nRF52840 debugging, breakpoint doesn't stop LFCLK

RE: nRF52840 debugging, breakpoint doesn't stop LFCLK

Jørgen Holmefjord — Mon, 05 Nov 2018 14:22:10 GMT

I do not know any ways to do this, except maybe adding code to stop LFCLK before any breakpoint and start it afterwards. You could check with Segger if they have some way to run code when hitting breakpoints. This is not related to the chip, it would be a debugger/IDE feature.

RE: nRF52840 debugging, breakpoint doesn't stop LFCLK

Ondrej — Thu, 01 Nov 2018 18:09:21 GMT

Hi Jørgen, thanks for answer. I'm no hardware expert. So let me simplify my question. When I use timers + scheduler from nRF52 SDK and I hit breakpoint in my timer handler and then I execute code again, the code crashes (app_fault_handler) with NRF_ERROR_NO_MEM (scheduler queue is full). What options do I have to avoid this?

RE: nRF52840 debugging, breakpoint doesn't stop LFCLK

Jørgen Holmefjord — Thu, 01 Nov 2018 14:44:08 GMT

Hi,

This is expected. A breakpoint will only halt CPU execution at the given address. LFCLK is a HW peripheral in the chip, it is not affected by breakpoints (other than that generated events are not being handled).

I'm not aware of any ways to automatically stop clocks when you hit breakpoints, and I'm not sure I understand what the purpose of this would be.

Best regards,
Jørgen