Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934511AbcCKHz1 (ORCPT ); Fri, 11 Mar 2016 02:55:27 -0500 Received: from mail-wm0-f42.google.com ([74.125.82.42]:36105 "EHLO mail-wm0-f42.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934343AbcCKHzS (ORCPT ); Fri, 11 Mar 2016 02:55:18 -0500 Date: Fri, 11 Mar 2016 08:55:15 +0100 From: Michal Hocko To: Johannes Weiner Cc: Andrew Morton , Vladimir Davydov , linux-mm@kvack.org, cgroups@vger.kernel.org, linux-kernel@vger.kernel.org, kernel-team@fb.com Subject: Re: [PATCH] mm: memcontrol: reclaim when shrinking memory.high below usage Message-ID: <20160311075514.GB27701@dhcp22.suse.cz> References: <1457643015-8828-1-git-send-email-hannes@cmpxchg.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1457643015-8828-1-git-send-email-hannes@cmpxchg.org> User-Agent: Mutt/1.5.24 (2015-08-30) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1619 Lines: 54 On Thu 10-03-16 15:50:13, Johannes Weiner wrote: > When setting memory.high below usage, nothing happens until the next > charge comes along, and then it will only reclaim its own charge and > not the now potentially huge excess of the new memory.high. This can > cause groups to stay in excess of their memory.high indefinitely. > > To fix that, when shrinking memory.high, kick off a reclaim cycle that > goes after the delta. This has been the case since the knob was introduce but I wouldn't bother with the CC: stable # 4.0+ as this was still in experimental mode. I guess we want to have it in 4.5 or put it to 4.5 stable. > > Signed-off-by: Johannes Weiner Acked-by: Michal Hocko Thanks! > --- > mm/memcontrol.c | 6 ++++++ > 1 file changed, 6 insertions(+) > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c > index 8615b066b642..f7c9b4cbdf01 100644 > --- a/mm/memcontrol.c > +++ b/mm/memcontrol.c > @@ -4992,6 +4992,7 @@ static ssize_t memory_high_write(struct kernfs_open_file *of, > char *buf, size_t nbytes, loff_t off) > { > struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of)); > + unsigned long nr_pages; > unsigned long high; > int err; > > @@ -5002,6 +5003,11 @@ static ssize_t memory_high_write(struct kernfs_open_file *of, > > memcg->high = high; > > + nr_pages = page_counter_read(&memcg->memory); > + if (nr_pages > high) > + try_to_free_mem_cgroup_pages(memcg, nr_pages - high, > + GFP_KERNEL, true); > + > memcg_wb_domain_size_changed(memcg); > return nbytes; > } > -- > 2.7.2 -- Michal Hocko SUSE Labs