From mboxrd@z Thu Jan 1 00:00:00 1970 From: Harald Welte Subject: Re: iptables is complaining with bogus unknown error 18446744073709551615 Date: Thu, 27 Apr 2006 10:51:19 -0300 Message-ID: <20060427135119.GB5177@rama> References: <200604210738.k3L7cBGO010103@mailgw.aecom.yu.edu> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="yNb1oOkm5a9FJOVX" Return-path: Content-Disposition: inline In-Reply-To: List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: netfilter-bounces@lists.netfilter.org Errors-To: netfilter-bounces@lists.netfilter.org To: Maurice Volaski Cc: netfilter@lists.netfilter.org, linux-kernel@vger.kernel.org --yNb1oOkm5a9FJOVX Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Wed, Apr 26, 2006 at 09:12:38PM -0400, Maurice Volaski wrote: > Automatic kernel module loading! That is an option and it's off by > default. When it's off, attempts to load kernel modules are ignored > internally, and that's why iptables was failing. It tried to load=20 > xt_tcpudp, but was ignored by the kernel. What do you mean by "it's an option" and "is off by default". I would claim that any major linux distribution that I've seen in the last ten years has support for module auto loading (enabled by default). There are many userspace programs that try to autoload modules, such as device-mapper, ipsec, etc. =20 If you disable module autoloading, it's your own responsibility to load modules manually. So the only thing that I really consider a bug is that bogus error message of iptables. This has been fixed in SVN, case closed. --=20 - Harald Welte http://netfilter.org/ =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D "Fragmentation is like classful addressing -- an interesting early architectural error that shows how much experimentation was going on while IP was being designed." -- Paul Vixie --yNb1oOkm5a9FJOVX Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) iD8DBQFEUMxXXaXGVTD0i/8RAk4rAJsHQ3b+gPR4259+l1AUEIeS1fdAlgCcDD3b a9ytrRZZ11XgkS75AVTn9/s= =efeV -----END PGP SIGNATURE----- --yNb1oOkm5a9FJOVX--