| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
If the client tests are run from within a desktop environment, the
XDG_CURRENT_DESKTOP variable may be set, causing the client to try to
load a platform theme for the user's desktop.
The tests, however, are running against a mock Wayland compositor, so
launching a platform theme that expects to connect to a certain
compositor implementation is probably not a good idea.
And furthermore, if the gtk3 platform theme is used, it will start
binding to interfaces and create wayland objects, confusing our existing
test code.
Setting XDG_CURRENT_DESKTOP to qtwaylandtests will prevent detection of
the (outer) desktop environment, making tests more predictable.
Change-Id: I57fa76e51cecdd0cbe8be6bd075ce67e9d2892bc
Reviewed-by: David Edmundson <davidedmundson@kde.org>
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- Replaced the usages of deprecated QWheelEvent::delta() and
QWheelEvent::orientation() with QWheelEvent::angleDelta().
- Removed the tests for deprecated APIs.
Task-number: QTBUG-76491
Change-Id: I2f9a53d3236bce8ba6cee66ec1b0b933d50518aa
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| |
| |
| |
| |
| | |
Change-Id: I35aec950da0ac0d10cf1fa0c4bc1f56ba232cf89
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This doesn't test much, but adds the infrastructure needed for more
thorough tests later on.
More or less exactly matches the mocking for wl_keyboard and wl_pointer.
Change-Id: I563bb9be7ccaaf0aa83611e84d051cc307455ccb
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
|\ \
| | |
| | |
| | | |
Change-Id: Iaf121f1db2fea08dd4ce16136b8d68c94b182084
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We already have a client-side implementation in the QPA plugin,
and now we add the server-side implementation.
In order to have a nice declarative API, we let users create a
XdgOutputV1 instance and then associate it with the zxdg_output_v1
object when it is requested by the client.
If the user forgets to create XdgOutputV1 beforehand we post an error to
the client.
To anticipate protocol version 3, we send zxdg_output_v1.done when we
send wl_output.done.
The multi-output example is extended using this protocol.
[ChangeLog][Compositor] Added support for xdg-output unstable v1
Wayland extension.
Change-Id: I1ec5913d8330cc01d7d634d05a289f4dc8b4fd22
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
- Make the code related to deprecated QWaylandWlScaler compile
conditionally, only when QWaylandWlScaler is enabled.
- Replace the usages of deprecated APIs by the corresponding
alternatives.
- Fix the tests to compile when the deprecated APIs are disabled.
Task-number: QTBUG-76491
Task-number: QTBUG-76541
Change-Id: Ieba98858e970868a2cbc97df2d06bae346e30d7a
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Reduces unnecessary coupling between the compositor and client modules.
[ChangeLog][Compositor] The compositor module no longer depends on the
client module.
Change-Id: I108e56960835b6c31c450d5ae08a0a101af9422b
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | | |
Change-Id: I4e35e86f489941dcce986ba416f3fe55d968d186
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | | |
Change-Id: Ia085dbe34b6e3b7cf83b4f55c265e12b1145ab8a
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | | |
Change-Id: If5d435a227b54f566f121331385e849b615fb615
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| | |
This means that pointer events need to be followed by frame events.
wl_seat version 4 is tested by the "tst_seatv4" autotest.
Change-Id: Ifa8e6d6edc998853be7cd901003e619029fc6f68
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
|\ \ |
|
| |\|
| | |
| | |
| | | |
Change-Id: Ief5f30fc1dbd0e6531b19e723057b06e9bd419b4
|
| | |\
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Conflicts:
src/compositor/configure.json
sync.profile
tests/auto/auto.pro
Change-Id: Ia6d1512aa9ad49ac7f92ae88f23026dc0ee2ccc5
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Change-Id: I641a1a80edd0861049d079422eef28b095fc26b3
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Change-Id: I08072c03f7283fae16294b8c9dd6df0c12d6496f
Reviewed-by: Pier Luigi Fiorini <pierluigi.fiorini@liri.io>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Allows disabling of screen blanking on a per-surface basis.
[ChangeLog][Compositor] Added support for idle-inhibit
unstable v1 Wayland extension.
