From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+w=401wt.eu-S1755482AbXL1QZX@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1755482AbXL1QZX (ORCPT <rfc822;w@1wt.eu>);
	Fri, 28 Dec 2007 11:25:23 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754352AbXL1QZJ
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Fri, 28 Dec 2007 11:25:09 -0500
Received: from e3.ny.us.ibm.com ([32.97.182.143]:43848 "EHLO e3.ny.us.ibm.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1754075AbXL1QZI (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Fri, 28 Dec 2007 11:25:08 -0500
Subject: Re: [PATCH] jfs: clear PAGECACHE_TAG_DIRTY for no-write pages
From: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Fengguang Wu <wfg@mail.ustc.edu.cn>,
       Andrew Morton <akpm@linux-foundation.org>,
       Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
       jfs-discussion@lists.sourceforge.net,
       Nick Piggin <nickpiggin@yahoo.com.au>
In-Reply-To: <1198841626.6821.92.camel@twins>
References: <E1J8AC5-0002BE-RP@localhost>  <1198841626.6821.92.camel@twins>
Content-Type: text/plain
Date: Fri, 28 Dec 2007 10:25:01 -0600
Message-Id: <1198859101.13075.10.camel@norville.austin.ibm.com>
Mime-Version: 1.0
X-Mailer: Evolution 2.12.1 
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org


On Fri, 2007-12-28 at 12:33 +0100, Peter Zijlstra wrote:
> On Fri, 2007-12-28 at 16:03 +0800, Fengguang Wu wrote:
> > Andrew,
> > 
> > This patch fixed the 'pdflush stuck in D state' bug
> >         http://bugzilla.kernel.org/show_bug.cgi?id=9291
> > and should be pushed to mainline ASAP.
> > ---
> > 
> > When JFS decides to drop a dirty metapage, it simply clears the META_dirty bit
> > and leave alone the PG_dirty and PAGECACHE_TAG_DIRTY bits.
> > 
> > When such no-write page goes to metapage_writepage(), the `relic'
> > PAGECACHE_TAG_DIRTY tag should be cleared, to prevent pdflush from
> > repeatedly trying to sync them.
> > 
> > Also, avoid the redirty when a bio submission is planned.
> > 
> > Tested-by: Markus Rehbach <Markus.Rehbach@gmx.de>
> > Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn> 
> > ---
> 
> > +++ b/fs/jfs/jfs_metapage.c
> 
> > @@ -449,9 +450,15 @@ static int metapage_writepage(struct page *page, struct writeback_control *wbc)
> >  			goto dump_bio;
> >  
> >  		submit_bio(WRITE, bio);
> > -	}
> > -	if (redirty)
> > +	} else if (redirty) {
> >  		redirty_page_for_writepage(wbc, page);
> > +	} else {
> > +		write_lock_irq(&mapping->tree_lock);
> > +		radix_tree_tag_clear(&mapping->page_tree,
> > +				page_index(page),
> > +				PAGECACHE_TAG_DIRTY);
> > +		write_unlock_irq(&mapping->tree_lock);
> > +	}
> 
> I'm not liking this open-coded tag_clear, although I currently fail to
> come up with a nice solution.

I'm looking at __block_write_full_page() for guidance.  The situation
here is similar to writing a page where all the bufferheads are clean.
__block_write_full_page() always calls set_page_writeback(page), and
then, if no I/O is initiated, immediately calls
end_page_writeback(page).

I'll put together a patch for testing.

Thanks,
Shaggy
-- 
David Kleikamp
IBM Linux Technology Center