On Sat, Feb 25, 2017 at 03:13:40PM +0000, Nick Lowe wrote: > The 802.11n, ac and ax specs expect devices to support 802.11e QoS in > order to use HT/VHT/HE rates. This is because the TID field is key to > aggregation mechanisms, including block ACKs, that enable high > throughput rates. WMM != 802.11e QoS > Where WMM is configured to be disabled and HT/VHT/HE support are > configured hostapd should fail to start therefore so that this > explicitly fails and the configuration can be corrected. > > We could silently enable WMM where HT/VHT/HE support are configured > and a WMM is not defined. I think, however, that it is cleaner and > more logically consistent to require it to be explicitly enabled in > that case. This would just make it more difficult for users without any clear benefit. At most, the check could verify that wmm_enabled=0 is not explicitly used with HT/VHT/HE. That said, I still do not see why hostapd would need to explicitly mandate WMM to be enabled. IEEE 802.11 QoS is fine for VHT/HE, no need for WMM there. In addition, if other HT/VHT/HE functionality works without QoS and only subset of functionality needs to be disable (e.g., everything related to aggregation), it is not clear why this possibility should be prevented. If for no other use, this could be of some use for testing purposes. -- Jouni Malinen PGP id EFC895FA _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
