How to change the STM32 family in a project using HAL?

[DocAlex]

Hello,

I am currently using a F4 processor in HAL together with the Stm32cubeIDE.

Thus the IDE generates F4 hal files which later are included in the main program. 

 

To access the specific functions I need to #include "stm32f4xx_hal.h" in my own programs and libraries to make the HAL available in my programs. 

 

When I later want to change the platform to an F7 processor I would need to change all the code to #include "stm32f7xx_hal.h". This is very inconvenient. 

How can I avoid this problem?

 

Thanks in advance

 

Alexander

 

[Imen.D]

Hello @DocAlex ,

To update the application code from STM32F4 to STM32F7 Library, you need to :

• Replace the stm32f4x_conf.h file with stm32F7xx_conf.h
• Replace the existing stm32f4x_it.c/stm32f4x_it.h files with stm32F7xx_it.c/Stm32F7xx_it.h
• Check and update the toolchain startup files (Project files, Linker configuration and vector table location files)

Please refer to this application note, which can help you on your project as it provides a guidance on hardware and peripheral migration : AN4660 Application note Migration of microcontroller applications from STM32F42xxx/F43xxx devices to STM32F7 Series devices.

 

[DocAlex]

Hello @Imen.D,

thank you for the migration guide. 

 

It is quite clear to replace the stm32f4x_conf.h file with stm32F7xx_conf.h files.

In reality I would set up a new F7 project where the necessary files would be generated anyway. 

My question was pointing in a different direction:

 

In my files e.g. Mylibrary.c and Mylibrary.h I need to  #include "stm32f4xx_hal.h". When I now change to project to a F7 family I need to change Mylibrary.c and Mylibrary.h and replace in these files  #include "stm32f4xx_hal.h" with  #include "stm32f7xx_hal.h".

 

Is there any way to include the HAL more general so that it includes automatically F4 or F7 HAL in Mylibrary.c and Mylibrary.h?

 

regards

 

Alexander

 

[Ozone]

You could define the target plaform (e.g. F4 and F7) on make file level, and thus pass it to each built call.
And of course make those specific includes conditional.
This way you can manage software base for multiple platforms and toolchains, although effort increases.

On a related note, you might need to compare some sections of the reference manuals carefully.
Trying to port code between F1, F3 and F4 devices, I found some subtle (and not so subtle) differences between peripherals that required substantial changes.

[Andrew Neil]

---

@DocAlex wrote:

To access the specific functions I need to #include "stm32f4xx_hal.h" in my own programs and libraries to make the HAL available in my programs. 

---

No, you don't need to do that.

CubeMX (whether standalone or as part of CubeIDE) places the  #include "stm32f4xx_hal.h" within  the auto-generated main.h.

So you just need to #include "main.h" and that will automatically give you the correct version of the HAL header