Subject: Re: ooops report 2.4.20-xfs-CVS-2003-02-21_06:00_UTC
From: Steve Lord <lord@sgi.com>
To: "Jeffrey E. Hundstad" <jeffrey.hundstad@mnsu.edu>
Cc: linux-xfs@oss.sgi.com
In-Reply-To: <3ECBB68E.2030306@mnsu.edu>
References: <3ECBB68E.2030306@mnsu.edu>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Organization: 
Message-Id: <1053550982.21472.0.camel@jen.americas.sgi.com>
Mime-Version: 1.0
Date: 21 May 2003 16:03:02 -0500
Sender: linux-xfs-bounce@oss.sgi.com
Errors-to: linux-xfs-bounce@oss.sgi.com
Precedence: bulk
Content-Length: 4523
Lines: 127

On Wed, 2003-05-21 at 12:25, Jeffrey E. Hundstad wrote:
> Hello,

There have been major changes in the sync code in the last few
weeks (after the code you are running). Please try a new kernel from 
CVS.

Steve


> 
> We had a machine hang on us twice since upgrading from 2.4.18-xfs.  It 
> ran for months without any problems.  We upgraded to 
> 2.4.20-xfs-CVS-2003-02-21_06:00_UTC and patched it for the ptrace 
> vurnerablility.  When the machine hangs it has a VERY high load average 
> ~200 and as soon as you issue a disk-io request of any kind your process 
>   hangs.  The hang has been happening during a cpio backup of the system.
> 
> This last time we caught some/most of an ooops.  I'll include the 
> ksymoops output and the raw input. I think we may be able to glean more 
> information from the report, but I'm not really familiar with the ooops 
> output.  Let me know if I can do anything to make the report better.
> 
> Here's some info. about the computer.
> 
> CPU: Pentium III (Coppermine) x 2 x 667 MHz
> Compiler: gcc version 2.95.4 20011002 (Debian prerelease)
> Drives:  software raid1 - 2xMaxtor 4G160J8-ide drives (with the write 
> cache turned off)
> 
> ksymoops'ed report:
> 
> ksymoops 2.4.9 on i686 2.4.20-xfs.  Options used
>       -V (default)
>       -k /var/log/ksymoops/20030521062552.ksyms (specified)
>       -l /var/log/ksymoops/20030521062552.modules (specified)
>       -o /lib/modules/2.4.20-xfs (specified)
>       -m /boot/System.map-2.4.20-xfs (specified)
> 
> Code: 39 70 30 0f 85 9d 00 00 00 39 78 34 0f 85 94 00 00 00 8b 50
> Using defaults from ksymoops -t elf32-i386 -a i386
> 
> 
> Code;  00000000 Before first symbol
> 00000000 <_EIP>:
> Code;  00000000 Before first symbol
>     0:   39 70 30                  cmp    %esi,0x30(%eax)
> Code;  00000003 Before first symbol
>     3:   0f 85 9d 00 00 00         jne    a6 <_EIP+0xa6> 000000a6 Before 
> first symbol
> Code;  00000009 Before first symbol
>     9:   39 78 34                  cmp    %edi,0x34(%eax)
> Code;  0000000c Before first symbol
>     c:   0f 85 94 00 00 00         jne    a6 <_EIP+0xa6> 000000a6 Before 
> first symbol
> Code;  00000012 Before first symbol
>    12:   8b 50 00                  mov    0x0(%eax),%edx
> 
> <1> unable to handle kernel paging request at virtual address a4446e2b
> c01ce450
> *pde = 00000000
> CPU: 0
> EIP: 0010:[<c01ce450>] Not tainted
> EFLAGS: 00010286
> Stack: ded82c00 ded82c48 00000000
>         00000000
>         d6054840
> Call Trace: [<c01d3900>] [<c01ce315>] [<c01e0c73>]
>          [<c013d5da>] [<c013c7b0>]
>          [<c013ca9a>] [<c0107134>]
> Code: 83 7b 14 00 0f 84 79 07 00 00 8b 6b 1c 85 ed 0f 85 db 00 00
> 
> 
>  >>EIP; c01ce450 <xfs_syncsub+134/b78>   <=====
> 
> Trace; c01d3900 <xfs_inode_flush+1b0/1cc>
> Trace; c01ce315 <xfs_sync+15/1c>
> Trace; c01e0c73 <linvfs_write_super+27/2c>
> Trace; c013d5da <sync_supers+f6/148>
> Trace; c013c7b0 <sync_old_buffers+34/9c>
> Trace; c013ca9a <kupdate+fe/120>
> Trace; c0107134 <arch_kernel_thread+28/38>
> 
> Code;  c01ce450 <xfs_syncsub+134/b78>
> 00000000 <_EIP>:
> Code;  c01ce450 <xfs_syncsub+134/b78>   <=====
>     0:   83 7b 14 00               cmpl   $0x0,0x14(%ebx)   <=====
> Code;  c01ce454 <xfs_syncsub+138/b78>
>     4:   0f 84 79 07 00 00         je     783 <_EIP+0x783> c01cebd3 
> <xfs_syncsub+8b7/b78>
> Code;  c01ce45a <xfs_syncsub+13e/b78>
>     a:   8b 6b 1c                  mov    0x1c(%ebx),%ebp
> Code;  c01ce45d <xfs_syncsub+141/b78>
>     d:   85 ed                     test   %ebp,%ebp
> Code;  c01ce45f <xfs_syncsub+143/b78>
>     f:   0f 85 db 00 00 00         jne    f0 <_EIP+0xf0> c01ce540 
> <xfs_syncsub+224/b78>
> 
> 
> Raw ooops hand copied from screen:
> 
> Code: 39 70 30 0f 85 9d 00 00 00 39 78 34 0f 85 94 00 00 00 8b 50
> <1> unable to handle kernel paging request at virtual address a4446e2b
> printing eip:
> c01ce450
> *pde = 00000000
> Ooops: 0000
> st sg appletalk eepro100 mii lvm-mod raid5 xor raid1 raid0 linear md aic7xxx
> CPU: 0
> EIP: 0010:[<c01ce450>] Not tainted
> EFLAGS: 00010286
> eax:
> es:
> ds:
> Process kupdated (pid: 7 stackpage = c1621000
> Stack: ded82c00 ded82c48 00000000
>         00000000
>         d6054840
> Call Trace: [<c01d3900>] [<c01ce315>] [<c01e0c73>]
> 	[<c013d5da>] [<c013c7b0>]
> 	[<c013ca9a>] [<c0107134>]
> Code: 83 7b 14 00 0f 84 79 07 00 00 8b 6b 1c 85 ed 0f 85 db 00 00
-- 

Steve Lord                                      voice: +1-651-683-3511
Principal Engineer, Filesystem Software         email: lord@sgi.com