From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751331AbeAWP1m (ORCPT ); Tue, 23 Jan 2018 10:27:42 -0500 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:34122 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750994AbeAWP1l (ORCPT ); Tue, 23 Jan 2018 10:27:41 -0500 Date: Tue, 23 Jan 2018 15:27:00 +0000 From: Roman Gushchin To: Andrey Grodzovsky CC: Andrey Grodzovsky , , , , , Subject: Re: [RFC] Per file OOM badness Message-ID: <20180123152659.GA21817@castle.DHCP.thefacebook.com> References: <1516294072-17841-1-git-send-email-andrey.grodzovsky@amd.com> <20180118170006.GG6584@dhcp22.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20180118170006.GG6584@dhcp22.suse.cz> User-Agent: Mutt/1.9.1 (2017-09-22) X-Originating-IP: [2620:10d:c092:200::1:c6d4] X-ClientProxiedBy: DB6PR0501CA0010.eurprd05.prod.outlook.com (2603:10a6:4:8f::20) To BL2PR15MB1074.namprd15.prod.outlook.com (2603:10b6:201:17::8) X-MS-PublicTrafficType: Email X-MS-Office365-Filtering-Correlation-Id: 52e9ea98-4b7f-4d9f-5fb0-08d56275c96b X-Microsoft-Antispam: UriScan:;BCL:0;PCL:0;RULEID:(7020095)(4652020)(5600026)(4604075)(4534165)(4627221)(201703031133081)(201702281549075)(2017052603307)(7153060)(7193020);SRVR:BL2PR15MB1074; X-Microsoft-Exchange-Diagnostics: 1;BL2PR15MB1074;3:HraSMpXR3tftqzMcVeCQH4M1Nt0MA9uQjqjkgvDdKFo3e9b3tR3rR0GGPnbcEphIzl3BeuIu/yyWBSX9KMSzgbCk5cQmKaodaChe5v3IBz6JCCOX0K9wJLEE2hrpBnbEY93npPX7z1wbEhKiTHKhY/5k3o9eqSkNAhuFAhzLpDUQH8NM49H534Ti2UNSWU6wuBURDHEt8R411xxHDgdGzLNo4IAiLY0C1hszKJNCqmljA7Mzm8OweWXFPRvahRAp;25:OLMAOkyEoUkx5uGtVxz3aCDayllOtx6dhKVAevc/u4uUcsdDP1L3Rra8NC92sylfcBljvC+LrySQOrCpvLDZ9upyy/D3TmQ76PZ5J5qLyVEj1SPzI+4azjhnAk0Zi52+e4GYzEA7yN9jADrHz0NHVQsg1sHbe4Q0SXkDdDI++q+ErrJ7x3k2zlfx1mQHjozGVv0VBGgoTIrOpu+lyFQYAcRhGfAeLDNO8oy3GBgTTUDDt+unx1Vu2ZtexS9ulozh9HbjGty3qMhenqcLM7k7mUxWDoMdzgrkAs8GRbC384JnRs/HqdqSO1MEiHGZkHqhwgr9Ftt9kPgRHDYydnJ93Q==;31:NFDrY8dKddg8pdF1NBTNlimvcX+d28qYYlFh6Bifl0zLKpap4T+JHEGVvr2peZwsI3yp0Fn+uSmjSoxyPFNSis1n92IivI0vM1xHccYwn+64YCBLy8mLpyUxDlVTMZQPmyyjLxRO5H7hw7R5vAbhrLPXQQDqrbujvg3xE6OhTbzhJkfsSS/aWAAbKHQjNbGhph3aFYh1w5da0JICAilofyFRJcCcGkVtUdw5RexJ28s= X-MS-TrafficTypeDiagnostic: BL2PR15MB1074: X-Microsoft-Exchange-Diagnostics: 1;BL2PR15MB1074;20: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;4:aCTn6inff8oEJL5DgR1RFjBq7wKnn0VKQKOumkNIZESLNWL2A7JPlJBgJrQAj+D0th3cI4VVTUf/ozQrVh5woqKFEV3sWz7ASkILrzPuD+m1xJpa512Tgw0m+3DsyJ8PamldrYowVLS7hHxBcAp6ynTs66XIyXRxqgOxb59tNRIw/O6B+HV+rcTeUXgVLjdFK8PYCYwKCnTgWsiczJbNvOl7Zz/F1qqDaYA9HN+iFZhgF7CFRivXFWVDJcUQOQC/ySnusAHv3vVV6p+c2A/vFpT4DC3eCG+5dJMsVH6m9bFjTvjXamFQVrvduALUL2fy