This replaces check. The code is a copy of pwtest which I wrote years
ago for pipewire but adjusted for us here the last few days.
There are a few advantages over check:
- Ability to SKIP tests or mark them as NOT_APPLICABLE, the latter
of which is used for early checks if a device doesn't meet
requirements.
- it captures stdout/stderr separately
- colors!
- YAML output format makes it a lot easier to read the results and
eventually parse them for e.g. "restart failed tests"
Less abstraction: we set up the tests, pass them to the runner and run
them with the given number of forks. This is an improvement over before
where we forked into N test suites which each called check which then
forked again. Since we're now keeping track of those processes
ourselves we can also write tests that are expected to fail with
signals.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1067>
list_append() came later than list_insert() and there's an argument to
be made that tests added later should be run first since they're less
likely to succeed. But it's a lot harder to read test logs when they are
in reverse order, and with the TEST_COLLECTION() macro the order of the
test suite is not obvious anyway.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1067>
Instead of extracting the suite name from the test's file name use the
current suite that is being parsed. This way we pave the way for
multiple suites in the same file.
This uses a global because otherwise we'd have to redo all the
litest_add() functions but it does the job here.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1065>
This is the first step in switching away from the check framework.
Our litest macros already do almost exactly the same anyway so most of
this is a simple sed with a few compiler fixes where things mismatch
(nonnull -> notnull) and (_tol -> _epsilon).
This now generates a whole bunch of integer mismatch warnings: check
casts everything to intmax_t whereas we use typeof, so lots of warnings
especially for enums.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1059>
Kernel commit 206f533a0a7c
"Input: uinput - reject requests with unreasonable number of slots"
limits the number of slots to 99 - let's manually adjust that so we can
keep creating uinput devices.
Since these are just a test device and we don't use the slots here
anyway (they're all fake MT devices) we can manually work around this.
The real devices won't be affected by this since this is a limitation
in uinput, not the input subsystem.
Also move the comment one line up in the ms-surface device, the previous
comment referred to the wrong event code.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1061>
Wraps libinput_dispatch() with a location which will make things a bit
easier to track. Output (in --verbose) is something like:
gestures_swipe_3fg_unaccel_fn():1346 - dispatching
Which makes it easier to associate the various calls to libinput
dispatch with the other output from libinput.
This patch switches all uses of libinput_dispatch() in test cases over
but not the litest functions that may call dispatch too. Remains to be
seen if that is necessary.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1048>
This was always intended but a bug prevented the actual abort.
strstr returns NULL when we cannot find the substring so we always
triggered the first noop condition on bugs.
Fixes: bd7b91065b ("evdev: warn if our event processing lags by 10ms or more")
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/1048>
In touchpad_2fg_scroll_return_to_motion we sometimes fail because the
timeout is too close to the actual timeout expiry, creating a race
condition on whether the scroll stop event (triggered by the gesture
end) is sent before or after the timeout expiry.
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/984>
Some tablets such as those in the XP-PEN PRO series use "dials" which
are actually scrollwheels and emit EV_REL events. These should not be
emulated as rings (which are absolute) so we must expose them as a new
tablet event.
Adds LIBINPUT_EVENT_TABLET_PAD_DIAL that work largely identical as our
high-resolution wheel events (i.e. the values are in multiples or
fractions of of 120). Currently supports two dials.
This is a lot of copy/paste from the ring axes because the interface is
virtually identical. The main difference is that dials give us a v120
value in the same manner as our scroll axes.
Notes:
- REL_DIAL is mutually exclusive with REL_WHEEL, we assume the kernel
doesn't (at this point) give us devices with both. If this changes for
devices with three dials (wheel + hwheel + dial) we need to add code
for that.
- REL_DIAL does not have a high-resolution axis and we assume that any
device with REL_WHEEL_HI_RES will also have REL_HWHEEL_HI_RES (if the
second wheel exists).
- With dials being REL_DIAL or REL_WHEEL there is no possibility of
detecting a finger release (the kernel does not route EV_RELs with a
value of zero). Unless this is implemented via a side-channel - and it
doesn't look like any hardware that supports dials does that - we
cannot forward any information here. So unlike absolute rings we
cannot provide a source information here.
Closes#600
Co-authored-by: Peter Hutterer <peter.hutterer@who-t.net>
Part-of: <https://gitlab.freedesktop.org/libinput/libinput/-/merge_requests/967>
If a key is still down when the tablet mode switch goes on, make sure we
release the key before the switch goes in effect.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
