On Mon, 2002-11-18 at 23:29, Austin Gonyou wrote:
> I got this output from ksymoops on a 2.4.19-aa1 kernel. Does anyone know
> if this stack looks corrupt? Looking at each compared to each other, I
> believe so, but I just wanted some confirmation of this. Your urgent
> help is most appreciated.
>
> The recent output is taken from a production box that won't stay up for
> more than 7 days at a time. The older output, is from an identical box,
> running the same kernel, under different circumstances
Austin, the xfs code in this kernel is so old it is scary. I know you
are sticking with aa kernels for other reasons, but we really cannot
help with stuff this old.
The iodone code is no longer in interrupt context in xfs, and the
cvs tree has changes in this area which fix a long standing crash
which happened in exactly that function. Having said that the
cvs code is about to get another revamp in this area for performance
reasons with that fix.
Steve
> TIA.
>
>
>
> <recent ksymoops output>
> ksymoops 2.4.1 on i686 2.4.19. Options used
> -V (specified)
> -k /proc/ksyms (specified)
> -l /proc/modules (default)
> -o /lib/modules/2.4.19/ (default)
> -m /boot/System.map-2.4.19 (default)
>
> Oops: 0000 2.4.19 #1 SMP Wed Oct 16 17:02:35 CDT 2002
> CPU: 0
> EIP: 0010:[<c01164ee>] Not tainted
> Using defaults from ksymoops -t elf32-i386 -a i386
> EFLAGS: 00010007
> eax: c031e000 ebx: c034d8a8 ecx: 00000000 edx: ffffffd4
> esi: c034d880 edi: 00000080 ebp: c031ffc8 esp: c031ff98
> ds: 0018 es: 0018 ss: 0018
> Process swapper (pid: 0, stackpage=c031f000)
> Stack: c034d880 00000019 c031e000 c031e000 ffffffd4 0008e000 00000000
> c0100018
> c0100018 c031e000 c0106e60 c0105000 0008e000 c0106ee9 0002080e
> 00098700
> c03207a5 c031e000 00000000 c02a6da0 00200000 00200000 00200000
> 00038000
> Call Trace: [<c0106e60>] [<c0105000>] [<c0106ee9>]
> Code: 8b 72 58 8b 40 5c 85 f6 89 45 d4 75 35 89 42 5c f0 ff 40 14
>
> >>EIP; c01164ee <schedule+1ae/390> <=====
> Trace; c0106e60 <default_idle+0/40>
> Trace; c0105000 <_stext+0/0>
> Trace; c0106ee9 <cpu_idle+29/30>
> Code; c01164ee <schedule+1ae/390>
> 00000000 <_EIP>:
> Code; c01164ee <schedule+1ae/390> <=====
> 0: 8b 72 58 mov 0x58(%edx),%esi <=====
> Code; c01164f1 <schedule+1b1/390>
> 3: 8b 40 5c mov 0x5c(%eax),%eax
> Code; c01164f4 <schedule+1b4/390>
> 6: 85 f6 test %esi,%esi
> Code; c01164f6 <schedule+1b6/390>
> 8: 89 45 d4 mov %eax,0xffffffd4(%ebp)
> Code; c01164f9 <schedule+1b9/390>
> b: 75 35 jne 42 <_EIP+0x42> c0116530
> <schedule+1f0/390>
> Code; c01164fb <schedule+1bb/390>
> d: 89 42 5c mov %eax,0x5c(%edx)
> Code; c01164fe <schedule+1be/390>
> 10: f0 ff 40 14 lock incl 0x14(%eax)
>
> <0> Kernel panic: Attempted to kill the idle task!