Change-Id: I3cd392f9e1bb3604691f2a496579aad3b87383b6
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
|/ / /
| | |
| | |
| | |
| | |
| | |
| | | |
Increase coverage by implementing a test for the available geometry.
Change-Id: Idaba38043dadea69d7494ff9caa944e224e4c228
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Make sure we test exactly the size we expect.
We used frameGeometry() to take client-side decoration margins into
account, but actually frameMargins() returns exactly that, this
means we can compute the actual size that we want to test.
Change-Id: I2f41ffdb2aa2e3a4253a1ff7038c4bc2e43cdf98
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This makes the code more compact and avoids any doubts regarding
accesses to the underlying container under iteration, if any.
Change-Id: I6cf8a6113f182ae5c4401cb44fce2a3904c67489
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
In this patch, we port Q_FOREACH loops to C++11 ranged-for loops. All cases are
trivial in the sense that either the argument is already const or is trivially
marked as const, either by qAsConst(), or, in the case of rvalues, by storing
to a const auto temporary first.
In addition, all loop bodies are clear enough to confirm that the container we
iterate over is not changed under iteration. This does not exclude cases where
a loop is prematurely exited just after calling a modifier on the container, as
that is safe, if not especially elegant.
Change-Id: I87a63f07797437d421567d60e52305391a3c4f21
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | | |
Change-Id: I5f04676e0090bcb849a765ae5365845f199de987
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
[ChangeLog][QPA plugin] Added support for middle mouse pasting through the
primary-selection-unstable-v1 protocol.
Fixes: QTBUG-66008
Change-Id: I7c8fb9aa2c856f5b6794aeab1ee75d80cad05dcd
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
|\| |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Conflicts:
src/client/qwaylanddisplay_p.h
src/client/qwaylandwindow.cpp
Change-Id: I50eb5c83a8b81e4bdb032b68d41f429b17d0a74d
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
[ChangeLog][QPA plugin] Fixed a bug where QGuiApplication::screens() and
primaryScreen() would return initial screens in the reverse order they were
added by the compositor. QGuiApplication::primaryScreen() will now return the
first output added by the compositor.
Calling forceRoundTrip in registry_global() meant it would call itself
recursively if there were additional wl_output events in the queue. This in
turn meant the screens got added in the reverse order. Instead we now add the
screen to a list of not yet initialized screens and add it properly when we've
received the required done events (wl_output and possibly zdg_output_v1).
This also has the added benefit of wl_output hot plugging not calling
forceRoundTrip().
Fixes: QTBUG-72828
Change-Id: I35c6959d6c219f65fd19d571a25b5a6cdb3f741b
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
It's not clear to me why it was failing in CI on Ubuntu 14.04 back in
2015. I can't get it to fail on my machine, and the linked logs in the
bug report are no longer available.
Let's try to remove the blacklist and fix it if it starts failing again.
Fixes: QTBUG-45108
Change-Id: I7a73bdbcee2e544480af71db135ecd73122f747c
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The existing solution was parsing compose tables on startup, it is
better to lazy initialize the compose table/state on a first key press,
instead of doing it on an application startup. This logic is inside of
the compose input plugin.
The existing code did not utilize correctly how Qt handles complex text
input. It used libxkbcommon-compose APIs to compose user input and then
passed the same input again to QPlatformInputContext (from
QWaylandInputDevice::Keyboard::sendKey), which was erroneous. This also
means that code was forcing "xkb compose", and did not respect QT_IM_MODULE
at client-side.
From commit that added compose key handling (57c4af2b18c0fb1d266b245a107fa6cb876b9d9e):
"We should expand on it in the future to handle things like resetting
the compose state on text field switching".
This is now handled by properly utilizing Qt IM framework.
Converted QWaylandInputDevice::Keyboard::sendKey into a class member function
to avoid adding one more arg (mXkbContext) to the already long argument list.
That whole function should be simplified, but that is out-of-scope for this
patch.
The reworked code uses qxkbcommon support library to reduce code duplication
between platforms and to unify behavior.
