From mboxrd@z Thu Jan 1 00:00:00 1970 From: Simon Horman Subject: Re: [PATCH net-next v3] MPLS: Add limited GSO support Date: Mon, 20 May 2013 12:20:49 +0900 Message-ID: <20130520032049.GG16811@verge.net.au> References: <1368773452-32310-1-git-send-email-horms@verge.net.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "dev@openvswitch.org" , netdev , Pravin B Shelar , Jarno Rajahalme , Eric Dumazet , Maciej =?utf-8?Q?=C5=BBenczykowski?= , Ben Hutchings To: Jesse Gross Return-path: Received: from kirsty.vergenet.net ([202.4.237.240]:43953 "EHLO kirsty.vergenet.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755255Ab3ETDUx (ORCPT ); Sun, 19 May 2013 23:20:53 -0400 Content-Disposition: inline In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Fri, May 17, 2013 at 02:42:01PM -0700, Jesse Gross wrote: > On Thu, May 16, 2013 at 11:50 PM, Simon Horman wrote: > > * Set skb->mac_len and skb->network_header to correspond to the > > end of the L3 header, including the MPLS label stack. > > I believe that the treatment of skb->mac_len is different from what > you are proposing in the OVS patch. I think this is because there is a > call to skb_reset_mac_len() in __skb_gso_segment() but it is somewhat > confusing. The GSO code is moving away from being able to reset all > header pointers and reparse the packet (obviously that's not possible > in the case of MPLS) so maybe we can just remove that call. I agree that is likely. I will see about making it so. > > > A new NETIF_F_GRE_GSO feature is added for devices which support > > hardware MPLS GSO offload. Currently no devices support this > > and MPLS GSO always falls back to software. > > I assume that this is supposed to refer to MPLS and not GRE. Sorry about that, yes, I meant to type MPLS.