From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755148Ab0BOQhT (ORCPT ); Mon, 15 Feb 2010 11:37:19 -0500 Received: from cantor.suse.de ([195.135.220.2]:38047 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754258Ab0BOQhR (ORCPT ); Mon, 15 Feb 2010 11:37:17 -0500 Date: Tue, 16 Feb 2010 03:37:12 +1100 From: Nick Piggin To: Christoph Hellwig Cc: Jan Kara , Andrew Morton , LKML , fengguang.wu@intel.com Subject: Re: [PATCH 2/3] mm: Implement writeback livelock avoidance using page tagging Message-ID: <20100215163712.GV5723@laptop> References: <1265929584-5080-1-git-send-email-jack@suse.cz> <1265929584-5080-3-git-send-email-jack@suse.cz> <20100212113955.4c023130.akpm@linux-foundation.org> <20100215154751.GG3434@quack.suse.cz> <20100215162127.GU5723@laptop> <20100215162453.GA6477@infradead.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20100215162453.GA6477@infradead.org> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Feb 15, 2010 at 11:24:53AM -0500, Christoph Hellwig wrote: > On Tue, Feb 16, 2010 at 03:21:27AM +1100, Nick Piggin wrote: > > Also, I wonder what you think about leaving the tags dangling when > > the loop bails out early? I have a *slight* concern about this > > because previously we never have a tag set when radix_tree_delete > > is called. I actually had a bug in that code in earlier versions > > of rcu radix tree that only got found by the user test harness. > > And another slight concern that it is just a bit ugly to leave the > > tag. But I can accept that lower CPU overhead trumps ugliness :) > > The XFS inode cache calls radix_tree_delete with a tag set, and > interestingly enough we're trying to catch a very weird bug in that > area currently, which seems more or less directly related to the use > of tags. But we also had a real locking bug related to tags, so I'm > not yet sure what the issue is back until I hear back from the bug > reported and reproducer. Upon looking at the radix_tree_delete code again, the tag clearing there seems rather simple so perhaps I misremembered some detail of that bug I had.