From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932272AbdJJPjv (ORCPT ); Tue, 10 Oct 2017 11:39:51 -0400 Received: from mga05.intel.com ([192.55.52.43]:45663 "EHLO mga05.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755866AbdJJPjt (ORCPT ); Tue, 10 Oct 2017 11:39:49 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.42,505,1500966000"; d="scan'208";a="1180573414" Subject: Re: [PATCH v3] mm, sysctl: make NUMA stats configurable To: Michal Hocko References: <1506579101-5457-1-git-send-email-kemi.wang@intel.com> <20171003092352.2wh2jbtt2dudfi5a@dhcp22.suse.cz> <221a1e93-ee33-d598-67de-d6071f192040@intel.com> <20171009075549.pzohdnerillwuhqo@dhcp22.suse.cz> <20171010054902.sqp6yyid6qqhpsrt@dhcp22.suse.cz> <20171010143113.gk6iqcrguefhhlmr@dhcp22.suse.cz> <20171010145728.q2levvekbpwlg57q@dhcp22.suse.cz> Cc: kemi , "Luis R . Rodriguez" , Kees Cook , Andrew Morton , Jonathan Corbet , Mel Gorman , Johannes Weiner , Christopher Lameter , Sebastian Andrzej Siewior , Vlastimil Babka , Tim Chen , Andi Kleen , Jesper Dangaard Brouer , Ying Huang , Aaron Lu , Proc sysctl , Linux MM , Linux Kernel , Aaron Lu From: Dave Hansen Message-ID: <4949ccef-6b7f-c2d6-f500-92eadb2ba649@linux.intel.com> Date: Tue, 10 Oct 2017 08:39:47 -0700 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.3.0 MIME-Version: 1.0 In-Reply-To: <20171010145728.q2levvekbpwlg57q@dhcp22.suse.cz> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 10/10/2017 07:57 AM, Michal Hocko wrote: >> But, let's be honest, this leaves us with an option that nobody is ever >> going to turn on. IOW, nobody except a very small portion of our users >> will ever see any benefit from this. > But aren't those small groups who would like to squeeze every single > cycle out from the page allocator path the targeted audience? They're the reason we started looking at this. They also care the most. But, the cost of these stats, especially we get more and more cores in a NUMA node is really making them show up in profiles. It would be nice to get rid of them there, too. Aaron, do you remember offhand how much of the allocator overhead was coming from NUMA stats? From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Subject: Re: [PATCH v3] mm, sysctl: make NUMA stats configurable To: Michal Hocko References: <1506579101-5457-1-git-send-email-kemi.wang@intel.com> <20171003092352.2wh2jbtt2dudfi5a@dhcp22.suse.cz> <221a1e93-ee33-d598-67de-d6071f192040@intel.com> <20171009075549.pzohdnerillwuhqo@dhcp22.suse.cz> <20171010054902.sqp6yyid6qqhpsrt@dhcp22.suse.cz> <20171010143113.gk6iqcrguefhhlmr@dhcp22.suse.cz> <20171010145728.q2levvekbpwlg57q@dhcp22.suse.cz> Cc: kemi , "Luis R . Rodriguez" , Kees Cook , Andrew Morton , Jonathan Corbet , Mel Gorman , Johannes Weiner , Christopher Lameter , Sebastian Andrzej Siewior , Vlastimil Babka , Tim Chen , Andi Kleen , Jesper Dangaard Brouer , Ying Huang , Aaron Lu , Proc sysctl , Linux MM , Linux Kernel , Aaron Lu From: Dave Hansen Message-ID: <4949ccef-6b7f-c2d6-f500-92eadb2ba649@linux.intel.com> Date: Tue, 10 Oct 2017 08:39:47 -0700 MIME-Version: 1.0 In-Reply-To: <20171010145728.q2levvekbpwlg57q@dhcp22.suse.cz> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: On 10/10/2017 07:57 AM, Michal Hocko wrote: >> But, let's be honest, this leaves us with an option that nobody is ever >> going to turn on. IOW, nobody except a very small portion of our users >> will ever see any benefit from this. > But aren't those small groups who would like to squeeze every single > cycle out from the page allocator path the targeted audience? They're the reason we started looking at this. They also care the most. But, the cost of these stats, especially we get more and more cores in a NUMA node is really making them show up in profiles. It would be nice to get rid of them there, too. Aaron, do you remember offhand how much of the allocator overhead was coming from NUMA stats? -- 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 From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pf0-f197.google.com (mail-pf0-f197.google.com [209.85.192.197]) by kanga.kvack.org (Postfix) with ESMTP id 066246B0260 for ; Tue, 10 Oct 2017 11:39:51 -0400 (EDT) Received: by mail-pf0-f197.google.com with SMTP id j64so62341700pfj.6 for ; Tue, 10 Oct 2017 08:39:50 -0700 (PDT) Received: from mga01.intel.com (mga01.intel.com. [192.55.52.88]) by mx.google.com with ESMTPS id p27si5055772pgd.44.2017.10.10.08.39.49 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 10 Oct 2017 08:39:50 -0700 (PDT) Subject: Re: [PATCH v3] mm, sysctl: make NUMA stats configurable References: <1506579101-5457-1-git-send-email-kemi.wang@intel.com> <20171003092352.2wh2jbtt2dudfi5a@dhcp22.suse.cz> <221a1e93-ee33-d598-67de-d6071f192040@intel.com> <20171009075549.pzohdnerillwuhqo@dhcp22.suse.cz> <20171010054902.sqp6yyid6qqhpsrt@dhcp22.suse.cz> <20171010143113.gk6iqcrguefhhlmr@dhcp22.suse.cz> <20171010145728.q2levvekbpwlg57q@dhcp22.suse.cz> From: Dave Hansen Message-ID: <4949ccef-6b7f-c2d6-f500-92eadb2ba649@linux.intel.com> Date: Tue, 10 Oct 2017 08:39:47 -0700 MIME-Version: 1.0 In-Reply-To: <20171010145728.q2levvekbpwlg57q@dhcp22.suse.cz> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Michal Hocko Cc: kemi , "Luis R . Rodriguez" , Kees Cook , Andrew Morton , Jonathan Corbet , Mel Gorman , Johannes Weiner , Christopher Lameter , Sebastian Andrzej Siewior , Vlastimil Babka , Tim Chen , Andi Kleen , Jesper Dangaard Brouer , Ying Huang , Aaron Lu , Proc sysctl , Linux MM , Linux Kernel Aaron Lu On 10/10/2017 07:57 AM, Michal Hocko wrote: >> But, let's be honest, this leaves us with an option that nobody is ever >> going to turn on. IOW, nobody except a very small portion of our users >> will ever see any benefit from this. > But aren't those small groups who would like to squeeze every single > cycle out from the page allocator path the targeted audience? They're the reason we started looking at this. They also care the most. But, the cost of these stats, especially we get more and more cores in a NUMA node is really making them show up in profiles. It would be nice to get rid of them there, too. Aaron, do you remember offhand how much of the allocator overhead was coming from NUMA stats? -- 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