From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1423374AbXDYSYq (ORCPT ); Wed, 25 Apr 2007 14:24:46 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1423372AbXDYSYq (ORCPT ); Wed, 25 Apr 2007 14:24:46 -0400 Received: from ebiederm.dsl.xmission.com ([166.70.28.69]:49430 "EHLO ebiederm.dsl.xmission.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2992883AbXDYSYp (ORCPT ); Wed, 25 Apr 2007 14:24:45 -0400 From: ebiederm@xmission.com (Eric W. Biederman) To: "H. Peter Anvin" Cc: Andrew Morton , Jeremy Fitzhardinge , Andi Kleen , Zachary Amsden , Linux Kernel Mailing List Subject: Re: [PATCH] i386: For debugging, make the initial page table setup less forgiving. References: <200704132149.l3DLnvUY012097@tazenda.hos.anvin.org> <20070425044813.f83dde4e.akpm@linux-foundation.org> <462F73A0.40003@zytor.com> <462F9665.1090602@zytor.com> Date: Wed, 25 Apr 2007 12:23:18 -0600 In-Reply-To: <462F9665.1090602@zytor.com> (H. Peter Anvin's message of "Wed, 25 Apr 2007 10:56:53 -0700") Message-ID: User-Agent: Gnus/5.110006 (No Gnus v0.6) Emacs/21.4 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org "H. Peter Anvin" writes: > Eric W. Biederman wrote: >>> This patch only affects the initial page tables, which should have been >>> thrown out *way* long ago at this point. >> >> Yes. I noticed this was happening a few days ago. >> I must not have mentioned it loudly enough. > > You mentioned the continued use of init_mm. This is *very* different. > > What we're seeing here is that ON PSE-CAPABLE HARDWARE, we continue to > not just use the init_mm page directory, but the actual page *tables*, > which should all have been replaced with PSE large pages to begin with. That is what I meant if not what I communicated. If you read the code that is exactly what it is trying to do. > Reusing the initial page tables on non-PSE-capable hardware *sort of* > makes sense, but his hardware should not fall in that category, I don't > think? (Unless it's one of these machines that fall over if you map the > bottom 4 MB with PSE pages?) I agree that I don't think the current behaviour makes sense. I think the code has accumulated so many small modifications that is very far from making sense in the corner cases. arch/i386 either needs to be frozen as a legacy only thing or it needs to be cleaned up so we can continue to enhance the code. Eric