Received: by 2002:a25:b323:0:0:0:0:0 with SMTP id l35csp119568ybj; Thu, 19 Sep 2019 11:36:25 -0700 (PDT) X-Google-Smtp-Source: APXvYqw1y+M6+mdaoC3TlsBlYPJ2CTOUAJ1X/0qMFwtvpt32GoCeFq2d+W/hRMOkzomL1ftJyGbo X-Received: by 2002:a50:aa96:: with SMTP id q22mr17525458edc.179.1568918185185; Thu, 19 Sep 2019 11:36:25 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1568918185; cv=none; d=google.com; s=arc-20160816; b=Z9iOjREZBfDR8HwJ1IkimCO8ZFEIVLmmndBwJTXmjkBbFwTzktIVtimPPoLH2t5WxK ZdiqONxyIaySAooLO/IpEmYw3JLubKFPzL6pqBcX9VMqKoBs6y3vgWs8fjl7XfByjN/R QVn+u6dJB7X5PiEuKFFIC+4b4RXjGXf7wBMkyKYz5lgPmRK4bfhIx5IJyzI3tYvsGcR5 JWm+MGU5QvkasI8eHLLpYp0fCB4oDYsiLpPxRPZc6725YW3rLNDdLTDuuQfOX+dZpvLI qdMfr5OfQya86g3TYlJVUPZvVWcHX0pjdctPVk6iFJBz4avgQixjkvahmVnoM9PZdgVf 8T9w== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:content-transfer-encoding :content-language:in-reply-to:mime-version:user-agent:date :message-id:autocrypt:openpgp:from:references:cc:to:subject; bh=y5YSVfQsKJaP5kiC7VNjkWq66eaXu8AINhJnfOaJCdM=; b=d5YZnFztvet+61VdlOUoQ4Atq+7TYXBoMmezjl3N4ud8ESdXNke+Ji/n0exaWuc7cb L8EDbAJH9IkV3hC9geuzwf+Fhf0IfcSLgSdbvRXXdNWGpAWBqxO2IdTXTo4gcI28javp 9ToubuoN6iEsk0eBlDjdU7A4gZVKldJp+Bo9nNDf76w+8sRLkLARKrybiDbQz59tEStW ok+hNKkeuzBaEkITcHh71vd6YcbS8+qH2GWtQ4JzoP2FSoS4vgBmlkdoTp2FaQ1Zumum 2iXXyIVb1HOsm5oslVYtlJZy4m/+J4r742DdWCHyCmPEHIoslSk6GdE+JLNEU5D1Wzai ViRw== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=intel.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id h19si5107047ejd.142.2019.09.19.11.36.01; Thu, 19 Sep 2019 11:36:25 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=intel.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2404337AbfISQ1V (ORCPT + 99 others); Thu, 19 Sep 2019 12:27:21 -0400 Received: from mga07.intel.com ([134.134.136.100]:36371 "EHLO mga07.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2404330AbfISQ1V (ORCPT ); Thu, 19 Sep 2019 12:27:21 -0400 X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga004.jf.intel.com ([10.7.209.38]) by orsmga105.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 19 Sep 2019 09:27:20 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.64,524,1559545200"; d="scan'208";a="338711859" Received: from schen9-desk.jf.intel.com (HELO [10.54.74.162]) ([10.54.74.162]) by orsmga004.jf.intel.com with ESMTP; 19 Sep 2019 09:27:20 -0700 Subject: Re: Usecases for the per-task latency-nice attribute To: Parth Shah , linux-kernel@vger.kernel.org, Peter Zijlstra , Patrick Bellasi , subhra mazumdar , Valentin Schneider Cc: mingo@redhat.com, morten.rasmussen@arm.com, dietmar.eggemann@arm.com, pjt@google.com, vincent.guittot@linaro.org, quentin.perret@arm.com, dhaval.giani@oracle.com, daniel.lezcano@linaro.org, tj@kernel.org, rafael.j.wysocki@intel.com, qais.yousef@arm.com