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 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 smtp.lore.kernel.org (Postfix) with ESMTPS id 0FB04E9A04A for ; Wed, 18 Feb 2026 15:27:24 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 94B0C10E0F8; Wed, 18 Feb 2026 15:27:24 +0000 (UTC) Authentication-Results: gabe.freedesktop.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.b="gPk62qfh"; dkim-atps=neutral Received: from mgamail.intel.com (mgamail.intel.com [192.198.163.19]) by gabe.freedesktop.org (Postfix) with ESMTPS id 5C9FE10E0F8 for ; Wed, 18 Feb 2026 15:27:22 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1771428443; x=1802964443; h=message-id:date:subject:to:cc:references:from: in-reply-to:mime-version; bh=NbDV04Bbulp2PeIbVbCIuWX/cwUMCPWWswJ8/8YRzBM=; b=gPk62qfhoHvH/5Sa+RjkgHj7APfbUwCVyFulBCHo6L6IbeJ4nyB9bogt wO7BuVdljS9k1KONamUjuJVVs+F03B087yEwMTcB9Bw4xDpcOUzpH4WkH 2CmJCOV5Xak9t78wDUolZ1MqYOV5RUfaTNcQsPAbRuPP9rqNVKo9U+jCD 0Uq8slgYZ8selsDisKPegBd3PZNrlbuO+uAiu+Pdrttg3VBQBh2IHw+U7 JsqLmr/uRhM2Ti3KAESUc90PDjzmskw94DA/b0X6hJnlkDtaI4c5CwyhN qrN9H8jWWa3+tKBv+SgNwlYeIlvOz2CUUTpG9hGyLaAKcybTniIWNGg4/ w==; X-CSE-ConnectionGUID: XXuW6VzlQY2GUW+0mTqd6Q== X-CSE-MsgGUID: aYYFQTZcTKK+gSwKD4N1rw== X-IronPort-AV: E=McAfee;i="6800,10657,11705"; a="71526029" X-IronPort-AV: E=Sophos;i="6.21,298,1763452800"; d="scan'208,217";a="71526029" Received: from orviesa005.jf.intel.com ([10.64.159.145]) by fmvoesa113.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 18 Feb 2026 07:27:22 -0800 X-CSE-ConnectionGUID: b1zhac6BRYqtKNNGab3EUw== X-CSE-MsgGUID: Tlnh577CQS6dI1z+xjnvpg== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="6.21,298,1763452800"; d="scan'208,217";a="219218994" Received: from fmsmsx903.amr.corp.intel.com ([10.18.126.92]) by orviesa005.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 18 Feb 2026 07:27:21 -0800 Received: from FMSMSX902.amr.corp.intel.com (10.18.126.91) by fmsmsx903.amr.corp.intel.com (10.18.126.92) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.2562.35; Wed, 18 Feb 2026 07:27:21 -0800 Received: from fmsedg902.ED.cps.intel.com (10.1.192.144) by FMSMSX902.amr.corp.intel.com (10.18.126.91) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.2562.35 via Frontend Transport; Wed, 18 Feb 2026 07:27:21 -0800 Received: from BYAPR05CU005.outbound.protection.outlook.com (52.101.85.70) by edgegateway.intel.com (192.55.55.82) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.2562.35; Wed, 18 Feb 2026 07:27:20 -0800 ARC-Seal: i=1; a=rsa-sha256; s=arcselector10001; d=microsoft.com; cv=none; b=mx25ieB70pmglr4bFeEqc3oN9B+ziugTWxwXfNK8wwZRrYDuvKTS3CAF2sgYwDnpAco7in4jzqz52lEGczcVonZhPmVtqR6bJuxhAUNTjDsZNQJON9iQg5eIz8uQ+znRtNZfC5sgRG411hHceOzTq/SaFa7OiGPc/097C/BREJ9ImrN0bLRZfRjNh9sSdgh4zxBevm7CHTsI+zIo/5wWFkxDDXfFhyN/IxhSbl30OdZk1wVXx5T0Sh8mtYgGl2EjtoGvgdtXV5vD4TCzukx1QCPb44NW+AHEbzr8T9OmiM1moEB4VUUSJ0celBNpq3JS3QnE0VDE3q2jKWGmO4jv6w== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector10001; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=iks8Ju1qPj87t3FiU5YKJv20lCkTSw3XFvkVNTVcahI=; b=JKbcFmkOTs3sG8YzEny///QdmsMtkPKFXo0qwnKaXJ0NT0bzNPQG2yj5QuAUcZPX2F5gDawBylcwyFV2zeZePjW3DDQPL+Y4uA6Pz7WNDlO2UwO/tSDzCd+7HXYTLmnrLvfOnuQKnZO0RoXuvm1t945ORpW9hUhIftE//eMZHL98/w1O0tm3D7fHXlDBuHjUbzPMnWut53GxtwZOKVk1CTX9g8btvjLZulM4/xf1XBxqWJ+md3CzmLek1KEMp25iNBT0rADon924rzpapTipdMRCR6QdgKVMaBSagFXsQ4QZVv3D5UXq4/QMSXExUiXg4HL9EPGojafO325MRNhoHA== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=intel.com; dmarc=pass action=none header.from=intel.com; dkim=pass header.d=intel.com; arc=none Authentication-Results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=intel.com; Received: from CH3PR11MB8706.namprd11.prod.outlook.com (2603:10b6:610:1d1::22) by PH8PR11MB6612.namprd11.prod.outlook.com (2603:10b6:510:1cf::14) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.9632.13; Wed, 18 Feb 2026 15:27:18 +0000 Received: from CH3PR11MB8706.namprd11.prod.outlook.com ([fe80::e419:ae5c:91ec:1e9d]) by CH3PR11MB8706.namprd11.prod.outlook.com ([fe80::e419:ae5c:91ec:1e9d%5]) with mapi id 15.20.9632.010; Wed, 18 Feb 2026 15:27:12 +0000 Content-Type: multipart/alternative; boundary="------------SB0wE4cPtzh0TDYjV2be5aXr" Message-ID: <3c816b8d-b7ce-4bc4-9e9c-fd8c2e32512c@intel.com> Date: Wed, 18 Feb 2026 20:57:04 +0530 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH i-g-t v7 