From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: VLAN rx acceleration bypasses bridge Date: Sat, 24 Oct 2009 11:11:52 -0700 Message-ID: <20091024111152.6299a203@nehalam> References: <1255564685.21615.131.camel@sakura.staff.proxad.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: Patrick McHardy , netdev@vger.kernel.org To: mbizon@freebox.fr Return-path: Received: from smtp1.linux-foundation.org ([140.211.169.13]:49150 "EHLO smtp1.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751790AbZJXSMF (ORCPT ); Sat, 24 Oct 2009 14:12:05 -0400 In-Reply-To: <1255564685.21615.131.camel@sakura.staff.proxad.net> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, 15 Oct 2009 01:58:05 +0200 Maxime Bizon wrote: > > Hi Patrick, Stephen & all, > > > If I do: > > # vconfig add eth0 100 > # brctl addbr br0 > # brctl addif br0 eth0 > > Then eth0.100 gets no more packet and br0.100 should be used instead to > get tagged packets. > > But, if vlan rx acceleration is enabled on eth0, then vlan 100 packets > no longer go through bridge, and eth0.100 is to be used instead. > > This is fixable, but I'm not sure which behavior is preferred ? > > > In case we prefer the first one, how could someone bridge untagged > packets from eth0 with some tagged packets from eth1, and some tagged > packets from eth0 with untagged packets from eth1 ? > > Something like this, which currently works when vlan rx accel is on: > > # vconfig add eth0 100 > # vconfig add eth1 100 > # brctl addbr br0 > # brctl addif br0 eth0 > # brctl addif br0 eth1.100 > # brctl addbr br1 > # brctl addif br1 eth0.100 > # brctl addif br1 eth1 > > > Regards, > The way I see it, if you define a vlan then you need to bridge that vlan. Mixing plain and vlanned interfaces is a mess that is hard to untangle and probably not worth fixing.