X-Microsoft-Antispam-PRVS: X-Exchange-Antispam-Report-Test: UriScan:(10436049006162); X-Exchange-Antispam-Report-CFA-Test: BCL:0;PCL:0;RULEID:(6040501)(2401047)(8121501046)(5005006)(3002001)(10201501046)(93006095)(93001095)(3231023)(11241501184)(2400081)(944501161)(6041288)(20161123558120)(20161123562045)(20161123564045)(201703131423095)(201702281528075)(20161123555045)(201703061421075)(201703061406153)(20161123560045)(6072148)(201708071742011);SRVR:BL2PR15MB1074;BCL:0;PCL:0;RULEID:(100000803101)(100110400095);SRVR:BL2PR15MB1074; X-Forefront-PRVS: 05610E64EE X-Forefront-Antispam-Report: SFV:NSPM;SFS:(10019020)(39860400002)(346002)(366004)(39380400002)(376002)(396003)(54094003)(76104003)(189003)(199004)(377424004)(2870700001)(8666007)(229853002)(81166006)(81156014)(106356001)(8676002)(8936002)(47776003)(6116002)(68736007)(33656002)(25786009)(4326008)(23756003)(50466002)(6246003)(1076002)(2906002)(105586002)(6666003)(2950100002)(6916009)(966005)(52116002)(58126008)(86362001)(76176011)(53936002)(575784001)(5660300001)(52396003)(59450400001)(316002)(97736004)(6506007)(386003)(7696005)(478600001)(9686003)(6306002)(7736002)(55016002)(305945005)(83506002)(18370500001)(42262002);DIR:OUT;SFP:1102;SCL:1;SRVR:BL2PR15MB1074;H:castle.DHCP.thefacebook.com;FPR:;SPF:None;PTR:InfoNoRecords;MX:1;A:1;LANG:en; X-Microsoft-Exchange-Diagnostics: =?iso-8859-1?Q?1;BL2PR15MB1074;23:SAWOB6jaeDkMymeWCtD6sptEIddyXXpIYFATcT4?= =?iso-8859-1?Q?o5FRhWNrZB8jwwPpJqSo2vGLpY3Bwb3bKi5mdD3t9ZJbTJiSKHezNwm973?= =?iso-8859-1?Q?mpcxpfluLXvSSXWb2kLTd5abEoTM4mGvfJ0KIuHqDnN84Y2X2NMZhKJhLt?= =?iso-8859-1?Q?/a7pZ2ekcdFZFZ/IqLJALO9WqFzmPNaojieiNLt9LC5uoK81oLy8Gxx95e?= =?iso-8859-1?Q?eeAOGo7WkXJiLI+HZGkC1E/mhI5Bnd0jLNivE6PDXVCRLEWJjksti5C7Er?= =?iso-8859-1?Q?jLCcMJtMQF/vnCJC0H9UmDVxj+LmZ5E432cbMqeMNDvPaAP2PZlM5r8G/7?= =?iso-8859-1?Q?TEmaWhNXDY07LLqBfZ7G0bKKoitGFZV+jD21knbZ70nefkHOEJvJM16GKw?= =?iso-8859-1?Q?2k8BkUTEjCO9iGMLABH1b3nebAqpxfZXCHFI8dvKo2cEychBuoiaaJrwu1?= =?iso-8859-1?Q?FM5WQ/q5DdBvBl8GzkxykYs/+q0zvdBUJo9lo+Y/rwHstJ6pl9SS9PgGOw?= =?iso-8859-1?Q?ydna+kZsYAOmANYAenyFZ51wv9iWbOTjiNffDdUbukvjKB08nKNuzaeexN?= =?iso-8859-1?Q?ZfMeEnfx1SD6bJXMAms9N0ku3XWvJEr1EbPoFLyZC0xnb92Me/P7ISG1MF?= =?iso-8859-1?Q?VJUe8m4jv+2I/MMnBOFMeD5FzbyywgIIn45B0jjgI8sU7tJEA16U9p2JdK?