Build STM32 mcu with yocto

[BBoge.2]

Hey everybody,

I am a little bit confused about the GNU Tools for STM32. (https://github.com/STMicroelectronics/gnu-tools-for-stm32) 

You are using this ST gnu tool inside the stm32cube IDE too compile all example applications.

But when I build the same example applications with yocto you are using in your example the official arm gnu tools:

https://github.com/STMicroelectronics/meta-st-stm32mp/blob/mickledore/recipes-extended/m4projects/m4projects.inc 

https://github.com/STMicroelectronics/meta-st-stm32mp/blob/mickledore/recipes-devtools/gcc-arm-none-eabi/gcc-arm-none-eabi-native_11.bb 

https://github.com/STMicroelectronics/meta-st-stm32mp/blob/mickledore/recipes-devtools/gcc-arm-none-eabi/gcc-arm-none-eabi_11.inc 

Why do you not use the gnu-tools-for-stm32 inside of the yocto build?

And what is the main difference between the arm gnu tools and the stm32 gnu tools?

Thank you 

[Pavel A.]

>  main difference between the arm gnu tools and the stm32 gnu tools?

Please see:

https://community.st.com/t5/stm32cubeide-mcus/difference-between-the-toolchain-quot-gnu-tool-for-stm32-quot/td-p/252561

> Why do you not use the gnu-tools-for-stm32 inside of the yocto build?

This is the purpose of  yocto: you don't want anything ready-made. Only the fresh, just built right here for you, from the very sources.

[BBoge.2 · Accepted solution]

Okay, I see. You created a "fork" of the ARM GNU tools, and you created all the build scripts by yourself. For me, this is not really practical to add your STM32 GNU tools to our Yocto to build our Cortex-M4 application with the same tools as my colleagues using STM32Cube IDE. For this reason, we all need to use the ARM GNU tools to ensure we're all using the same compiler and so on.