> </recent ksymoops output>
>
>
>
>
> <older ksymoops output>
> ksymoops 2.4.1 on i686 2.4.19. Options used
> -V (specified)
> -k /proc/ksyms (specified)
> -l /proc/modules (default)
> -o /lib/modules/2.4.19/ (default)
> -m /boot/System.map-2.4.19 (default)
>
> Oops: 0000 2.4.19 #1 SMP Mon Oct 7 11:01:05 CDT 2002
> CPU: 2
> EIP: 0010:[<c01a71b7>] Tainted: P
> Using defaults from ksymoops -t elf32-i386 -a i386
> EFLAGS: 00010202
> eax: d6f37b7c ebx: 00000001 ecx: 00000000 edx: e5c5f280
> esi: d6f37b7c edi: e5c5f280 ebp: c93b2400 esp: c706fcb0
> ds: 0018 es: 0018 ss: 0018
> Process swapper (pid: 0, stackpage=c706f000)
> Stack: c01a6c7b e5c5f280 c93b2400 d6f37b7c 00000202 c93b2400 d6f37b7c
> 00001636
> 00004ca8 c93b2400 d6f37b7c 00001636 00004ca8 c01d2f74 d6f37b7c
> 00000286
> 00000286 0000000a c93b2400 d041ceb8 00001636 00004ca8 00000000
> 00000000
> Call Trace: [<c01a6c7b>] [<c01d2f74>] [<c01d2d00>] [<c01c798e>]
> [<d9848583>]
> [<c01c7acb>] [<c01c6804>] [<c01dfc38>] [<c01e0013>] [<c023ad55>]
> [<c023ae87>]
> [<c023b232>] [<d984db87>] [<c024574f>] [<d9854d63>] [<c0234629>]
> [<c0234350>]
> [<c011ef8b>] [<c011ee31>] [<c011ebbb>] [<c010a97e>] [<c0106e60>]
> [<c010d248>]
> [<c0106e60>] [<c0106e89>] [<c0106ee4>] [<c011a589>]
> Code: 8b 41 20 89 42 44 85 c0 75 e0 8b 42 34 85 c0 74 07 c7 42 34
>
> >>EIP; c01a71b7 <xfs_buf_iodone_callbacks+b7/180> <=====
> Trace; c01a6c7b <xfs_buf_item_unlock+8b/a0>
> Trace; c01d2f74 <xfs_trans_chunk_committed+1a4/208>
> Trace; c01d2d00 <xfs_trans_committed+10/e0>
> Trace; c01c798e <xlog_state_do_callback+14e/260>
> Trace; d9848583 <END_OF_CODE+11796564/????>
> Trace; c01c7acb <xlog_state_done_syncing+2b/70>
> Trace; c01c6804 <xlog_iodone+44/a0>
> Trace; c01dfc38 <pagebuf_iodone+8/80>
> Trace; c01e0013 <_end_pagebuf_page_io_multi+d3/110>
> Trace; c023ad55 <scsi_merge_requests_fn_dc+5e5/880>
> Trace; c023ae87 <scsi_merge_requests_fn_dc+717/880>
> Trace; c023b232 <scsi_init_io_vd+b2/480>
> Trace; d984db87 <END_OF_CODE+1179bb68/????>
> Trace; c024574f <cdrom_read_block+8f/d0>
> Trace; d9854d63 <END_OF_CODE+117a2d44/????>
> Trace; c0234629 <scsi_partsize+69/140>
> Trace; c0234350 <print_req_sense+0/20>
> Trace; c011ef8b <bh_action+4b/80>
> Trace; c011ee31 <tasklet_hi_action+61/a0>
> Trace; c011ebbb <do_softirq+7b/e0>
> Trace; c010a97e <do_IRQ+fe/110>
> Trace; c0106e60 <default_idle+0/40>
> Trace; c010d248 <call_do_IRQ+5/d>
> Trace; c0106e60 <default_idle+0/40>
> Trace; c0106e89 <default_idle+29/40>
> Trace; c0106ee4 <cpu_idle+24/30>
> Trace; c011a589 <printk+119/140>
> Code; c01a71b7 <xfs_buf_iodone_callbacks+b7/180>
> 00000000 <_EIP>:
> Code; c01a71b7 <xfs_buf_iodone_callbacks+b7/180> <=====
> 0: 8b 41 20 mov 0x20(%ecx),%eax <=====
> Code; c01a71ba <xfs_buf_iodone_callbacks+ba/180>
> 3: 89 42 44 mov %eax,0x44(%edx)
> Code; c01a71bd <xfs_buf_iodone_callbacks+bd/180>
> 6: 85 c0 test %eax,%eax
> Code; c01a71bf <xfs_buf_iodone_callbacks+bf/180>
> 8: 75 e0 jne ffffffea <_EIP+0xffffffea>
> c01a71a1 <xfs_buf_iodone_callbacks+a1/180>
> Code; c01a71c1 <xfs_buf_iodone_callbacks+c1/180>
> a: 8b 42 34 mov 0x34(%edx),%eax
> Code; c01a71c4 <xfs_buf_iodone_callbacks+c4/180>
> d: 85 c0 test %eax,%eax
> Code; c01a71c6 <xfs_buf_iodone_callbacks+c6/180>
> f: 74 07 je 18 <_EIP+0x18> c01a71cf
> <xfs_buf_iodone_callbacks+cf/180>
> Code; c01a71c8 <xfs_buf_iodone_callbacks+c8/180>
> 11: c7 42 34 00 00 00 00 movl $0x0,0x34(%edx)
>
> <0>Kernel panic: Aiee, killing interupt handler!
> </older ksymoops output>
--
Stephen Lord <lord@xxxxxxx>
|