References: <3e5c3f36-b806-5bcc-e666-14dc759a2d7b@linux.ibm.com> <426c0513-4354-e085-5a5d-8073ab035030@linux.intel.com> <707610a5-3344-f0ac-5536-73c9565bc1c7@linux.ibm.com> From: Tim Chen Openpgp: preference=signencrypt Autocrypt: addr=tim.c.chen@linux.intel.com; prefer-encrypt=mutual; keydata= mQINBE6ONugBEAC1c8laQ2QrezbYFetwrzD0v8rOqanj5X1jkySQr3hm/rqVcDJudcfdSMv0 BNCCjt2dofFxVfRL0G8eQR4qoSgzDGDzoFva3NjTJ/34TlK9MMouLY7X5x3sXdZtrV4zhKGv 3Rt2osfARdH3QDoTUHujhQxlcPk7cwjTXe4o3aHIFbcIBUmxhqPaz3AMfdCqbhd7uWe9MAZX 7M9vk6PboyO4PgZRAs5lWRoD4ZfROtSViX49KEkO7BDClacVsODITpiaWtZVDxkYUX/D9OxG AkxmqrCxZxxZHDQos1SnS08aKD0QITm/LWQtwx1y0P4GGMXRlIAQE4rK69BDvzSaLB45ppOw AO7kw8aR3eu/sW8p016dx34bUFFTwbILJFvazpvRImdjmZGcTcvRd8QgmhNV5INyGwtfA8sn L4V13aZNZA9eWd+iuB8qZfoFiyAeHNWzLX/Moi8hB7LxFuEGnvbxYByRS83jsxjH2Bd49bTi XOsAY/YyGj6gl8KkjSbKOkj0IRy28nLisFdGBvgeQrvaLaA06VexptmrLjp1Qtyesw6zIJeP oHUImJltjPjFvyfkuIPfVIB87kukpB78bhSRA5mC365LsLRl+nrX7SauEo8b7MX0qbW9pg0f wsiyCCK0ioTTm4IWL2wiDB7PeiJSsViBORNKoxA093B42BWFJQARAQABtDRUaW0gQ2hlbiAo d29yayByZWxhdGVkKSA8dGltLmMuY2hlbkBsaW51eC5pbnRlbC5jb20+iQI+BBMBAgAoAhsD BgsJCAcDAgYVCAIJCgsEFgIDAQIeAQIXgAUCXFIuxAUJEYZe0wAKCRCiZ7WKota4STH3EACW 1jBRzdzEd5QeTQWrTtB0Dxs5cC8/P7gEYlYQCr3Dod8fG7UcPbY7wlZXc3vr7+A47/bSTVc0 DhUAUwJT+VBMIpKdYUbvfjmgicL9mOYW73/PHTO38BsMyoeOtuZlyoUl3yoxWmIqD4S1xV04 q5qKyTakghFa+1ZlGTAIqjIzixY0E6309spVTHoImJTkXNdDQSF0AxjW0YNejt52rkGXXSoi IgYLRb3mLJE/k1KziYtXbkgQRYssty3n731prN5XrupcS4AiZIQl6+uG7nN2DGn9ozy2dgTi smPAOFH7PKJwj8UU8HUYtX24mQA6LKRNmOgB290PvrIy89FsBot/xKT2kpSlk20Ftmke7KCa 65br/ExDzfaBKLynztcF8o72DXuJ4nS2IxfT/Zmkekvvx/s9R4kyPyebJ5IA/CH2Ez6kXIP+ q0QVS25WF21vOtK52buUgt4SeRbqSpTZc8bpBBpWQcmeJqleo19WzITojpt0JvdVNC/1H7mF 4l7og76MYSTCqIKcLzvKFeJSie50PM3IOPp4U2czSrmZURlTO0o1TRAa7Z5v/j8KxtSJKTgD lYKhR0MTIaNw3z5LPWCCYCmYfcwCsIa2vd3aZr3/Ao31ZnBuF4K2LCkZR7RQgLu+y5Tr8P7c e82t/AhTZrzQowzP0Vl6NQo8N6C2fcwjSrkCDQROjjboARAAx+LxKhznLH0RFvuBEGTcntrC 3S0tpYmVsuWbdWr2ZL9VqZmXh6UWb0K7w7OpPNW1FiaWtVLnG1nuMmBJhE5jpYsi+yU8sbMA 5BEiQn2hUo0k5eww5/oiyNI9H7vql9h628JhYd9T1CcDMghTNOKfCPNGzQ8Js33cFnszqL4I N9jh+qdg5FnMHs/+oBNtlvNjD1dQdM6gm8WLhFttXNPn7nRUPuLQxTqbuoPgoTmxUxR3/M5A KDjntKEdYZziBYfQJkvfLJdnRZnuHvXhO2EU1/7bAhdz7nULZktw9j1Sp9zRYfKRnQdIvXXa jHkOn3N41n0zjoKV1J1KpAH3UcVfOmnTj+u6iVMW5dkxLo07CddJDaayXtCBSmmd90OG0Odx cq9VaIu/DOQJ8OZU3JORiuuq40jlFsF1fy7nZSvQFsJlSmHkb+cDMZDc1yk0ko65girmNjMF hsAdVYfVsqS1TJrnengBgbPgesYO5eY0Tm3+0pa07EkONsxnzyWJDn4fh/eA6IEUo2JrOrex O6cRBNv9dwrUfJbMgzFeKdoyq/Zwe9QmdStkFpoh9036iWsj6Nt58NhXP8WDHOfBg9o86z9O VMZMC2Q0r6pGm7L0yHmPiixrxWdW0dGKvTHu/DH/ORUrjBYYeMsCc4jWoUt4Xq49LX98KDGN dhkZDGwKnAUAEQEAAYkCJQQYAQIADwIbDAUCXFIulQUJEYZenwAKCRCiZ7WKota4SYqUEACj