Re: 2.6.18-rc4-mm1 BUG null pointer deref while saving a file

From: Helge Hafting
Date: Fri Aug 18 2006 - 04:25:01 EST


Dave Kleikamp wrote:
On Wed, 2006-08-16 at 13:44 +0200, Helge Hafting wrote:
Aug 16 13:20:30 hh kernel: BUG: unable to handle kernel NULL pointer dereference at virtual address 00000000
Aug 16 13:20:30 hh kernel: printing eip:
Aug 16 13:20:30 hh kernel: c0468863
Aug 16 13:20:30 hh kernel: *pde = 00000000
Aug 16 13:20:30 hh kernel: Oops: 0002 [#1]
Aug 16 13:20:30 hh kernel: 4K_STACKS
Aug 16 13:20:30 hh kernel: last sysfs file: /devices/platform/i2c-9191/9191-0290/in0_input
Aug 16 13:20:30 hh kernel: CPU: 0
Aug 16 13:20:30 hh kernel: EIP: 0060:[<c0468863>] Not tainted VLI
Aug 16 13:20:30 hh kernel: EFLAGS: 00010002 (2.6.18-rc4-mm1 #7)
Aug 16 13:20:30 hh kernel: EIP is at __down+0x5c/0xed
Aug 16 13:20:30 hh kernel: eax: 00000000 ebx: d462e7e4 ecx: c14e7e40 edx: dde06f0c
Aug 16 13:20:30 hh kernel: esi: 00000286 edi: c3365550 ebp: d462e7ec esp: dde06efc
Aug 16 13:20:30 hh kernel: ds: 007b es: 007b ss: 0068
Aug 16 13:20:30 hh kernel: Process lyx (pid: 9353, ti=dde06000 task=c3365550 task.ti=dde06000)
Aug 16 13:20:30 hh kernel: Stack: dde06f0c 00000001 c3365550 c0116941 d462e7ec 00000000 00000008 d851a9a8
Aug 16 13:20:30 hh kernel: d462e7c0 d32a1200 c0466bc7 c14e7e40 dde06000 c01ed987 00000008 00000000
Aug 16 13:20:30 hh kernel: 00000000 00000000 d7a31c00 caf9be0c 00000000 00001dbb d462e7e4 00000008
Aug 16 13:20:30 hh kernel: Call Trace:
Aug 16 13:20:30 hh kernel: [<c0466bc7>] __down_failed+0x7/0xc
Aug 16 13:20:30 hh kernel: DWARF2 unwinder stuck at __down_failed+0x7/0xc
Aug 16 13:20:31 hh kernel: Leftover inexact backtrace:
Aug 16 13:20:31 hh kernel: [<c01ed987>] .text.lock.file+0x54/0x9d
Aug 16 13:20:31 hh kernel: [<c01516bd>] __fput+0xb2/0x163
Aug 16 13:20:31 hh kernel: [<c014ee9d>] filp_close+0x3e/0x62
Aug 16 13:20:31 hh kernel: [<c0150168>] sys_close+0x5c/0x6b
Aug 16 13:20:31 hh kernel: [<c01028c9>] sysenter_past_esp+0x56/0x79

By any chance would you be using cifs? This looks just like the bug
that is fixed by this patch:
http://www.kernel.org/git/gitweb.cgi?p=linux/kernel/git/sfrench/cifs-2.6.git;a=commitdiff;h=e466e4876bf39474e15d0572f2204578137ae7f5
Yes, this is on cifs. I should have mentioned that. Is this patch going
in anyway, or should I test it? I don't know a way of reproducing
the problem.


Helge Hafting
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/