Subject: Re: i386 MP panic: amap_pp_adjref: negative reference count
To: Manuel Bouyer <email@example.com>
From: Sami Kantoluoto <firstname.lastname@example.org>
Date: 12/16/2002 21:51:23
On Mon, Dec 16, 2002 at 09:19:14PM +0200, Sami Kantoluoto wrote:
> On Mon, Dec 16, 2002 at 08:07:23PM +0100, Manuel Bouyer wrote:
> > On Sat, Dec 14, 2002 at 10:26:04PM +0200, Sami Kantoluoto wrote:
> > > Hi,
> > >
> > > I'm getting that panic (amap_pp_adjref: negative reference count) when
> > > building openoffice. It seems to be repeatable. I don't have ddb/trace
> > > output yet (I'll post it when I get the panic next time [don't know how
> > > long it actually takes]).. What other information could be useful to
> > > catch the reason? Kernel is almost equal to GENERIC.MP (I'm going to compile
> > > a new kernel with MPDEBUG option [DEBUG & DIAGNOSTIC are enabled already]).
> > Maybe it's related to the panic I got this week end:
> > http://mail-index.netbsd.org/tech-kern/2002/12/16/0005.html
> Same though came up in my mind too when I saw the message.
> > My kernel is compiled with MPDEBUG and LOCKDEBUG.
> > If you can reproduce the problem can you try with MPDEBUG and LOCKDEBUG ?
> Actually I had MPDEBUG and LOCKDEBUG enabled already when previous panic
> happened. But now I'm waiting for a next panic. So let's wait and see.
This propably has nothing to do with MPDEBUG and/or LOCKDEBUG but I saw
(never seen before) the following message in my console:
Data modified on freelist: word 0 of object 0xc1599540 size 64 previous type
hosttbl (0xfffffffe != 0xdeadbeef)