Qt Wayland Compositor 是提供用于开发自定义显示服务器，方便且强大的 QML 和 C++ API 模块，基于 Wayland 协议。显示服务器 (通常称为 Compositor 合成器) 显示来自支持 Wayland 协议的客户端应用程序的内容。
Wayland's design philosophy is to keep the core protocol simple and minimal. Developers can then expand on this core protocol with use-case-specific extensions. Qt Wayland Compositor supports many common extensions by default, and also has APIs to enable the creation of new, custom extensions.
Typically, a compositor written with Qt Wayland Compositor becomes a subsystem inside a larger application manager process. Qt Wayland Compositor provides the APIs to communicate with clients and display their content on the screen. The QML APIs contain high-level APIs that easily integrate with the rest of Qt, enabling convenient animations, effects, and UI through Qt Quick. There are also C++ APIs available - if you need more low-level access.
An application manager would typically implement additional features such as application life cycle, virtual keyboard input, security, and Inter-Process Communication (IPC). Qt provides the APIs that can be used to develop the remaining parts of an application manager in other modules. The Qt Automotive Suite 提供 Qt Application Manager , which is a complete application manager that includes a compositor developed using Qt Wayland Compositor.
有关 Wayland 的更多信息，见 Wayland and Qt .
The Qt Wayland Compositor includes features necessary to create a compositor:
The Qt Wayland Compositor recognizes the following environment variables and command-line arguments:
--wayland-socket-name
Overrides the default socket name used for communicating with clients.
As long as it does not depend on any unavailable platform-specific features, the compositor can easily be tested on an X11-based desktop system. This can be useful during development, both for simplified debugging and efficient turn-around on trying out new features.
Qt Wayland supports several backends for sharing graphics buffers between clients and the server. The main ones are:
wayland-egl
: This is the default backend and should be preferred whenever possible. However, it requires support in the OpenGL driver on the system for this to work.
xcomposite-glx
: This uses the
XComposite
extension to the X11 display server for sharing buffers. It requires that an X11 server is running on the system and that the xcb platform plugin is used with the "GLX" backend. This is usually the default on X11 systems, but can be forced by setting the
QT_XCB_GL_INTEGRATION
environment variable to "xcb_glx".
xcomposite-egl
: This uses the
XComposite
extension to the X11 display server for sharing buffers. It requires that an X11 server is running on the system and that the xcb platform plugin is used with the "EGL" backend. This can be forced by setting
QT_XCB_GL_INTEGRATION
environment variable to "xcb_egl".
Select the backend by setting the
QT_WAYLAND_CLIENT_BUFFER_INTEGRATION
环境变量。
注意：
If Qt Wayland Compositor is unable to initialize the client buffer backend, then it will fall back to using the "shared memory" backend (based on
wl_shm
) as a fail-safe. This backend will use CPU memory for sharing the graphics buffers and copy the data back and forth as needed. This has performance implications, especially on high density screens and limited graphics hardware. When investigating performance issues with Qt Wayland Compositor, start by checking that the correct client buffer integration is used.
Also bear in mind that if your system is already running a Wayland compositor, you may have to set
XDG_RUNTIME_DIR
to point to a different location. If this is the case, you will see warnings when starting the compositor. The
XDG_RUNTIME_DIR
can point to any accessible location which is not already in use.
For instance, if you want to run the
pure-qml
example with the
wayland-egl
backend, you could use the following command line:
% XDG_RUNTIME_DIR=~/my_temporary_runtime QT_XCB_GL_INTEGRATION=xcb_egl QT_WAYLAND_CLIENT_BUFFER_INTEGRATION=wayland-egl ./pure-qml
The client can subsequently be run on the compositor by setting the same
XDG_RUNTIME_DIR
and by passing "-platform wayland" as a command line argument. The
QT_QPA_PLATFORM
environment variable can also be used to select the Wayland QPA plugin on the client side.
注意： In most cases, the client will adapt to the same OpenGL as the server when it connects. However, when running with the EGL backend on some specific drivers, it is required that the initialization happens earlier. If you encounter this problem, you may pass "-platform wayland-egl" instead to pre-initialize the client to EGL.
Sometimes, when you are developing a complex compositor, you may encounter issues that require further investigation.
设置
WAYLAND_DEBUG
environment variable to "1" will enable log output for the Wayland library itself. This can be very useful, for example, when debugging custom extensions to the Wayland protocol. It will show you exactly which events and requests are being passed between the client and the server, as well as timestamps for these.
In addition, Qt has logging categories
qt.waylandcompositor.*
and
qt.qpa.wayland.*
to enable additional logging. The latter should be set on the client side, as it enables logging from the Wayland QPA plugin.
查看 Qt Wayland Compositor 范例 to learn how these APIs can be used to write custom compositors.
可以使用 Qt Wayland Compositor，从 C++ 或 QML：
移植到 Qt 6 - Qt Wayland Compositor 列出了 Qt 的 Qt 6 系列在模块 API 和功能上所做出的重要改变。
Qt Wayland Compositor 和 Qt Wayland 集成插件在商业许可下是可用的来自 Qt 公司 .
此外，Qt Wayland Compositor 是可用的在 GNU GPL (一般公共许可) 第 3 版 ，而 Qt Wayland 集成插件可用于 GNU GPL (一般公共许可) 第 3 版 或 GNU GPL (一般公共许可) 第 2 版 .
见 Qt 许可 进一步了解细节。
Qt Wayland Compositor 和 Qt Wayland 集成插件所用协议定义遵循以下准许许可：