If a plugin adds events to an event frame that are not supported by the
target device we may eventually dereference a null pointer (for ABS_*
events) or, possibly, use an OOB index access (for buttons or keys).
Let's filter out any events that the device doesn't support immediately.
Fixes#1202
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1324>
A device may send axis events while the tool is out of proximity,
causing our plugin to force a proximity in for the pen. If the tool then
sends a proximity event for a different tool we ended up with two tools
in proximity.
The sequence in #1171 shows this:
- evdev:
- [ 1, 499608, 3, 27, 0] # EV_ABS / ABS_TILT_Y 0 (+30)
- [ 1, 499608, 0, 0, 0] # ------------ SYN_REPORT (0) ---------- +0ms
- evdev:
- [ 2, 199637, 1, 321, 1] # EV_KEY / BTN_TOOL_RUBBER 1
- [ 2, 199637, 4, 4, 30] # EV_MSC / MSC_SCAN 30 (obfuscated)
- [ 2, 199637, 1, 330, 1] # EV_KEY / BTN_TOUCH 1
- [ 2, 199637, 3, 0, 910] # EV_ABS / ABS_X 910 (+246)
- [ 2, 199637, 3, 1, 8736] # EV_ABS / ABS_Y 8736 (-105)
- [ 2, 199637, 3, 27, -25] # EV_ABS / ABS_TILT_Y -25 (-25)
- [ 2, 199637, 0, 0, 0] # ------------ SYN_REPORT (0) ---------- +700ms
Fix this by remembering that we forced the tool out of proximity so if
we see tool events for another tool we force the pen out of proximity
again.
This will have some interplay with the other tablet plugins but
hopefully none that affect real-world devices, e.g. forcing a proximity
out means the proximity out timer plugin gets disabled. Since devices
behave in unexpected manners anyway let's see if it affects a real-world
device.
Closes#1171
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1306>
Because our lua hooks don't expose the device-added callback we need to
cache any calls to disable a feature and then apply it quietly when the
device is actually added. Any other call can go straight through.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1249>
This patch adds support for Lua scripts to modify evdev devices and
event frames before libinput sees those events.
Plugins in Lua are sandboxed and restricted in what they can do: no IO,
no network, not much of anything else.
A plugin is notified about a new device before libinput handles it and
it can thus modify that device (changes are passed through to our libevdev
context). A plugin can then also connect an evdev frame handler which
gives it access to the evdev frames before libinput processes them. The
example plugin included shows how to swap left/right mouse buttons:
libinput:register({1})
function frame(device, frame)
for _, v in ipairs(frame.events) do
if v.usage == evdev.BTN_RIGHT then
v.usage = evdev.BTN_LEFT
elseif v.usage == evdev.BTN_LEFT then
v.usage = evdev.BTN_RIGHT
end
end
return frame
end
function device_new(plugin, device)
local usages = device:usages()
if usages[evdev.BTN_LEFT] and usages[evdev.BTN_RIGHT] then
device:connect("evdev-frame", frame)
end
end
libinput:connect("new-evdev-device", device_new)
A few other example plugins are included in this patch
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1192>
This patch adds a new API for enabling public "plugins" in libinput, in
addition to the exisitng internal ones. The API is currently limited to
specifying which paths should be loaded and whether to load them.
Public plugins are static, they are loaded before the context is initialized
and do not update after that.
If plugins are to be loaded, libinput will then run through those paths,
look up files and pass them to (future) plugins to load the actual
files.
Our debugging tools have an --enable-plugins and
--disable-plugins option that load from the default data paths
(/etc/libinput/plugins and /usr/lib{64}/libinput/plugins) and from
the $builddir/plugins directory.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1192>
Allows to keep track of a set of evdev masks, or at least the subset we
care about.
The EV_KEY range is split into two masks to save some memory. The future
use of this is for the plugins to use those masks and some of those will
set BTN_TOOL_PEN and friends. This would immediately create an 81 byte
mask of zeroes just to keep that one bit.
Splitting it into a key/btn mask with the latter starting at BTN_MISC
means we duplicate the infmask struct (2x16 bytes) but instead only use
8 bytes for the mask itself to keep the BTN_TOOL_PEN bits.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1271>
This test explicitly checks for devices only sending REL_WHEEL but *not*
REL_WHEEL_HI_RES. But that check only needs to run if the device has
the HI_RES axis.
The test device with REL_WHEEL_HI_RES disabled via quirk needs to be
special-cased since we cannot detect this in the test suite otherwise.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1284>
In commit 9a9466b6a9 ("evdev: discard any frame with EV_SYN SYN_REPORT 1")
all frames with a SYN_REPORT 1 were discarded on the assumption of those
being key repeat frames. Unfortunately the kernel uses the same sequence
to simply mark *any* injected/emulated event, regardless of the cause. Key repeat
events are merely the most numerous ones but as shown in commit
7140f13d82 ("evdev: track KEY_SYSRQ frames and pass them even as repeat frames")
Alt+PrintScreen is also an emulated event.
