This logs to a tracefs filesystem which can be viewed in tools such as
gpuvis to see precise timings of activities in relation to other trace
markers in X or graphic drivers.
This patch is loosely based on D23114. Though modified with thread
safety, support for string building, and a RAII pattern for durations.
Ultimately that expanded it somewhat.
Currently, we estimate the expected render time purely based on the
latency policy.
The problem with doing so is that the real render time might be larger,
this can result in frame drops.
In order to avoid frame drops, we need to take into account previous
render times while estimating the next render time. For now, we just
measure how long it takes to record rendering commands on the CPU.
In the future, we might want consider using OpenGL timer queries for
measuring the real render time, but for now, it's good enough.
At the moment, our frame scheduling infrastructure is still heavily
based on Xinerama-style rendering. Specifically, we assume that painting
is driven by a single timer, etc.
This change introduces a new type - RenderLoop. Its main purpose is to
drive compositing on a specific output, or in case of X11, on the
overlay window.
With RenderLoop, compositing is synchronized to vblank events. It
exposes the last and the next estimated presentation timestamp. The
expected presentation timestamp can be used by effects to ensure that
animations are synchronized with the upcoming vblank event.
On Wayland, every outputs has its own render loop. On X11, per screen
rendering is not possible, therefore the platform exposes the render
loop for the overlay window. Ideally, the Scene has to expose the
RenderLoop, but as the first step towards better compositing scheduling
it's good as is for the time being.
The RenderLoop tries to minimize the latency by delaying compositing as
close as possible to the next vblank event. One tricky thing about it is
that if compositing is too close to the next vblank event, animations
may become a little bit choppy. However, increasing the latency reduces
the choppiness.
Given that, there is no any "silver bullet" solution for the choppiness
issue, a new option has been added in the Compositing KCM to specify the
amount of latency. By default, it's "Medium," but if a user is not
satisfied with the upstream default, they can tweak it.
Plasma Mobile announced that they plan to drop support for Halium
devices, see the announcement blog post [1] for the reasons that led to
such a decision.
But just to summarize, here are some of the key points from the post:
* Some of our team members no longer have access to reference LG Nexus
5X device anymore
* After KDE Neon switched to using Ubuntu 20.04 we no longer are
updating the rootfs for halium devices
* After several important architecture changes in upstream KWin, the
hwcomposer backend might be broken and we have no way of verifying it
If the community members are interested in reviving the hwcomposer
backend,
* it pretty much needs rewrite/re-thinking given differences of hwc1
and hwc2 API for hwcomposer part of it, see also [2]
* It also needs removal of Android 5 based libhardware API as we don't
think code can be kept sane with 3 different levels of ifdefs
* This backend needs better way of fixing difference between
CAF/non-CAF devices then just recompiling with different headers,
maybe env vars?
* This backend does not support various things like transformation/
rotation etc, and is not exactly feature complete as the DRM backend
[1] https://www.plasma-mobile.org/2020/12/14/plasma-mobile-technical-debt.html
[2] 83f563c339
This change introduces a new component - ColorManager that is
responsible for color management stuff.
At the moment, it's very naive. It is useful only for updating gamma
ramps. But in the future, it will be extended with more CMS-related
features.
The ColorManager depends on lcms2 library. This is an optional
dependency. If lcms2 is not installed, the color manager won't be built.
This also fixes the issue where colord and nightcolor overwrite each
other's gamma ramps. With this change, the ColorManager will resolve the
conflict between two.
We need the multimedia component only to play preview videos in a KCM.
The find_package(Qt5Multimedia) is not needed because we already check
if Qt5Multimedia QML module is present using ecm_find_qmlmodule().
The main motivation behind this change is to work around constant build
failures caused by buggy config file of the Qt5::Multimedia component.
Set -DLIBINPUT_HAS_TOTEM per file rather than per target so that all
targets that build it have the define.
This particular file is also used by some tests and this fixes the
warning that this ifdef resolves at once for all targets.
One of the annoying things about EGL headers is that they include
platform headers by default, e.g. on X11, it's Xlib.h, etc.
The problem with Xlib.h is that it uses the define compiler directive to
declare constants and those constants have very generic names, e.g.
'None', which typically conflict with enums, etc.
