From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [NET][TG3] Fwd: page allocation failure with linux 3.1.1 Date: Mon, 28 Nov 2011 20:28:14 +0100 Message-ID: <1322508494.2970.14.camel@edumazet-laptop> References: <1322462145.2826.6.camel@edumazet-laptop> <1322468762.2826.17.camel@edumazet-laptop> <4ED3DB91.9090409@hp.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: =?UTF-8?Q?=D0=9F=D0=BB=D0=B0=D0=BC=D0=B5=D0=BD_?= =?UTF-8?Q?=D0=9F=D0=B5=D1=82=D1=80=D0=BE=D0=B2?= , David Miller , netdev@vger.kernel.org To: Rick Jones Return-path: Received: from mail-lpp01m010-f46.google.com ([209.85.215.46]:42078 "EHLO mail-lpp01m010-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753436Ab1K1T2V (ORCPT ); Mon, 28 Nov 2011 14:28:21 -0500 Received: by lahl5 with SMTP id l5so661853lah.19 for ; Mon, 28 Nov 2011 11:28:19 -0800 (PST) In-Reply-To: <4ED3DB91.9090409@hp.com> Sender: netdev-owner@vger.kernel.org List-ID: Le lundi 28 novembre 2011 =C3=A0 11:05 -0800, Rick Jones a =C3=A9crit : > On 11/28/2011 12:26 AM, Eric Dumazet wrote: > > Since some devices might copy skb to a linear one in their > > ndo_start_xmit(), we could set a netdev limit so that TSO can still= be > > used on these devices, but limiting number of frags to > > 2^PAGE_ALLOC_COSTLY_ORDER >=20 > How much "TSO value" remains after such a copy? Instead of filling up to 64Kbytes + one head of data per skb, we'll fil= l 32kbytes (8*4Kbytes) + one head. Thats still better than 1460 bytes per skb :)