Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 19 Aug 2004 20:55:04 +0300
From:      Ruslan Ermilov <ru@FreeBSD.org>
To:        Andre Oppermann <andre@FreeBSD.org>
Cc:        John Baldwin <jhb@FreeBSD.org>
Subject:   Re: ipfw2 broken
Message-ID:  <20040819175504.GC44018@ip.net.ua>
In-Reply-To: <4124E1DA.B5341C04@freebsd.org>
References:  <1092881027.999.3.camel@server.mcneil.com> <412497B5.6040203@freebsd.org> <4124D1D4.5080307@samsco.org> <200408191251.32319.jhb@FreeBSD.org> <4124E1DA.B5341C04@freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help

--TYecfFk8j8mZq+dy
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Aug 19, 2004 at 07:22:34PM +0200, Andre Oppermann wrote:
> John Baldwin wrote:
> >=20
> > On Thursday 19 August 2004 12:14 pm, Scott Long wrote:
> > > Andre Oppermann wrote:
> > > > Sean McNeil wrote:
> > > >> How do I get the ipfw2 module to compile with divert?  It doesn't
> > > >> recognize the following in my config file when building the module:
> > > >>
> > > >> options         IPFIREWALL_FORWARD
> > > >> options         IPDIVERT
> > > >>
> > > >> Also, the /etc/rc.d/ipfw script is looking for an invalid sysctl v=
ar:
> > > >>
> > > >> net.inet.ip.fw.enable
> > > >>
> > > >> and it will fail if I have the IPFIREWALL option which compiles th=
e code
> > > >> into the kernel because it will try to load and return 1 on a fail=
ure.
> > > >
> > > > I'm looking into it and will have a fix later today.
> > >
> > > This, and all of the rc.d, module loading, and kernel option problems
> > > are now a blocking issue for BETA1.  We had planned to start the BETA1
> > > no later than 2200 UTC today.  What is your schedule for getting all =
of
> > > this fixed?
> >=20
> > It looks like fixing the rc.d script is simply a matter of checking for=
 the fw
> > node rather than fw.enable.  The pfil(9) requirement is just a matter of
> > documenting the new requirement.  The IPDIVERT thing is probably larger
> > though. :(  We may need to just tell people to compile ipfw into the ke=
rnel
> > for now if they want divert sockets, much as they do if they want 'defa=
ult to
> > accept'.
>=20
> IPDIVERT wasn't compiled into the module before.  It's surrounded by
> #if !defined(KLD_MODULE).  However if the kernel was compiled with option
> IPDIVERT but w/o IPFIREWALL is was working anyway.
>=20
Will this produce a working "ipfw divert" now:

1.  Kernel compiled with IPDIVERT but w/o IPFIREWALL.
2.  ipfw.ko compiled with IPDIVERT in CFLAGS.


Cheers,
--=20
Ruslan Ermilov
ru@FreeBSD.org
FreeBSD committer

--TYecfFk8j8mZq+dy
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (FreeBSD)

iD8DBQFBJOl4qRfpzJluFF4RAshzAKCV4KhAEObggGbJeXRo93m7vvxzXQCgmsaG
SWUU/t7AFSINxBMx4CPPNWY=
=ZJ18
-----END PGP SIGNATURE-----

--TYecfFk8j8mZq+dy--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040819175504.GC44018>