P/GMnWbaG6s4TPM5Dg6lkiSjFLWWJi74m34I19vaX2CAJDxPXoTU6ya8KwNgXU4yhVq7TMId keQGTIw/fnCv3RLNRcTAapLarxwDPRzzq2snkZKIeNh+WcwilFjTpTRASRMRy9ehKYMq6Zh7 PXXULzxblhF60dsvi7CuRsyiYprJg0h2iZVJbCIjhumCrsLnZ531SbZpnWz6OJM9Y16+HILp iZ77miSE87+xNa5Ye1W1ASRNnTd9ftWoTgLezi0/MeZVQ4Qz2Shk0MIOu56UxBb0asIaOgRj B5RGfDpbHfjy3Ja5WBDWgUQGgLd2b5B6MVruiFjpYK5WwDGPsj0nAOoENByJ+Oa6vvP2Olkl gQzSV2zm9vjgWeWx9H+X0eq40U+ounxTLJYNoJLK3jSkguwdXOfL2/Bvj2IyU35EOC5sgO6h VRt3kA/JPvZK+6MDxXmm6R8OyohR8uM/9NCb9aDw/DnLEWcFPHfzzFFn0idp7zD5SNgAXHzV PFY6UGIm86OuPZuSG31R0AU5zvcmWCeIvhxl5ZNfmZtv5h8TgmfGAgF4PSD0x/Bq4qobcfaL ugWG5FwiybPzu2H9ZLGoaRwRmCnzblJG0pRzNaC/F+0hNf63F1iSXzIlncHZ3By15bnt5QDk l50q2K/r651xphs7CGEdKi1nU0YJVbQxJQ== Message-ID: <34a63eac-ecd2-9bf5-74a1-7100faf55abf@linux.intel.com> Date: Thu, 19 Sep 2019 09:27:20 -0700 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.7.0 MIME-Version: 1.0 In-Reply-To: <707610a5-3344-f0ac-5536-73c9565bc1c7@linux.ibm.com> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 9/19/19 1:37 AM, Parth Shah wrote: > >> >> $> Separating AVX512 tasks and latency sensitive tasks on separate cores >> ------------------------------------------------------------------------- >> Another usecase we are considering is to segregate those workload that will pull down >> core cpu frequency (e.g. AVX512) from workload that are latency sensitive. >> There are certain tasks that need to provide a fast response time (latency sensitive) >> and they are best scheduled on cpu that has a lighter load and not have other >> tasks running on the sibling cpu that could pull down the cpu core frequency. >> >> Some users are running machine learning batch tasks with AVX512, and have observed >> that these tasks affect the tasks needing a fast response. They have to >> rely on manual CPU affinity to separate these tasks. With appropriate >> latency hint on task, the scheduler can be taught to separate them. >> > > Thanks for listing out your usecase. > > This is interesting. If scheduler has the knowledge of AVX512 tasks then > with these interface the scheduler can refrain from picking such core > occupying AVX512 tasks for the task with "latency-nice = -19". > > So I guess for this specific use-case, the value for such per-task > attribute should have range (most probably [-19,20]) and the name > "latency-nice" also suits the need. Yes. > > Do you have any specific values in mind for such attr? Not really. I assume a [-19 20] range that the user who launch the task will set. Probably something towards the -19 end for latency sensitive task and something towards the 20 end for AVX512 tasks. And 0 as default for most tasks. Tim