2026-05-16 1:40 AM
I normally use 1.19.0 because of the MX sillyness...
When compiling you get warnings - let's take stm32h723xx.h as an example.
Line 2516 is -
#define GPIOB ((GPIO_TypeDef *) GPIOB_BASE)
In 1.19.0 you get a warning -
Description Resource Path Location Type
(near initialization for 'output_port[9]') stm32h723xx.h /Cube-Wave-723/Drivers/CMSIS/Device/ST/STM32H7xx/Include line 2516 C/C++ Problem
In 2.1.1 you get an error -
Description Resource Path Location Type
initialization of 'long unsigned int' from 'GPIO_TypeDef *' makes integer from pointer without a cast [-Wint-conversion] stm32h723xx.h /Cube-Wave-723/Drivers/CMSIS/Device/ST/STM32H7xx/Include line 2516 C/C++ Problem
As it is actually your generated code I do wonder how this has slipped through your 'testing' process.
I was awaiting a news item reporting on ST employees being treated for burns as their ears caught fire after I found this ;)
Andy.
2026-05-16 2:09 AM - edited 2026-05-16 2:13 AM
Please include more log lines and some source (C) lines. Where is that 'long unsigned int' ? initialization for 'output_port[9]' ?
2026-05-16 2:18 AM
Is this sufficient?
2026-05-16 2:47 AM - edited 2026-05-16 2:48 AM
This is a bit better. This shows that the error is in Core/code/src/numstr.c, lines 63 and 187 - but we still do not see these lines. In any case the problem is not in stm32h723xx.h but in the .c file. Feel free to fix the errors.
2026-05-16 2:52 AM
These is also the same problem in num2str.c
I thought it would be easier to see in an ST originated file.
The same project when compied with 1.19.0 shows warnings.
When compiled with 2.1.1 it shows errors.
2026-05-16 3:20 AM
Indeed, newer versions of C compilers find more errors and are more strict.
(I tried to search for "Cube-Wave-723" in ST software but got no results.)