cancel
Showing results for 
Search instead for 
Did you mean: 

STM32U073 bug in SystemCoreClockUpdate()

SZano
Associate III

I'm using STM32CubeMX 6.12.0, with the firmware package U0_V1.1.0.
When generating code for a STM32U073, the file system_stm32u0xx.c is generated, which contains the implementation of the SystemCoreClockUpdate() function.
This file seems to be a mere copy of the firmware package's file "<...>\STM32Cube\Repository\STM32Cube_FW_U0_V1.1.0\Drivers\CMSIS\Device\ST\STM32U0xx\Source\Templates\system_stm32u0xx.c"
I'm attaching it for reference.
On line 330 we have:

tmp = AHBPrescTable[((RCC->CFGR & RCC_CFGR_HPRE) >> 4U) & 0xFU];

It seems that this should actually be
tmp = AHBPrescTable[((RCC->CFGR & RCC_CFGR_HPRE) >> RCC_CFGR_HPRE_Pos) & 0xFU];

since RCC_CFGR_HPRE_Pos is 8, not 4 (as defined in stm32u073xx.h:5524).
The RM agrees with the value 8: §5.4.3 shows HPRE[3:0] in position [11:8] in RCC_CFGR.

Also note that stm32u0xx_hal_rcc.c:396 has the correct value (it uses the macro RCC_CFGR_HPRE_Pos, in HAL_RCC_OscConfig()).

 

Note also that the same bug potentially affects the system_stm32 file for every other MCU family too, since the shift amount is always hardcoded as 4 rather than use the macro, though other families might actually be fine with the value 4.

2 REPLIES 2
SofLit
ST Employee

Hello @SZano 

I'm analyzing it and get back to you as soon as possible.

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.
SofLit
ST Employee

Internal ticket number for follow up: 190058 (not accessible by the community users).

I will get back to you for any feedback.

 

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.