If a touch moves by more than 20mm within a single frame, reset the motion
history, effectively discarding the movement. This is a relatively common bug
and almost always needs a kernel fix, so add an explanatory page to the docs.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Don't move across the touchpad in one single event, it looks like a cursor
jump that we're trying to detect in future patches.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
25a9f39 changed the range to [-1, 1] but that's incorrect for the distance
values. Split the normalization up into two functions and make sure our
distance range is correct.
https://bugs.freedesktop.org/show_bug.cgi?id=95074
And while we're at it, sneak in a test for pressure ranges too.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
C guarantees that an enum value is previous value+1, so we might as well start
at -1000 and not have to worry about the actual values.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
Reviewed-by: Carlos Garnacho <carlosg@gnome.org>
This interface handles the buttons on the physical tablet itself, including
the touch ring and the strip.
A notable difference to other libinput interfaces here is that we do not use
linux/input.h event codes for buttons. Instead, the buttons are merely
numbered sequentially, starting at button 1. This means:
* the API is different, instead of get_button() we have get_button_number() to
drive the point home
* there is no seat button count. pads are inherently different devices and
compositors should treat them as such. The seat button count makes sense
when you want to know how many devices have BTN_LEFT down, but it makes no
sense for buttons where all the semantics are handled by the compositor
anyway.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
Reviewed-by: Carlos Garnacho <carlosg@gnome.org>
If a touchscreen has a fuzz value use it for motion hysteresis similar to how
we do it for a touchpad. This stops pointer wobbles as seen in
https://bugs.freedesktop.org/show_bug.cgi?id=94918
It's up to the system to override or set the kernel's fuzz value correctly,
i.e. a udev hwdb entry is required where the kernel driver does not set it.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jonas Ådahl <jadahl@gmail.com>
We already handle the case where we have MB emulation active and a middle
button is pressed because we often don't know if we have a middle button on
the device.
But the other way round makes little sense, when a physical middle button is
down emulation should not engage. Test for this.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Supposed to be [-1, 1] but we only generated [0, 1]
Reported-by: Carlos Garnacho <carlosg@gnome.org>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Tested-by: Carlos Garnacho <carlosg@gnome.org>
Make sure that changing the accel profile doesn't affect the default
accel profile.
Signed-off-by: Jonas Ådahl <jadahl@gmail.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
The libinput_*_get_time() returns a 32 bit unsigned integer, but in the
tests we compared them to a 64 bit unsigned integer. This means that
when the 32 bit integer overflowed, we'd still compare to a
non-overflowed 64 bit integer, causing the tests to fail.
This commit fixes this by always casting the millisecond 64 bit unsigned
integer to a 32 unsigned integer, triggering the same overflow.
Signed-off-by: Jonas Ådahl <jadahl@gmail.com>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
This path is hit when we set up an abs device without setting specific axis
ranges. Usually because we only care that the axis is there, not the values,
see the tests of invalid devices.
0xffff is fine as max for most axes but setting it for ABS_MT_SLOT can cause
allocation errors. Reduce to 100 because we don't care about the range anyway
and 100 is still big enough for basic tests.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Middle button interaction is most commonly to paste and it is a single-event
interaction (button press). We provided middle button in software button mode
by emulating it with a two-finger press with L+R down at the same time. This
is also what many touchpads are spectacularly bad at, it is very common to
detect the physical button down event before the second finger registers,
resulting in left or right clicks where a middle button should be triggered.
Unless the fingers are resting on the touchpad for at least one scanout, the
success rate for middle button emulation is only at 70% or so.
This patch adds a 25%-width middle button area between the left and the right
software button, everything else stays the same. To avoid immediate breakage,
the middle button emulation remains but may be removed in the future.
The doc is updated to only refer to the middle button area now.
https://bugs.freedesktop.org/show_bug.cgi?id=94755
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
The previous code would swap the top software buttons depending on the
touchpad's left-handed setting, not the trackpoint setting. Changing both
devices to left-handed resulted in a double-swap, i.e. the trackpoint was
always right-handed.
https://bugs.freedesktop.org/show_bug.cgi?id=94733
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Resetting the motion history has the side-effect of swallowing movements, we
don't calculate deltas until we have 4 motion events. During a finger release,
we're likely to get a large pressure change between two events, resetting the
motion history prevents the cursor from jumping on release.
The value of 7 found by trial-and-error, tested on the T440 and T450 hardware.
The absolute value is highly variable but recordings show that the pressure
changes only by 1 or 2 units during normal interaction. Higher pressure
changes are during finger position changes but since those should not cause a
jump anyway, we tend to win there too.
