WARNING: THIS PROJECT IS NO LONGER MAINTAINED AND IS INCOMPATIBLE WITH NEW VERSION OF CUBEMX. AN ARTICLE ON USING CUBE ALONG WITH CMAKE IS ON ITS WAY

From this repository:

git clone https://github.com/eugene-babichenko/cubemx2cmake.git

cd cubemx2cmake

pip install .

Or you can install it from PyPI:

pip install cubemx2cmake

Generate your initialization code from STM32CubeMX with the following Code Generation Options:

• Toolchain: SW4STM32
• [X] Generate under root

Then open your project directory and run cubemx2cmake <your_cube_mx_file>.ioc. If your working directory contains only one CubeMX project file (.ioc), you can simply run *cubemx2cmake and it will find your project file automatically. This will generate CMakeLists.txt, STM32Toolchain.cmake, openocd_debug.cfg and openocd_flash.cfg files.

To get all the things done right use -DCMAKE_TOOLCHAIN_FILE=STM32Toolchain.cmake command line option with CMake.

CMakeLists contains the separate target called flash, which compiles your code and then flashes it to the target processor with OpenOCD. By default stlink-v2 is specified as the debugging interface. You can specify any other interface with --interface command line option. You can also specify flash memory start address with --memory_start option (0x08000000 by default). Telnet and GDB ports can be changed with --telnet-port and --gdb-port (4444 and 3333 are the defaults).

This scripts also generates a bunch of shell script files:

• openocd_flash.sh - flashes the target MCU;
• openocd_debug.sh - open debugging port for the target MCU;
• gdb.sh - connect to the target MCU after openocd_debug.sh was started.