add reboot on Softdevice assert

[ivnn]

Hi,
Is any problem could be if we add reboot on softdevice assert?
For example, if we have PCB covered with compound, etc. without any assess

__WEAK void app_error_fault_handler(uint32_t id, uint32_t pc, uint32_t info)
{
#endif
    switch (id)
    {
        case NRF_FAULT_ID_SD_ASSERT:
            __LOG(LOG_SRC_APP, LOG_LEVEL_ERROR, "Softdevice assert: %u:%u\n", pc, info);
            break;
        case NRF_FAULT_ID_APP_MEMACC:
            __LOG(LOG_SRC_APP, LOG_LEVEL_ERROR, "Application memory access: %u:%u\n", pc, info);
            break;
        case NRF_FAULT_ID_SDK_ASSERT:
        {
            assert_info_t * p_info = (assert_info_t *)info;
            UNUSED_VARIABLE(p_info);
            __LOG(LOG_SRC_APP, LOG_LEVEL_ERROR, "Mesh assert at 0x%08x (%s:%u)\n",
                                                pc,
                                                p_info->p_file_name,
                                                p_info->line_num);
            break;
        }
        case NRF_FAULT_ID_SDK_ERROR:
        {
            error_info_t * p_info = (error_info_t *)info;
            UNUSED_VARIABLE(p_info);
            __LOG(LOG_SRC_APP, LOG_LEVEL_ERROR, "Mesh error %u at 0x%08x (%s:%u)\n",
                                                p_info->err_code,
                                                pc,
                                                p_info->p_file_name,
                                                p_info->line_num);
            break;
        }
        default:
            __LOG(LOG_SRC_APP, LOG_LEVEL_ERROR, "UNKNOWN FAULT at 0x%08X\n", pc);
            break;
    }
    NRF_POWER->SYSTEMOFF = 1;
   
    //sleep_forever();
}

[Susheel Nuguru]

Sure you can, as a last option for a recovery mechanism, but before that , it is in your best interest to find out why softdevice asserted. It is not a normal behavior that softdevice asserts.

[ivnn]

Hi, I see an interesting case, after a softdevice assert has happened then it's some time before the device is back to life,

What do you think maybe will be better to use NVIC_SystemReset() instead of NRF_POWER->SYSTEMOFF = 1;

[Susheel Nuguru]

depends on your wakeup source, If you do not want to sleep intentionally after this assert, then you should not use SYSTEMOF, then a system reset will be a better option..