In order to work around bad things coming from Xlib.h, we include
fixx11.h file that contains some workarounds to redefine some Xlib's
types.
There's a flag or rather two flags (EGL_NO_PLATFORM_SPECIFIC_TYPES and
EGL_NO_X11) that are cross-vendor and they can be used to prevent EGL
headers from including platform specific headers, such as Xlib.h [1]
The benefit of setting those two flags is that you can simply include
EGL/egl.h or epoxy/egl.h and the world won't explode due to Xlib.h
MESA_EGL_NO_X11_HEADERS is set to support older versions of Mesa.
[1] https://github.com/KhronosGroup/EGL-Registry/pull/111
Night Color adjusts the color temperature based on the current time in
your location. It's not a generic color correction module per se.
We need a central component that can be used by both night color and
colord integration to tweak gamma ramps and which will be able to
resolve conflicts between the two. The Night Color manager cannot be
such a thing because of its very specific usecase.
This change converts Night Color into a plugin to prepare some space for
such a component.
The tricky part is that the dbus api of Night Color has "ColorCorrect"
in its name. I'm afraid we cannot do that much about it without breaking
API compatibility.
krunner stuff doesn't really belong in kwin, it has nothing to do with
compositing or any other things that are the domain of compositors.
Given that, being as a plugin suits the krunner integration stuff best.
This change introduces basic colord integration in wayland session. It
is implemented as a binary plugin.
If an output is connected, the plugin will create the corresponding
colord device using the D-Bus API and start monitoring the device for
changes.
When a colord devices changes, the plugin will read the VCGT tag of the
current ICC color profile and apply it.
The scripting api is not suitable for implementing all features that
should not be implemented in libkwin. For example, the krunner
integration or screencasting are the things that don't belong to be
compiled right into kwin and yet we don't have any other choice.
This change introduces a quick and dirty plugin infrastructure that
can be used to implement things such as colord integration, krunner
integration, etc.
Without the KWindowSystem integration plugin, Wayland experience will be
negatively affected because windows created by kwin itself won't behave
as desired. Therefore it makes little sense to load this plugin at runtime.
On wayland, we know we're always going to load our internal QPA. Instead
of shipping a plugin and loading it dynamically we can use Qt static
plugins.
This should result in slightly faster load times, but also reduce the
number of moving pieces for kwin.
This also prevents anyone outside kwin loading our QPA which wouldn't
have made any sense and just crashed.
Currently, we use glFinish() to ensure that stream consumers don't see
corrupted or rather incomplete buffers. This is a serious issue because
glFinish() not only prevents the gpu from processing new GL commands,
but it also blocks the compositor.
This change addresses the blocking issue by using native fences. With
the proposed change, after finishing recording a frame, a fence is
inserted in the command stream. When the native fence is signaled, the
pending pipewire buffer will be enqueued.
If the EGL_ANDROID_native_fence_sync extension is not supported, we'll
fall back to using glFinish().
VirtualKeyboard class does not implement the relevant VirtualKeyboard
protocol but rather implements the InputMethod protcol and can in theory
be used by other input method like e.g. ibus.
Make class name consistent with what it does to avoid confusion in
future.
For now only rename of main class is done and dbus service is kept as-is
to provide retro-compatibility, when input method protocol is
implemented fully, we can think of what to do wrt the dbus interface
later when we fully implement zwp_input_method_unstable_v1 protocol.
This change introduces a shell integration class for input panel
surfaces. This effectively breaks the direct dependency between our
virtual keyboard component in kwin and the input_panel protocol,
which means that an input method server could use the layer-shell
protocol instead of the input_panel protocol.
The layer-shell protocol allows wayland clients to create surfaces that
can be used for building desktop environment components such as panels,
notifications, etc.
The support for the plasma-shell protocol will be dropped once plasma in
all its entirety is ported to the layer-shell protocol.
in XdgSurfaceClient setFrameGeometry is async,
so we can't rely on it having the final value immediately.
make setVirtualKeyboardGeometry a virtual.
in the implementation on setVirtualKeyboardGeometry
use requestedFrameGeometry() instead of frameGeometry()
When the Xwayland process crashes, we will print a warning message, but
that warning message won't show up anywhere in the terminal because the
current message severity for KWIN_XWL is way too high.
