Re: New XFS Survey Response

[Eric Sandeen <sandeen@sgi.com>]

Hi - 

Can you try running your oops through ksymoops manually?  The oops output
below looks a little odd.  For starters, you said this happened after
several minutes, but you show a couple log recovery functions in your stack,
which doesn't make sense unless the kernel oopsed during mount.

also, the mailing list at linux-xfs@oss.sgi.com is the best place
to send these reports.

Thanks,
-Eric

On Fri, 5 Jul 2002, hisaak@email.cz wrote:

> Name: hisaak 
> Email: hisaak@email.cz
> Company: 
> XFS version: SGI_XFS_1.1
> Number of machines: 1
> Std. Linux distribution: Redhat
> Other Linux distribution: 
> Hardware: other
> Total Number Systems: 1
> Total Number Server: 0
> Total Number Workstation: 1
> Average CPU count: 1
> Hard Drives: ide  
> Manufacturer(s): ibm, seagate
> Storage amount: 51 - 100 GB 
> LVM:  MD: true OTHER:  NONE: 
> Tape Drive: 
> Comments:
> I use xfs patches from http://oss.sgi.com/projects/xfs/patchlist.html for quite long time. When I tried them with 2.4.19-pre10 and 2.4.19-rc1 my system goes to trouble after several minutes of using xfs partition.
> This partition is 10G raid1 md device. With linux 2.4.17 I saw no problem for many months. Now it hangs after two or three minutes with this error in /var/log/messages:
> Jul  5 20:38:08 mrkvoslav kernel: Unable to handle kernel NULL pointer dereference at virtual address 0000010e
> Jul  5 20:38:08 mrkvoslav kernel:  printing eip:
> Jul  5 20:38:08 mrkvoslav kernel: c01bd6fd
> Jul  5 20:38:08 mrkvoslav kernel: pde = 00000000
> Jul  5 20:38:08 mrkvoslav kernel: Oops: 0000
> Jul  5 20:38:08 mrkvoslav kernel: CPU:    0
> Jul  5 20:38:08 mrkvoslav kernel: EIP:    0010:[xfs_inobt_get_rec+3501/4688]    Not tainted
> Jul  5 20:38:08 mrkvoslav kernel: EIP:    0010:[c01bd6fd]    Not tainted
> Jul  5 20:38:08 mrkvoslav kernel: EFLAGS: 00010246
> Jul  5 20:38:08 mrkvoslav kernel: eax: c02ecb80   ebx: ffffffe8   ecx: c02e5530   edx: c02ecb80
> Jul  5 20:38:08 mrkvoslav kernel: esi: 00000000   edi: de55a838   ebp: de55a824   esp: dcd69e38
> Jul  5 20:38:08 mrkvoslav kernel: ds: 0018   es: 0018   ss: 0018
> Jul  5 20:38:08 mrkvoslav kernel: Process kdeinit (pid: 4165, stackpage=dcd69000)
> Jul  5 20:38:08 mrkvoslav kernel: Stack: 00000000 00000000 f7228800 0000004b f22e1148 f70d64f0 c01d2d7c f7228800 
> Jul  5 20:38:08 mrkvoslav kernel:        00000000 0100c9af 00000000 00000000 dcd69ec0 00000000 00000000 00000000 
> Jul  5 20:38:08 mrkvoslav kernel:        00000008 00000004 dcc4741c c0123e71 e8e883c0 f70d6508 00000008 f70d64f0 
> Jul  5 20:38:08 mrkvoslav kernel: Call Trace: [xlog_recover_do_trans+37916/70656] [c0123e71] [xlog_recover_do_trans+55661/70656] [xfs_symlink+37461/71072] [c013b6ed] 
> Jul  5 20:38:08 mrkvoslav kernel: Call Trace: [c01d2d7c] [c0123e71] [c01d72cd] [c01e2a75] [c013b6ed] 
> Jul  5 20:38:08 mrkvoslav kernel:    [c013be71] [c013b43d] [c013c493] [c01393f4] [c0125212] [c010887c] 
> Jul  5 20:38:08 mrkvoslav kernel:    [c010878b] 
> Jul  5 20:38:08 mrkvoslav kernel: 
> Jul  5 20:38:08 mrkvoslav kernel: Code: 66 83 bb 26 01 00 00 00 75 0f 66 83 a3 18 01 00 00 f7 53 e8 
> 
> After this I must shutdown to get control over this partition again. :-(
> 
> I didn't find better place to post this bug so sorry for this. 
> I hope I will get some answer to hisaak@email.cz.
>