Ashish Kumar wrote:
> Now my problem is,
>
> When I insert the module and then call strace on any executable ( e.g.
> i simple "Hello World" program).... strace fails and kernel gives
> oops ...
> Below is the dmesg log....
>
> Can someone has any idea wt might be the reason?
> It is showing NULL pointer dereference at virtual address
> 00000000 ...
> My module is used for providing buffer overflow protection.
> It's doing this by hooking into syscalltable...
>
>
>
> <1>Unable to handle kernel NULL pointer dereference at virtual address
> 00000000
> printing eip:
> 00000000
> *pde = 00000000
> Oops: 0000 [#24]
> CPU: 0
> EIP: 0060:[<00000000>] Tainted: GF U
> EFLAGS: 00010213 (2.6.5-7.97-default)
> EIP is at 0x0
> eax: 00000000 ebx: bffff070 ecx: cdc56690 edx: 00000000
> esi: bffff070 edi: bfffefe4 ebp: c9eca000 esp: c9ecbfc0
> ds: 007b es: 007b ss: 0068
> Process a.out (pid: 30280, threadinfo=c9eca000 task=cdc56690)
> Stack: c0107e27 bffff070 00000000 400144d5 bffff070 bfffefe4 bfffeeb4
> ffffffda
> 0000007b 0000007b 00000005 40011c64 00000073 00000212 bfffee20
> 0000007b
> Call Trace:
> [] syscall_call+0x7/0xb
>
> Code: Bad EIP value.

Just an educated guess: As it's your application that causes the Oops
and noone else, maybe your module isn't correctly hooking into the
syscall table. As I understand it (never been there, never done that),
doing that is a nontrivial thing.
How about you dump some part of the syscall table before and after you
hook into it and see if the hook is been done correctly.

--
These are my personal views and not those of Fujitsu Siemens Computers!
Josef Möllers (Pinguinpfleger bei FSC)
If failure had no penalty success would not be a prize (T. Pratchett)
Company Details: http://www.fujitsu-siemens.com/imprint.html

On Dec 12, 12:42 pm, Josef Moellers siemens.com> wrote:
> Ashish Kumar wrote:
> > Now my problem is,
>
> > When I insert the module and then call strace on any executable ( e.g.
> > i simple "Hello World" program).... strace fails and kernel gives
> > oops ...
> > Below is the dmesg log....
>
> > Can someone has any idea wt might be the reason?
> > It is showing NULL pointer dereference at virtual address
> > 00000000 ...
> > My module is used for providing buffer overflow protection.
> > It's doing this by hooking into syscalltable...
>
> > <1>Unable to handle kernel NULL pointer dereference at virtual address
> > 00000000
> > printing eip:
> > 00000000
> > *pde = 00000000
> > Oops: 0000 [#24]
> > CPU: 0
> > EIP: 0060:[<00000000>] Tainted: GF U
> > EFLAGS: 00010213 (2.6.5-7.97-default)
> > EIP is at 0x0
> > eax: 00000000 ebx: bffff070 ecx: cdc56690 edx: 00000000
> > esi: bffff070 edi: bfffefe4 ebp: c9eca000 esp: c9ecbfc0
> > ds: 007b es: 007b ss: 0068
> > Process a.out (pid: 30280, threadinfo=c9eca000 task=cdc56690)
> > Stack: c0107e27 bffff070 00000000 400144d5 bffff070 bfffefe4 bfffeeb4
> > ffffffda
> > 0000007b 0000007b 00000005 40011c64 00000073 00000212 bfffee20
> > 0000007b
> > Call Trace:
> > [] syscall_call+0x7/0xb
>
> > Code: Bad EIP value.
>
> Just an educated guess: As it's your application that causes the Oops
> and noone else, maybe your module isn't correctly hooking into the
> syscall table. As I understand it (never been there, never done that),
> doing that is a nontrivial thing.
> How about you dump some part of the syscall table before and after you
> hook into it and see if the hook is been done correctly.
>
> --
> These are my personal views and not those of Fujitsu Siemens Computers!
> Josef Möllers (Pinguinpfleger bei FSC)
> If failure had no penalty success would not be a prize (T. Pratchett)
> Company Details:http://www.fujitsu-siemens.com/imprint.html- Hide quoted text -
>
> - Show quoted text -

Thanks , i see that ..