2005-04-22 23:54:17

by Christian Kujau

[permalink] [raw]
Subject: NFS oops with 2.6.12-rc2-mm3

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

hello,

upon unmounting some NFSv3 shares, the following oops happened and i could
not remount any nfs shares after this:

lockd: cannot unmonitor 192.168.10.10
Unable to handle kernel NULL pointer dereference at virtual address 00000000
printing eip:
00000000
*pde = 00000000
Oops: 0000 [#1]
PREEMPT
Modules linked in: pdc202xx_old snd_ens1371 snd_rawmidi snd_ac97_codec
usbmouse usbkbd uhci_hcd via82cxxx via_agp agpgart usbcore autofs4
snd_pcm_oss snd_pcm snd_timer snd_page_alloc snd_mixer_oss snd soundcore
ide_cd cdrom ide_disk ide_core
CPU: 0
EIP: 0060:[<00000000>] Not tainted VLI
EFLAGS: 00010286 (2.6.12-rc2-mm3)
EIP is at 0x0
eax: c1617ec0 ebx: c1617ec0 ecx: 00000000 edx: c1771a50
esi: 00000000 edi: c1617f34 ebp: df0fee40 esp: dec6ff2c
ds: 007b es: 007b ss: 0068
Process rpciod/0 (pid: 8166, threadinfo=dec6f000 task=c1771a50)
Stack: c03976db c1771a50 c1771b78 00000296 c1617f3c 00000293 c1617f40 c012a0ae
00000000 00000000 dffc6550 dec6f000 df0fee58 df0fee48 df0fee50 dec6f000
c1617ec0 c0397810 dec6f000 ffffffff ffffffff 00000001 00000000 c0116ac0
Call Trace:
[<c03976db>] __rpc_execute+0x14b/0x250
[<c012a0ae>] worker_thread+0x1ae/0x280
[<c0397810>] rpc_async_schedule+0x0/0x10
[<c0116ac0>] default_wake_function+0x0/0x10
[<c0116b07>] __wake_up_common+0x37/0x60
[<c0116ac0>] default_wake_function+0x0/0x10
[<c0129f00>] worker_thread+0x0/0x280
[<c012e3a5>] kthread+0x95/0xd0
[<c012e310>] kthread+0x0/0xd0
[<c010132d>] kernel_thread_helper+0x5/0x18
Code: Bad EIP value.


i rarely use -mm kernels so i don't have much experience yet. some more
details are here: http://nerdbynature.de/bits/prinz/2.6.12-rc2-mm3/

thank you,
Christian.

- --
BOFH excuse #67:

descramble code needed from software company
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFCaY4k+A7rjkF8z0wRAsXGAKCTO1hlydgUbSoeklNOPttsZ3uUJQCdH3SS
bw4wvNW8wzWeAqPTzJObwZ8=
=VolN
-----END PGP SIGNATURE-----


2005-04-29 11:05:22

by Andrew Morton

[permalink] [raw]
Subject: Re: NFS oops with 2.6.12-rc2-mm3

Christian Kujau <[email protected]> wrote:
>
> upon unmounting some NFSv3 shares, the following oops happened and i could
> not remount any nfs shares after this:
>
> lockd: cannot unmonitor 192.168.10.10
> Unable to handle kernel NULL pointer dereference at virtual address 00000000
> printing eip:
> 00000000
> *pde = 00000000
> Oops: 0000 [#1]
> PREEMPT
> Modules linked in: pdc202xx_old snd_ens1371 snd_rawmidi snd_ac97_codec
> usbmouse usbkbd uhci_hcd via82cxxx via_agp agpgart usbcore autofs4
> snd_pcm_oss snd_pcm snd_timer snd_page_alloc snd_mixer_oss snd soundcore
> ide_cd cdrom ide_disk ide_core
> CPU: 0
> EIP: 0060:[<00000000>] Not tainted VLI
> EFLAGS: 00010286 (2.6.12-rc2-mm3)
> EIP is at 0x0
> eax: c1617ec0 ebx: c1617ec0 ecx: 00000000 edx: c1771a50
> esi: 00000000 edi: c1617f34 ebp: df0fee40 esp: dec6ff2c
> ds: 007b es: 007b ss: 0068
> Process rpciod/0 (pid: 8166, threadinfo=dec6f000 task=c1771a50)
> Stack: c03976db c1771a50 c1771b78 00000296 c1617f3c 00000293 c1617f40 c012a0ae
> 00000000 00000000 dffc6550 dec6f000 df0fee58 df0fee48 df0fee50 dec6f000
> c1617ec0 c0397810 dec6f000 ffffffff ffffffff 00000001 00000000 c0116ac0
> Call Trace:
> [<c03976db>] __rpc_execute+0x14b/0x250
> [<c012a0ae>] worker_thread+0x1ae/0x280
> [<c0397810>] rpc_async_schedule+0x0/0x10
> [<c0116ac0>] default_wake_function+0x0/0x10
> [<c0116b07>] __wake_up_common+0x37/0x60
> [<c0116ac0>] default_wake_function+0x0/0x10
> [<c0129f00>] worker_thread+0x0/0x280
> [<c012e3a5>] kthread+0x95/0xd0
> [<c012e310>] kthread+0x0/0xd0
> [<c010132d>] kernel_thread_helper+0x5/0x18
> Code: Bad EIP value.
>
>
> i rarely use -mm kernels so i don't have much experience yet. some more
> details are here: http://nerdbynature.de/bits/prinz/2.6.12-rc2-mm3/

If this is at all repeatable, please enable CONFIG_DEBUG_SLAB, see if we
get some more info.