From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:36311) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TElw8-00029D-VT for qemu-devel@nongnu.org; Thu, 20 Sep 2012 14:57:21 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1TElw4-000358-P7 for qemu-devel@nongnu.org; Thu, 20 Sep 2012 14:57:20 -0400 Received: from mx1.redhat.com ([209.132.183.28]:38593) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TElw4-00034x-GC for qemu-devel@nongnu.org; Thu, 20 Sep 2012 14:57:16 -0400 Message-ID: <505B6706.1000104@redhat.com> Date: Thu, 20 Sep 2012 14:57:10 -0400 From: Rik van Riel MIME-Version: 1.0 References: <1348149875-29678-1-git-send-email-mgorman@suse.de> <1348149875-29678-7-git-send-email-mgorman@suse.de> In-Reply-To: <1348149875-29678-7-git-send-email-mgorman@suse.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 6/6] mm: compaction: Restart compaction from near where it left off List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Mel Gorman Cc: Richard Davies , KVM , QEMU-devel , LKML , Linux-MM , Avi Kivity , Shaohua Li On 09/20/2012 10:04 AM, Mel Gorman wrote: > This is almost entirely based on Rik's previous patches and discussions > with him about how this might be implemented. > > Order > 0 compaction stops when enough free pages of the correct page > order have been coalesced. When doing subsequent higher order allocations, > it is possible for compaction to be invoked many times. > > However, the compaction code always starts out looking for things to compact > at the start of the zone, and for free pages to compact things to at the > end of the zone. > > This can cause quadratic behaviour, with isolate_freepages starting at > the end of the zone each time, even though previous invocations of the > compaction code already filled up all free memory on that end of the zone. > This can cause isolate_freepages to take enormous amounts of CPU with > certain workloads on larger memory systems. > > This patch caches where the migration and free scanner should start from on > subsequent compaction invocations using the pageblock-skip information. When > compaction starts it begins from the cached restart points and will > update the cached restart points until a page is isolated or a pageblock > is skipped that would have been scanned by synchronous compaction. > > Signed-off-by: Mel Gorman Together with patch 5/6, this has the effect of skipping compaction in a zone if the free and isolate markers have met, and it has been less than 5 seconds since the "skip" information was reset. Compaction on zones where we cycle through more slowly can continue, even when this particular zone is experiencing problems, so I guess this is desired behaviour... Acked-by: Rik van Riel