Bug#491731: madwifi-source: 0.9.4+r3698 and later no longer work in master mode with hostapd and WPA

[Kel Modderman]

Hi Andrew,

On Tuesday 22 July 2008 17:25:37 Andrew Baumann wrote:
> Hi Kel,
> 
> On Tuesday 22 July 2008 08:29:29 Kel Modderman wrote:
> > Is wireless interface in a bridge?
> 
> No, it's not. The relevant stanza from /etc/network/interfaces is:
> 
> iface ath0 inet static
>         madwifi-base wifi0
>         madwifi-mode ap
>         wireless-channel 9
>         wireless-essid foo
>         address 192.168.0.1
>         netmask 255.255.255.0
> 
> For the record, hostapd.conf has:
> interface=ath0
> driver=madwifi
> ssid=foo
> country_code=CH
> hw_mode=g
> channel=9
> beacon_int=100
> dtim_period=2
> max_num_sta=255
> rts_threshold=2347
> fragm_threshold=2346
> macaddr_acl=0
> auth_algs=3
> ignore_broadcast_ssid=0
> wme_enabled=1
> wme_ac_bk_cwmin=4
> wme_ac_bk_cwmax=10
> wme_ac_bk_aifs=7
> wme_ac_bk_txop_limit=0
> wme_ac_bk_acm=0
> wme_ac_be_aifs=3
> wme_ac_be_cwmin=4
> wme_ac_be_cwmax=10
> wme_ac_be_txop_limit=0
> wme_ac_be_acm=0
> wme_ac_vi_aifs=2
> wme_ac_vi_cwmin=3
> wme_ac_vi_cwmax=4
> wme_ac_vi_txop_limit=94
> wme_ac_vi_acm=0
> wme_ac_vo_aifs=2
> wme_ac_vo_cwmin=2
> wme_ac_vo_cwmax=3
> wme_ac_vo_txop_limit=47
> wme_ac_vo_acm=0
> eapol_key_index_workaround=0
> eap_server=0
> own_ip_addr=127.0.0.1
> wpa=1
> wpa_passphrase=bar
> wpa_key_mgmt=WPA-PSK
> wpa_pairwise=TKIP CCMP
> wpa_group_rekey=600

My working configuration is very similar, except I have disabled WME
completely.

Could you try commenting out all wme configuration key/values?

> 
> > I had no problem setting up ath0 in ap mode using hostapd for wpa2 auth,
> > using current madwifi + hostapd packages, dnsmasq for DNS + dhcp and
> > iptables script for NAT of ethernet connection.
> 
> Well, I'm using dhcp3-server and bind, but this seems to be a lower-level 
> problem. It seems only to have directly affected the squeezebox stuff (my 
> laptop is still able to connect), which is odd, but looking at packet 
> captures I see a lot of short packets and other weirdness, which leads me to 
> suspect that the new version is corrupting or dropping packets.

Understood.

> 
> > Regressions like this really suck. Sadly if we go back to 0.9.4 many users
> > have no hardware support at all... And its crunch time for Lenny release...
> 
> I guess I'll just have to hang onto 0.9.4 for now and hope they figure it out 
> eventually upstream.

Let's hope ath5k/ath9k improve rapidly.

Thanks, Kel.