From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: ** X-Spam-Status: No, score=2.4 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,HTML_MESSAGE,MAILING_LIST_MULTI,SPF_HELO_NONE, SPF_PASS autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9DEC1C2BA2B for ; Mon, 13 Apr 2020 20:13:00 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 6EB1D20656 for ; Mon, 13 Apr 2020 20:13:00 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (1024-bit key) header.d=amdcloud.onmicrosoft.com header.i=@amdcloud.onmicrosoft.com header.b="RoKjdlSX" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 6EB1D20656 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=amd.com Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=dri-devel-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id EB39F89D87; Mon, 13 Apr 2020 20:12:59 +0000 (UTC) Received: from NAM04-CO1-obe.outbound.protection.outlook.com (mail-eopbgr690049.outbound.protection.outlook.com [40.107.69.49]) by gabe.freedesktop.org (Postfix) with ESMTPS id 5E8A089D87; Mon, 13 Apr 2020 20:12:59 +0000 (UTC) ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=ZuZVJf5HsoWJLevVnzpIFprjE15GjkDEP1a40nj4OZ2aq+GnBuu496kptw0YUEAopHWcu6d6RwQq3M1rPEcrDJn5XFjJi1DgVtOzsFmRhd5kyxvRWSfTQ7hK2pRu0NCcJ755u3H7kO06ldkFDM0oJsaReixCRnQNXss1dRene/WlP57i3Zw2B6xs+MX4H+3Jmgwmnx9a2bWtCXM3UmlpdyTMnvYDG970VhdZYQRcLQPytsKUx9rB5d39xhCen8NPNg8B2L+k0rL4zWwW0wMWxwcy3nrLUF2VxizYupqexvlBZmnTxz73H1ayiJch34NNp8AF4G4eBJbSBmdkQs12Vw== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=4HDN964L3IYmQaJjJGdyQvWLrXuHRuqLvwxwrAWYjWo=; b=XAna71dLzxRaQ7eUW7fH3ORh8fEh56mBFw+AgvFYHzG9v1xJNQ4SQHGWTGsqKhfAWZucOh4UYX60HuSfQcxB8+cagcGA/MUONuQFmp8F4nrjFEj55oD/jgfftHlO9U32HU1BaaryFpru7kZWGYK5iCCoe9laNKSOdoOAoZb/WN960bUO3udu7gT87eQHi33BsI8OvHwtXOstxfuCb/OE1ZJD+1UvBlVUROW0gQcMxm6cElW30Vvkip+GoTGNwihMW8RgcIoz0uYp/ivK6A+gyH/0GSjB3qQpb+JtJ9pjGCvrJSdySV/lvUZTJ9aaGn4zR7t+GQnZjP2wg+nVz88ufA== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=amd.com; dmarc=pass action=none header.from=amd.com; dkim=pass header.d=amd.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amdcloud.onmicrosoft.com; s=selector2-amdcloud-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=4HDN964L3IYmQaJjJGdyQvWLrXuHRuqLvwxwrAWYjWo=; b=RoKjdlSX/85TNGBx54CqcsItI5AELnK0kOzAbqKW/Dv9StLWKe4Zot2yiCO+RLy9ENgdak42TCcJXDjJ0WuXNQPtaYt3F+siKXIb3HM7ErrhHhcQ0DvAz97h4xAelf7q6plaVGjmhXcbT3m9Q4lDFrvRhC64gBndpbr1BoYhHEQ= Received: from DM6PR12MB3996.namprd12.prod.outlook.com (2603:10b6:5:1c1::31) by DM6PR12MB3418.namprd12.prod.outlook.com (2603:10b6:5:116::31) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.2900.28; Mon, 13 Apr 2020 20:12:56 +0000 Received: from DM6PR12MB3996.namprd12.prod.outlook.com ([fe80::8111:a00b:8293:ba1c]) by DM6PR12MB3996.namprd12.prod.outlook.com ([fe80::8111:a00b:8293:ba1c%4]) with mapi id 15.20.2900.028; Mon, 13 Apr 2020 20:12:56 +0000 From: "Ho, Kenny" To: Tejun Heo , Kenny Ho Subject: Re: [PATCH v2 00/11] new cgroup controller for gpu/drm subsystem Thread-Topic: [PATCH v2 00/11] new cgroup controller for gpu/drm subsystem Thread-Index: AQHV7NdAtJDkPVX4/kS2rnKQbqOvE6hNEXkAgAst7ICAAADPgIAfdNEAgAAAkXw= Date: Mon, 13 Apr 2020 20:12:56 +0000 Message-ID: References: <20200226190152.16131-1-Kenny.Ho@amd.com> <20200324184633.GH162390@mtj.duckdns.org> , <20200413191136.GI60335@mtj.duckdns.org> In-Reply-To: <20200413191136.GI60335@mtj.duckdns.org> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: msip_labels: MSIP_Label_76546daa-41b6-470c-bb85-f6f40f044d7f_Enabled=True; MSIP_Label_76546daa-41b6-470c-bb85-f6f40f044d7f_SiteId=3dd8961f-e488-4e60-8e11-a82d994e183d; MSIP_Label_76546daa-41b6-470c-bb85-f6f40f044d7f_SetDate=2020-04-13T20:12:55.333Z; MSIP_Label_76546daa-41b6-470c-bb85-f6f40f044d7f_Name=Internal Distribution Only; MSIP_Label_76546daa-41b6-470c-bb85-f6f40f044d7f_ContentBits=0; MSIP_Label_76546daa-41b6-470c-bb85-f6f40f044d7f_Method=Standard; authentication-results: spf=none (sender IP is ) smtp.mailfrom=Kenny.Ho@amd.com; x-originating-ip: [165.204.84.11] x-ms-publictraffictype: Email x-ms-office365-filtering-ht: Tenant x-ms-office365-filtering-correlation-id: 72b6710b-de19-49d1-9b6a-08d7dfe70e58 x-ms-traffictypediagnostic: DM6PR12MB3418:|DM6PR12MB3418: x-ms-exchange-transport-forked: True x-microsoft-antispam-prvs: x-ms-oob-tlc-oobclassifiers: OLM:10000; x-forefront-prvs: 037291602B x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:; IPV:NLI; SFV:NSPM; H:DM6PR12MB3996.namprd12.prod.outlook.com; PTR:; CAT:NONE; SFTY:; SFS:(10009020)(4636009)(136003)(376002)(346002)(366004)(396003)(39860400002)(5660300002)(66946007)(33656002)(66476007)(66556008)(64756008)(66446008)(478600001)(316002)(9686003)(110136005)(52536014)(55016002)(76116006)(54906003)(4326008)(8936002)(71200400001)(2906002)(86362001)(26005)(81156014)(8676002)(186003)(19627405001)(6506007)(7696005)(53546011); DIR:OUT; SFP:1101; received-spf: None (protection.outlook.com: amd.com does not designate permitted sender hosts) x-ms-exchange-senderadcheck: 1 x-microsoft-antispam: BCL:0; x-microsoft-antispam-message-info: waoyhu8QydWyMvyZYOOHoE5+D8EWnMo3Qhu09AQoh7uiYaTebc7hG4mryD4DWIciO6evpBRucy5cOl1V+3zcL9hoDpqQAFskJnySj6yxMcScLw6MSpSOs4Qla7i5LeExH8ye6nIKWXPopD2PSyXZA/TQjIz/0wwUpbV82DwtID6d4j7fgol2RpOMq8lrvUdJ91cL6VCyVesgJayZiPkVONsvpfMfJJ26+huD9cdW9yDS8gQCM4ceZltcPUzedgG43r41ltLvSc9Sbancb+rX86DvXF2U8C7ur2RcqajANRo97FbP8mKDfMIGq46ccianG9ciU3s/EjUFt7HQg9f97gKxEnLVp+0m1FQxKNpt/QNrdm4ZqYLjOhTsr065SemdalcJqhqaYx6baY6eQBSUhcOtbiwDlTbQW4oL69VAESFQjXa8D+GRHjn4qUZgFBrX x-ms-exchange-antispam-messagedata: JbkFTTH3tNPNaHw7CSPyugkZ0K4mnlhd/HG/RMavqqPOourQu0fLRJMuZProoMrm8GYqTJG97NaIaQw16U0nyk56eliIJd0v/TITuArUB/Q063huV7IksOocVx2uiyJZORxH3rpuYzdmHlgo6QuWyw== MIME-Version: 1.0 X-OriginatorOrg: amd.com X-MS-Exchange-CrossTenant-Network-Message-Id: 72b6710b-de19-49d1-9b6a-08d7dfe70e58 X-MS-Exchange-CrossTenant-originalarrivaltime: 13 Apr 2020 20:12:56.5203 (UTC) X-MS-Exchange-CrossTenant-fromentityheader: Hosted X-MS-Exchange-CrossTenant-id: 3dd8961f-e488-4e60-8e11-a82d994e183d X-MS-Exchange-CrossTenant-mailboxtype: HOSTED X-MS-Exchange-CrossTenant-userprincipalname: Qf4d7SAQRd60TCHZAH07mbMjXuxmCqZXkgw0LzJJ8biyseJyuiPJ3u7tBJfol/op X-MS-Exchange-Transport-CrossTenantHeadersStamped: DM6PR12MB3418 X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: "Kuehling, Felix" , "jsparks@cray.com" , amd-gfx list , Larry Kaplan , dri-devel , "Greathouse, Joseph" , "Deucher, Alexander" , "cgroups@vger.kernel.org" , "Koenig, Christian" Content-Type: multipart/mixed; boundary="===============1338284564==" Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" --===============1338284564== Content-Language: