2.6.5 panic while unloading cisco_ipsec

2.6.5 panic while unloading cisco_ipsec

[Geoff Mishkin]

Please let me know if I'm posting this in the wrong place.  I get a
kernel panic with kernel 2.6.5 (compiled with gcc 2.95.3) when I try to
unload the Cisco VPN client's (4.0.3b) cisco_ipsec module.  I know it's
a binary-only module, so I'm not sure how much help this will be, but
here's the text I managed to pull off the screen when I tried rmmod'ing
it:

(Not sure if there is anything above this, this is all that fit on the
screen, and I didn't find anything in the logs)
PREEMPT
CPU:	0
EIP:	0060:[<c02a9d26>]    Tainted: PF
EFLAGS:	00010202   (2.6.5)
EIP is at qdisc_destroy+0x6/0xc4
eax: 00000064	ebx: c6daa000	ecx: c6dabf20	edx: c6dabf28
esi: 00000064	edi: ce5fd028	ebp: c6dabf6c	esp: c6dabf28
ds: 007b   es: 007b   ss:0068
Process rmmod (pid: 7134, threadinfo=c6daa000 task=c6db19c0)
Stack: c6daa000 d0abb600 c02a9ff3 00000064 d0abb600 d0abb600 c02a1426
d0abb600
       d0abb600 00000000 c0398a9c c02440c8 d0abb600 d0ae8f80 d0a99aec
d0abb600
       d0ae8f80 c6daa000 c012f724 bffff730 00000000 bffff730 00000000
63736963
Call Trace:
 [<c02a9ff3>] dev_shutdown+0x57/0xe4
 [<c02a1426>] unregister_netdevice+0x192/0x288
 [<c02440c8>] unregister_netdev+0x10/0x28
 [<d0a99aec>] cleanup_module+0x1c/0x30 [cisco_ipsec]
 [<c012f724>] sys_delete_module+0x140/0x174
 [<c0143803>] sys_munmap+0x37/0x54
 [<c0108f2f>] syscall_call+0x7/0xb

Code: 8b 5e 0c ff 4e 18 0f 94 c0 84 c0 0f 84 a8 00 00 00 8b 56 2c
 <0>Kernel panic: Fatal exception in interrupt
In interrupt handler - not syncing

			--Geoff Mishkin <gmishkin@comcast.net>

Re: 2.6.5 panic while unloading cisco_ipsec

[Helge Deller]

Hi Geoff,

I don't have a fix for you, but I would propose that you try an open source alternative instead:
http://www.unix-ag.uni-kl.de/~massar/vpnc/

It worked for me like a charm, and the most important issue is, that it's a complete user-space application without any need for kernel modules at all :-)

