kern/111458: [panic] Panic on 6.2-RELEASE AMD in kern_mutex

[Remco Bressers]

On Tue, 2007-04-24 at 17:58 -0400, Kris Kennaway wrote:
> On Tue, Apr 24, 2007 at 11:47:37PM +0200, Remco Bressers wrote:
> > On Tue, 2007-04-24 at 17:42 -0400, Kris Kennaway wrote:
> > > On Tue, Apr 24, 2007 at 11:38:05PM +0200, Remco Bressers wrote:
> > > > On Tue, 2007-04-24 at 17:26 -0400, Kris Kennaway wrote:
> > > > > On Tue, Apr 24, 2007 at 11:21:39PM +0200, Remco Bressers wrote:
> > > > > > On Tue, 2007-04-24 at 17:14 -0400, Kris Kennaway wrote:
> > > > > > > On Tue, Apr 24, 2007 at 09:10:09PM +0000, Remco Bressers wrote:
> > > > > > > 
> > > > > > > >  To update this bugreport and to keep it 'warm'
> > > > > > > >  
> > > > > > > >  I've got the very same problem overhere.
> > > > > > > >  
> > > > > > > >  Our box : 
> > > > > > > >  
> > > > > > > >  # uname -r
> > > > > > > >  6.2-RELEASE-p2
> > > > > > > >  
> > > > > > > >  This is an amd64 release
> > > > > > > >  
> > > > > > > >  Updates on GENERIC : 
> > > > > > > >  
> > > > > > > >  options         QUOTA 
> > > > > > > >  device          pf
> > > > > > > >  device          pflog
> > > > > > > >  options		SMP
> > > > > > > >  
> > > > > > > >  Kernel messages : 
> > > > > > > >  
> > > > > > > >  cpuid = 0; apic id = 00
> > > > > > > >  fault virtual address = 0x18c
> > > > > > > >  fault code = supervisor read, page not present
> > > > > > > >  current process = 5 (thread taskq)
> > > > > > > >  trap number = 12
> > > > > > > >  panic: page fault
> > > > > > > >  cpuid = 0
> > > > > > > 
> > > > > > > How do you know it is "the very same problem"?  In order to determine
> > > > > > > this you need to compare backtraces from the panic, which you didn't
> > > > > > > provide.
> > > > > > 
> > > > > > I contacted the submitter for this problem and compared hardware and
> > > > > > software. The symptoms are the same
> > > > > 
> > > > > Which just means "it crashed"
> > > > > 
> > > > > , the kernel panic is the same and
> > > > > 
> > > > > The panic message also just means "it crashed"
> > > > > 
> > > > > > the installed FreeBSD version is exactly the same.
> > > > > 
> > > > > You're both running the most recent version, no real surprises there.
> > > > > 
> > > > > > Sounds fair enough to me.  Ofcourse i cannot be 100% sure, but it
> > > > > > sounds too obvious to me.
> > > > > 
> > > > > Well, maybe, but I respectfully submit that you don't understand the
> > > > > issue well enough to conclude that :) Please follow up with a
> > > > > backtrace and then we'll see where things stand.
> > > > 
> > > > You didn't write anything about the fact that software combinations +
> > > > hardware is almost identical! That's no coincidence in my humble
> > > > opinion :). Anyway..
> > > 
> > > I'd prefer not to prematurely jump to conclusions before you have any
> > > supporting evidence.  It is of course an obviously true statement that
> > > two identical systems may panic in two completely different and
> > > unrelated ways.
> > > 
> > > > The problem in this case is, that the backtrace isn't written to disk. I
> > > > must wait for the next opportunity to get that backtrace.
> > > 
> > > OK, let us know.
> > 
> > Is it right that i have to configure the dump device with 'dumpdev' in
> > rc.conf ? I'm a little lost here :S 
> 
> See the developers handbook chapter on kernel debugging.

I just found out how to configure it. Thanks,

Remco