From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH 6/7] mlx4_en: Adding rxhash support Date: Tue, 18 Oct 2011 03:48:45 +0200 Message-ID: <1318902525.2571.24.camel@edumazet-laptop> References: <4E9C8D8B.90606@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-ww0-f44.google.com ([74.125.82.44]:40903 "EHLO mail-ww0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757165Ab1JRBsu (ORCPT ); Mon, 17 Oct 2011 21:48:50 -0400 Received: by wwe6 with SMTP id 6so138711wwe.1 for ; Mon, 17 Oct 2011 18:48:49 -0700 (PDT) In-Reply-To: <4E9C8D8B.90606@mellanox.co.il> Sender: netdev-owner@vger.kernel.org List-ID: Le lundi 17 octobre 2011 =C3=A0 22:18 +0200, Yevgeny Petrilin a =C3=A9c= rit : > Moving to Toeplitz function in RSS calculation. > Reporting rxhash in skb. >=20 > Signed-off-by: Yevgeny Petrilin > rss_context->flags =3D rss_mask; > + rss_context->hash_fn =3D 1; > + for (i =3D 0; i < 10; i++) > + rss_context->rss_key[i] =3D random32(); > =20 Thats bit of a problem : Two NICS will have different seeds, and thus provide different rxhash for a given flow. A bonding of two NICS will not be able to provide a consistent rxhash. drivers/net/ethernet/intel/igb/igb_main.c uses a static table to avoid this problem.