= =?iso-8859-1?Q?Xt1h+XyBZ8i0h5CG1i7UVgmGmED8B+3l+diR6hXLSm4VRWEAbsdYBAV+Lu?= =?iso-8859-1?Q?YDHhTaERZoM231X3Lj1V/8zq1WmXDp/CEszucCMbg8l+Ptxvtb2oWSK3J5?= =?iso-8859-1?Q?wHkxHJVG+DlKE0gsFtqUh5SrigdjqdtJQI3dPKqXRzdpqC6FrrKXplcbuc?= =?iso-8859-1?Q?GmOchDnHc4VwpyzDbW8oHLmqoyv2H5y7DxVL3yZ6g/hoFvUi3cK95QCyrZ?= =?iso-8859-1?Q?VwU7jSwQddnfh/DfhAPxGw6ijfbqDEcG8357DE6/JR3f6Z8dcLKik8TjkU?= =?iso-8859-1?Q?DjEUMJrsBPTzUBxC6j4pwxhqQCka9Q02gGA/22y5EtH0lY8deITCvWfYuU?= =?iso-8859-1?Q?zPRAnYzEoUR8vJRgiSF9UzWG/CU1pi+caNITLochWHlMyPNpqx/kl2aTIi?= =?iso-8859-1?Q?HpQAT2gOtmdMVWUcoLBsukkxim5l2TvPdI7QI9FQqQUBcHrBHTgG4zKA6n?= =?iso-8859-1?Q?JFjjE7eXjO4GUD3LAGLvt4TiQKsXxqJQMc0mFW7IXmf0ujsyGJLgzNq7nn?= =?iso-8859-1?Q?rWmn4Vq9AhXAr2THWB9MhjP5CRTkIZUPQZdDWfB66dVrYtZiHURiX4L3gM?= =?iso-8859-1?Q?uU6AD26ZOCnUqFHcqcE3E4EldkG/cVn1KU2T/lfuIp/PT6bAwp8Jnf9lif?= =?iso-8859-1?Q?wU6AiNABEI5YBxSuuvCYn3vD9yOPi?= X-Microsoft-Exchange-Diagnostics: 1;BL2PR15MB1074;6:X6aUPIbBLay07kZoZTkktT4jiNrXc7gBGUtkjMcNOpiNHXbOUb/LqAvEVOeQrjrL/KyzBBgxdVyoIlSSq3vcpPvVAudWqv4M2p6WifsQm+bCxeKRz/xbPtg98AnSCfTIgAG1Qii/nls0BxrmwErxOugYu6NLDs7dyPoO4+0w1aaQuQ0WSDJs2oVd/FDkrNBkh4btIEcgFz1ZAUw5FWiS5SHuw1viM++0pxEmBd7MRBxbnayaJobLlrJs37IjnUVlkdBNX5/uc4Rgn5pTsXxbwbwtO/SbxCvQ6d9sG89Tgg3Znx8arjcLxW7exTlPQ88aT3C70wtWsfLqYTaTQE7KzEnXxq3uSHR/HG0PQ8zTN/I=;5:BLYes4YE7tzANVUmvhB+hUKIkLGRP7nXzqV3B2bHYAVHfQVO5VvvjveMkHJenRT3VJPSnr3otJeMjjo4hCJyepJZlLEyuU+CzTVs0kvzbzW14oPynsgPR88a8+R7TTEHpcf89xIXpkJsfrizGwqZaUx65ur13Yv8n3BY3mGYoUA=;24:k9hKH2nL1AjYCLN7GzLtfW9OGKeKZD2FJEY6xm6yV1g3+BERcvGaXROxy0VvAf+9r2Kik4d82wcvLd7wrpQrNWvS2BLWTHPjDLPJx0kjNtE=;7:fOMSe2Wbp2qA6UBXfd0/qrPRak2NxEv+DSUEwcassvp5zkZp0xsIoRgG02qBYGQXyMzxoQOcVc++EhDhTyFQX4UCZDlgJgFBY43u49HSKM9zE9NVz0tmhNjYtrX3DTUFjSLjsgJ2xGiq/de6vnc9AV7ENaZOWnHyuT+2FqU09FhFgnyGeNF7NEPp12ZPBQTMNOolHpHUccldAhTp2aOGkGlsIa8zkR4F8K08QIvpTFybhAzn6aEUGzFu+abrg+ZG SpamDiagnosticOutput: 1:99 SpamDiagnosticMetadata: NSPM X-Microsoft-Exchange-Diagnostics: 1;BL2PR15MB1074;20:R9yH5MuLW/28/4wHPB6wFYTHopmpoKKBeL6eeeH+n2RxVu+j9BvudwMewFeVnlTfOuXdAkFBzwNE4W2FwAEBxhEOQPzzVLTHVwGSMuOuutUXOxmSRsqvUa9xeG5XtcPjisa8A1CAyyNDfTPQJZBvTTGatiiS45q5lYmpmUkN9LE= X-MS-Exchange-CrossTenant-OriginalArrivalTime: 23 Jan 2018 15:27:11.3055 (UTC) X-MS-Exchange-CrossTenant-Network-Message-Id: 52e9ea98-4b7f-4d9f-5fb0-08d56275c96b X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 8ae927fe-1255-47a7-a2af-5f3a069daaa2 X-MS-Exchange-Transport-CrossTenantHeadersStamped: BL2PR15MB1074 X-OriginatorOrg: fb.com