>> Would it make more sense to notify link down *after* the WLAN leave >> notification? > > Why? The notification is asynchronous, so you wouldn't really gain any > advantage. I may be wrong, but if_link_state_change() will queue a link down notification too, that will be sent so received before the WLAN leave one by one subscriber listening to both. The asynchronous aspect is no problem indeed, but the order could be, no? Semantically, the link can be seen as down in BSS mode only *after* WLAN dissociation, not before. > Sounds like you had fun! Oh yes. Meanwhile, my brain could be still damaged by some tough but fun kite crash, so be kind with previous remarks :-) Bye, Philippe.