3/5] tools/gputop.src/xe_gputop: Refactor xe_populate_engines to close card_fd and use direct ioctl calls To: "Belgaumkar, Vinay" , , , , , CC: , References: <20260130095318.644256-1-soham.purkait@intel.com> <20260130095318.644256-4-soham.purkait@intel.com> <735a3190-0757-4cf5-ba1f-f6260647626b@intel.com> Content-Language: en-US From: "Purkait, Soham" In-Reply-To: <735a3190-0757-4cf5-ba1f-f6260647626b@intel.com> X-ClientProxiedBy: MA5P287CA0056.INDP287.PROD.OUTLOOK.COM (2603:1096:a01:1d3::11) To CH3PR11MB8706.namprd11.prod.outlook.com (2603:10b6:610:1d1::22) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: CH3PR11MB8706:EE_|PH8PR11MB6612:EE_ X-MS-Office365-Filtering-Correlation-Id: cb9ff43f-f500-49cf-c8ec-08de6f02302a X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0;ARA:13230040|1800799024|366016|376014|8096899003; X-Microsoft-Antispam-Message-Info: =?utf-8?B?dlVhaHZPUC9PSW9RMEJzUWgxNjVPbllWOGtROVdENEZVY2dGZG9iNTI0MEhi?= =?utf-8?B?SVJRM0prbTBOdnBCVEJndGY2U01GWmxUL3ZQNm5qQ0RPVUtUZENrSjZKbFEy?= =?utf-8?B?WW1STlBLTytjV3BVVUZxTXdLNmFTVXVWVWxBZ0pUK2NtdzBNckV2dHU1YWli?= =?utf-8?B?eWd4RmdId3NDZTFzVmlvWUdIRUhFM2l5djNLZmVpLzFXWVJDR3ZlcVJOYjl5?= =?utf-8?B?UERGd3VXd3lLaTBxL3NFS1NQWWdNazBHbkxyTFBRaDVGbGtObG1SYWdGN0FR?= =?utf-8?B?V1B3Ri94OVV2SkYxUnBiVm9oc2xLSnBVNDR2NzEzbTdYNWl6ZkN1VHNpTUhy?= =?utf-8?B?QXI5dm5HM1pBbWk1bk1tcXpqZUgxL3RhaUFQVk4rNC9BQkZaRGV1WXc5Slps?= =?utf-8?B?OEsybVdNVE80Z0FMaG9HVGROQWo5bWVyMGlhUWVkU1FLbTJERFVyYURadFdH?= =?utf-8?B?d3ZQUDhabnJyYUZjOTNQaXZlRUxqb1lmeUM3U0dJY3lWNFJxSC9DOUhLTlRI?= =?utf-8?B?U1VZaDdDSDBhcW9XSENUL1VXWG5NaGpCRVAwYmVWemFFMHc0MXZrSTYxaklj?= =?utf-8?B?Sm5lZlN6RzF5bFMvWWhSZ0s4Yy9XVHhkWGFmclArU3ljdmgxSlB1SDc3Qldj?= =?utf-8?B?emRUbDdFZ1h1L3lNL3VNMVNOWE1PK1M5Vm9OVW10NXRNVXdVaFFRUE9lMUtm?= =?utf-8?B?bjAzZnhLQVdIQXE4bjlnbElFUVYyQ3Q3L3hlcHBLUWUxNUl4OW1ScHNDQnI0?= =?utf-8?B?RVhGSFl5aWtsbXZjLytrWWw5YzlJMUpKVWM4Nys3VlBzeXFNUUwyU0FlZTJB?= =?utf-8?B?ZWxwaURlQjNKYjRGSHUvNFVVZmQyMGZvRHF1aHZtZUlVR2h0UG1uVGZ3SjdW?= =?utf-8?B?MEFKTWFrY0Z1RlA1Rnl3NzBQcmFJSmFxbmhlUUlEMXBVYjVxM1l0OGVtVkt1?= =?utf-8?B?bTRMMm5JUTRtN2dHd3h6NlNFS09GUkFkSFRSNEhZb1BaclgrS3RscXFDVURO?= =?utf-8?B?aDZleUthTmgwMFhMWmp0Mk1NbXFPekRJdDByVEhSc09zRlkrOTJkekJHcUFC?