Re: IBSS lockdep failure

[Johannes Berg <johannes@sipsolutions.net>]

[-- Attachment #1: Type: text/plain, Size: 1222 bytes --]


> wlan0: beacon TSF higher than local TSF - IBSS merge with BSSID
> XX:XX:XX:XX:XX:XD
> phy0: Adding new IBSS station XX:XX:XX:XX:XX:X1 (dev=wlan0)
> phy0: Allocated STA XX:XX:XX:XX:XX:X1
> phy0: Inserted STA XX:XX:XX:XX:XX:X1
> wlan0: RX ProbeReq SA=XX:XX:XX:XX:XX:X1 DA=ff:ff:ff:ff:ff:ff
> BSSID=ff:ff:ff:ff:ff:ff (tx_last_beacon=1)
> wlan0: Sending ProbeResp to XX:XX:XX:XX:XX:X1
> RX beacon SA=XX:XX:XX:XX:XX:XC BSSID=XX:XX:XX:XX:XX:XD TSF=0x14dc0bd5c
> BCN=0x14dc0be19 diff=-189 @1071562

That value is suspiciously low, you might want to double-check you're
setting the correct RX timestamp in the driver, mac80211 expects the
timestamp to be at the first symbol of the frame.


> phy0: Destroyed STA XX:XX:XX:XX:XX:XB
> phy0: Destroyed STA XX:XX:XX:XX:XX:XC

Ok so at least it's not trying to remove the same STA twice :)

> BUG: unable to handle kernel NULL pointer dereference at 00000004
> IP: [<c043d4eb>] zap_class+0x22/0x64
> *pde = 00000000
> Oops: 0002 [#1] SMP
> Modules linked in: mac80211 cfg80211 netconsole i915 drm autofs4 hidp
> rfcomm l2cap sunrpc cpufreq_ondemand acpi_cpufreq dm_multipath sbs

Any chance of getting a stack trace? zap_class seems weird.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 836 bytes --]