This patch lowers the message severity for the KWIN_XWL logging category
so warning messages can be seen from the terminal.
Currently in order to load an Xcursor theme, kwin uses libwayland api,
which looks really awkward because of the way how the compositor talks
to itself via the internal connection.
The main motivation behind this change is to limit the usage of kwayland
client api in kwin.
- The name of tty devices is different; put detection into
the CMakeLists and pass that as a define when building
virtual_terminal.cpp -- this means no ifdeffery for the path.
- The name of dri devices is different; FreeBSD just has one.
- On FreeBSD epoll.h if provided through a library (from a port, epoll-shim)
that emulates the Linux behavior. Make a CMake module for it to hide
differences in systems behind a module.
- Find epoll on FreeBSD and link to it
Now on FreeBSD, tries to compile virtual_terminal.cpp, which will fail because
that needs more patches.
There are several ways to handle unmapping of a wl_surface. The first
one is to destroy the associated AbstractClient instance. The second one
is to transition the AbstractClient in a special state.
The problem with the second approach is that it makes animations such as
fade out more difficult to handle since effects in kwin are geared more
towards the first approach (destroying AbstractClient).
Summary:
This change splits the XdgShellClient class to better match existing
abstractions in the xdg-shell protocol and fix a few issues related to
sending configure events.
In the new client classes, configure events are handled differently.
Instead of blocking configure events, we try to send them as late as
possible. Delaying configure events will let us merge changeMaximize()
for X11 clients and Wayland clients and it also fixes the bug where
we don't send the final configure event when user has finished resizing
a window.
Given that configure events are not sent immediately, XdgSurfaceClient
keeps the last requested frame geometry and the last requested client
geometry.
This patch doesn't intend to fix all issues in kwin's implementation of
the xdg-shell protocol. For example, we still handle surface unmapping
very poorly.
Test Plan: Tests pass.
Reviewers: #kwin
Subscribers: kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D27861
Currently, the focus tracking functionality in the zoom effect does not
work because it relies on kaccessibleapp, which is dead. Luckily for us,
there is a library called libqaccessibilityclient that provides a way
to monitor focus changes.
BUG: 421234
Sub-surface contribute to the window quads cache. So, we must re-build
it, if any sub-surface has been changed.
Differential Revision: https://phabricator.kde.org/D29131
In case Xwayland does not use multiple buffers, the currently attached
buffer is going to be destroyed if the frame window is resized. It may
render the previous and the current window pixmap invalid and thus result
in visual artifacts when an X11 client is being interactively resized.
In order to avoid the visual artifacts, this change disables support for
synchronized resizing for X11 clients if the version of Xwayland is less
than the version in which Xwayland started using multiple buffers, i.e.
1.21.
Differential Revision: https://phabricator.kde.org/D29250
Given that we now query the current X11 time stamp on Wayland, we can
enable synchronized resizing for Xwayland clients.
Differential Revision: https://phabricator.kde.org/D29250
Summary: Recent changes prevented it to build on KDE Neon.
Test Plan: Built it on Neon
Reviewers: #kwin, bshah, davidedmundson
Reviewed By: #kwin, bshah, davidedmundson
Subscribers: kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D28511
Summary:
We build some objects several times which makes it uncomfortable to develop KWin
since every time we modify something a lot of things get rebuilt. This should
help a bit although it doesn't solve all the problems.
Test Plan: Builds, tests pass
Reviewers: #kwin, zzag
Reviewed By: #kwin, zzag
Subscribers: davidedmundson, zzag, anthonyfieroni, iasensio, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D28445
Summary:
Uses the tablet classes introduced in kwayland.
Depends on D26858
Test Plan:
Scratched my tablet with a magic stick and it did things depending on the pressure.
https://youtu.be/GGx0TlNJlzs
Reviewers: #kwin, #plasma, zzag, davidedmundson
Reviewed By: #kwin, #plasma, zzag, davidedmundson
Subscribers: davidedmundson, zzag, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D26859
Summary: No need to keep them around for no reason.
Test Plan: Tested the plugins I thought could be affected. Have been using it for a couple of days without problems
Reviewers: #kwin, zzag
Reviewed By: #kwin, zzag
Subscribers: zzag, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D28062
Summary:
This functionality will instead be implemented in KScreen such that manual and
automatic output rotation can be used and configured through a single UI in
unison together.
