From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH 0/1] RFC: poll/select performance on datagram sockets Date: Sat, 30 Oct 2010 00:08:12 +0200 Message-ID: <1288390092.2680.80.camel@edumazet-laptop> References: <20101029191857.5f789d56@chocolatine.cbg.collabora.co.uk> <1288380431.2680.3.camel@edumazet-laptop> <1288386348.2680.25.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Alban Crequy , "David S. Miller" , Stephen Hemminger , Cyrill Gorcunov , Alexey Dobriyan , netdev@vger.kernel.org, Linux Kernel Mailing List , Pauli Nieminen , Rainer Weikusat To: Davide Libenzi Return-path: Received: from mail-ww0-f44.google.com ([74.125.82.44]:49525 "EHLO mail-ww0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755031Ab0J2WIS (ORCPT ); Fri, 29 Oct 2010 18:08:18 -0400 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: Le vendredi 29 octobre 2010 =C3=A0 14:57 -0700, Davide Libenzi a =C3=A9= crit : > It is, though I am not sure that shuffling code around to chase false= =20 > sharing on very specific cases, is an approach worth following. > The new position of the table on the stack might share with another h= ot=20 > line, under a different load. >=20 Sure. Its only a start as I said. We should identify the shared portion and make sure it sits in a dedicated cache line.