Hi,

dev->queue_lock is taken in a scenario like below: always after
dev->ingress_lock and p->tcfc_lock, so just like on this last
backtrace with info about held locks. But this report shows that
lockdep for some reason forgot the history before dev->queue_lock,
and recorded it again. It seems, even if there is something wrong
with init lockdep shouldn't report it like this.

I CC this report to lockdep maintainers and linux-kernel.

Regards,
Jarek P.

Original lockdep report for 2.6.24.2:
http://permalink.gmane.org/gmane.linux.network/86896 


--

[ message continues ]

...Hmmm... On the other hand, despite misleading dependency chain on
this report, lockdep seems to be right: dev->queue_lock and
dev->ingress_lock are really taken in a different order from
qdisc_lock_tree() and while using act_mirred! Now I wonder why this
warning is so rare?

So, let's give a break to lockdep maintainers and linux-kernel, and
try to figure it out more in netdev...

Jarek P.
--

[ message continues ]