From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261467AbVFCSLI (ORCPT ); Fri, 3 Jun 2005 14:11:08 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261478AbVFCSLI (ORCPT ); Fri, 3 Jun 2005 14:11:08 -0400 Received: from mail25.sea5.speakeasy.net ([69.17.117.27]:59020 "EHLO mail25.sea5.speakeasy.net") by vger.kernel.org with ESMTP id S261467AbVFCSKy (ORCPT ); Fri, 3 Jun 2005 14:10:54 -0400 Date: Fri, 3 Jun 2005 11:10:49 -0700 (PDT) From: dwhite X-X-Sender: dave@willie To: linux-kernel@vger.kernel.org Subject: icmp redirects on one-armed linux routers Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Just wondering, is there any reason why the patch mentioned in the following post has not been implemented (as of kernel 2.6.11.11 it is not)? Is it a matter of testing? Is there a better alternative to solve the problem described in the thread? http://oss.sgi.com/projects/netdev/archive/2004-07/msg00512.html The current behavior (unchanged from that described in the post above) can be problematic, for example when I want to do stateful firewalling on my linux router. I've been applying the patch therein to recent kernels and it seems to be doing the right thing (sends redirects only when the redirect target is on the same subnet as the sending host). Any thoughts? TIA -dave (IANAP) ps. please cc me directly in your response :)