feat: add embedded compilation using arm-none-eabi for performance sanity

[gabrielfrasantos]

add embedded compilation using arm-none-eabi for performance sanity

[gabrielfrasantos]

It should analyze for cortex m4, m7 and m33.
If needed, create a new layer called "embedded" with targets for ALL required components.
An cation should be copied and adapt from e-foc to analyze the machine cycle spent for the calculations.