Tim Sullivan wrote:
>
> Hello,
>
> The following occurred during startup using test12-pre3. test12-pre2
> does not exhibit the problem.
>
> regards,
>
> -tim
>
> kernel: Unable to handle kernel paging request at virtual address
> fffffffc
> kernel: printing eip:
> kernel: c011a41f
> kernel: *pde = 00001063
> kernel: *pte = 00000000
> kernel: Oops: 0000
> kernel: CPU: 0
> kernel: EIP: 0010:[sys_setitimer+191/208]
> kernel: EFLAGS: 00010246
> kernel: eax: 00000000 ebx: cf4e1fb0 ecx: 00000000 edx: c027169d
> kernel: esi: bffffc68 edi: cf4e1fc0 ebp: 00000000 esp: cf4e1f88
> kernel: ds: 0018 es: 0018 ss: 0018
> kernel: Process ntpdate (pid: 451, stackpage=cf4e1000)
> kernel: Stack: cf4e0000 bffffd1c 00000009 bffffc70 00000000 cf4e0000
> 400538e8 00000000
> kernel: 00000000 40166214 00000000 00030d40 00000000 000186a0
> c010a847 00000000
> kernel: bffffc58 00000000 bffffd1c 00000009 bffffc70 00000068
> 0000002b 0000002b
> kernel: Call Trace: [system_call+51/56]
> kernel: Code: c8 5d 83 c4 28 c3 90 90 90 90 90 90 90 90 90 90 90 83 ec
> 44
Disassemblind the code bytes, it shows up as:
0: c8 5d 83 c4 enter $0x835d,$0xc4
4: 28 c3 subb %al,%bl
6: 90 nop
which is definately wrong. Moving one byte forward it becomes:
0: 5d popl %ebp
1: 83 c4 28 addl $0x28,%esp
4: c3 ret
5: 90 nop
which a normal sequence of instructions for the end of a function. It
looks like a branch may have been miscompiled off by one. What
compiler/binutils are you using?
--Brian Gerst - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Nov 30 2000 - 21:00:22 EST