Oops on 2.6.14

Oops on 2.6.14

[Zorglub]

Hi list,

my machine suddenly froze. It's been working just fine for about a week or so. 
No unusual load. I read in Documentation/oops-tracing.txt that I should 
probably send the log (var/log/messages) to this list. So here goes:

Nov 20 23:07:52 asimov Unable to handle kernel NULL pointer dereference at 
virtual address 00000008
Nov 20 23:07:52 asimov printing eip:
Nov 20 23:07:52 asimov c0114096
Nov 20 23:07:52 asimov *pde = 00000000
Nov 20 23:07:52 asimov Oops: 0000 [#1]
Nov 20 23:07:52 asimov PREEMPT
Nov 20 23:07:52 asimov Modules linked in: nvidia
Nov 20 23:07:52 asimov CPU:    0
Nov 20 23:07:52 asimov EIP:    0060:[<c0114096>]    Tainted: P      VLI
Nov 20 23:07:52 asimov EFLAGS: 00010282   (2.6.14-gentoo-r2)
Nov 20 23:07:52 asimov EIP is at copy_mm+0x226/0x3f0
Nov 20 23:07:52 asimov eax: 00000000   ebx: cdca1bcc   ecx: df549040   edx: 
dc5c88e0
Nov 20 23:07:52 asimov esi: d50b343c   edi: cdca1c24   ebp: d50b33e4   esp: 
dbae9ec8
Nov 20 23:07:52 asimov ds: 007b   es: 007b   ss: 0068
Nov 20 23:07:52 asimov Process sh (pid: 12595, threadinfo=dbae8000 
task=d60ea580)
Nov 20 23:07:52 asimov Stack: cdca1bcc 000000d0 cdca15f4 cdca1244 dbae8000 
df93b4f0 cdca160c cdca1614
Nov 20 23:07:52 asimov cdca1600 df93b4c0 df549040 c751b5e0 c835f878 00000000 
01200011 c0114be4
Nov 20 23:07:52 asimov 01200011 c751b5e0 00000500 00000000 df93b4c0 00000001 
df7b30c0 dbae8000
Nov 20 23:07:52 asimov Call Trace:
Nov 20 23:07:52 asimov [<c0114be4>] copy_process+0x424/0xd80
Nov 20 23:07:52 asimov [<c0115649>] do_fork+0x69/0x182
Nov 20 23:07:52 asimov [<c0209512>] copy_to_user+0x42/0x60
Nov 20 23:07:52 asimov [<c0122900>] sys_rt_sigprocmask+0xa0/0x100
Nov 20 23:07:52 asimov [<c01019de>] sys_clone+0x3e/0x50
Nov 20 23:07:52 asimov [<c0102e27>] sysenter_past_esp+0x54/0x75
Nov 20 23:07:52 asimov Code: 00 89 ee f3 a5 81 60 14 ff df ff ff 8b 4c 24 28 
c7 40 0c 00 00 00 00 89 08 89 04 24 e8 04 63 03 00 8b 53 4c 85 d2 74 40 8b 42 
08
 <8b> 40 08 ff 42 14 f6 43 15 08 74 06 ff 88 30 01 00 00 8b 44 24

Re: Oops on 2.6.14

[Con Kolivas]

On Mon, 21 Nov 2005 09:26, Zorglub wrote:
> Hi list,
>
> my machine suddenly froze. It's been working just fine for about a week or
> so. No unusual load. I read in Documentation/oops-tracing.txt that I should
> probably send the log (var/log/messages) to this list. So here goes:

Thank you for your bug report.

> Nov 20 23:07:52 asimov Modules linked in: nvidia
> Nov 20 23:07:52 asimov EIP:    0060:[<c0114096>]    Tainted: P      VLI

Unfortunately we can only provide support for code we wrote and have access 
to. The nvidia and other proprietary binary only drivers cause bugs that we 
are unable to debug. Please try to reproduce the problem without any binary 
only drivers.

Regards,
Con

Re: Oops on 2.6.14

[Zorglub Olsen]

Hi,

Con Kolivas wrote:
>Unfortunately we can only provide support for code we wrote and have access
>to. The nvidia and other proprietary binary only drivers cause bugs that we
>are unable to debug. Please try to reproduce the problem without any binary
>only drivers.
>

Thank you for your answer. I understand about the binary, I just wasn't able 
to see that it was the sole cause of the oops.


//Niels

Re: Oops on 2.6.14

[Con Kolivas]

On Mon, 21 Nov 2005 12:41, Zorglub Olsen wrote:
> Hi,
>
> Con Kolivas wrote:
> >Unfortunately we can only provide support for code we wrote and have
> > access to. The nvidia and other proprietary binary only drivers cause
> > bugs that we are unable to debug. Please try to reproduce the problem
> > without any binary only drivers.
>
> Thank you for your answer. I understand about the binary, I just wasn't
> able to see that it was the sole cause of the oops.

That's an understandable misconception. Drivers can cause corruption anywhere, 
even if the stack trace does not look related to the driver.

Regards,
Con