disable those codes so we
have a keyboard with all keys and pass the tests. The rest needs better
fixing.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 4c4d400bcb)
A touchpad that was disabled by toggling the sendevents option would come back
normally after a lid resume, despite still being nominally disabled.
https://bugzilla.redhat.com/show_bug.cgi?id=1448962
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 53d53dd18e)
Devices tagged as accelerometers may also be other devices like tablet pads.
Only ignore pure accelerometer devices but disable the accelerometer axes for
devices that have multiple types.
https://bugs.freedesktop.org/show_bug.cgi?id=102100
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
(cherry picked from commit 7d97bb0087)
The current tests worked because all rings had the same range, so our error
margin covered for that. With the upcoming MobileStudio Pro 16 pad device, the
range is half and our error margins don't work anymore. Switch to a more
reliable approach that tests every integer value the wheel can send, even
though it relies on kernel filtering.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit e3e6406c4f)
We don't know the position of the third finger on 2-slot touchpads, differing
between swipe and pinch is reliable. Simply disable 3-finger pinch and always
use swipe; 3fg pinch is uncommon anyway and it's better to have one of the
gestures working reliably than both unreliably.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
(cherry picked from commit 6d435cda06)
The lid dispatch interface is a one-trick pony and can only handle SW_LID. It
ignores other switches but crashes on any event type other than EV_SW and
EV_SYN. Disable those types so we just ignore the event instead of asserting.
https://bugs.freedesktop.org/show_bug.cgi?id=101853
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
(cherry picked from commit 6bb05c594a)
If a timer_func causes the removal or addition of a different timer, our tmp
pointer from the list_for_each_safe may not be valid anymore.
This was triggered by having the debounce code trigger a middle button state
change, which caused that timer to be cancelled.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 6d0edf9d07)
Race conditions may happen where code that cancels a timer is called just
as that timer triggers. If we cancel a timer, we assume that we've put the
code into a state where the timer firing will trigger a bug.
This could be observed with the middle button code if the release event was
held back just long enough. The button release code cancelled the timer, set
the state back to idle and then complained when the timeout handling sent a
'timeout' event while being in idle.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 696fdff247)
If the kernel sends us a button press for a button that is thought to be down
we have lost track of the state of the button. Ignore the button press event,
in the hope that the next release makes things right again.
A release event may be masked if another process grabs the device for some
period of time, e.g. libinput debug-events --grab.
https://bugs.freedesktop.org/show_bug.cgi?id=101796
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 399c50dbeb)
It seems the unit tests rely on another part of <linux/input.h> which I
missed in the previous commit (5cf4b35b).
Signed-off-by: Philip Withnall <withnall@endlessm.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Various files use #include <linux/input.h> and, if the system input.h is
too old, will fail to compile. Use the internal copy by adding -Iinclude
to the build command lines. This was the case in the old autotools build
system.
Signed-off-by: Philip Withnall <withnall@endlessm.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
exec/data distinction is done based on install dir so compat scripts
must be moved in exec hook.
This should fix this occasional failure:
| install: cannot change permissions of
| ‘/usr/bin/libinput-debug-events.compat’: No such file or directory
Signed-off-by: Jussi Kukkonen <jussi.kukkonen@intel.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
If the keyboard is removed while dwt thinks it is in active state, that state
is never reset and subsequent touches are ignored.
https://bugs.freedesktop.org/show_bug.cgi?id=101743
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 9e5d1df4bb)
This isn't currently hooked up to the fdo repo but it's hooked up to my
github mirror. I had SemaphoreCI hooked up to that before but it only
supports ubuntu 14.04 and the recent meson switch made it a bit hard to setup.
CircleCI supports running docker containers, so let's do that and run against
the most recent released Fedora and a recent Ubuntu. I'm not bothering with
rawhide, it's likely to increase the work for little gain when it's in a
semi-broken state.
Run the default build with a few permutations to test meson options. Run
scan-build too but that's just for the logs, eventually this may turn into a
hard failure.
Ubuntu 17.04's meson is too old, so we have to clone that from git. Install
arguments are taken from the meson.deb package.
Most of this effort was done by Benjamin Tissoires.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Cc: Benjamin Tissoires <benjamin.tissoires@gmail.com>
(cherry picked from commit eb36957625)
Triggered an error because we still used dep_libwacom unconditionally:
Meson encountered an error in file meson.build, line 76, column 0:
Unknown variable "dep_libwacom".
https://bugs.freedesktop.org/show_bug.cgi?id=101693
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit c6e0c96e74)
Most modern touchpads are around 100mm wide, so this provides a ca 8mm edge
zone on each side. The extra 3mm should provide for more reliable palm
detection, a few touches happen to be just on the edge of the 5mm mark.
https://bugs.freedesktop.org/show_bug.cgi?id=101433
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
(cherry picked from commit 381cce8ddd)
Make the libinput page more generic but suitable for short attention spans and
most importantly, point to the xf86-input-libinput man page in a more obvious
manner since we're now shadowing that.
The rest of the man pages have punctuation and formatting cleanups only.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Meson does not have a single style but the "foo : bar" style is more common in
the docs and in our meson.build file. Make it consistent.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
The check framework takes and stores the pointer and expects it to be live for
the livetime of the test but it doesn't strdup it. We have to keep those
pointers around ourselves.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
We had one shared parsing function for all config options so tools parse
options that don't actually make sense (e.g. --quiet or --show-keycodes for
libinput-list-devices).
This patch splits the actual libinput device configuration out and reshuffles
everything to make use of that. One large patch, because splitting this up is
more confusing than dumping it all.
This means the actual option parsing is partially duplicated between debug-gui
and debug-events but hey, not everything in life is perfect.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Do so on the synaptics serial touchpads at least, they're known to cause
cursor jumps when the third finger is down. Not detecting a tap move means
three-finger taps get more reliable on these touchpads.
This change affects gestures who now effectively have to wait for the tap
timeout to happen. It's a trade-off.
https://bugs.freedesktop.org/show_bug.cgi?id=101435https://bugzilla.redhat.com/show_bug.cgi?id=1455443
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Since meson commit ae9b238 "ninja: De-dup libraries and use --start/end-group"
we get linker errors with the tools. The duplication is apparently a bit too
agressive, swapping the order here make sure libinput isn't removed.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reported-by: Michel Dänzer <michel@daenzer.net>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
The full information is now in the man page, the usage() now just tells you
how to use it. This way there's only one place to maintain it.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