en-US Content-Type: multipart/alternative; boundary="_000_DM6PR12MB399645C60BE2F5E3CC4F852883DD0DM6PR12MB3996namp_" --_000_DM6PR12MB399645C60BE2F5E3CC4F852883DD0DM6PR12MB3996namp_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable [AMD Official Use Only - Internal Distribution Only] Hi Tejun, Thanks for taking the time to reply. Perhaps we can even narrow things down to just gpu.weight/gpu.compute.weigh= t as a start? In this aspect, is the key objection to the current implemen= tation of gpu.compute.weight the work-conserving bit? This work-conserving= requirement is probably what I have missed for the last two years (and hen= ce going in circle.) If this is the case, can you clarify/confirm the followings? 1) Is resource scheduling goal of cgroup purely for the purpose of throughp= ut? (at the expense of other scheduling goals such as latency.) 2) If 1) is true, under what circumstances will the "Allocations" resource = distribution model (as defined in the cgroup-v2) be acceptable? 3) If 1) is true, are things like cpuset from cgroup v1 no longer acceptabl= e going forward? To be clear, while some have framed this (time sharing vs spatial sharing) = as a partisan issue, it is in fact a technical one. I have implemented the= gpu cgroup support this way because we have a class of users that value lo= w latency/low jitter/predictability/synchronicity. For example, they would= like 4 tasks to share a GPU and they would like the tasks to start and fin= ish at the same time. What is the rationale behind picking the Weight model over Allocations as t= he first acceptable implementation? Can't we have both work-conserving and= non-work-conserving ways of distributing GPU resources? If we can, why no= t allow non-work-conserving implementation first, especially when we have u= sers asking for such functionality? Regards, Kenny ________________________________ From: Tejun Heo on behalf of Tejun Heo Sent: Monday, April 13, 2020 3:11 PM To: Kenny Ho Cc: Ho, Kenny ; cgroups@vger.kernel.org ; dri-devel ; amd-gfx list ; Deucher, Alexander ; K= oenig, Christian ; Kuehling, Felix ; Greathouse, Joseph ; jsparks@cray.co= m Subject: Re: [PATCH v2 00/11] new cgroup controller for gpu/drm subsystem Hello, Kenny. On Tue, Mar 24, 2020 at 02:49:27PM -0400, Kenny Ho wrote: > Can you elaborate more on what are the missing pieces? Sorry about the long delay, but I think we've been going in circles for qui= te a while now. Let's try to make it really simple as the first step. How abou= t something like the following? * gpu.weight (should it be gpu.compute.weight? idk) - A single number per-device weight similar to io.weight, which distributes computation resources in work-conserving way. * gpu.memory.high - A single number per-device on-device memory limit. The above two, if works well, should already be plenty useful. And my guess= is that getting the above working well will be plenty challenging already even though it's already excluding work-conserving memory distribution. So, let'= s please do that as the first step and see what more would be needed from the= re. Thanks. -- tejun --_000_DM6PR12MB399645C60BE2F5E3CC4F852883DD0DM6PR12MB3996namp_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

