Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 04 May 2003 13:41:04 -0700
From:      Terry Lambert <tlambert2@mindspring.com>
To:        "Cliff L. Biffle" <cbiffle@safety.net>
Cc:        current@freebsd.org
Subject:   Re: Reproduceable kernel panic on boot
Message-ID:  <3EB57ADF.49E5A115@mindspring.com>
References:  <200305040004.40214.cbiffle@safety.net> <3EB56147.A4CBEB0A@mindspring.com> <200305041321.07162.cbiffle@safety.net>

next in thread | previous in thread | raw e-mail | index | archive | help
"Cliff L. Biffle" wrote:
> On Sunday 04 May 2003 11:51 am, Terry Lambert wrote:
> > If you have a debug kernel available, you probably want to gdb -k
> > it, and then look at the code at 0xc1c1bd75; that will give you
> > the exact line with the error.
> 
> Believe it or not, no debug kernel was generated, either.
> So for some reason, a kernel config that has always generated a kernel.debug
> and an install process that have always backed up my kernel did neither.
> Something must be seriously hosed here.  Any ideas as to what that could be?

You have to either add the specific option in for it to do
a debug kernel, or you have to use the command "config -g"
instead of just "config".

As far as the install process not working, it beats me.  I
generally install manually, and manually back up the kernel,
so I can name it something other than "kernel.old".

If you had an install problem, it may be that you are running
a new kernel with old kernel modules.  That can also cause the
problems you are seeing (panic on boot), particularly acpi.ko,
and ipfw.ko.  You may want to make sure you are not loading
any kernel modules at all (break into the boot and unload all
and manually load only the kernel itself).  That might get you
past the panic to being able to build the kernel and install it
correctly.


> > It's generally a bad idea to set the CPU type (see the recent
> > discussion on -current about the guy who shot his foot off).
> > In general, GCC tends to generate buggy code if you set the
> > CPU type; of course, if you can live with buggy code...
> 
> I've been running with the closest CPUTYPE I can since I switched to FreeBSD
> (which was ~2yrs ago), and haven't had any bugginess to speak of, much less
> panics.  I'll try backing that out all the same.

Basically, expect that anything you have which is different is
probably not good, particularly if you're having problems.


> > (to know for sure, you would have to post a traceback).
> 
> I've got a semiworking kernel on the machine now, and I'm working on
> recompiling the errant kernel with DDB enabled.  I'll post the results.

OK; if you can't panic any more after you do that, you are
probably fighting mismatched modules.

-- Terry



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3EB57ADF.49E5A115>