Re: [Users] super-freeswan-1.99kb2 !pluto failure!

From: Ken Bantoft (ken_at_freeswan.ca)
Date: Sun Dec 15 2002 - 05:36:02 CET


-----BEGIN PGP SIGNED MESSAGE-----

Never seen/head of this before.. and I'm running RH 8.0 on this very
laptop.

On 13 Dec 2002, James P. Kinney III wrote:

> I'm running RedHat 8 with the super-freeswan-1.99kb2 and the patched
> kernel source rpm (_that_ took a while to compile :) of
> 2.4.18-18.8.0.2foo.
>
> When ipsec is started, pluto fails, sleeps 10 seconds, reloads and
> continues this process forever. I turned on full debugging ad have
> included a short snippet:
>
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_remove_socket: .
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_remove_socket:
> succeeded.
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_destroy_socket:
> pfkey_remove_socket called.
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_destroy_socket:
> sk(c45b1140)->(&c45b1188)receive_queue.{next=c45b1188,prev=c45b1188}.
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_destroy_socket:
> destroyed.
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_list_remove_socket:
> removing sock=cb7a2ac4
> Dec 13 09:28:39 castle last message repeated 7 times
> Dec 13 09:28:39 castle ipsec__plutorun: !pluto failure!: exited with
> error status 136 (signal 8)
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_list_remove_socket:
> removing sock=cb7a2ac4
> Dec 13 09:28:39 castle ipsec__plutorun: restarting IPsec after pause...
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_list_remove_socket:
> removing sock=cb7a2ac4
> Dec 13 09:28:39 castle last message repeated 3 times
> Dec 13 09:28:39 castle kernel: klips_debug:pfkey_release: succeeded.
> Dec 13 09:28:43 castle ipsec_setup: Stopping FreeS/WAN IPsec...
>
> The last line is where I killed it. In fact, the /etc/init.d/ipsec stop
> won't stop it. I added a procedure that would rename ipsec/_plutorun to
> force it to die after it's 10 second time out.
>
> Is this a config issue or something else I don't understand?

Could be... you didn't include any config bits, so it's a shot in the
dark.

- From "man 7 signal":

SIGFPE 8 C Floating point exception

So it's apparently a floating point exception.

This has been seen earlier with straight FreeS/WAN + X.509 patch:

http://lists.freeswan.org/pipermail/users/2002-September/013981.html

However no result was ever posted.

- --
Ken Bantoft The Unoffical FreeS/WAN Site:
ken_at_freeswan.ca http://www.freeswan.ca
                           PGP Key: finger ken_at_bantoft.org
"Anyone who considers arithmetical methods of producing
random digits is, of course, in a state of sin."
                    -- John Von Neumann, 1951

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3ia
Charset: noconv

iQCVAwUBPfwGtFiWUusaxGxpAQEBEQP8C3xRGosGZ8OpLyUjooshiazq7fjLCu+c
zXvoRIYfbZo/j3GVoKBPbJVQ7wzVOve5hBbKL5Hcmaq7v0CpmsJ70pDOUMBjgD/s
gDVg7nWFBeb2FFWTAF8fbAufBln0iFa/y5OUVKiAp5PICTctfgcRKZmo67zDrc9Q
FKe7j0ilqA8=
=khIy
-----END PGP SIGNATURE-----

_______________________________________________
Users mailing list
Users_at_lists.freeswan.org
http://lists.freeswan.org/mailman/listinfo/users



This archive was generated by hypermail 2.1.5 : Tue Dec 17 2002 - 05:21:05 CET