Re: kernel module debugging question

From: Chuck Ebbert
Date: Fri Jun 30 2006 - 14:53:43 EST


In-Reply-To: <1151661642.11434.23.camel@xxxxxxxxxxxxxxxxxxxxx>

On Fri, 30 Jun 2006 12:00:42 +0200, Arjan van de Ven wrote:

> On Fri, 2006-06-30 at 12:28 +0000, s.a. wrote:
> > Hi,
> > I have got the following fault , can you provide me with more details
> > about the problem ?
> > Best Regards
>
> you disabled CONFIG_KALLSYMS in your kernel configuration, which means
> the backtrace isn't really useful for anyone to look at

The only thing on the stack is the regs, thus no backtrace.

Someone modified entry.S and added code just before the sysexit. There is
no 'sysenter_exit' in stock 2.6.16 for one, and there's new code before
the sysexit:

3: e8 81 ca 03 00 call 3ca89 <_EIP+0x3ca89> <== new
8: fb sti <== new
9: 8b 4d 08 mov 0x8(%ebp),%ecx
c: 66 f7 c1 ff fe test $0xfeff,%cx
11: 0f 85 4e 01 00 00 jne 165 <_EIP+0x165>
17: e8 ed ea 00 00 call eb09 <_EIP+0xeb09> <== new
1c: 8b 44 24 18 mov 0x18(%esp),%eax <== new
20: 8b 54 24 28 mov 0x28(%esp),%edx
24: 8b 4c 24 34 mov 0x34(%esp),%ecx
28: 31 ed xor %ebp,%ebp
2a: fb sti
0: 0f 35 sysexit <===== oops

But that doesn't explain the oops. Maybe the SYSENTER MSR got screwed up?

--
Chuck
"You can't read a newspaper if you can't read." --George W. Bush
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/