Some users might mistakenly think that this patch introduces a regression
with Qt on KDE, but it is actually a KWin/Wayland compositor bug:
https://bugs.kde.org/show_bug.cgi?id=405388
The work around on KDE is to use QT_IM_MODULE at client-side to select
input method, as KWin compositor over the wire supports only the qtvirtualkeyboard
module. Setting this envvar is not someting out of the ordinary for users
on Linux. Input method handling at compositor-side is new feature and
clearly not very well supported yet.
Task-number: QTBUG-65503
Change-Id: Ie511d950396fa2fb6cbe6672996cee9791f3ab11
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The shell integration is initialized lazily, so it's not possible to send ping
events to the client before the first window has been initialized.
This adds a simple window to the pong test.
Change-Id: I13b4a9cb802b7abe18bfc23cf8c75eb873ded3ca
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The surface(), may be a leftover cursor surface from earlier tests. Check for
an xdgSurface instead, as those are verified to be cleaned up between tests.
Change-Id: I6a2b402130814e896d335787fcb90fd8d57cafb7
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
[ChangeLog][QPA plugin] Added support for animated cursors.
Previously we would just show the first frame of the animation.
Fixes: QTBUG-48181
Change-Id: Ie06bff8950678b5ff7b7e2e50915c85905a1200b
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This used to crash the client.
Task-number: QTBUG-73801
Change-Id: I04d1c6fcc45b6d90f2ac28844e753da2cdf3ab08
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
[ChangeLog][Compositor API] The compositor API now works without OpenGL support.
This makes the compositor API work with shared memory clients only. If OpenGL
clients connect they will currently punch holes in the compositor window, but
fixing that is out of scope for this patch.
Fixes: QTBUG-74896
Change-Id: I6c1ba82f28ba9edecf380e471124e15d16f9518e
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This adds a higher level protocol on top of the server buffer extension,
providing an easy way for Qt Quick applications to use the shared
textures.
[ChangeLog] Added protocol and Qt Quick image provider for sharing
textures in graphics memory between compositor and multiple clients.
Task-number: QTBUG-73822
Change-Id: Idc41b3479d6ca37be35d9ccd7b89e9994ff17f8f
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
|\| |
| | |
| | |
| | | |
Change-Id: I3dc204fcaa71c01a80b0c622443012eb07964431
|
| | |
| | |
| | |
| | |
| | |
| | | |
Task-number: QTBUG-73524
Change-Id: Ie9a13aeae52a7576699147e5515e2ed32a3a4d1c
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When the cursor focus' wl_surface is destroyed, or the pointer leaves a
surface, we have to reset the enter serial to avoid sending illegal set_cursor
requests.
Change-Id: I0c886e4123acb4aebd325b07bf15b9d3fa8589da
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Let's simplify all those pointer()->surfaceCursor() calls.
Change-Id: I83c51f460fa2313a0b84c8c64509d48027156443
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Not all setups or themes have cursors with a matching DPI. In such cases,
wl_cursor_load_theme will then return a theme that is the closest resolution it
can get.
With the previous implementation, cursors themes without a high dpi version
would become become really tiny on high DPI displays.
This patch prevents it by setting a lower wl_surface.scale for those themes.
This also implements proper tracking of the cursor surface's entered outputs
(i.e. if the entered surface is destroyed, the scale is reset, and similarly
the following sequence of events should also be handled:
wl_surface.enter(wl_output@1)
wl_surface.enter(wl_output@2)
wl_surface.leave(wl_output@2)
In the old implementation, we would be stuck with the scale from wl_output@2,
but now we now should correctly get the scale of wl_output@1.
[ChangeLog][QPA plugin] Cursors on high DPI screens are now scaled up if the
cursor theme does not have an appropriate high resolution version.
Change-Id: Ic87d00e35612b5afdf8c2e3a4463fcfef1f1f09d
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Also refactors the mocking code for data device manager and implements its
isClean method to verify there are no leaks after each test.
MockCompositor::DataDevice now persists across get_data_device requests.
Task-number: QTBUG-73825
Change-Id: Ib5866e0c54d021e12557f9a96f27950010f2611b
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch is mostly a cleanup to prepare for implementations of
xcursor-configuration, but also fixes a couple of issues.
