From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from e8.ny.us.ibm.com (e8.ny.us.ibm.com [32.97.182.138]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "e8.ny.us.ibm.com", Issuer "Equifax" (verified OK)) by ozlabs.org (Postfix) with ESMTPS id 9EB25B7B8F for ; Sat, 10 Oct 2009 07:23:08 +1100 (EST) Received: from d01relay02.pok.ibm.com (d01relay02.pok.ibm.com [9.56.227.234]) by e8.ny.us.ibm.com (8.14.3/8.13.1) with ESMTP id n99KKNgL012810 for ; Fri, 9 Oct 2009 16:20:23 -0400 Received: from d01av02.pok.ibm.com (d01av02.pok.ibm.com [9.56.224.216]) by d01relay02.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id n99KN5Wq255888 for ; Fri, 9 Oct 2009 16:23:05 -0400 Received: from d01av02.pok.ibm.com (loopback [127.0.0.1]) by d01av02.pok.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id n99KJe1V013729 for ; Fri, 9 Oct 2009 16:19:41 -0400 Date: Fri, 9 Oct 2009 15:23:04 -0500 From: Robert Jennings To: Andrew Morton Subject: Re: [PATCH 1/2][v2] mm: add notifier in pageblock isolation for balloon drivers Message-ID: <20091009202304.GB19114@austin.ibm.com> References: <20091002184458.GC4908@austin.ibm.com> <20091008163449.00dce972.akpm@linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20091008163449.00dce972.akpm@linux-foundation.org> Cc: linux-mm@kvack.org, Mel Gorman , Gerald Schaefer , linux-kernel@vger.kernel.org, linuxppc-dev@ozlabs.org, Martin Schwidefsky , Badari Pulavarty , Brian King , Paul Mackerras , Ingo Molnar , KAMEZAWA Hiroyuki List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , * Andrew Morton (akpm@linux-foundation.org) wrote: > On Fri, 2 Oct 2009 13:44:58 -0500 > Robert Jennings wrote: > > > Memory balloon drivers can allocate a large amount of memory which > > is not movable but could be freed to accomodate memory hotplug remove. > > > > Prior to calling the memory hotplug notifier chain the memory in the > > pageblock is isolated. If the migrate type is not MIGRATE_MOVABLE the > > isolation will not proceed, causing the memory removal for that page > > range to fail. > > > > Rather than failing pageblock isolation if the the migrateteype is not > > MIGRATE_MOVABLE, this patch checks if all of the pages in the pageblock > > are owned by a registered balloon driver (or other entity) using a > > notifier chain. If all of the non-movable pages are owned by a balloon, > > they can be freed later through the memory notifier chain and the range > > can still be isolated in set_migratetype_isolate(). > > The patch looks sane enough to me. > > I expect that if the powerpc and s390 guys want to work on CMM over the > next couple of months, they'd like this patch merged into 2.6.32. It's > a bit larger and more involved than one would like, but I guess we can > do that if suitable people (Mel? Kamezawa?) have had a close look and > are OK with it. > > What do people think? I'd love to get it in 2.6.32 if that's possible. I have gone over the comments from Mel and Kamezawa I produced a new patchset. I just finished testing it (and I also tested with CONFIG_MEMORY_HOTPLUG_SPARSE=n) and it will be posted shortly. > > Has it been carefully compile- and run-time tested with > CONFIG_MEMORY_HOTPLUG_SPARSE=n? Yes, I have compiled the kernel CONFIG_MEMORY_HOTPLUG_SPARSE=n and made sure that we didn't have any problems. --Robert Jennings