From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267320AbUHPB6x (ORCPT ); Sun, 15 Aug 2004 21:58:53 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267323AbUHPB6x (ORCPT ); Sun, 15 Aug 2004 21:58:53 -0400 Received: from mx1.redhat.com ([66.187.233.31]:38604 "EHLO mx1.redhat.com") by vger.kernel.org with ESMTP id S267320AbUHPB6v (ORCPT ); Sun, 15 Aug 2004 21:58:51 -0400 Date: Sun, 15 Aug 2004 18:56:44 -0700 From: "David S. Miller" To: Christoph Lameter Cc: linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: page fault fastpath: Increasing SMP scalability by introducing pte locks? Message-Id: <20040815185644.24ecb247.davem@redhat.com> In-Reply-To: References: <20040815130919.44769735.davem@redhat.com> <20040815165827.0c0c8844.davem@redhat.com> X-Mailer: Sylpheed version 0.9.12 (GTK+ 1.2.10; sparc-unknown-linux-gnu) X-Face: "_;p5u5aPsO,_Vsx"^v-pEq09'CU4&Dc1$fQExov$62l60cgCc%FnIwD=.UF^a>?5'9Kn[;433QFVV9M..2eN.@4ZWPGbdi<=?[:T>y?SD(R*-3It"Vj:)"dP Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sun, 15 Aug 2004 17:11:53 -0700 (PDT) Christoph Lameter wrote: > pgd/pmd walking should be possible always even without the vma semaphore > since the CPU can potentially walk the chain at anytime. munmap() can destroy pmd and pte tables. somehow we have to protect against that, and currently that is having the VMA semaphore held for reading, see free_pgtables().