From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932672AbeBVOJg (ORCPT ); Thu, 22 Feb 2018 09:09:36 -0500 Received: from mx2.suse.de ([195.135.220.15]:53051 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753657AbeBVOJe (ORCPT ); Thu, 22 Feb 2018 09:09:34 -0500 Date: Thu, 22 Feb 2018 15:09:32 +0100 From: Michal Hocko To: Andrey Ryabinin Cc: Andrew Morton , Shakeel Butt , Cgroups , LKML , Linux MM , Johannes Weiner , Vladimir Davydov Subject: Re: [PATCH v5 2/2] mm/memcontrol.c: Reduce reclaim retries in mem_cgroup_resize_limit() Message-ID: <20180222140932.GL30681@dhcp22.suse.cz> References: <20171220102429.31601-1-aryabinin@virtuozzo.com> <20180119132544.19569-1-aryabinin@virtuozzo.com> <20180119132544.19569-2-aryabinin@virtuozzo.com> <20180119133510.GD6584@dhcp22.suse.cz> <20180119151118.GE6584@dhcp22.suse.cz> <20180221121715.0233d34dda330c56e1a9db5f@linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.9.3 (2018-01-21) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu 22-02-18 16:50:33, Andrey Ryabinin wrote: > On 02/21/2018 11:17 PM, Andrew Morton wrote: > > On Fri, 19 Jan 2018 16:11:18 +0100 Michal Hocko wrote: > > > >> And to be honest, I do not really see why keeping retrying from > >> mem_cgroup_resize_limit should be so much faster than keep retrying from > >> the direct reclaim path. We are doing SWAP_CLUSTER_MAX batches anyway. > >> mem_cgroup_resize_limit loop adds _some_ overhead but I am not really > >> sure why it should be that large. > > > > Maybe restarting the scan lots of times results in rescanning lots of > > ineligible pages at the start of the list before doing useful work? > > > > Andrey, are you able to determine where all that CPU time is being spent? > > > > I should have been more specific about the test I did. The full script looks like this: > > mkdir -p /sys/fs/cgroup/memory/test > echo $$ > /sys/fs/cgroup/memory/test/tasks > cat 4G_file > /dev/null > while true; do cat 4G_file > /dev/null; done & > loop_pid=$! > perf stat echo 50M > /sys/fs/cgroup/memory/test/memory.limit_in_bytes > echo -1 > /sys/fs/cgroup/memory/test/memory.limit_in_bytes > kill $loop_pid > > > I think the additional loops add some overhead and it's not that big by itself, but > this small overhead allows task to refill slightly more pages, increasing > the total amount of pages that mem_cgroup_resize_limit() need to reclaim. > > By using the following commands to show the the amount of reclaimed pages: > perf record -e vmscan:mm_vmscan_memcg_reclaim_end echo 50M > /sys/fs/cgroup/memory/test/memory.limit_in_bytes > perf script|cut -d '=' -f 2| paste -sd+ |bc > > I've got 1259841 pages (4.9G) with the patch vs 1394312 pages (5.4G) without it. So how does the picture changes if you have multiple producers? -- Michal Hocko SUSE Labs From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wr0-f199.google.com (mail-wr0-f199.google.com [209.85.128.199]) by kanga.kvack.org (Postfix) with ESMTP id AFDB46B02DC for ; Thu, 22 Feb 2018 09:09:35 -0500 (EST) Received: by mail-wr0-f199.google.com with SMTP id r29so3637869wra.13 for ; Thu, 22 Feb 2018 06:09:35 -0800 (PST) Received: from mx2.suse.de (mx2.suse.de. [195.135.220.15]) by mx.google.com with ESMTPS id i130si297992wmf.178.2018.02.22.06.09.34 for (version=TLS1 cipher=AES128-SHA bits=128/128); Thu, 22 Feb 2018 06:09:34 -0800 (PST) Date: Thu, 22 Feb 2018 15:09:32 +0100 From: Michal Hocko Subject: Re: [PATCH v5 2/2] mm/memcontrol.c: Reduce reclaim retries in mem_cgroup_resize_limit() Message-ID: <20180222140932.GL30681@dhcp22.suse.cz> References: <20171220102429.31601-1-aryabinin@virtuozzo.com> <20180119132544.19569-1-aryabinin@virtuozzo.com> <20180119132544.19569-2-aryabinin@virtuozzo.com> <20180119133510.GD6584@dhcp22.suse.cz> <20180119151118.GE6584@dhcp22.suse.cz> <20180221121715.0233d34dda330c56e1a9db5f@linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: owner-linux-mm@kvack.org List-ID: To: Andrey Ryabinin Cc: Andrew Morton , Shakeel Butt , Cgroups , LKML , Linux MM , Johannes Weiner , Vladimir Davydov On Thu 22-02-18 16:50:33, Andrey Ryabinin wrote: > On 02/21/2018 11:17 PM, Andrew Morton wrote: > > On Fri, 19 Jan 2018 16:11:18 +0100 Michal Hocko wrote: > > > >> And to be honest, I do not really see why keeping retrying from > >> mem_cgroup_resize_limit should be so much faster than keep retrying from > >> the direct reclaim path. We are doing SWAP_CLUSTER_MAX batches anyway. > >> mem_cgroup_resize_limit loop adds _some_ overhead but I am not really > >> sure why it should be that large. > > > > Maybe restarting the scan lots of times results in rescanning lots of > > ineligible pages at the start of the list before doing useful work? > > > > Andrey, are you able to determine where all that CPU time is being spent? > > > > I should have been more specific about the test I did. The full script looks like this: > > mkdir -p /sys/fs/cgroup/memory/test > echo $$ > /sys/fs/cgroup/memory/test/tasks > cat 4G_file > /dev/null > while true; do cat 4G_file > /dev/null; done & > loop_pid=$! > perf stat echo 50M > /sys/fs/cgroup/memory/test/memory.limit_in_bytes > echo -1 > /sys/fs/cgroup/memory/test/memory.limit_in_bytes > kill $loop_pid > > > I think the additional loops add some overhead and it's not that big by itself, but > this small overhead allows task to refill slightly more pages, increasing > the total amount of pages that mem_cgroup_resize_limit() need to reclaim. > > By using the following commands to show the the amount of reclaimed pages: > perf record -e vmscan:mm_vmscan_memcg_reclaim_end echo 50M > /sys/fs/cgroup/memory/test/memory.limit_in_bytes > perf script|cut -d '=' -f 2| paste -sd+ |bc > > I've got 1259841 pages (4.9G) with the patch vs 1394312 pages (5.4G) without it. So how does the picture changes if you have multiple producers? -- Michal Hocko SUSE Labs -- 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