= =?utf-8?B?UFJqVjJkV1hEY2RPWTNKdVhCL2JtUjJWZ0JGclVlQ2RJalozRlNmTlgzQ2dC?= =?utf-8?B?MXI2TFJLSE1GNUg0UFRPTGYwaVJjTGFVWUFOQU5ESmQwdUZhUGdiaWFiUG04?= =?utf-8?B?eU1FYjBkK01mNGswY3dvNmg2ekozcWljTWNXRVZHbHNERkVVelY5VXZKV1FH?= =?utf-8?B?VHd1QXZLRFJKOUhVQ2JIcmt2NW9JRHpSSHJ2bDVXbEE1SlpCRFh2ZERYYWZw?= =?utf-8?B?dVVKWjhxWExSU0RmeFhaUFdtUkRJSjFFZjhwdHRUQXpIV1FUQ2dyL0lrM3l1?= =?utf-8?B?cXFiYWVxNmtLOWVXdjhnOFNVSzVvblh4T01HL2ZVa0J6bmtrZmFSUXc5QXFz?= =?utf-8?B?clZjY3JuZndDTTZFT0RzR2ZGU2hyc01MWUNVYVpVZlR3bHRKRTFVKzl4SUIz?= =?utf-8?B?cDNuTDhJdXVJSXZUZlRWSWEzU2hReGJ0T3Rrdjd5VVVFdVhOakxsUW9RRmIy?= =?utf-8?B?NmxoMmpmRlkyOHA3aDJJNTFCTWpmN2hLMTAxZDVWL0laakF6NmlEUnFvNmxz?= =?utf-8?B?eTB0NzY0MU9LVTRDRzNNRWkzRXl1cE1XREUrYjdnTDBjWEN4dHJxcmQyNzNk?= =?utf-8?B?U3pXdmtRWDc3Zkd6WHZrWGg0ZzFsKzFTb3dNUWZVblMxQmJXbEdjZ0E5NmVO?= =?utf-8?B?T1FNR3cxclF0TnR3OUloYXdwRkUvTkhKSVViQ3dvMVJNdWFjUkpvZVVjSVVR?= =?utf-8?B?ZGI3bmFjRmp4QXZzSXRhSzF3Z3V5dURydm9VakhLMWRlejNjMUdGdzZFb3NV?= =?utf-8?B?ZXFWazRacE5aQ3c2aHpCdUNBb1FGT0taUTNPbjVVZGxNelU5OW9YRWpKbDVV?= =?utf-8?B?U3J6aU90WVJjZkFGQlhkVGp2cVFoQ2o4ZFRoTmdsQ1hjTUxVRHlJcHloZnZS?= =?utf-8?B?c2dsNXJWWlFjY3R0UTM0T05xV3BEY2JHSjE3SXVpcytzNGxrbXhaU09iNHNz?= =?utf-8?B?Ykk5RHVGd1ZRTXlkcW45UlcyalFUSkpSb2JhS1BPRjZWNXk1ME02TjlOT3BY?= =?utf-8?Q?5M2Q=3D?= X-Forefront-Antispam-Report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:; IPV:NLI; SFV:NSPM; H:CH3PR11MB8706.namprd11.prod.outlook.com; PTR:; CAT:NONE; SFS:(13230040)(1800799024)(366016)(376014)(8096899003); DIR:OUT; SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?utf-8?B?RDlxUnhyZkw4TStQejhPbFU1NVRKTERVU2FMRVRuQU91S01Qem96bDJZYkM2?= =?utf-8?B?dXlCVHF3VDY3U2JpSUl2RWxvbkRvSFZTUnF6aTlteGd4UEF3aWpSeDZMNlNC?= =?utf-8?B?aVd6bGRjazhhNnNMQ0VFS2lsS1lsOHNxdm5qM3FpelozMWVocGd4U2wrckox?= =?utf-8?B?d0w1M3ExL1JkeU9YT0NHUVNNNHFFSEEvbWtRN0VVQ0t6bC9VMkU1UVhlQmRy?= =?utf-8?B?eFUra2ZjemVXWDk4WVFrWVFyZ3ZLYUNKbHhqUmxZaEtwMldkQ0x4WnlhVWZx?= =?utf-8?B?QkRlbmpYem80MExXWkFoYXpCMFBqQVZwSFZ5blhpT3c1TW50aDR4SE40elQr?= =?utf-8?B?eWgxZzNyL3dEQWNRbU1TM1lxRkhyRTZCcHd6VWdjYkpDM0FoMnpUcnlhTmVx?= =?utf-8?B?Z3paclNJR2p5MGNTQ0F1QTdoRTc5SUp3OWxwOGZYUVo1Umt0VUhrWlFmWlJX?= =?utf-8?B?VXFtQ2NvMERWY3k4dlFuSjFJbjRablhzVXF2NG1lZkg2bjNBRWdFMVM2QzBz?