Qt for QNX
From the perspective of an application development platform, Qt runs on top of QNX Neutrino RTOS. System integrators and application developers use a QNX SDP (Software Development Platform) which includes QNX Neutrino RTOS components for various targets and a full set of development tools.
Qt 5 is currently tested and supported on QNX SDP 7.0, with support for both 32-bit and 64-bit target architectures:
Support for 64-bit architectures was added in QNX SDP 7.0.
Older QNX SDP releases and other architectures may also work, but are not subject to regular builds and testing.
For a specific list of available boards, displays and input devices, contact QNX Software Systems (firstname.lastname@example.org) or KDAB (email@example.com). Qt on the QNX Neutrino OS is also supported by Qt for Device Creation, contact The Qt Company for more details.
All the Qt Essentials modules are supported in Qt for QNX.
See Qt Add-Ons for information about which Qt Add-Ons modules are supported in Qt for QNX.
If you build Qt from a released source code distribution archive, the
configure tool ensures that the build will be made only for modules which are known to build for the target specified. Additionally, you can explicitly exclude unsupported or not needed modules from the build via the
-skip <module> option when running the
If you build Qt directly from the Git repository, you should initialize the repository only with modules which are supported or build and install modules separately. This helps to prevent unexpected build problems.
Building Qt 5 requires the following packages available from the QNX website (registration required for access to downloads):
The compiler and other parts of the tool chain are provided in the SDP packages. You do not need any additional installations on your host PC. However, it is advisable to ensure that you can build Qt for the desktop target on the host as well. See Building Qt Sources for more details. All supported QNX Neutrino RTOS targets are fully integrated in the makespecs provided in the Qt sources.
You need to specify the target name for cross-compilation when configuring a Qt build:
|Architecture||Configure cross-compilation target|
./configure -xplatform qnx-armle-v7-qcc
./configure -xplatform qnx-aarch64le-qcc
./configure -xplatform qnx-x86-qcc
./configure -xplatform qnx-x86-64-qcc
configure tool runs various tests to detect capabilities of the SDP and its add-ons. It sets related flags according the test results. Check the configure summary on the console or in
./qtbase/config.summary if needed and set other flags manually as required.
Independent of the platform you are building for, consider using the
-prefix option with
configure as well, for example,
-prefix<Qt-install-path>. This specifies the target folder where Qt will be installed and places all build artifacts outside of the source code tree.
See Qt Configure Options for more details.
QNX Software Systems delivers software solutions for making embedded devices. This includes creation of QNX Neutrino RTOS boot images in a very flexible way. Even though QNX provides several reference scripts for the OS image creation, sooner or later you will need to make changes or you might decide to create an OS image from scratch. Qt relies on the existence of several third-party components and selected OS services. Due to this, QNX Neutrino RTOS boot images for a target device running Qt must meet a couple of requirements to ensure that Qt works as expected. The sections below list the crucial parts for Qt.
Before any Qt application can be started, the QNX Screen Graphics Subsystem (also called just "Screen") must be running. The Screen consists not only of a driver, but also includes several utilities and other services which are used by Qt. This includes processing of mouse and keyboard events as well. You can ensure that the Screen is configured and running correctly by verifying that a graphics application, like
gles2-gears, can be started and runs without problems. Most of the recent sample OS build scripts in the SDP include sections with all components and start commands for the Screen. Use them as a reference.
Qt's networking stack requires IPv6 support to be enabled, independent of whether the actual networking uses IPv4 or IPv6. That means, that
io-pkt-v6-hc must be running, not
/dev/random to be present and functional. Start it before starting any Qt application or during the system startup:
random -p waitfor /dev/random
QNX provides the slog2 logging framework tailored to specifics of embedded systems. The major advantage of it is a much better performance than a text output to files. Qt uses this framework by default for any logging output going through QDebug on QNX Neutrino RTOS. The
slog2 service should be started during the system startup. The
slog2 also relies on the existence of the
/tmp folder which has to be configured as well, for example:
[type=link] /tmp=/dev/shmem display_msg "Starting slogger2..." slogger2 -U1001:1000 & waitfor /dev/slog2
There is a set of environment variables you should set when starting a Qt application. Most of them are actually not specific to QNX Neutrino RTOS. They are still mentioned here, since knowing them just saves a lot of time in case of problems.
