EEPROM Emulation issue on STM32G0B0CET6 in RTOS task

[saikrishna_jarsh]

Hi,

I'm currently using an STM32G0B0CET6 and encountering an issue with writing a uint32_t variable to EEPROM using Write_U32_Variable(). The function works correctly during task initialization and in normal bare metal while loop too, but it fails when I try to save the variable at runtime (EE_WRITE_ERROR).

This code was previously working without issues on an STM32G0B0KET6. I have configured the EEPROM emulation with START_PAGE_ADDRESS set to 0x08040000U and NB_OF_VARIABLES defined as 781.

Could you help review or suggest changes to ensure compatibility with STM32G0B0CET6?

void DaQ(void *argument)
{
  /* USER CODE BEGIN DaQ */
	Write_U32_Variable(__OPERATING_MODE, 12); // Working
	/* Infinite loop */
	for(;;)
	{
	  Write_U32_Variable(__OPERATING_MODE, 14); // Non-working
	  osDelay(30000);
	}
  /* USER CODE END DaQ */
}

bool Write_U32_Variable(uint32_t variable_id, uint32_t value){
	/* Declaring 8-bit unsigned integer to store the EEPROM read/Write Status */
	bool return_variable = true;

	uint8_t EE_Status = 0x00;

	/* Computing the Half Word to be put into  */
	uint16_t half_word_position = PERSISTENCE_FLASH_MEMORY_U32_VARIABLES_START_INDEX + (uint16_t)(variable_id*2);

	/* Declaring a variable to store the computed half word */
	uint16_t original_half_word = 0x0000;

	/* Declaring a variable to store the read half word */
	uint16_t modified_half_word = 0x0000;

	/* ----Adding LSB of Data into Slot 1---- */
	original_half_word = (uint16_t)(value&0x0000FFFF);

	if(__is_hardware_initialized == 1){
		__disable_irq();
		taskENTER_CRITICAL();
	}
	
	HAL_FLASH_Unlock();

	/* Writing the half word into the memory */
	EE_Status = EE_WriteVariable16bits(half_word_position, original_half_word);			// STOPPING HERE


	/* Locking the Flash memory */
	HAL_FLASH_Lock();

	if(__is_hardware_initialized == 1){
		taskEXIT_CRITICAL();
		__enable_irq();
	}

	/* Reading the memory contents to confirm the data written */
	EE_Status = EE_ReadVariable16bits(half_word_position, &modified_half_word);


	/* Checking for Successful Write operation */
	if(original_half_word!=modified_half_word){

		/* Setting Write Operation Error */
		__ERROR_SET(__FLASH_EEPROM_WRITE_ERROR);
	}

	/* ----Adding MSB of Data into Slot 2---- */
	original_half_word = (uint16_t)((value&0xFFFF0000)>>16);


	if(__is_hardware_initialized == 1){
		__disable_irq();
		taskENTER_CRITICAL();
	}

	/* Unlocking the Flash Memory */
	HAL_FLASH_Unlock();

	/* Writing the half word into the memory */
	EE_Status = EE_WriteVariable16bits(half_word_position+1, original_half_word);

	/* Locking the Flash memory */
	HAL_FLASH_Lock();

	if(__is_hardware_initialized == 1){
		taskEXIT_CRITICAL();
		__enable_irq();
	}

	/* Reading the memory contents to confirm the data written */
	EE_Status = EE_ReadVariable16bits(half_word_position+1, &modified_half_word);

	/* Checking for Successful Write operation */
	if(original_half_word!=modified_half_word){

		/* Setting Write Operation Error */
		__ERROR_SET(__FLASH_EEPROM_WRITE_ERROR);
	}

	/* Returning values to the calling function */
	return return_variable;
}

[Tesla DeLorean]

It should be basically the same IC, one in a 48-pin package, the other in 32-pin

Would check the Option Bytes to see if the FLASH is configured / banked differently, but otherwise assume the problem is elsewhere.

Does __is_hardware_initialized change at any point? I don't expect the EE code to be multi-threaded or re-entrant, so you're like going to need to serialize access to the resource.