From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+willy=40w.ods.org-S268977AbUJEKP1@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S268977AbUJEKP1 (ORCPT <rfc822;willy@w.ods.org>);
	Tue, 5 Oct 2004 06:15:27 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268980AbUJEKP0
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Tue, 5 Oct 2004 06:15:26 -0400
Received: from mx1.redhat.com ([66.187.233.31]:63167 "EHLO mx1.redhat.com")
	by vger.kernel.org with ESMTP id S268977AbUJEKOv (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Tue, 5 Oct 2004 06:14:51 -0400
From: David Howells <dhowells@redhat.com>
In-Reply-To: <24616.1096914276@redhat.com> 
References: <24616.1096914276@redhat.com> 
To: Linux filesystem caching discussion list 
	<linux-cachefs@redhat.com>
Cc: linux-kernel@vger.kernel.org
Subject: [RFC] Splitting CacheFS into general manager and cache backend bits 
User-Agent: EMH/1.14.1 SEMI/1.14.5 (Awara-Onsen) FLIM/1.14.5 (Demachiyanagi) APEL/10.6 Emacs/21.3 (i386-redhat-linux-gnu) MULE/5.0 (SAKAKI)
MIME-Version: 1.0 (generated by SEMI 1.14.5 - "Awara-Onsen")
Content-Type: text/plain; charset=US-ASCII
Date: Tue, 05 Oct 2004 11:14:45 +0100
Message-ID: <6260.1096971285@redhat.com>
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org


I've whipped up a patch to split CacheFS into two:

 (1) FS-Cache

     This is a general cache manager that the interested netfs's talk to. It
     is supplied with a number of cache backends over which it distributes the
     operations from the netfs's.

 (2) CacheFS

     This is then made into a cache backend. The act of mounting a block
     device as cachefs causes the space contained therein to be made available
     through FS-Cache.

To give you a clearer idea of what I'm thinking, this diagram might help:

	+---------+
	|         |                        +-----------+
	|   NFS   |--+                     |           |
	|         |  |                 +-->|  CacheFS  |
	+---------+  |   +----------+  |   | /dev/hda5 |
	             |   |          |  |   +-----------+
	+---------+  +-->|          |  |
	|         |      |          |--+   +-------------+
	|   AFS   |----->| FS-Cache |      |             |
	|         |      |          |----->| Cache Files |
	+---------+  +-->|          |      | /var/cache  |
	             |   |          |--+   +-------------+
	+---------+  |   +----------+  |
	|         |  |                 |   +-------------+
	|  ISOFS  |--+                 |   |             |
	|         |                    +-->| ReiserCache |
	+---------+                        | /           |
	                                   +-------------+

So, for example, you might have a block device mounted with CacheFS as one of
your backends, the traditional bunch of cache files as another backend and
some sort of Reiser4 plug-in as a third backend.

A netfs only has to be modified to take account of FS-Cache. It then shouldn't
need to be modified further to support each different type of cache.

David