Re: ath0: stuck beacon; resetting (bmiss count 4)

Hugo Silva wrote:
> Mar  5 19:27:48 daedalus kernel: ath0: stuck beacon; resetting (bmiss 
> count 4)
> Mar  5 19:28:19 daedalus last message repeated 95 times
> Mar  5 19:30:20 daedalus last message repeated 359 times
> Mar  5 19:32:54 daedalus last message repeated 481 times
> Mar  5 19:32:54 daedalus kernel:
> Mar  5 19:32:55 daedalus kernel: ath0: stuck beacon; resetting (bmiss 
> count 4)
> Mar  5 19:33:23 daedalus last message repeated 80 times
> Mar  5 19:33:23 daedalus kernel: ath0: stuck beacon; resetting (bmkss 
> ccunt 4e
> Mar  5 19:33:23 daedalus kernel:
> Mar  5 19:33:23 daedalus kernel: ath0: stuck beacon; resetting (bmiss 
> count 4)
> Mar  5 19:33:27 daedalus last message repeated 11 times
>
>
> ath0_at_pci0:5:1:0:        class=0x020000 card=0x5a001385 chip=0x0013168c 
> rev=0x01 hdr=0x00
>    vendor     = 'Atheros Communications Inc.'
>    device     = 'AR5212, AR5213 802.11a/b/g Wireless Adapter'
>    class      = network
>    subclass   = ethernet

dmesg|grep ath is more useful in identifying the device

>
>
> wlan0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 
> 2290
>        ether xxx
>        inet xxx netmask 0xffffff00 broadcast xxx.255
>        media: IEEE 802.11 Wireless Ethernet autoselect mode 11g <hostap>
>        status: running
>        ssid xxx channel 5 (2432 Mhz 11g) bssid xxx
>        regdomain ETSI country XX ecm authmode WPA privacy MIXED 
> deftxkey 2
>        TKIP 2:128-bit txpower 30 scanvalid 60 pureg protmode CTS wme 
> burst ff
>        dturbo -apbridge dtimperiod 1 -dfs
>
>
> After some time, the AP dies - my laptop scanning for an access point 
> will no longer show on tcpdump -pntttiath0 (on the AP).

"AP dies" is not helpful

>
> The system itself is still up, it's just that it's no longer possible 
> to associate with the AP.
> Rebooting makes the AP work again.
>
>
> Suggestions?
"stuck beacon" means the ap tried to xmit a beacon frame but for various 
reasons (where "various" is VERY large) it wasn't able to.  After N 
consecutive attempts, 4 in this case) it gives up and resets the device 
(which should bring it back to life in case it's stuck because of a 
hardware bug).  Often this occurs because you are using a slow platform 
and the time provided to prepare the beacon frame is insufficient to 
push it to the hardware fast enough to meet the realtime requirements.  
This can be exacerbated by operating on a busy channel.  Unfortunately 
you've provided very little info so I can't advise you.  I suggest you 
search for this subject and look at the discussions that have occurred 
for the last 5-6 years...

    Sam