Most of the logic has now been moved out of QWaylandDisplay and QWaylandCursor
and into QWaylandInputDevice and QWaylandInputDevice::Pointer. QWaylandDisplay
now only contains mechanisms for avoiding loading the same theme multiple
times.
There is now only one setCursor method on QWaylandInputDevice, accepting a
QCursor and storing its values so changing scale factor doesn't require calling
setCursor again. QWaylandInputDevice::Pointer::updateCursor() is called
instead.
Cursor buffer scale is now set according to enter/leave events of the cursor
surface itself instead of the current window, this fixes incorrect buffer
scales for cursors on windows that span multiple outputs. The window buffer
scale can still be passed into the seat as a fallback until the first enter
event is received.
This also fixes a bug where the QWaylandBuffer of a bitmap cursor could be
deleted while it was being used as a cursor.
[ChangeLog][QPA plugin] Fixed a bug where the DPI of bitmap cursors were not
sent to the compositor, leading to the compositor incorrectly scaling the
cursor up or down.
[ChangeLog][QPA plugin] Fixed a bug where bitmap cursor hotspots were off when
the screen scale factor was different from the bitmap cursor device pixel
ratio.
Task-number: QTBUG-68571
Change-Id: I747a47ffff01b7b5f6a0ede3552ab37884c4fa60
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The QVector m_surface is modified in both threads.
Change-Id: I1818a5e1307d191f1613513b86703eaa7bda3c6e
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Version 5 adds frame events, which groups multiple pointer events together,
enabling diagonal scrolling, leave and enter without an intermediate state, and
together with the new axis_source, axis_discrete and axis_stop also adds what
needed to differentiate between wheel events and touchpad scrolling.
This patch adds scroll phases and pixel deltas to QWaylandInputDevice, and
makes sure handleWheelEvent is called accordingly.
[ChangeLog][QPA plugin] Pixel delta is now set for mouse scrolling events if
originating from an appropriate device such as a touch pad (requires compositor
support for wl_seat version 5 or later).
Fixes: QTBUG-69876
Fixes: QTBUG-63720
Change-Id: I094a1ef0365893bee135cae7e6df89fafdafa2f2
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This used to cause protocol errors.
Task-number: QTBUG-71734
Change-Id: Ic937210fc42c93f1d411fb0fb4f269de01f07b5b
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
[ChangeLog][QPA plugin] Fixed a bug where pasting from the clipboard would
prefer ASCII over UTF-8, causing loss of special characters.
The bug was fixed in QtBase, this is just the test.
Task-number: QTBUG-54786
Change-Id: Ibe08bf455ad0be8fdd7a3e082dec6131f19c25d6
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| | |
Change-Id: Id89ec40a3f484808385401829a09a43781a1bed7
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| | |
Change-Id: Ib28be5277af9145834c7808f993c747e21845616
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
[ChangeLog][Compositor] Add deprecated support for the deprecated wl_scaler.
Some clients are stuck on the unofficial Weston protocol wl_scaler instead of
the standardized and stable wp_viewporter. This implements optional support for
wl_scaler.
This patch has been written to be easily removable in Qt 6, and touches a
minimal amount of other code.
The implementation can be tested using the new manual test compositor:
tests/manual/wlscaler.
Weston 1.11.1 has a demo client named weston-resizor, which can be used to
test. Note that although newer versions of Weston also have weston-resizor,
those versions use wp_viewporter instead of wl_scaler, so you really need
1.11.1 or earlier to test.
Fixes: QTBUG-69635
Change-Id: Id572e28044cd764eaf63b1259a92bc589711c39d
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The plugins subdir which includes the platform plugins was not built without
the feature wayland-server. Moves the checks inside the plugins folder to make
it build.
Also disables some tests that won't pass without the compositor API.
Fixes: QTBUG-73169
Change-Id: Ie68badd2ff62ee544c31edaff7c732082e0a4115
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|
| |
| |
| |
| |
| | |
Change-Id: Ib9dbad2e7312dab94cecbc1a14315dcb47c8c46a
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
|