From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eli Cohen Subject: Re: [ofa-general] [PATCH] net/inet_lro: remove setting skb->ip_summed when not LRO-able Date: Wed, 25 Jun 2008 15:30:05 +0300 Message-ID: <1214397005.23583.60.camel@mtls03> References: <4862304D.2050306@voltaire.com> Reply-To: eli@dev.mellanox.co.il Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: Jan-Bernd Themann , OpenFabrics General , netdev@vger.kernel.org, Roland Dreier , Vladimir Sokolovsky To: Or Gerlitz Return-path: Received: from nf-out-0910.google.com ([64.233.182.188]:41220 "EHLO nf-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753916AbYFYM2k (ORCPT ); Wed, 25 Jun 2008 08:28:40 -0400 Received: by nf-out-0910.google.com with SMTP id d3so129843nfc.21 for ; Wed, 25 Jun 2008 05:28:38 -0700 (PDT) In-Reply-To: <4862304D.2050306@voltaire.com> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 2008-06-25 at 14:47 +0300, Or Gerlitz wrote: > Jan-Bernd Themann wrote: > > no, what I meant is that it is only not needed at that particular > > place as the packet is not handled by LRO. Without this line the > > driver can set an individual value for each SKB that is not > > aggregated if wished. For example when the packet is not a valid IP > > packet. However, removing all ip_summed fields impacts the fragment > > lro mode. There we have to set some value for not aggregated packets. > > The SKBs are generated within the LRO engine. If desired (and if there > > is HW that wants to use that) we can pass that value for each provided > > fragment. This would add one additional paramter to the already 8 > > parameters of __lro_proc_segment. That is of course possible. > OK, understood, both points. > > Eli, lets add to this patch a comment in inet_lro.h saying that the > value of lro_mgr->ip_summed is ignored by the core lro code for drivers > that use the non fragmented mode. We already have this comment: struct net_lro_mgr { ... u32 ip_summed; /* Set in non generated SKBs in page mode */ I think we should use have it something like that: ... /* * Set for generated SKBs in that are not added to * the frag list in fragmented mode */ u32 ip_summed; What do you think? > Also for the ipoib patch, lets not set > this value. Agree.