Re: [PATCH] to deal with bad dev->refcnt in unregister_netdevice()

[Thomas Hood <jdthoodREMOVETHIS@yahoo.co.uk>]

> > Here is a patch which may not solve the underlying 
> 
> This does not. refcnt cannot be <1 at this point. 

The refcnt shouldn't be less than 1, but it is in fact
less than 1.  (As I'm sure you understand.)

> > assuming that the latter messages aren't serious? 
> 
> They are fatal. Machine must be rebooted after them. 

True.  I found that with testing---lots of ifups and ifdowns,
etc.---the kernel becomes unstable.

> > I hope the networking gurus can find the real bugs here. 
> 
> Well, someone forgets to grab refcnt or makes redundant dev_put. 
> Try to catch this, f.e. adding BUG() to the places where fatal 
> messages are generated to get backtraces. 
> Alexey

I think that the ipx driver makes an inappropriate dev_put
in its notifier callback.  However that is for people better
acquainted with the come than I to judge.  Removing the ipx
driver does work around the problem though.

Thomas
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/