Toradex Linux images are built using the Yocto Project/OpenEmbedded-core. The same build system can be used to generate SDKs specific to these images. A standard Yocto Project SDK includes:
For more information, refer to the Yocto Project documentation.
Whether building an SDK for the standard evaluation Toradex image or for your own custom Toradex image, the method of generating the SDK is the same. After building the image as described in the OpenEmbedded-core article, you can use bitbake to generate the SDK:
bitbake <image_name> -c populate_sdk
The generated SDK will be located at
Note: In earlier build system configurations, the deploy directory was here: 'oe-core/build/out-glibc/deploy/sdk/'.
Note: There are a number of recipes (meta-toolchain-xxx.bb) which are alternative ways to build an SDK. Among them meta-toolchain.bb, meta-toolchain-qte, and meta-toolchain-qt5 targeting use on command line, Qt4e, and Qt5 targets. They can be built as follows: 'bitbake meta-toolchain-xxx'. Note that these SDKs do not necessarily include all libraries and headers for the packages deployed in your image nor are all libraries for which the SDK does contain stuff necessarily installed in your image.
Note: The SDK filename does not change when you build for a different image_name or a different machine leading to errors like this: 'The recipe meta-toolchain-qt5 is trying to install files into a shared area when those files already exist. ...' Delete the existing file from an earlier run and restart bitbake to get around the issue.
Execute the generated SDK script file to initiate installation:
When prompted, accept the default installation path (or enter an alternate path) for the SDK.
Begin by changing into the SDK's installation directory and sourcing the environment script:
cd /usr/local/oecore-x86_64 . environment-setup-armv7at2hf-neon-angstrom-linux-gnueabi
Note: Source the environment script in every shell session which you work with the SDK.
The newly exported environment variables expand the PATH variable with the bin directory of the native sysroots (containing cross compiler etc.) and define the SDK's associated cross development tools and flags. To make use of this environment in development, use the environment variables. For example, to compile a C hello world application, use the C compiler variable
$CC hello.c -o hello
Create a new C or C++ project. Select
Empty Project and
Cross GCC toolchain.
Then, in the toolbar, click
Project and select
Properties from the drop down menu.
C/C++ Build -->
Tool Settings tab, select
[All configurations] from the configurations list.
Cross GCC Compiler from the Tool Settings menu.
Cross GCC Compiler, select
Other flags, enter:
Cross G++ Compiler from the Tool Settings menu.
Cross G++ Compiler, select
Other flags, enter:
Cross G++ Linker from the Tool Settings menu.
Cross G++ Linker, select
Linker flags, enter:
Cross GCC Assembler from the Tool Settings menu.