Re: LOR (re0 and user map) + PANIC

On 2004-09-10 13:15 -0400, Robert Watson wrote:
> 
> On Fri, 10 Sep 2004, John-Mark Gurney wrote:
> 
> > Marian Cerny wrote this message on Fri, Sep 10, 2004 at 12:22 +0200:
> > > Heh, I noticed one more LOR. This one happens when I turn the laptop off
> > > using 'halt -p' after syncing disks. It's *hand*-rewritten, because the
> > > laptop turned itself off after 10 seconds, despite the fact, that I was
> > > inside kernel debugger (I took a shot with my digital photo camera).
> > > 
> > > lock order reversal
> > >  1st 0xc177b6e8 re0 (network driver) _at_ /usr/src/sys/dev/re/if_re.c:1752
> > >  2nd 0xc08adee4 user map (user map) _at_ /usr/src/sys/vm/vm_map.c:2997
> > 
> > What version of if_re.c are you running?  neither RELENG_5 nor HEAD has
> > a lock on line 1752. 

It is v 1.28 from BETA3.

1752:        RL_LOCK(sc);

> While this would be useful to know, the actual bug (per my earlier e-mail)
> is a NULL pointer deref:
> 
> > > --- trap 0xc, eip = 0xc0575b76, esp = 0xceef9cc0, ebp = 0xceef9cdc ---
> > > re_rxeof(c177b000) at re_rxeof+0x2ae
> 
> It would be very useful to know the revision of the file, as well as what
> line in the code re_rxeof+0x2ae is.

Is there any way how to find it out? re_rxoef() is called 4 times in
that file.

-- 
Marian Cerny <jojo_at_matfyz.cz>
Jabber: jojo_at_njs.netlab.cz

[ UNIX is user friendly. It's just selective about who its friends are. ]