From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752191Ab1AQCwH (ORCPT ); Sun, 16 Jan 2011 21:52:07 -0500 Received: from mga01.intel.com ([192.55.52.88]:55612 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751515Ab1AQCwC (ORCPT ); Sun, 16 Jan 2011 21:52:02 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.60,332,1291622400"; d="scan'208";a="647584092" Subject: [PATCH 0/4]x86: allocate up to 32 tlb invalidate vectors -resend From: Shaohua Li To: lkml Cc: Ingo Molnar , Andi Kleen , "hpa@zytor.com" , Andrew Morton , Eric Dumazet Content-Type: text/plain; charset="UTF-8" Date: Mon, 17 Jan 2011 10:51:59 +0800 Message-ID: <1295232719.1949.706.camel@sli10-conroe> Mime-Version: 1.0 X-Mailer: Evolution 2.30.3 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org last post is lost, resent. Hi, In workload with heavy page reclaim, flush_tlb_page() is frequently used. We currently have 8 vectors for tlb flush, which is fine for small machines. But for big machines with a lot of CPUs, the 8 vectors are shared by all CPUs and we need lock to protect them. This will cause a lot of lock contentions. please see the patch 3 for detailed number of the lock contention. Andi Kleen suggests we can use 32 vectors for tlb flush, which should be fine for even 8 socket machines. Test shows this reduces lock contention dramatically (see patch 3 for number). One might argue if this will waste too many vectors and leave less vectors for devices. This could be a problem. But even we use 32 vectors, we still leave 78 vectors for devices. And we now have per-cpu vector, vector isn't scarce any more, but I'm open if anybody has objections. Thanks, Shaohua