From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267869AbUIVEzQ (ORCPT ); Wed, 22 Sep 2004 00:55:16 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266233AbUIVEzP (ORCPT ); Wed, 22 Sep 2004 00:55:15 -0400 Received: from peabody.ximian.com ([130.57.169.10]:42884 "EHLO peabody.ximian.com") by vger.kernel.org with ESMTP id S263448AbUIVEzK (ORCPT ); Wed, 22 Sep 2004 00:55:10 -0400 Subject: Re: Does ZONE_HIGHMEM exist on machines with 1G memeory From: Robert Love To: Ronghua Zhang Cc: linux-kernel@vger.kernel.org In-Reply-To: References: <1095826387.2454.101.camel@localhost> Content-Type: text/plain Date: Wed, 22 Sep 2004 00:55:10 -0400 Message-Id: <1095828910.2454.124.camel@localhost> Mime-Version: 1.0 X-Mailer: Evolution 1.5.94.1 (1.5.94.1-1) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2004-09-22 at 00:39 -0400, Ronghua Zhang wrote: > But why cannot we simply map the whole physical memory into the 1GB > kernel virtual address space? In this case, we don't need to reserve any > address space for the permanent or temporary mapping. Am I missing > something here? Thanks It is because of how the mappings work and where PAGE_OFFSET is. With some hacking you can definitely use all 1GB without HIGHMEM, but you would not be able to use more than 1GB. And it would break ABI, at least for modules. Robert Love