From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH -next 0/3] tc state machinery cleanups Date: Sun, 17 May 2015 23:34:05 -0400 (EDT) Message-ID: <20150517.233405.601002797695700706.davem@davemloft.net> References: <1431679850-31896-1-git-send-email-fw@strlen.de> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, jhs@mojatatu.com, alexei.starovoitov@gmail.com, daniel@iogearbox.net To: fw@strlen.de Return-path: Received: from shards.monkeyblade.net ([149.20.54.216]:51308 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750780AbbERDeJ (ORCPT ); Sun, 17 May 2015 23:34:09 -0400 In-Reply-To: <1431679850-31896-1-git-send-email-fw@strlen.de> Sender: netdev-owner@vger.kernel.org List-ID: From: Florian Westphal Date: Fri, 15 May 2015 10:50:47 +0200 > This series prepares removal of tc_verd member from sk_buff. > > It simplifies tc state machinery to what is required to keep current > mirred/ifb combinations working. > > I tested a few scenarios, namely: > > 1 - htb based shaping on egress > 2 - netem attached to ifb with mirred redirect from ingress qdisc > 3 - mirred to different egress device > 4 - mirred to ifb egress device with qdiscs set up on ifb > to provide illusion of 'single' transmit interface for traffic shaping > > After this series tc_verd is only used by ifb to skip actions on egress. > > Part #2 of this series will remove tc_verd completely. > > motivation is two-fold: > 1) make states and state transitions more obvious > 2) provide a way to later reduce skb size by 8 bytes > (s/u16 mac_len/u8 mac_len/ would result in > two 2 byte and one 4 byte hole, i.e. 8 byte reduction with > minor reshuffling). It looks like there will be changes to this series. My only comment from my perspective is that, in patch #1, you should probably add a "TC_FROM_UNKNOWN = 0" to the enumeration and test/set using that instead of a magic constant. Thanks.