Re: Panic in _mtx_lock_flags+0x40

At Thu, 22 May 2003 16:35:33 +0000 (UTC),
John Baldwin wrote:
> > db> trace
> > _mtx_lock_flags(0,0,c0401b81,11d,e11edb18) at _mtx_lock_flags+0x40
> 
> Trying to lock a NULL pointer is definitely going to panic, yes.
> I'm curious what vm_fault() line you are at.
> 
> > vm_fault(c082f000,deadc000,1,0,c3affab0) at vm_fault+0x2cc
> > trap_pfault(e11edc10,0,deadc0de,cf74d400,deadc0de) at trap_pfault+0x161
> > trap(c0410018,10,c03e0010,c7a73c9c,c7e8f600) at trap+0x3ad
> > calltrap() at calltrap+0x5
> > --- trap 0xc, eip = 0xc015c580, esp = 0xe11edc50, ebp = 0xe11edc68 ---
> > ad_detach(c7a73c9c,ffffffff,c03d5de3,0,1) at ad_detach+0x40
> 
> The real bug is probably here, probably a NULL pointer dereference.
> I would use gdb to figure out what this file:line is and send the
> report to Søren.
> 
> > ata_reinit(c7a73c00,cacf5100,c03d1747,0,0) at ata_reinit+0x9b
> > ad_timeout(cacf5100,0,c03ec2fe,bf,4feaef) at ad_timeout+0x136
> > softclock(0,0,c03e9271,233,c3afe780) at softclock+0x19c
> > ithread_loop(c3afd200,e11edd48,c03e9122,2f8,0) at ithread_loop+0x182
> > fork_exit(c021bab0,c3afd200,e11edd48) at fork_exit+0xc0
> > fork_trampoline() at fork_trampoline+0x1a
> > --- trap 0x1, eip = 0, esp = 0xe11edd7c, ebp = 0 ---

Is this help you?

(kgdb) l *ad_detach+0x40
0xc015c580 is in ad_detach (../../../dev/ata/ata-disk.c:218).
213     
214         atadev->flags |= ATA_D_DETACHING;
215         ata_prtdev(atadev, "removed from configuration\n");
216         ad_invalidatequeue(adp, NULL);
217         TAILQ_FOREACH(request, &atadev->channel->ata_queue, chain) {
218             if (request->softc != adp)
219                 continue;
220             TAILQ_REMOVE(&atadev->channel->ata_queue, request, chain);
221             biofinish(request->bp, NULL, ENXIO);
222             ad_free(request);


-- 
Jun Kuriyama <kuriyama_at_imgsrc.co.jp> // IMG SRC, Inc.
             <kuriyama_at_FreeBSD.org> // FreeBSD Project