How to handle HardFault?

[AAgar.2]

There is a function called HardFault_Handler in the STM32 MCUs (full code below). This function I usually work with when debugging but am unsure how to use this for handling error. What is the correct protocol to handle errors in this function? If a fault has happened ideally I would like to clear the MCU ram and reset the MCU. However what if the handler code causes more faults. Is there a standard protocol for handling faults (not just HardFaults)?

void HardFault_Handler(void)
{
  /* USER CODE BEGIN HardFault_IRQn 0 */
 
  /* USER CODE END HardFault_IRQn 0 */
  while (1)
  {
    /* USER CODE BEGIN W1_HardFault_IRQn 0 */
    /* USER CODE END W1_HardFault_IRQn 0 */
  }
}

[Tesla DeLorean · Accepted solution]

You can technically return if you remediate the issue it faulted over, but that requires a very deep understanding of the CPU/ISA

What I normally do is output enough diagnostic information about registers, and perhaps a partial stack dump, so I can provide useful answers about the failure to the boss/customers rather than shrug my shoulders.

If you don't do anything complicated it shouldn't double-fault. Zeroing the imternal RAM using simple code (looping with internal registers, no function calls) should work fine.

[Tesla DeLorean]

https://community.st.com/s/question/0D50X0000BxwhEhSQI/my-stm32h743vi-application-runs-to-a-debug-instruction-in-the-startup-code

[AAgar.2]

That makes sense. What registers are useful for diagnostic info? For clearing RAM I can just memset to NULL for all vars?

Whether the app can continue would be determined on the nature of the failure. I would think in buffer overflow or peripheral failure the app cannot continue after fault.

Appreciate the support clive.