From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 2 Oct 2001 09:22:57 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 2 Oct 2001 09:22:37 -0400 Received: from e23.nc.us.ibm.com ([32.97.136.229]:11506 "EHLO e23.nc.us.ibm.com") by vger.kernel.org with ESMTP id ; Tue, 2 Oct 2001 09:22:17 -0400 Date: Tue, 2 Oct 2001 18:58:16 +0530 From: Dipankar Sarma To: linux-kernel@vger.kernel.org Subject: [PATCH] cacheline align rt_cache_stat struct Message-ID: <20011002185816.A8643@in.ibm.com> Reply-To: dipankar@in.ibm.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org The per-cpu rt_cache_stat structure should be padded to cacheline to avoid sharing of cachelines between different CPUs. Here is a patch to do that for SMP kernels. Thanks Dipankar -- Dipankar Sarma Project: http://lse.sourceforge.net Linux Technology Center, IBM Software Lab, Bangalore, India. diff -urN linux-2.4.10/include/net/route.h linux-2.4.10+rt/include/net/route.h --- linux-2.4.10/include/net/route.h Thu Sep 27 11:59:45 2001 +++ linux-2.4.10+rt/include/net/route.h Tue Oct 2 18:34:45 2001 @@ -105,7 +105,7 @@ unsigned int out_hit; unsigned int out_slow_tot; unsigned int out_slow_mc; -}; +} ____cacheline_aligned_in_smp; extern struct ip_rt_acct *ip_rt_acct;