No, it's something in the ath driver and ath_hal code. I'm sorry, I've been busy debugging other things in my limited spare time; I just haven't had the chance to sit down and look at the rfkill code. :( -a