If you did not build Qt with RPATH enabled, you can place the Qt runtime components in any folder in the file system. The following environment variables should point to valid locations assuming that the Qt installation is located in
LD_LIBRARY_PATHshould contain the path to the Qt libraries in
QT_PLUGIN_PATHdefines where the Qt plug-ins are located. It should be set to
QML2_IMPORT_PATHdefines where the Qt Quick plug-ins are located. It should be set to
- If Qt does not use fontconfig, you must either provide the path to the fonts with
QT_QPA_FONTDIR, or place the fonts under
<Qt-install-path>/lib/fontswhich is the default font location.
Qt needs information about the physical dimensions of the attached display to determine DPI values and thus set correct font sizes. Usually, this information is provided by the Screen. In some cases you may see an invalid screen size such as 0mmx0mm. In such a case Qt requires an environment variable
QQNX_PHYSICAL_SCREEN_SIZE set to according values to get the required information. If Qt applications exit with an error message saying that the physical screen size couldn't be determined, set the variable like this before starting the Qt application:
to inform Qt that the screen has a width of 150mm and a height of 100mm in this particular case. The physical screen size is measured in millimeters (mm) and should correspond to the actual physical size of the display in the final configuration.
The only drawback of the flexibility provided by QNX Neutrino RTOS is a risk that the run environment is slightly different on your target than what is expected by Qt. This is a common reason why a Qt application on a new target fails to start. A few generic environment variables can help locate the root cause for problems:
1helps to see why the QPA plugin cannot load.
1helps to see where and how shared libraries are loaded. This can be used in combination with
QT_DEBUG_PLUGINSwhen a plugin failed to load because it could not load other shared libraries.
QT_LOGGING_TO_CONSOLEwill force sending all logging messages to
slog2. This is useful for analyzing application startup problems, as you do not need to use another tool to see the messages.
Make sure that your system image contains the following additional 3rd-party libraries which are not always included in minimal QNX Neutrino RTOS images:
Note: This is not a complete list of shared libraries used by Qt on QNX Neutrino OS. Multiple shared libraries are already available in a typical OS image, or included by other parts of the system, for example, the
getconf are required to access POSIX environment variables, especially for the host name and time and date.
A QNX Neutrino RTOS target used for development with Qt Creator should provide a few additional command line tools and services.
Deployment and remote application startup relies on SSH. The SSH daemon (
sshd) must be installed and running during the system startup. Please note that it requires
/dev/random which was already mentioned. It also expects specific access rights to be set correctly to its configuration files. The latter makes it impossible starting SSH from a file system which does not have full access rights support, for example, FAT.
sftp utilities are used by Qt Creator for deployment and should be available on the target. Additionally,
inetd daemon has to be started at the system startup which then launches according services, including FTP when needed.
Remote debugging requires the
pdebug (process-level debugger) to be running on the development device. It is usually started by the
qconn process which also provides various process information to Qt Creator. Generally, Qt Creator uses the same remote debugging interfaces as the QNX Momentics IDE provided in the SDP. See the related QNX documentation for more details.
QNX provides board support packages (BSPs) for many common platforms and their individual variants. However, if you need a BSP for a board that QNX does not provide, you can modify an existing one or develop your own.
When you develop a BSP for your board, you should include the following utilities to enable Qt Creator to deploy, run, debug, and stop applications on QNX devices:
The utilities need to be in a standard location, such as
/usr/bin. Alternatively, you can use the
~/.ssh/environment file to set variables for remote commands, such as
PATH. In addition, you must set
enable PermitUserEnvironment in the SSH Daemon configuration.
Run the test when you configure a new target device in Qt Creator to verify that all required components are available and running.
For more information about modifying or developing a BSP for you board, see Customizing a BSP in the QNX developer documentation.
Qt Creator deploys an application to the target device before starting it remotely. On most embedded devices, it is just copying application binaries and related data files to a folder on the target device. However, there is no standard for how this folder is called and where it is located. Quite a few configurations use
/tmp for this, which is often mapped to a RAM disk. This can cause some confusion as deployed applications disappear on the next device shutdown.
Another popular target is
/opt. This folder is located in a read-only partition on many systems, making deployment impossible. Make sure that you use a destination folder which meets your needs. You can set it in the
install.path variable in the
.pro file of your project, see Installing Files in the qmake documentation for more details.
© 2019 The Qt Company Ltd. Documentation contributions included herein are the copyrights of their respective owners. The documentation provided herein is licensed under the terms of the GNU Free Documentation License version 1.3 as published by the Free Software Foundation. Qt and respective logos are trademarks of The Qt Company Ltd. in Finland and/or other countries worldwide. All other trademarks are property of their respective owners.