From mboxrd@z Thu Jan  1 00:00:00 1970
From: Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH 3/3 v2] mm: Avoid resetting wb_start after each
 writeback round
Date: Thu, 5 Aug 2010 17:21:07 -0700
Message-ID: <20100805172107.703ccf77.akpm@linux-foundation.org>
References: <1281034399-13055-1-git-send-email-jack@suse.cz>
	<1281034399-13055-4-git-send-email-jack@suse.cz>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Cc: linux-fsdevel@vger.kernel.org, hch@infradead.org,
	jaxboe@fusionio.com, Wu Fengguang <fengguang.wu@intel.com>
To: Jan Kara <jack@suse.cz>
Return-path: <linux-fsdevel-owner@vger.kernel.org>
Received: from smtp1.linux-foundation.org ([140.211.169.13]:46240 "EHLO
	smtp1.linux-foundation.org" rhost-flags-OK-OK-OK-OK)
	by vger.kernel.org with ESMTP id S1751662Ab0HFAWB (ORCPT
	<rfc822;linux-fsdevel@vger.kernel.org>);
	Thu, 5 Aug 2010 20:22:01 -0400
In-Reply-To: <1281034399-13055-4-git-send-email-jack@suse.cz>
Sender: linux-fsdevel-owner@vger.kernel.org
List-ID: <linux-fsdevel.vger.kernel.org>

On Thu,  5 Aug 2010 20:53:19 +0200
Jan Kara <jack@suse.cz> wrote:

> WB_SYNC_NONE writeback is done in rounds of 1024 pages so that we don't write
> out some huge inode for too long while starving writeout of other inodes. To
> avoid livelocks, we record time we started writeback in wbc->wb_start and do
> not write out inodes which were dirtied after this time. But currently,
> writeback_inodes_wb() resets wb_start each time it is called thus effectively
> invalidating this logic and making any WB_SYNC_NONE writeback prone to
> livelocks.

data point: that was eight-year-old code that we broke.  Five months ago.

How did you discover this?  You have test cases which cover this
scenario and things locked up?