> >> + /* IBSS uses different parameters for Beacon sending */ > >> + qparam.cw_min++; > >> + qparam.cw_min *= 2; > >> + qparam.cw_min--; > >> + local->ops->conf_tx(local_to_hw(local), > >> + IEEE80211_TX_QUEUE_BEACON, &qparam); > > > > And shouldn't we do that only for IBSS? I'm in favour of removing it > > completely, however, since the beacon queue is a way to construct > > hardware, not do things in software... > > Good question; I've just moved this part of code out of > ieee80211_sta_set_ssid as that is not the best place for it. I don't > object to removing that code completely; seems that noone uses > IEEE80211_TX_QUEUE_BEACON anyway. On the other hand, it can be a good > hint for those implementing beaconing on hw where it has to be set > explicitly.. I disagree with the latter part, but let's just leave it for now, I think my QoS patch (the one Ron is adopting) will kill it anyway. johannes
Attachment:
signature.asc
Description: This is a digitally signed message part
