Problems with GPIO P1.xx on nRF52833

RE: Problems with GPIO P1.xx on nRF52833

Kenneth — Fri, 17 Dec 2021 10:50:24 GMT

Difficult to comment on, if it only impacted twi it may seem a bit strange.

Kenneth

RE: Problems with GPIO P1.xx on nRF52833

picatostas — Fri, 17 Dec 2021 10:04:23 GMT

Hi Kenneth,
I made further investigation and I realized that I parametrized the make system wrong and even when I was compiling for nrf52833 for some reason system_nrf52832.c was chosen instead of system_nrf52833.c my bad. Would that explaing that the error was undefined behavior ?

RE: Problems with GPIO P1.xx on nRF52833

Kenneth — Thu, 16 Dec 2021 10:31:53 GMT

Hi,

I think your lookup of the error code is wrong. The error code is 8, which is NRF_ERROR_INVALID_STATE, in comparison NRFX_ERROR_FORBIDDEN would be 0x0BAD0008. It should be possible to look at the implementation of the twi() call to find the reason for this error. I think you should check that the parameters passed to the twi() api, e.g. the twi instance, is static and global.

You migt also build your project with DEBUG, such that you can find the exact error code, file name and line number of the assert in app_error_fault_handler() during development, if not the fault handler will simply execute NVIC_SystemReset() to recover (which is a good idea for an end product to safely recover, but not during development to identify errors).

Kenneth