>> Hi Dan, I'd say thats a false positive. >> There exist HW configuration where PTP is disabled or not available. >> >> PTR_ERR(NULL) is 0, so eventually driver now functions correctly, allowing to >> proceed but marking ptp functionality as disabled. >> >> I agree that's a bit counterintuitive but correct. > > I have looked at it again and the "!clock" check should be removed > because it is dead code. It's not possible for it to be NULL there > because aq_ptp_init() is a dummy function if the PTP clock is not > enabled. Hm, seems you are right. I've added exclusion code with a later commit and yes, this means NULL will never be a retval here. Thanks again, I'll clean it up. -- Regards, Igor
