lkml.org 
[lkml]   [1998]   [Jul]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectQuota related oops in 2.1.110

Hi,

I seem to be having problems with quotas. I had some processes in the D
state (rm -r of a large directory, among others), so I decided to reboot
my machine. When the scripts executed quotaoff I received the following
oops (copied by hand, there could be errors in it):

Unable to handle kernel NULL pointer dereference at virtual address 00000008
current->tss.cr3=07280000, %cr3=07280000
*pde=00000000
Oops: 0000
CPU: 1
EIP: 0010:[<c0141ae3>]
EFLAGS: 00010282
eax: 00000000 ebx: c7edd800 ecx: 00000346 edx: 00000346
esi: c66cfd20 edi: c0366b80 ebp: 00000000 esp: c6659f74
ds: 0018 es: 0018 ss: 0018
Process quotaoff (pid: 2060, process nr: 52, stackpage=c665900)
Stack: 00006000 00000200 400a811c 00000000 00000000 03460000 c0142100 00000346
00000000 c6658000 00000000 400aa11c bffffd30 ffffffea 00000200 c74903e0
03468000 00000000 c6659fc4 c010b508 00020000 400a811c 00000000 00000000
Call Trace: [<c0142100>] [<c010b508>]
Code: 8b 58 08 85 db 74 38 0f b7 4c 24 16 66 39 4b 20 75 2d 80 7b

Using `/boot/System.map21-test' to map addresses to symbols.

>>EIP: c0141ae3 <quota_off+9b/1b8>
Trace: c0142100 <sys_quotactl+24c/428>
Trace: c010b508 <system_call+38/40>
Code: c0141ae3 <quota_off+9b/1b8>
Code: c0141ae3 <quota_off+9b/1b8> 8b 58 08 movl 0x8(%eax),%ebx
Code: c0141ae6 <quota_off+9e/1b8> 85 db testl %ebx,%ebx
Code: c0141ae8 <quota_off+a0/1b8> 74 38 je c0141b22 <quota_off+da/1b8>
Code: c0141aea <quota_off+a2/1b8> 0f b7 4c 24 16 movzwl 0x16(%esp,1),%ecx
Code: c0141af5 <quota_off+ad/1b8> 66 39 4b 20 cmpw %cx,0x20(%ebx)
Code: c0141af9 <quota_off+b1/1b8> 75 2d jne c0141b22 <quota_off+da/1b8>
Code: c0141afb <quota_off+b3/1b8> 80 7b 00 90 cmpb $0x90,0x0(%ebx)
Code: c0141b05 <quota_off+bd/1b8> 90 nop
Code: c0141b06 <quota_off+be/1b8> 90 nop

After this oops I couldn't correctly shut down my machine. So I tried to
use SysRQ to sync, unmount, reboot. But the sync was only partial:

SysRq: Emergency Sync 0
Syncing device 03:46 ... ok
...
Syncing device 03:46 ...

The sync of hda6 never finished, hda6 is the quota filesystem on which I
was doing the rm.

In february I already reported problems with quotas (at that time it was
2.1.70 - 2.1.84). If I remember correctly there was some discussion then
about this, but the problem was never resolved.

I have not had this problem since about that time. I did recreate my
.quota.user files recently, so it could be a problem with the state of
those files? Either that, or some changes to the quota system have been
reversed?

I hope we can get this solved. If you have any more questions, feel free
to ask. Or if you have any further suggestions to further investigate this
problem when it happens again.

Rafael D'Halleweyn
rdhall@mail.dma.be


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html

\
 
 \ /
  Last update: 2005-03-22 13:43    [W:0.059 / U:0.036 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site