From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+willy=40w.ods.org@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id <S268568AbTCCQq5>; Mon, 3 Mar 2003 11:46:57 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org
	id <S268577AbTCCQq5>; Mon, 3 Mar 2003 11:46:57 -0500
Received: from mx1.elte.hu ([157.181.1.137]:4009 "HELO mx1.elte.hu")
	by vger.kernel.org with SMTP id <S268568AbTCCQqz>;
	Mon, 3 Mar 2003 11:46:55 -0500
Date: Mon, 3 Mar 2003 17:56:57 +0100 (CET)
From: Ingo Molnar <mingo@elte.hu>
Reply-To: Ingo Molnar <mingo@elte.hu>
To: Linus Torvalds <torvalds@transmeta.com>
Cc: Andrew Morton <akpm@zip.com.au>, <linux-mm@kvack.org>,
       <linux-kernel@vger.kernel.org>
Subject: Re: [patch] remap-file-pages-2.5.63-A0
In-Reply-To: <Pine.LNX.4.44.0303030849050.11244-100000@home.transmeta.com>
Message-ID: <Pine.LNX.4.44.0303031755160.13116-100000@localhost.localdomain>
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


On Mon, 3 Mar 2003, Linus Torvalds wrote:

> > the attached patch, against BK-curr, is a preparation to make
> > remap_file_pages() usable on swappable vmas as well. When 'swapping out'
> > shared-named mappings the page offset is written into the pte.
> > 
> > it takes one bit from the swap-type bits, otherwise it does not change the
> > pte layout - so it should be easy to adapt any other architecture to this
> > change as well. (this patch does not introduce the protection-bits-in-pte
> > approach used in my previous patch.)
> 
> One question: Why?
> 
> What's wrong with just using the value we use now (0), and just
> calculating the page from the vma/offset information? Why hide the
> offset in the page tables, when there is no need for it?

you mean why in the normal, linear mapping case? No reason, just for
testing. At zap-time we could test whether that pte is linear or not, and
only store it in the swap entry as a file-pte if it's nonlinear.

for swappable sys_file_remap_pages() support it's necessary though, as
they break up the linear mapping without splitting up the vma into
zillions of pieces.

	Ingo