Best regards,
Helge

On Thursday 13 May 2004 23:12, Geoff Mishkin wrote:
> Please let me know if I'm posting this in the wrong place.  I get a
> kernel panic with kernel 2.6.5 (compiled with gcc 2.95.3) when I try to
> unload the Cisco VPN client's (4.0.3b) cisco_ipsec module.  I know it's
> a binary-only module, so I'm not sure how much help this will be, but
> here's the text I managed to pull off the screen when I tried rmmod'ing
> it:
> 
> (Not sure if there is anything above this, this is all that fit on the
> screen, and I didn't find anything in the logs)
> PREEMPT
> CPU:	0
> EIP:	0060:[<c02a9d26>]    Tainted: PF
> EFLAGS:	00010202   (2.6.5)
> EIP is at qdisc_destroy+0x6/0xc4
> eax: 00000064	ebx: c6daa000	ecx: c6dabf20	edx: c6dabf28
> esi: 00000064	edi: ce5fd028	ebp: c6dabf6c	esp: c6dabf28
> ds: 007b   es: 007b   ss:0068
> Process rmmod (pid: 7134, threadinfo=c6daa000 task=c6db19c0)
> Stack: c6daa000 d0abb600 c02a9ff3 00000064 d0abb600 d0abb600 c02a1426
> d0abb600
>        d0abb600 00000000 c0398a9c c02440c8 d0abb600 d0ae8f80 d0a99aec
> d0abb600
>        d0ae8f80 c6daa000 c012f724 bffff730 00000000 bffff730 00000000
> 63736963
> Call Trace:
>  [<c02a9ff3>] dev_shutdown+0x57/0xe4
>  [<c02a1426>] unregister_netdevice+0x192/0x288
>  [<c02440c8>] unregister_netdev+0x10/0x28
>  [<d0a99aec>] cleanup_module+0x1c/0x30 [cisco_ipsec]
>  [<c012f724>] sys_delete_module+0x140/0x174
>  [<c0143803>] sys_munmap+0x37/0x54
>  [<c0108f2f>] syscall_call+0x7/0xb
> 
> Code: 8b 5e 0c ff 4e 18 0f 94 c0 84 c0 0f 84 a8 00 00 00 8b 56 2c
>  <0>Kernel panic: Fatal exception in interrupt
> In interrupt handler - not syncing
> 
> 			--Geoff Mishkin <gmishkin@comcast.net>

Re: 2.6.5 panic while unloading cisco_ipsec

[Felipe Alfaro Solana]

On Thu, 2004-05-13 at 23:12, Geoff Mishkin wrote:
> Please let me know if I'm posting this in the wrong place.  I get a
> kernel panic with kernel 2.6.5 (compiled with gcc 2.95.3) when I try to
> unload the Cisco VPN client's (4.0.3b) cisco_ipsec module.  I know it's
> a binary-only module, so I'm not sure how much help this will be, but
> here's the text I managed to pull off the screen when I tried rmmod'ing
> it:

I'm thinking the new register arguments syscall convention is the
culprit. Please, make sure you compile the kernel with CONFIG_REGPARMS
set to NO. You can check that by grepping into the .config file before
starting the build.

> (Not sure if there is anything above this, this is all that fit on the
> screen, and I didn't find anything in the logs)
> PREEMPT
> CPU:	0
> EIP:	0060:[<c02a9d26>]    Tainted: PF
> EFLAGS:	00010202   (2.6.5)
> EIP is at qdisc_destroy+0x6/0xc4
> eax: 00000064	ebx: c6daa000	ecx: c6dabf20	edx: c6dabf28
> esi: 00000064	edi: ce5fd028	ebp: c6dabf6c	esp: c6dabf28
> ds: 007b   es: 007b   ss:0068
> Process rmmod (pid: 7134, threadinfo=c6daa000 task=c6db19c0)
> Stack: c6daa000 d0abb600 c02a9ff3 00000064 d0abb600 d0abb600 c02a1426
> d0abb600
>        d0abb600 00000000 c0398a9c c02440c8 d0abb600 d0ae8f80 d0a99aec
> d0abb600
>        d0ae8f80 c6daa000 c012f724 bffff730 00000000 bffff730 00000000
> 63736963
> Call Trace:
>  [<c02a9ff3>] dev_shutdown+0x57/0xe4
>  [<c02a1426>] unregister_netdevice+0x192/0x288
>  [<c02440c8>] unregister_netdev+0x10/0x28
>  [<d0a99aec>] cleanup_module+0x1c/0x30 [cisco_ipsec]
>  [<c012f724>] sys_delete_module+0x140/0x174
>  [<c0143803>] sys_munmap+0x37/0x54
>  [<c0108f2f>] syscall_call+0x7/0xb
> 
> Code: 8b 5e 0c ff 4e 18 0f 94 c0 84 c0 0f 84 a8 00 00 00 8b 56 2c
>  <0>Kernel panic: Fatal exception in interrupt
> In interrupt handler - not syncing
> 
> 			--Geoff Mishkin <gmishkin@comcast.net>
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/