Re: kernel panic with todays source

On Sunday 16 November 2003 05:49, Harald Schmalzbauer wrote:
> On Sunday 16 November 2003 05:34, Robert Watson wrote:
> > On Sun, 16 Nov 2003, Harald Schmalzbauer wrote:
> > > Fatal trap 12   :page fault while in kernel mode
> > > fault virtual address   =0x24
> > > fault code              =supervisor read, page not present
> > > instruction pointer     =0x8:0xc056c706
> > > stack pointer           =0x10:0xcdca4ca4
> > > frame pointer           =0x10:0xcdca4ca4
> > > code segment            =base 0x0, limit 0xfffff, type 0x1b
> > >                         =DPL 0, pres 1, def32 1, gran 1
> > > processor eflags        =resume, IOPL=0
> > > current process         =11 (idle)
> > > trap number             =12
> > > panic: page fault
> > >
> > > I do have compiled the kernel with makeoptions debug but I don't have a
> > > serial terminal nor firewire.
> >
> > Could you show the output from running the following command in "gdb -k
> > kernel.debug":
> >
> >   l *0xc056c706

Sorry, forgot to mention that I answered this in Craig Rodrigues mail. But 
while I'm here:

(kgdb) l *0xc056c706
0xc056c706 is in vsscanf (/usr/src/sys/kern/subr_scanf.c:224).
219
220                     case '[':
221                             fmt = __sccl(ccltab, fmt);
222                             flags |= NOSKIP;
223                             c = CT_CCL;
224                             break;
225
226                     case 'c':
227                             flags |= NOSKIP;
228                             c = CT_CHAR;

Thanks,

-Harry

> >
> > This will tell us where in the kernel the instruction pointer in question
> > was.  For whatever reason, your kernel panic doesn't seem to have dropped
> > you into DDB (at least, the output looks that way).  If you did get into
> > ddb, the results of the "trace" command would be very helpful.  As you
> > have no serial console, it's probably sufficient to just transcript the
>
> I think for this I need to set "options DDB" in my kernel, don't I? But
> without serial terminal this is pretty useless I think.
> But I'll do if you can get needed information which was otherwise not
> accessable.
>
> Thanks,
>
> -Harry
>
> > offsets at the end of each line in the trace (functioname+0xOFFSET)
> > without the argument entries.
> >
> > Robert N M Watson             FreeBSD Core Team, TrustedBSD Projects
> > robert_at_fledge.watson.org      Network Associates Laboratories