From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx173.postini.com [74.125.245.173]) by kanga.kvack.org (Postfix) with SMTP id CF58C6B13F0 for ; Sat, 11 Feb 2012 09:55:50 -0500 (EST) Message-ID: <4F36816A.6030609@redhat.com> Date: Sat, 11 Feb 2012 09:55:38 -0500 From: Rik van Riel MIME-Version: 1.0 Subject: Re: reclaim the LRU lists full of dirty/writeback pages References: <20120208093120.GA18993@localhost> <20120210114706.GA4704@localhost> <20120211124445.GA10826@localhost> In-Reply-To: <20120211124445.GA10826@localhost> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Wu Fengguang Cc: Greg Thelen , Jan Kara , "bsingharora@gmail.com" , Hugh Dickins , Michal Hocko , linux-mm@kvack.org, Mel Gorman , Ying Han , "hannes@cmpxchg.org" , KAMEZAWA Hiroyuki , Minchan Kim On 02/11/2012 07:44 AM, Wu Fengguang wrote: > Note that it's data for XFS. ext4 seems to have some problem with the > workload: the majority pages are found to be writeback pages, and the > flusher ends up blocking on the unconditional wait_on_page_writeback() > in write_cache_pages_da() from time to time... > > XXX: commit NFS unstable pages via write_inode() > XXX: the added congestion_wait() may be undesirable in some situations Even with these caveats, this seems to be the right way forward. > CC: Jan Kara > CC: Mel Gorman > CC: Rik van Riel > CC: Greg Thelen > CC: Minchan Kim > Signed-off-by: Wu Fengguang Acked-by: Rik van Riel -- All rights reversed -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: email@kvack.org