From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH 3/4 v2 net-next] net: make GRO aware of skb->head_frag Date: Wed, 02 May 2012 18:27:51 +0200 Message-ID: <1335976071.22133.581.camel@edumazet-glaptop> References: <1335523026.2775.236.camel@edumazet-glaptop> <1335809434.2296.9.camel@edumazet-glaptop> <4F9F21E2.3080407@intel.com> <1335835677.11396.5.camel@edumazet-glaptop> <1335854378.11396.26.camel@edumazet-glaptop> <4FA00C9F.8080409@intel.com> <1335891892.22133.23.camel@edumazet-glaptop> <4FA03D69.6060907@intel.com> <1335947084.22133.134.camel@edumazet-glaptop> <4FA15DDE.5090904@intel.com> <1335975578.22133.580.camel@edumazet-glaptop> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Alexander Duyck , David Miller , netdev , Neal Cardwell , Tom Herbert , Jeff Kirsher , Michael Chan , Matt Carlson , Herbert Xu , Ben Hutchings , Ilpo =?ISO-8859-1?Q?J=E4rvinen?= , Maciej =?UTF-8?Q?=C5=BBenczykowski?= To: Alexander Duyck Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:63552 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754535Ab2EBQ16 (ORCPT ); Wed, 2 May 2012 12:27:58 -0400 Received: by bkcji2 with SMTP id ji2so624136bkc.19 for ; Wed, 02 May 2012 09:27:57 -0700 (PDT) In-Reply-To: <1335975578.22133.580.camel@edumazet-glaptop> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 2012-05-02 at 18:19 +0200, Eric Dumazet wrote: > On Wed, 2012-05-02 at 09:16 -0700, Alexander Duyck wrote: > > > I was working with the out-of-tree ixgbe because I have the option there > > of stripping out FCoE and RSC via a couple of build flags. The problem > > is I don't know if the head frag stuff will work out very well with > > ixgbe because RSC and FCoE require that we have to use 1K aligned > > receive buffers. It would require us to make us have to bump up our > > allocation size by NET_SKB_PAD plus skb_shared_info which would likely > > force us up to order 1 pages on most platforms. > > What is RSC exactly, and why RSC is used in the build_skb() context ? > > It looks like e1000e would be a good candidate for build_skb() (without packet split)