Stopped by vector catch, Unknown function

RE: Stopped by vector catch, Unknown function

wpreska1 — Fri, 19 Mar 2021 17:54:52 GMT

Well, I never did get this totally figured out, but it seems the vector catch was triggered by incorrect flash sizing in the configuration. I tried playing with the settings in Project/Options fields, and I experimented with the flash_placement.xml file. It actually just created all kinds of other build errors (occurring during the linking phase), and I never got a running program. At this point I'm calling this project a "fail" and I'm starting over from scratch.

The standard Template example project in the SDK's Peripheral/Examples folder builds and runs just fine, so I'm just starting over with that as a base and adding in my application-specific code on top of it.

Thanks Vidar for your help with this.

RE: Stopped by vector catch, Unknown function

Vidar Berg — Thu, 18 Mar 2021 08:03:07 GMT

Thanks for confirming. So this is most likely the reason why program execution did not reach main(). I hope it works once you adjust the memory settings.

RE: Stopped by vector catch, Unknown function

wpreska1 — Wed, 17 Mar 2021 23:44:06 GMT

Wow - good catch Vidar.

The "Memory Segments" section in the Linker settings was correct, but the "Section Placement Macros" settings were wrong (they were set to the larger values of the 840 chip).

I think you may be onto something here - previously, I had indeed been tinkering with the flash_placement.xml file. I'll play with it and keep you posted.

Thanks!

RE: Stopped by vector catch, Unknown function

Vidar Berg — Wed, 17 Mar 2021 21:50:37 GMT

I think it must be the linker settings that are wrong then. Have you tried to debug the original "pca10040" project (\examples\ble_peripheral\ble_app_template\pca10040\s132\ses to see if that works?

Here is the original configuration for the ble_app_template_pca1040 project:

RE: Stopped by vector catch, Unknown function

wpreska1 — Wed, 17 Mar 2021 21:44:19 GMT

Thank you for looking into this Vidar.

For the hardware, I am using the nRF52 DK based on the 832 (not the 840). The "Target" is set in the Debugger section of the project options to nRF52832-xxAA. Where else within SES can I check to see why it's generating code for the 840?

RE: Stopped by vector catch, Unknown function

Vidar Berg — Wed, 17 Mar 2021 21:31:12 GMT

Hi,

Thanks for the update. I think it does help. Are you running this on a nRF52840 DK or a nRF 52 DK (nRF52832)? I'm asking because the stack pointer value shown in the image you posted is clearly for the nRF52840, not the nRF52832 which only has 64 K of RAM ( i.e. upper boundary is 0x20010000). This could explain why the hardfault exception is raised before the program even reaches main().

Note that pca10040 is the board name for the nRF52 DK targeting the nRF52832 chip and it is the project configuration you should select if you are using the nRF52 DK board.

Regarding the "Vector catch" message: it's reported because the breakpoint on "hardfault" is enabled by default in SES:

RE: Stopped by vector catch, Unknown function

wpreska1 — Wed, 17 Mar 2021 20:39:46 GMT

Additional info - in case it helps...

Here is a screenshot of the CPU registers at the time of the program halt:

Still dead in the water... hoping someone knows how to debug these vector-catches...