Subject: Re: Some IPF 4.1.22 regression tests fail
To: Martti Kuparinen <martti.kuparinen@iki.fi>
From: Darren Reed <darrenr@NetBSD.org>
List: current-users
Date: 05/17/2007 13:13:01
On Wed, May 16, 2007 at 02:35:30PM +0300, Martti Kuparinen wrote:
> 
....
> diff -u expected/$i results/$i; done
> --- expected/n4 2007-05-01 22:08:48.000000000 +0300
> +++ results/n4  2007-05-16 14:27:06.000000000 +0300
> @@ -53,14 +53,14 @@
>  ip #0 40(20) 6 10.3.3.3,12345 > 10.1.1.0,53
>  ip #0 40(20) 6 10.1.1.1,53 > 10.3.3.3,12345
>  -------------------------------
> -ip #0 40(20) 6 10.3.3.3,12345 > 10.2.2.1,23
> +ip #0 40(20) 6 10.3.3.3,12345 > 10.1.1.1,23

There is something wrong here...

...
> --- expected/ni23       2007-05-01 22:08:49.000000000 +0300
> +++ results/ni23        2007-05-16 14:27:06.000000000 +0300
> @@ -2,7 +2,7 @@
>  ip #0 28(20) 17 2.2.2.2,4500 > 3.3.3.1,6700
>  ip #0 28(20) 17 1.1.2.3,4500 > 3.3.3.1,6700
>  List of active MAP/Redirect filters:
> -rdr le0,bge0 1.1.0.0/16 -> 2.2.2.2 ip
> +rdr le0,bge0 1.1.0.0/16 -> 2.2.2.2 hopopt
>  map hme0,ppp0 3.3.3.0/24 -> 4.4.4.4/32

Hmm.... ip -> hopopt will be from a change to /etc/protocols.

I need to think about what this means.

If you revert to an older /etc/protocols file, do they all pass?

Darren