I am defining my own list of errors that my application will use. At the same time, I would like to report the NRF errors as well. What do you think about a solution where I define my own ERROR_BASE_NUM? In nrf_error.h there are some error ranges defined, like this:
// From nrf_error.h
#define NRF_ERROR_BASE_NUM (0x0) ///< Global error base
#define NRF_ERROR_SDM_BASE_NUM (0x1000) ///< SDM error base
#define NRF_ERROR_SOC_BASE_NUM (0x2000) ///< SoC error base
#define NRF_ERROR_STK_BASE_NUM (0x3000) ///< STK error base
What if i use those errors, and also expand it with my own range of errors starting at 0xA000, like this:
#define MY_ERROR_BASE_NUM (0xA000) ///< My own base num
#define MY_ERROR_OVERFLOW (MY_ERROR_BASE_NUM + 3) ///< Overflow Error
ret_code_t my_func(uint8_t x)
{
ret_code_t err_code;
// NRF error code will be returned
err_code = bsp_indication_set(BSP_INDICATE_ADVERTISING);
if (err_code != NRF_SUCCESS)
{
return err_code;
}
// My own error code will be returned
if(x>25)
return MY_ERROR_OVERFLOW;
return NRF_SUCCESS;
}
For this to work, I need to at least be sure that the range 0xA000 will continue to be reserved for me in the future, and there are probably more cases of trouble that I didn't think of.
Do you have other suggestions on how to handle my own errors as well as NRF errors?
