From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753968AbXLTSes (ORCPT ); Thu, 20 Dec 2007 13:34:48 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751668AbXLTSej (ORCPT ); Thu, 20 Dec 2007 13:34:39 -0500 Received: from mx1.redhat.com ([66.187.233.31]:45981 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751184AbXLTSei (ORCPT ); Thu, 20 Dec 2007 13:34:38 -0500 Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 From: David Howells In-Reply-To: <200712181800.23578.nickpiggin@yahoo.com.au> References: <200712181800.23578.nickpiggin@yahoo.com.au> <200712141508.23756.nickpiggin@yahoo.com.au> <20071205193909.24617.26538.stgit@warthog.procyon.org.uk> <643.1197930966@redhat.com> To: Nick Piggin Cc: dhowells@redhat.com, viro@ftp.linux.org.uk, hch@infradead.org, Trond.Myklebust@netapp.com, sds@tycho.nsa.gov, casey@schaufler-ca.com, linux-kernel@vger.kernel.org, selinux@tycho.nsa.gov, linux-security-module@vger.kernel.org Subject: Re: [PATCH 10/28] FS-Cache: Recruit a couple of page flags for cache management [try #2] X-Mailer: MH-E 8.0.3+cvs; nmh 1.2-20070115cvs; GNU Emacs 23.0.50 Date: Thu, 20 Dec 2007 18:33:39 +0000 Message-ID: <6481.1198175619@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Nick Piggin wrote: > > > I'd much prefer if you would handle this in the filesystem, and have it > > > set PG_private whenever fscache needs to receive a callback, and DTRT > > > depending on whether PG_fscache etc. is set or not. > > > > That's tricky and slower[*]. One of the things I want to do is to modify > > iso9660 to do be able to do caching, but PG_private is 'owned' by the > > generic buffer cache code. > > Maybe it is harder, but it is the right way to do it. You're wrong. It would mean that PG_private is the logical disjunction of PG_fscache and some condition not otherwise explicitly stored. I tried that with NFS and it was nasty. As you can no doubt see, it means that you can't distinguish all the states you used to be able to. > So you should modify the filesystems rather than core code. I think you missed what I said: but PG_private is 'owned' by the generic buffer cache code. That means more of the core code would have to change - or, at least, change more. David