HardFault Handler

[inspiringdev123]

I am dealing with a HardFault occuring in my code, and I am unable to trace the error. The output when the HardFault occurs is <error> hardfault: HARD FAULT at 0xD0D0D02A
<error> hardfault: R0: 0x20007A8C R1: 0x00000000 R2: 0x00000000 R3: 0x20007ACF
<error> hardfault: R12: 0x20007A8C LR: 0x00044343 PSR: 0x61000000
<error> hardfault: Cause: The processor has attempted to execute an undefined instruction.

The address of 0xD0D0D02A is garbage code.

[Vidar Berg]

Hello,

Yes, the PC address explains why the fault exception was raised. Did you try looking at the stack trace to trace back to where the error occurred? You can also check the LR address. Hopefully, that will show where the branch to this address was made. It's possible that a corrupted function pointer is being called.

Best regards,

Vidar

[inspiringdev123]

I enter the LR address (0x00044343) into the memory search and my Segger Embedded Studio crashes. 

[Turbo J]

Does not mean much. Stack based buffer overflow can also destroy the LR saved on the stack - we have no idea if that value was vaild or not.

[Vidar Berg]

inspiringdev123 said:

I enter the LR address (0x00044343) into the memory search and my Segger Embedded Studio crashes. 

That sounds like a bug in Segger embedded studio. Not sure what version you are using, but I haven't experienced this myself.

Turbo J said:

Does not mean much. Stack based buffer overflow can also destroy the LR saved on the stack - we have no idea if that value was vaild or not.

I didn't suspect a stack buffer overrun in this case because it's only the stacked PC value that is obviously corrupted, so I think checking the LR or inspecting the full call stack is a good first step.

[inspiringdev123]

Found a buffer overflow. Thanks guys!

[Vidar Berg]

Great to hear! Was this buffer on the stack or in static (global) memory?

[inspiringdev123]

Global Memory