Currently only enabled for negative pressure changes, let's see how we go with
that.
https://bugs.freedesktop.org/show_bug.cgi?id=94379
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Had this in a private bug report recently. Missing hooks for open/close just
segfault with little information to debug. Add an assert, this is definitely a
bug in the caller and we don't need to recover from that.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
One test for an actual file (so far we only tested /tmp, the directory) and
one for a kernel dev that returns a udev device and thus gets one step further
in the error handling code.
Plus, I saw test code doing this (opening /dev/uinput) and it crashed (for
other reasons), so we might as well test it.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
A nonzero resolution on the tilt axes is units/rad so we can calculate the
physical min/max based. Uneven min/max ranges are supported.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
We previously used the half-way mark of the touchpad's y axis to decide where
to ignore tapping. Move this down to the top edge of the software buttons
instead. Users may tap with a finger in the software button areas, on the rest
of the touchpad it's unlikely that they tap within 5% of the edge.
On touchpads with physical buttons or if clickfinger is enabled, the
no-tapping zone extends to the bottom of the touchpad. This required splitting
the tests into clickfinger, softbuttons and hardbuttons.
https://bugs.freedesktop.org/show_bug.cgi?id=93947
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
The newer Cintiqs have a minimum value of 400/400 advertised by the kernel but
the actual sensor goes past the 0/0 origin. Test this, make sure that a value
outside the boundaries generates negative mm values.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
One test now gets close enough to 100% of the axis value that
ck_assert_double_lt() is not fine-grained enough. Switch to a direct x < 100.0
check.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
If the tablet is already in proximity on startup, we used to immediately sent
a proximity event. We can't fetch MSC_SERIAL from the kernel, so that tool
always had a serial of 0, followed by events with the real serial. Since
clients are supposed to use the serial for the tracking of tools, this is
suboptimal.
When the tablet is added, merely set the internal proximity flags. This way we
wait until the first real event from the device (which includes the serial
number) and convert that into a proximity event.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
The Wacom tilt range is 64 degrees so we map everything into that until we
know otherwise.
This commit also switches the tilt axes around to align the angles with the
x/y orientation, i.e. tilting the top of the stylus towards the positive x
axis now generates a positive x tilt.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Jason Gerecke <jason.gerecke@wacom.com>
If dwt is disabled on the commandline, e.g. by setting an xinput property it
may be disabled before the release event comes in. This caused the timer to
refresh indefinitely since the key state mask was still on for that key.
Always updating the key state mask (even when dwt is disabled) fixes that.
If a key is held down while dwt is disabled, this can still cause a indefinite
timer refresh, so in the timer func, check if dwt is enabled before refreshing
the timer.
https://bugs.freedesktop.org/show_bug.cgi?id=94015
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
If a USB keyboard like the YubiKey is found before the internal keyboard, it
will be paired with the touchpad when it is seen. The internal keyboard is
seen later bug ignored because we already have a keyboard paired with the
touchpad.
This is obviously wrong. For now, give priority to serio keyboards, and
override existing dwt pairings with the new keyboard.
https://bugs.freedesktop.org/show_bug.cgi?id=93983
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
If a key enables dwt and is held down when the timeout expires, re-issue the
timeout.
There is a corner case where dwt may not work as expected:
1. key down and held down
2. dwt timer expires, dwt is re-issued
3. touch starts
4. key is released
5. dwt timer expires
6. touch now starts moving the pointer
This is an effect of the smart touch detection. A touch starting after the
last key press is released for pointer motion once dwt turns off again. This
is what happens in the above case, the dwt timer expiring is the last virtual
key press. This is a corner case and likely hard to trigger by a real user.
https://bugs.freedesktop.org/show_bug.cgi?id=93984
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
This was the reason for the valgrind test case failures whenever we
accumulated too many tests (see 9c2afae1 and 2a110104). The cause was simply
that we ran out of fds which caused libevdev to fail the scandir() searching
for the event node. That resulted in a NULL devnode and an abort in litest.
Close the fd before freeing the evdev device.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Last argument is a boolean whether we want to have a tap timeout. It used to
be the ms to sleep, obsolete since e4adbff919
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
If a finger is resting in the software button area, it must not be counted
towards the gesture. So a two-finger movement must be a scroll event, not a
three-finger pinch.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
This button sends a release N, press N+1 on each press, cycling through the
three event codes supported. This causes a stuck button since the current mode
is never released.
Long-term this better served by a set of switches that toggle accordingly, for
now disable the button codes.
https://bugs.freedesktop.org/show_bug.cgi?id=92127
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
