From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753879Ab2HAUxk (ORCPT ); Wed, 1 Aug 2012 16:53:40 -0400 Received: from mx1.redhat.com ([209.132.183.28]:33945 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752815Ab2HAUxi (ORCPT ); Wed, 1 Aug 2012 16:53:38 -0400 Message-ID: <501996CD.70007@redhat.com> Date: Wed, 01 Aug 2012 16:51:25 -0400 From: Rik van Riel User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:13.0) Gecko/20120605 Thunderbird/13.0 MIME-Version: 1.0 To: Minchan Kim CC: Rafael Aquini , linux-mm@kvack.org, linux-kernel@vger.kernel.org, virtualization@lists.linux-foundation.org, Rusty Russell , "Michael S. Tsirkin" , Mel Gorman , Andi Kleen , Andrew Morton , Konrad Rzeszutek Wilk , Rafael Aquini Subject: Re: [PATCH v4 1/3] mm: introduce compaction and migration for virtio ballooned pages References: <49f828a9331c9b729fcf77226006921ec5bc52fa.1342485774.git.aquini@redhat.com> <20120718054824.GA32341@bbox> <20120720194858.GA16249@t510.redhat.com> <20120723023332.GA6832@bbox> In-Reply-To: <20120723023332.GA6832@bbox> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 07/22/2012 10:33 PM, Minchan Kim wrote: > IMHO, better approach is that after we can get complete free pageblocks > by compaction or reclaim, move balloon pages into that pageblocks and make > that blocks to unmovable. It can prevent fragmentation and it makes > current or future code don't need to consider balloon page. I believe this is the wrong thing to do. In a KVM guest, getting applications in transparent huge pages can be a 10-25% performance benefit. Therefore, we need to make all the 2MB pageblocks we can available for use by userland. Using 2MB blocks for the balloon (which is never touched) is extremely wasteful and could result in a large performance penalty, if we cannot defragment the remaining memory enough to give 2MB pages to applications. The 2MB blocks are prime real estate. They should remain available for applications. From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx125.postini.com [74.125.245.125]) by kanga.kvack.org (Postfix) with SMTP id 797656B004D for ; Wed, 1 Aug 2012 16:53:32 -0400 (EDT) Message-ID: <501996CD.70007@redhat.com> Date: Wed, 01 Aug 2012 16:51:25 -0400 From: Rik van Riel MIME-Version: 1.0 Subject: Re: [PATCH v4 1/3] mm: introduce compaction and migration for virtio ballooned pages References: <49f828a9331c9b729fcf77226006921ec5bc52fa.1342485774.git.aquini@redhat.com> <20120718054824.GA32341@bbox> <20120720194858.GA16249@t510.redhat.com> <20120723023332.GA6832@bbox> In-Reply-To: <20120723023332.GA6832@bbox> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Minchan Kim Cc: Rafael Aquini , linux-mm@kvack.org, linux-kernel@vger.kernel.org, virtualization@lists.linux-foundation.org, Rusty Russell , "Michael S. Tsirkin" , Mel Gorman , Andi Kleen , Andrew Morton , Konrad Rzeszutek Wilk , Rafael Aquini On 07/22/2012 10:33 PM, Minchan Kim wrote: > IMHO, better approach is that after we can get complete free pageblocks > by compaction or reclaim, move balloon pages into that pageblocks and make > that blocks to unmovable. It can prevent fragmentation and it makes > current or future code don't need to consider balloon page. I believe this is the wrong thing to do. In a KVM guest, getting applications in transparent huge pages can be a 10-25% performance benefit. Therefore, we need to make all the 2MB pageblocks we can available for use by userland. Using 2MB blocks for the balloon (which is never touched) is extremely wasteful and could result in a large performance penalty, if we cannot defragment the remaining memory enough to give 2MB pages to applications. The 2MB blocks are prime real estate. They should remain available for applications. -- 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/ . Don't email: email@kvack.org