Return-path: Received: from nbd.name ([46.4.11.11]:54991 "EHLO nbd.name" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751943Ab1ARPC4 (ORCPT ); Tue, 18 Jan 2011 10:02:56 -0500 Message-ID: <4D35AB9D.6040004@openwrt.org> Date: Tue, 18 Jan 2011 16:02:53 +0100 From: Felix Fietkau MIME-Version: 1.0 To: Johannes Berg CC: linux-wireless@vger.kernel.org, linville@tuxdriver.com Subject: Re: [PATCH v2] mac80211: drop non-auth 3-addr data frames when running as a 4-addr station References: <1295362128-2294-1-git-send-email-nbd@openwrt.org> <1295362547.3563.22.camel@jlt3.sipsolutions.net> In-Reply-To: <1295362547.3563.22.camel@jlt3.sipsolutions.net> Content-Type: text/plain; charset=UTF-8 Sender: linux-wireless-owner@vger.kernel.org List-ID: On 2011-01-18 3:55 PM, Johannes Berg wrote: > On Tue, 2011-01-18 at 15:48 +0100, Felix Fietkau wrote: >> When running as a 4-addr station against an AP that has the 4-addr VLAN >> interface and the main 3-addr AP interface bridged together, sometimes >> frames originating from the station were looping back from the 3-addr AP >> interface, causing the bridge code to emit warnings about receiving frames >> with its own source address. >> I'm not sure why this is happening yet, but I think it's a good idea to >> drop all frames (except 802.1x/EAP frames) that do not match the configured >> addressing mode, including 4-address frames sent to a 3-address station. >> User test reports indicate that the problem goes away with this patch. > > This looks better, thanks. But did they really test _this_ patch? :) A few users tested this and told me that it didn't make things worse compared to my first patch - which did fix some, but not all issues related to this. - Felix