From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [PATCH 1/2][RFC] vlan: use pskb_copy() when inserting a vlan tag by hand Date: Mon, 07 Jul 2008 23:15:48 +0200 Message-ID: <48728784.2030202@trash.net> References: <20080707205646.GB12997@xi.wantstofly.org> <4872841E.1030609@trash.net> <20080707210701.GD12997@xi.wantstofly.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Cc: Nicolas Pitre , Dale Farnsworth , Ashish Karkare , Jesper Dangaard Brouer , netdev@vger.kernel.org, "David S. Miller" To: Lennert Buytenhek Return-path: Received: from stinky.trash.net ([213.144.137.162]:38822 "EHLO stinky.trash.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756745AbYGGVPv (ORCPT ); Mon, 7 Jul 2008 17:15:51 -0400 In-Reply-To: <20080707210701.GD12997@xi.wantstofly.org> Sender: netdev-owner@vger.kernel.org List-ID: Lennert Buytenhek wrote: > On Mon, Jul 07, 2008 at 11:01:18PM +0200, Patrick McHardy wrote: > >>> When inserting a vlan tag into an skb by hand (in the case the >>> target network device hardware does not support inserting a vlan >>> tag by passing it in via the TX descriptor), making a copy of the >>> skb to be modified by calling skb_copy() is overkill, since >>> skb_copy() will needlessly linearize the skb, copying of all of >>> the fragmented data around, and checksumming the paket in software >>> even if the hardware is perfectly capable of doing that by itself. >>> >>> For this case, pskb_copy() does exactly what is needed and no more, >>> so use that instead. >>> >>> Signed-off-by: Lennert Buytenhek >> This looks fine to me. > > OK. Since the mv643xx_eth patch (2/2) depends on both this patch > and the one you/I sent previously, could you merge the mv643xx_eth > patch (2/2) via your tree as well? I already sent the network_header fix to Dave, so its probably easiest to have it go through him directly. Your patch applies cleanly on top of my VLAN update from this morning.