From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH net-next 1/8] flow_keys: Record IP layer protocol in skb_flow_dissect() Date: Wed, 11 Jun 2014 15:06:36 -0700 (PDT) Message-ID: <20140611.150636.857251759787443666.davem@davemloft.net> References: <539703F5.8030707@redhat.com> <1402410390.3645.429.camel@edumazet-glaptop2.roam.corp.google.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: _govind@gmx.com, dborkman@redhat.com, netdev@vger.kernel.org, ssujith@cisco.com, gvaradar@cisco.com, benve@cisco.com To: eric.dumazet@gmail.com Return-path: Received: from shards.monkeyblade.net ([149.20.54.216]:47477 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752657AbaFKWGh (ORCPT ); Wed, 11 Jun 2014 18:06:37 -0400 In-Reply-To: <1402410390.3645.429.camel@edumazet-glaptop2.roam.corp.google.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Eric Dumazet Date: Tue, 10 Jun 2014 07:26:30 -0700 > On Tue, 2014-06-10 at 19:38 +0530, Govindarajulu Varadarajan wrote: >> > >> determining IPv4/IPv6 is important because this can be used in dissecting flow >> in Accelerated RFS. Adaptor does not support IPv6 filters. Since Accelerated >> RFS is supported for IPV6, using skb_flow_dissect will return true with non-zero >> values in src/dst for IPv6. >> >> If I am going to write separate function for getting IP address and port >> numbers, its definition is going to be somewhat same as skb_flow_dissect. >> Why not improve whats already written and reuse it? >> >> Is there any significant downside of adding u16 n_proto and increasing >> size of qdisc_skb_cb by 4 bytes? > > You can avoid this increase (might be bad for IB, hard to tell), by > changing sch_choke.c to only store a part of the struct flow_keys. I think this is fine, IPOIB's control block will need still just 44 bytes after these changes, so there will still be 4 bytes to spare. I'm going to apply this series.