The main motivation behind this change is to unify render target
representation across opengl and software renderers and avoid accessing
the render backend directory in order to get the render target.
GLRenderTarget doesn't provide a generic abstraction for framebuffer
objects, so let's call GLRenderTarget what it is - a framebuffer.
Renaming the GLRenderTarget class allows us to use the term "render
target" which abstracts fbos or shm images without creating confusion.
With these two actions being separate, RenderLoop can record the time spent
in endFrame (for example for multi-gpu transfers) without risking also recording
blocking swapbuffer calls, and endFrame can later be moved to output layer
In the desktop grid effect, dragging an empty area of a desktop to
another desktop swaps all windows on the two desktops. This change makes
sure that this matches only windows in the current activity.
BUG: 386769
FIXED-IN: 5.24.5
When setting-up the plasma window interface we need to call
`setOnAllDesktops()` after the actual virtual desktops have
been set.
Otherwise, the current desktop would be always added to the
interface (as if the user were unsetting the "onAllDesktops"
flag) and the plasma interface and KWin can get un-synced,
specially if there are specific window rules.
BUG: 452171
FIXED-IN: 5.24.5
On x11 when multiscreen is enabled often kwin crashes on
laptop lid close, for two reasons:
* m_outputLayer going dangling in RenderLayer: fixed by using a QPointer
* m_activeOutput going dangling in Workspace
For the latter, the output disabling was done correctly, but x11client did set the old deleted output
again in X11Client::moveResizeInternal
This is fixed by moving desktopResized() at the bottom of
Workspace::slotOutputDisabled() after the reassign of outputs to
the toplevels
In 5.24, the same code path is used for testing direct scanout, so that
causes false negatives. Generally though, the user setting shouldn't be
touched, it's not really proper feedback for the driver or KWin having
problems.
Some modes can have the exact frame timings but different flags.
Currently, the mode comparison function doesn't take that into account
which can result in the drm backend setting the current mode flag
incorrectly.
Since the Toplevel.output property is properly synchronized with the
frame geometry, Workspace::clientArea() can use the output property
instead of looking up the screen. It makes code intuitive and removes
unnecessary lookups.
Currently, if geometry updates are blocked, the Toplevel.output property
won't be updated. On the other hand, it's reasonable to use the output
property instead of manually looking up the output in window management
code, e.g. Workspace::clientArea().
In other words, using the Toplevel.output property is like walking on a
mine field, things can blow up. You can't use Toplevel.output even if it
makes perfect sense.
This change ensures that Toplevel.output property is always kept in sync
with the frame geometry. Unfortunately, this means that the output
property no longer can be updated when the frameGeometryChanged() signal
is emitted. It has to be done in moveResizeInternal() method.
If the blob is fetched while there is no kernel-visible reference to it,
the driver may re-use the blob ID. When DrmProperty is created or updated,
KWin holds a reference on the blob via drmModeObjectProperties, so this
should prevent any possible issues.
CCBUG: 449285
Using the global coordinate system when specifying output layer damage
regions would be very confusing. In order to make the coordinate system
comprehensible, use the layer-local coordinate system.
The infinite region is used to tell the Compositor when it needs to
repaint the entire layer.
In xdg_activation_v1, if a process fails to negotiate activation, at
least decorate it as demanding attention.
This will give apps ways to indicate attention is required. It can can also
help detect clients doing weird things.
Right now Workspace and SessionManagement were entangled with
SessionManager being a dumb class that just signalled to Workspace to do
the work.
The code, however, lived in sm.cpp despite being part of workspace.
This commit is an initial cleanup to strip some X11 specific features
from Workspace and move towards having the right code in the right
files. It should hopefully help get us in the right direction we get
some wayland session management.
Behaviour should be unchanged.
- added option to remove the frametime graph
- added option to remove the "this is a benchmark" message
- location of the fps counter is now on the "active" monitor by default
- removed the hard-limit of 100 for the FPS counter
- added option to color the text based off the FPS value
When changing global settings like switching between light and dark
themes the result is currently quite messy. Apps update in a scattered
order and a jarring change.
The solution is heavily inspired from Gnome, we cache the window for X
milliseconds, then perform a crossfade between the cache and current
contents.
It does mean any video is paused for half a second, but pragmatically
it's not really bothersome.
On receipt of an explicit DBus method call we create a cache and start
animating. The explicit method call is needed so that we can guarantee
kwin has a fully copy before a client might change.
Use of an offscreen texture is used so that we include any window
decorations. The DeformEffect effect is repurposed as it has most the
relevnat logic if we remove the live updating.
Doing that breaks when the input device has a very high resolution, which
is the case with touchscreens for example. The minimum delta of 5 is an
arbitrarily chosen value that can be changed if needed
When the system goes to suspend and screen locking for suspend is enabled,
suspend is inhibited until ScreenLocker::KSldApp::self()->lockScreenShown()
gets called, in order to make sure that the lockscreen is shown before the
system goes to standby, and thus also when the system wakes (instead of
potentially sensitive user information).
However, signalling that when the lockscreen gets mapped can't work reliably,
as it's then a matter of timing whether or not KWin actually presents an
image with the lockscreen before suspending. To fix that, this commit replaces
that logic with only calling lockScreenShown once every output actually got a
lockscreen presented.
* Makes the close button on thumbnails always visible: besides making it usable on touch, makes it more usable with mouse as well
* add a drag down gesture to close windows, moving opacity to zero while dragigng down to signal the window will be destroyed
* restore behavior animations: animate changes in thumbnails size or positions when not in drag mode or drag gesture to not make jarring sudden changes after closing a window or at end of drag mode
With this change, the Workspace would provide clientArea() overloads
that take only AbstractOutput and VirtualDesktop. integer ids are
obsolete as they are unstable.
