[Openswan Users] - Building module saga (was: kernel panic )

[Paul Wouters]

On Thu, 10 Mar 2005 mario.lobo at ipad.com.br wrote:

> Machine 10.2.1.98 is auto=add
> Machine 10.2.1.99 is auto=start
> 
> When negociations start, I get what you see bellow:
> 
> ========================================
> 
> [ /var/log/secure ]
> 
> Mar 10 07:41:13 Spyket ipsec__plutorun: Starting Pluto subsystem...
> Mar 10 07:41:14 Spyket pluto[2595]: Starting Pluto (Openswan Version 2.CVSHEAD X.509-1.5.4 
> PLUTO_SENDS_VENDORID PLUTO_USES_KEYRR; Vendor ID OEEFBy\177du|_[)

That line looks a bit bad, Michael? Is there a problem in head with our
vendorid?

> "/etc/spyket/conf/vpn/secrets/centro.secrets"
> Mar 10 07:43:37 Spyket pluto[2595]: packet from 10.2.1.99:500: received Vendor ID payload [Openswan 
> (this version) 2.CVSHEAD  X.509-1.5.4 PLUTO_SENDS_VENDORID PLUTO_USES_KEYRR]
> Mar 10 07:43:37 Spyket pluto[2595]: packet from 10.2.1.99:500: received Vendor ID payload [Dead 
> Peer Detection]
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: responding to Main Mode
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: transition from state STATE_MAIN_R0 to state 
> STATE_MAIN_R1
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: transition from state STATE_MAIN_R1 to state 
> STATE_MAIN_R2
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: Main mode peer ID is ID_FQDN: '@ponta1-right'
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: I did not send a certificate because I do not have 
> one.
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: transition from state STATE_MAIN_R2 to state 
> STATE_MAIN_R3
> Mar 10 07:43:37 Spyket pluto[2595]: "centro" #1: sent MR3, ISAKMP SA established
> Mar 10 07:43:38 Spyket pluto[2595]: "centro" #2: responding to Quick Mode {msgid:30366b6f}

So far so good, but there should be more then this.
 
> ===> Here is when I get the kernel messages bellow
> 
> Mar 10 07:43:48 Spyket ipsec__plutorun: Restarting Pluto subsystem...
> 
> 
> [ Kernel Messages: ]
> 
> Mar 10 07:43:38 Spyket kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000000

yes, unloading klips causes this failure. I thought we removed unloading the
module from cvs. Are you running a fresh cvs checkout or could this tree be
a few weeks old?

Paul