> > yes, i noticed that. the prints are coming of course from the MLME > > itself that can be called while no session is in progress and checks > > the status when called. > > i'll do a quick check to decide which way is better (outside check of > > the state machine or just a silent discard of non active TIDs > > requests) and fix it. > > Ron > > ok, i checked it. i think it is the same as the Rx issue we had. those > prints can be eliminated by .config file's MAC80211_HT_DEBUG. if we > don't use this flag what's the use of it? if we do use it, i would > like to insert the Rx prints back as well. Well, yes, but I think most debugging features in mac80211 should enable us to see things going wrong, not totally flood us in messages. YMMV, but if these things drown you in messages then even developers will not enable them by default which means that they've effectively been made useless [1]. Hence, I think _DEBUG symbols should be ok to enable by default, not print too many messages but enable just some important messages and further debugging checks. Now, maybe that message is actually one of such debugging checks, in that it tells you when you did a wrong (i.e. while not enabled) call. But then mac80211 itself shouldn't do such calls. johannes [1] as, incidentally, you can see in the current wireless tree from a few days ago, iwlwifi with debugging doesn't even compile!
Attachment:
signature.asc
Description: This is a digitally signed message part