When modesets are necessary, they are attempted when an output on the given
GPU gets presented. With multi-gpu setups however, the situation can arise
where there is only one disabled output on a GPU; in that case KWin eternally
waits and never properly turns off the display.
In order to work around this, explicitly call DrmGpu::maybeModeset when
an output gets disabled.
BUG: 449878
FIXED-IN: 5.24.4
When using a "legacy" colorscheme, the QPalette will be loaded twice.
With this change, the QPalette is going to be created only once. It also
cleans up variable naming.
Swipe with three fingers
- left to switch to the previous virtual desktop
- right to switch to the next virtual desktop
- up and down to toggle the overview
CCBUG: 439925
The .clang-format file is based on the one in ECM except the following
style options:
- AlwaysBreakBeforeMultilineStrings
- BinPackArguments
- BinPackParameters
- ColumnLimit
- BreakBeforeBraces
- KeepEmptyLinesAtTheStartOfBlocks
The anonymous variant of qmlRegisterType() is deprecated in Qt5 and
gone in Qt6.
This however now requires us to specify a module URI and major version
number. I used the names and (lowest) versions used already for other
types nearby where available, and fell back to "org.kde.kwin" and 0
otherwise.
In conjunction with a buffer format with alpha channel this allows
to have KWin's backdrop transparent to see through to a display
layer below.
Signed-off-by: Eike Hein <eike.hein@mbition.io>
The tablet mode code checks whether we are currently in tablet mode
based on whether or not we have a touch screen and a pointer. For some
devices, a touch screen or pointer device may be detected but we would
like to ignore it, since it incorrectly influences the tablet mode.
This reads the udev tag "kwin_ignore_tabletmode" from libinput devices
and will skip devices with that tag when trying to determine whether we
are in tablet mode or not.
ScreenLockerWatcher was written whislt overthinking DBus.
An interface does not have to have the same lifespan as the service the
other end. DBus connections works in terms of established "match rules",
these can be set up before a service is created the other side and they
remain if a service closes. The only part that needs doing is re-fetch
any cached properties.
There's also need to check the intial service names at startup. We can
just let our call to GetActive fail. It's far faster.
This class was getting convoluted in both ifdefs and Qt6 porting all to
solve something that isn't real.
There should be no behavioural differences.
It was added to work around Xwayland crashing at startup due to a mesa
issue, see commit 58fa92e40b.
It's been almost 6 years since that change. No other major wayland
compositor forces EGL_PLATFORM, so lets remove the workaround.
The main motivation behind this change is to get rid of code that's not
straightforward and should be unnecessary in general.
These are two conceptually different tasks that were intertwined.
On it's own it doesn't accomplish anything but is an important refactor
for longer term goals, namely:
- moving xwayland into kwin_wayland_wrapper with our wayland restart
handling support
- having multiple X connections
Behaviour should be the same.
support realtime activation for screenedges gestures, making it possible
for effects to show half-triggered states while dragging from
the edge with a finger, making them much more usable
In 52bc46069e, some code was shuffled
around to improve code readability. However, it was overlooked that
spa_meta_bitmap->offset is initialized too late, after QImage dest is
constructed. That's the reason why the left edge of the cursor is
wrapped around horizontally.
This change fixes the cursor glitch by ensuring that
spa_meta_bitmap->offset is initialized to proper value before getting
pointer to bitmap data. While on this, this change also moves
spa_meta_bitmap initialization code around to make spa_meta_bitmap setup
look less like a bowl of spaghetti, i.e. spa_meta_bitmap is initialized
first, and QImage dest is created with spa_meta_bitmap's values.
Virtual machines aren't properly supporting atomic mode setting yet, which
causes the cursor to be offset, and will cause more issues with overlay
planes. In order to prevent that from impacting users, fall back to legacy,
unless KWIN_DRM_NO_AMS is set.
BUG: 427060
FIXED-IN: 5.24.4
Some effects (AnimationEffect) transform windows without setting
PAINT_SCREEN_WITH_TRANSFORMED_WINDOWS flag. Make the scene disable
render optimizations if that's the case.
Whether AnimationEffect does a right thing is up to debate.
Window painting is no longer split in two phases - PAINT_WINDOW_OPAQUE
and PAINT_WINDOW_TRANSLUCENT.
PAINT_WINDOW_TRANSLUCENT is used as a hint to the occlusion culling
logic to ignore the opaque region.
Given that, the handling of the opaque region can be simplified. If no
effect sets the PAINT_WINDOW_TRANSLUCENT flag, then the opaque region
can be used as is.
It's not possible to get the surface damage before calling
Scene::paint(), which is a big problem because it blocks proper surface
damage and buffer damage calculation when walking render layer tree.
This change reworks the scene compositing stages to allow getting the
next surface damage before calling Scene::paint().
The main challenge is that the effects can expand the surface damage. We
have to call prePaintWindow() and prePaintScreen() before actually
starting painting. However, prePaintWindow() is called after starting
rendering.
This change makes Scene call prePaintWindow() and prePaintScreen() so
it's possible to know the surface damage beforehand. Unfortunately, it's
also a breaking change. Some fullscreen effects will have to adapt to
the new Scene paint order. Paint hooks will be invoked in the following
order:
* prePaintScreen() once per frame
* prePaintWindow() once per frame
* paintScreen() can be called multiple times
* paintWindow() can be called as many times as paintScreen()
* postPaintWindow() once per frame
* postPaintScreen() once per frame
After walking the render layer tree, the Compositor will poke the render
backend for the back buffer repair region and combine it with the
surface damage to get the buffer damage, which can be passed to the
render backend (in order to optimize performance with tiled gpus) and
Scene::paint(), which will determine what parts of the scene have to
repainted based on the buffer damage.
