http://pastebin.com/qX3SLy2H
This appears to be a legitimate lockdep (when trying to re-connect to
an AP whose password I just changed).
These take the locks in reverse order:
ieee80211_work_work() -> ieee80211_authenticate() (Inlined in the
lockdep print).
ieee80211_sta_rx_queued_mgmt() -> ieee80211_rx_mgmt_disassoc()
I don't know the state machine well enough and it's getting kind of
late here. I'll probably have time to take a closer look in a couple
of days (unless Johannes finds this one trivial).
Arik
On Wed, Oct 19, 2011 at 00:12, Arik Nemtsov <[email protected]> wrote:
> http://pastebin.com/qX3SLy2H
>
> This appears to be a legitimate lockdep (when trying to re-connect to
> an AP whose password I just changed).
> These take the locks in reverse order:
>
> ieee80211_work_work() -> ieee80211_authenticate() (Inlined in the
> lockdep print).
> ieee80211_sta_rx_queued_mgmt() -> ieee80211_rx_mgmt_disassoc()
>
Please disregard this one. The problem was with an internal patch I
was using for testing (as can be seen in the stack trace as well).
Arik