Yep, we have seen this internally, I have a fix - I just escaped from a
2 hour long meeting, maybe I can actually check it in now.
Steve
> Hi,
>
> a coworker here told me he got a kernel oops while installing something
> with rpm. The problem is the best description I got out of him was
> exactly that: "installing something using rpm". Anyways, if this helps
> somehow:
>
> invalid operand: 0000
> CPU: 0
> EIP: 0010:[<c01e9d1d>]
> Using defaults from ksymoops -t elf32-i386 -a i386
> EFLAGS: 00010282
> eax: 0000001a ebx: de29ba80 ecx: df39a000 edx: 00000001
> esi: ffffffff edi: c189fbe0 ebp: 00000001 esp: c189f9e0
> ds: 0018 es: 0018 ss: 0018
> Process kupdated (pid: 7, stackpage=c189f000)
> Stack: c029717c c029724e 00000030 c01cd166 c0290fc0 c0290e1b 0000014f c189fbe
> 0
> ffffffff c016924f de29ba80 00000004 ffffffff dfe73400 00000000 c016c06
> 2
> c189fbe0 00022d14 dfe73400 00000002 c189fbe0 0000000c dfe735d8 0000000
> 4
> Call Trace: [<c01cd166>] [<c016924f>] [<c016c062>] [<c017e48e>] [<c0162844>]
> [<c
> Warning (Oops_read): Code line not seen, dumping what data is available
>
> >>EIP; c01e9d1d <assfail+2d/40> <=====
> Trace; c01cd166 <xfs_trans_mod_sb+146/2b0>
> Trace; c016924f <xfs_alloc_ag_vextent+26f/290>
> Trace; c016c062 <xfs_alloc_vextent+3e2/520>
> Trace; c017e48e <xfs_bmap_alloc+1e4e/2230>
> Trace; c0162844 <avl_remove+d4/f0>
>
> Code: 0f 0b 83 c4 0c c3 8d b6 00 00 00 00 8d bc 27 00 00 00 00 83
>
> Code; c01e9d1d <assfail+2d/40>
> 00000000 <_EIP>:
> Code; c01e9d1d <assfail+2d/40>
> 0: 0f 0b ud2a
> Code; c01e9d1f <assfail+2f/40>
> 2: 83 c4 0c add $0xc,%esp
> Code; c01e9d22 <assfail+32/40>
> 5: c3 ret
> Code; c01e9d23 <assfail+33/40>
> 6: 8d b6 00 00 00 00 lea 0x0(%esi),%esi
> Code; c01e9d29 <assfail+39/40>
> c: 8d bc 27 00 00 00 00 lea 0x0(%edi,1),%edi
> Code; c01e9d30 <random+0/50>
> 13: 83 00 00 addl $0x0,(%eax)
>
> I'll try to reproduce this. Code is from CVS, arround Thr Apr 5 02:00 GMT.
>
> --
> Marcelo
|