From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx132.postini.com [74.125.245.132]) by kanga.kvack.org (Postfix) with SMTP id 697556B0002 for ; Mon, 18 Feb 2013 21:51:42 -0500 (EST) Received: by mail-vc0-f178.google.com with SMTP id m8so4022342vcd.9 for ; Mon, 18 Feb 2013 18:51:41 -0800 (PST) Message-ID: <5122E8C0.4020404@gmail.com> Date: Mon, 18 Feb 2013 21:51:44 -0500 From: KOSAKI Motohiro MIME-Version: 1.0 Subject: Re: [Lsf-pc] [LSF/MM TOPIC][ATTEND] a few topics I'd like to discuss References: <20130218145018.GJ4365@suse.de> In-Reply-To: <20130218145018.GJ4365@suse.de> Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Mel Gorman Cc: KOSAKI Motohiro , lsf-pc@lists.linux-foundation.org, "linux-mm@kvack.org" >> * Hugepage migration ? Currently, hugepage is not migratable and can?t >> use pages in ZONE_MOVABLE. It is not happy from point of CMA/hotplug >> view. >> > > migrate_huge_page() ? > > It's also possible to allocate hugetlbfs pages in ZONE_MOVABLE but must > be enabled via /proc/sys/vm/hugepages_treat_as_movable. Oops. I missed that. Sorry for noise. But but... I wonder why this knob is disabled by default. I don't think anybody get a benefit form current default. >> * Remove ZONE_MOVABLE ?Very long term goal. Maybe not suitable in this year. > > Whatever about removing it totally I would like to see node memory hot-remove > not depending on ZONE_MOVABLE. Yup. I also think so. the first target is page table and icache/dcache because they use a lot of memory. second step is to implement slab migration callback likes shrinker_slab. etc. it is very long term story. -- 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