Test Plan: Compiles.
Reviewers: #kwin, davidedmundson, zzag
Reviewed By: #kwin, zzag
Subscribers: kwin, bshah, davidedmundson, zzag
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D26036
Test Plan:
KScreen now shows "Dell Inc." instead of DEL and
"Eizo Nano Corporation" instead of ENC in output names, which
matches closer to what's written on my monitors.
Reviewers: graesslin, davidedmundson, #plasma
Reviewed By: davidedmundson, #plasma
Subscribers: apol, feverfew, ngraham, davidedmundson, mart, kwin, sebas
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D10041
Summary:
To successfully build kwin needs 13e6459a3baeebe6a7b594efd995b88f6dd7524e
in kdecoration. Otherwise it fails with: "previewclient.h:93:11: error:
'QSize KDecoration2::Preview::PreviewClient::size() const' marked
'override', but does not override QSize size() const override;"
So this is not a new requirement but just an update to the build system
to reflect the actual requirements.
Test Plan: Builds fine with the newly required version
Reviewers: #kwin
Subscribers: kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D27242
Summary:
This wasn't used on startup (ksmserver calls KProcess)
Nor for the rules dialog (kwin invokes it with QProcess)
We may as well simplify the code and drop the dependency
See also T12140
Test Plan:
Restarted
Opened the rules dialog
Reviewers: #kwin, zzag
Reviewed By: #kwin, zzag
Subscribers: kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D26835
Summary:
The ClockSkewNotifier provides a convenient way for monitoring system
clock changes. One of the key ideas was to hide platform details from
users of the class. This allows us to add a QTimer fallback path for
operating systems that don't provide anything to detect system clock
changes.
In long term, I would like to move the new class to Frameworks.
Reviewers: #kwin, davidedmundson
Reviewed By: #kwin, davidedmundson
Subscribers: davidedmundson, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D25962
Summary:
This includes support for them on libinput and turns it into fake
pointer actions.
This doesn't implement zwp_tablet, this will have to happen in an
iteration later.
Test Plan:
Been playing around with it, see video.
https://www.youtube.com/watch?v=GF1WbO8FVvU
Reviewers: #plasma, #kwin, romangg
Reviewed By: #plasma, #kwin, romangg
Subscribers: zzag, davidedmundson, romangg, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D25663
Summary:
This allows devs to run: "make clang-format" and format all files easily
using the preset KDE clang format style
This patch adds support so devs can easily test the intended formatting
now, it doesn't change the code yet.
Doing an initial run to commit everything will happen later.
Test Plan:
Ran script
git diff was full of amazingly cleaned code
See D25134
Summary:
Currently kwin opens a second ICE connection to ksmserver in order to
tell the state of kwin's whether we're logging out and saving clients or
not.
This requires that kwin launches after ksmserver to have the connection
which is a dependency I want to break.
Practically this code is already ksmserver specific as it relies on some
custom code that sends the first saveState request to kwin first.
Instead we can replace it with a bespoke IPC over DBus and siplify the
code both end. This will allow several other future enhancements that we
want with regards to handling the session state, as well as make an
effort platform agnostic session management, as well as cleaning up some
complex code.
Ksmserver calls into kwin, rather than having kwin watch ksmserver state
to allow us make sure it's race free.
Reviewers: #kwin, zzag
Reviewed By: #kwin, zzag
Subscribers: romangg, zzag, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D24862
Summary:
The EGL platform might go away at any time through reconfiguration or because
of a graphic error. KWin then resets the graphics. The dmabuf implementation
must respect that and recover from a graphics reset by recreating all EGL
images for existing buffer.
This assumes that we won't change our graphics API mid-session and that
supported plane and modifier configuration stays constant.
In practise we remember all current dmabufs in a single map and only remove
them if the client did destroy the resource.
BUG: 411980
CCBUG: 413403
FIXED-IN: 5.17.2
Test Plan: Applied screenedge configuration without crash.
Reviewers: #kwin, zzag
Reviewed By: #kwin, zzag
Subscribers: fvogt, kwin
Tags: #kwin
Differential Revision: https://phabricator.kde.org/D24954