Issue #1165 details another case: keyboards with n-key rollover can
exceed the kernel-internal event buffer, typically 8 events for devices
without EV_REL/EV_ABS. Those events will be broken up by the kernel into
multiple frames - once nevents == buffer_size the current state is
flushed as SYN_REPORT 1 frame. Then, if any more events are pending
those are flushed as SYN_REPORT 0 frame. In the case of exactly 8
events, the second frame is never present, so we cannot easily detect if
another one is coming.
Issue #1145 only affects us in the touchpad code, the rest of the
backends seem to (so far) be fine. So let's move the discarding of
SYN_REPORT 1 to the touchpad backend and leave the rest of the code
as-is.
This effectively
Reverts: 7140f13d82 ("evdev: track KEY_SYSRQ frames and pass them even as repeat frames")
Reverts: 9a9466b6a9 ("evdev: discard any frame with EV_SYN SYN_REPORT 1")
Closes#1165
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1282>
config.set10 is much more convenient and nicer to read but can provide
false positive if the value is 0 and #ifdef is used instead of #if. So
let's switch everything to use #ifdef instead, that way we cannot get
false positives if the value is unset.
Closes#1162
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1277>
Make sure we drop any potential high-resolution wheel events from a
device that isn't supposed to have them.
Where the device's axes were disabled due to a quirk, re-enabling the
axes means the device's events won't be filtered anymore. Our wheel
emulation plugin thus emulates high-resolution wheel events in addition
to the hardware events.
Fix this by simply filtering out any high-resolution wheel events on any
device that uses this plugin.
Closes#1160
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1279>
Same approach as chosen in libinput-record, this leaves the F1-F10 out
but otherwise prints every other "normal" key (including modifiers) as
KEY_A.
In the future we may need some more specific approach but for now this
will do. For the use-cases where we do need some specific approach,
libinput record and libinput debug-events will still show the full
keycode on request anyway.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1276>
Our CI pipeline fails 9 times out of 10 on the valgrind tests. The tests
seem to finish in either 35 min or exceed the 60 min timeout limit, with
nothing in between. To avoid this let's split into more groups so we can
a) run those more in parallel and b) are less likely to hit the
timeout when run slowly.
Analysis of recent logs shows the eraser button tests to be the worst
offender, taking 752s (due to the combinatorial explosion) alone. The
various tip and proximity tests together also take some time so let's
group those out.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1274>
Accumulating (and partially discarding) scroll wheel events serves to
debounce a scroll wheel and provide for more fluid scrolling where
scroll wheels can send events going in the wrong direction.
This is unlikely to happen on devices with low resolution multipliers
(i.e. where some significant physical movement by the wheel is required
to trigger events) so let's make it contingent on devices more likely to
have flaky wheels.
The magic threshold picked is 30 (HID resolution multiplier of 4) as a
guess. The resolution multiplier isn't accessible in userspace so we
have to heuristically get to it - typical interaction with a mouse will
have that value set within the first two, three scroll wheel events
though.
Closes#1150
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1261>
The previous approach had a fixed threshold of 60 (half a detent)
below which scroll events were ignored. Reduce this threshold to have a
threshold of one device-specific delta. That threshold adjusts over time
to the device's individual minimum delta so after a few scroll event it
should settle on the lowest value possible.
The result is that fine-grained scrolling is possible on those devices
and only the very first scroll event is held back/swallowed, two events
in the same direction release scrolling.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1258>
When the kernel inserts a repeat frame it does that with EV_KEY code
value of 2 and the frame itself is a SYN_REPORT with value 1. Nothing in
libinput wants those repeat values, so let's discard them here before
anything tries to process them.
This inserted frame causes bugs on touchpads with EV_REP (rare enough)
because while the key event itself is dropped, the timestamp of the
frame still causes the next real frame's delta time to shorten,
resulting in wrong acceleration values.
Closes#1149
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1255>
mtdev is used only for MT Protocol A device of which there are quite
few. But that protocol is also a perfect example for event frames in ->
different event frame out so let's move this into the plugin pipeline.
Because the plugin doesn't really have full access to the device's
internals we set up mtdev base on the libevdev information rather than
just handing it the fd and letting it extract the right info.
A minor functionality change: previously mtdev-backed devices returned
zero on libinput_device_touch_get_touch_count(). Now it is hardcoded to
10 - the number of callers that care about this is likely near zero.
Because it's now neatly factored out into a plugin we can also make
mtdev no longer a strict requirement.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1245>
We have one test device that only has a horizontal scroll wheel but not
a vertical one, causing these tests to run unexpectedly.
One test needs both enabled (not strictly so but let's not bother) and
the other one only needs the vertical wheel.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1251>
This device was added before high-res scroll events existing in the
kernel and it's used in a test to verify that a device that has
ABS_MT_POSITION_X but not _Y doesn't get automatically ignored.
Said test (device_quirks_no_abs_mt_y) uses a wheel event to verify that
we do get events from this device.
Since then we've long had kernels that support hi-res scrolling and the
kernel takes care of those events for us. So let's update the device
description and the events we send to include the high-resolution
events. That doesn't change the validity of the test but stops it from
becoming a false positive.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1251>