X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10432:,, definitions=2018-01-23_04:,, signatures=0 X-Proofpoint-Spam-Reason: safe X-FB-Internal: Safe Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Jan 18, 2018 at 06:00:06PM +0100, Michal Hocko wrote: > On Thu 18-01-18 11:47:48, Andrey Grodzovsky wrote: > > Hi, this series is a revised version of an RFC sent by Christian König > > a few years ago. The original RFC can be found at > > https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.freedesktop.org_archives_dri-2Ddevel_2015-2DSeptember_089778.html&d=DwIDAw&c=5VD0RTtNlTh3ycd41b3MUw&r=jJYgtDM7QT-W-Fz_d29HYQ&m=R-JIQjy8rqmH5qD581_VYL0Q7cpWSITKOnBCE-3LI8U&s=QZGqKpKuJ2BtioFGSy8_721owcWJ0J6c6d4jywOwN4w& > Here is the origin cover letter text > : I'm currently working on the issue that when device drivers allocate memory on > : behalf of an application the OOM killer usually doesn't knew about that unless > : the application also get this memory mapped into their address space. > : > : This is especially annoying for graphics drivers where a lot of the VRAM > : usually isn't CPU accessible and so doesn't make sense to map into the > : address space of the process using it. > : > : The problem now is that when an application starts to use a lot of VRAM those > : buffers objects sooner or later get swapped out to system memory, but when we > : now run into an out of memory situation the OOM killer obviously doesn't knew > : anything about that memory and so usually kills the wrong process. > : > : The following set of patches tries to address this problem by introducing a per > : file OOM badness score, which device drivers can use to give the OOM killer a > : hint how many resources are bound to a file descriptor so that it can make > : better decisions which process to kill. > : > : So question at every one: What do you think about this approach? > : > : My biggest concern right now is the patches are messing with a core kernel > : structure (adding a field to struct file). Any better idea? I'm considering > : to put a callback into file_ops instead. Hello! I wonder if groupoom (aka cgroup-aware OOM killer) can work for you? We do have kmem accounting on the memory cgroup level, and the cgroup-aware OOM selection logic takes cgroup's kmem size into account. So, you don't need to introduce another accounting mechanism for OOM. You can find the current implementation in the mm tree. Thanks! Roman