[AMD Official Use Only - Internal Distribution Only]


Hi Tejun,

Thanks for taking the time to reply.

Perhaps we can even narrow things down to just gpu.weight/gpu.compute.weigh= t as a start?  In this aspect, is the key objection to the current imp= lementation of gpu.compute.weight the work-conserving bit?  This work-= conserving requirement is probably what I have missed for the last two years (and hence going in circle.)

If this is the case, can you clarify/confirm the followings?

1) Is resource scheduling goal of cgroup purely for the purpose of throughp= ut?  (at the expense of other scheduling goals such as latency.)
2) If 1) is true, under what circumstances will the "Allocations"= resource distribution model (as defined in the cgroup-v2) be acceptable?
3) If 1) is true, are things like cpuset from cgroup v1 no longer acceptabl= e going forward?

To be clear, while some have framed this (time sharing vs spatial sharing) = as a partisan issue, it is in fact a technical one.  I have implemente= d the gpu cgroup support this way because we have a class of users that val= ue low latency/low jitter/predictability/synchronicity.  For example, they would like 4 tasks to share a GPU and they would like th= e tasks to start and finish at the same time.

What is the rationale behind picking the Weight model over Allocations as t= he first acceptable implementation?  Can't we have both work-conservin= g and non-work-conserving ways of distributing GPU resources?  If we c= an, why not allow non-work-conserving implementation first, especially when we have users asking for such functionality?

Regards,
Kenny


From: Teju= n Heo <htejun@gmail.com> on behalf of Tejun Heo <tj@kernel.org>=
Sent: Monday, April 13, 2020 3:11 PM
To: Kenny Ho <y2kenny@gmail.com>
Cc: Ho, Kenny <Kenny.Ho@amd.com>; cgroups@vger.kernel.org <= cgroups@vger.kernel.org>; dri-devel <dri-devel@lists.freedesktop.org&= gt;; amd-gfx list <amd-gfx@lists.freedesktop.org>; Deucher, Alexander= <Alexander.Deucher@amd.com>; Koenig, Christian <Christian.Koenig@= amd.com>; Kuehling, Felix <Felix.Kuehling@amd.com>; Greathouse, Joseph <Jos= eph.Greathouse@amd.com>; jsparks@cray.com <jsparks@cray.com>
Subject: Re: [PATCH v2 00/11] new cgroup controller for gpu/drm subs= ystem
 
Hello, Kenny.

On Tue, Mar 24, 2020 at 02:49:27PM -0400, Kenny Ho wrote:
> Can you elaborate more on what are the missing pieces?

Sorry about the long delay, but I think we've been going in circles for qui= te
a while now. Let's try to make it really simple as the first step. How abou= t
something like the following?

* gpu.weight (should it be gpu.compute.weight? idk) - A single number
  per-device weight similar to io.weight, which distributes computatio= n
  resources in work-conserving way.

* gpu.memory.high - A single number per-device on-device memory limit.

The above two, if works well, should already be plenty useful. And my guess= is
that getting the above working well will be plenty challenging already even=
though it's already excluding work-conserving memory distribution. So, let'= s
please do that as the first step and see what more would be needed from the= re.

Thanks.

--
tejun
--_000_DM6PR12MB399645C60BE2F5E3CC4F852883DD0DM6PR12MB3996namp_-- --===============1338284564== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel --===============1338284564==--