From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH 3/7] mlx4_en: Incoming traffic alignment optimizations Date: Tue, 18 Oct 2011 03:53:19 +0200 Message-ID: <1318902799.2571.26.camel@edumazet-laptop> References: <4E9C8D53.5040006@mellanox.co.il> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: davem@davemloft.net, netdev@vger.kernel.org To: Yevgeny Petrilin Return-path: Received: from mail-wy0-f174.google.com ([74.125.82.174]:38388 "EHLO mail-wy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757165Ab1JRBxY (ORCPT ); Mon, 17 Oct 2011 21:53:24 -0400 Received: by wyg36 with SMTP id 36so89828wyg.19 for ; Mon, 17 Oct 2011 18:53:23 -0700 (PDT) In-Reply-To: <4E9C8D53.5040006@mellanox.co.il> Sender: netdev-owner@vger.kernel.org List-ID: Le lundi 17 octobre 2011 =C3=A0 22:17 +0200, Yevgeny Petrilin a =C3=A9c= rit : > Packet headers are copied to skb linear part (which is IP aligned), s= o there is no reason for > the scatter entry to be IP aligned. >=20 > Signed-off-by: Yevgeny Petrilin > --- > @@ -91,7 +91,7 @@ > /* Receive fragment sizes; we use at most 4 fragments (for 9600 byte= MTU > * and 4K allocations) */ > enum { > - FRAG_SZ0 =3D 512 - NET_IP_ALIGN, > + FRAG_SZ0 =3D 2048, > FRAG_SZ1 =3D 1024, > FRAG_SZ2 =3D 4096, > FRAG_SZ3 =3D MLX4_EN_ALLOC_SIZE Is the 512 -> 2048 change really wanted ? Its not mentioned in changelo= g and is confusing. This means mlx4 lost the ability to use a small frag (512 bytes) to store small frames.