= =?utf-8?B?c1ZzVWoyVmZhQy92RUxxOFVrNm5adVVoY3pzck5wRERSWUR4RUFDakhBZWhN?= =?utf-8?B?eHdaUXM2Y2YrY01NN1l5elFVMmtBRTRYYldKQUZzb29zWGpsQTRkUkJvK1RL?= =?utf-8?B?dWVORUV4YjRNN2RIbVhabXY2UWlqNG1mQmV6aGVmQUJtRzZHWHVPekJRTkhs?= =?utf-8?B?SnROblVJTUNiU3AzVTFtdFVqZWMyc2dzSG9ydUgrZnFhVzhpM2J0MSsrekNX?= =?utf-8?B?dkt0d2YyN3AyQ1lwVGNHeHg0ajNhU3BMN2d6UDJ3bTBCSHpKdkdmUDZGb2cy?= =?utf-8?B?YlR3eTh5WThtbTlaL25TaWQya0kwQWJKZmhaa01KSmxtdExhQnJQajYrZlMy?= =?utf-8?B?RldiTEdCSGRoVWlKcGYzRktNM0dnVVdPMElUTzhMeG5za0VLQk1NS3NxdUxN?= =?utf-8?B?ck1VdGhGK3FUUjkvQjg5YlE2MTRoSW14UjVpZVoxak9YeFhwOFAxWlNjemlI?= =?utf-8?B?ZUFrQUg1L2xYYTJDOUxPVjUwNTIremxDeEM0eVExUkY0WlFlRUR1TGlNRXQv?= =?utf-8?B?bGgyR3c2VFc4L2FnR29JZzFxa1d6RFJNNWJPM2UxY1daSDRuOFZpQy9Yei8v?= =?utf-8?B?RkE4aVZNaGZkbFFXY2ViNmpVSmliMUpnODJWTEJITGM0M2Q1eHdvMUtKOC84?= =?utf-8?B?bGdQS2NBcVBCbXdXMkZiMU9GSi8yL28xRG5DSS9mWlNxYUtKMnVobHlna0Z6?= =?utf-8?B?eURJamVJMFFGM3RZS0MxUmFTYmF2NWtaYS9wQ3lOeU9aR1dpRDJmVFNzUjBP?= =?utf-8?B?dzB5RUtscVFNL3pNaExsUEtselBiNThUYU9ZenNITUt6elFUazM3VFBQSC9Z?= =?utf-8?B?ZUdVRTZCMWxZVDN1d2lORkFrTGRyalA0cThZRml4QTVtVENjb0lPYXlBQXR0?= =?utf-8?B?c0wwVStpSFNlRndCQUtERUQ3NkVhYlNqTXdxNmNQbUQxU0F1L3RPZGZLeFBI?= =?utf-8?B?Y0xUajJkL1ZTM2VrVkZmdWNYbVB2YWVvSW9OckN0T0hnUTJ2ZW9LZXZxR3pX?= =?utf-8?B?WktscldhKzJ2cWtYZFR5eFdaZjBOYTJoZGJVclZDQUF3bENHVG9GUlp2Nllv?= =?utf-8?B?VkpqZDB6T3Y1dGJ1M3dpRXAzWFozNm1hU3FNTFBZaWtQN1Z3QklDVVREWnBS?= =?utf-8?B?WExyM0Z5enFsbiswTDVnd3ZIVEdlVWJUZjd6NzJnU1g3akNkb1VpWHM5MWNE?= =?utf-8?B?N09ENnlCdU1aaEt6VkhlSjA3UmI1QVNmMFZyZmxQZUhUMy9wUTVYbmF3OW5M?= =?utf-8?B?UFI0MjNkQVFKbTZ2a0ZLNVVzWXMvVVhwbW9NRDZicTRUcDl1Yjk0bEErUVFq?= =?utf-8?B?aVRRR2xVZ3NxU1ZMNUJzdlprbW1PM2I1NUMwR3lkeTJ5K2lXQ0tTa2FFQUN1?= =?utf-8?B?L2JpV05Wdi9ZUjVKcW5vZHFnMzlwczZvczRjMk9HU0phajFGd09NYXcyL0ha?= =?utf-8?B?Z01kUE12NU1vR0M1T2tSMmR3UUkyMEVaY1hRNmZBNWs5aE9ZbEFaVHNjZ3dP?= =?utf-8?B?T1k1NmtUeEE4bmxiWlRkaUFlWkNPRWNZdmZ1SExKL1l0L0FDQVAvQjVIRzVH?= =?utf-8?B?Yjl6cmZ1OU5YY2xaa0JXU25rN1ppUm0xMFR5T1BqLy9jd0FDREg0ckR0YkNC?= =?utf-8?B?M2NVQmpSbURYRlkrL2d3dEQ3eGxsK0kycTdGRk9TYVBFL2RxcUZ1L0tQbnFJ?= =?utf-8?B?OWJDTWNyVHMwL1dhZTJjM2o2dm5IeTU3dVFtQjc2bXNMa3V1RnFaUTNEem9V?= =?utf-8?B?NUJwWkFLTkNlSFdmTWVYM2xReEd2bGNUQ0UwVXFuanl0MlBkdnJrSlJ5dFla?= =?utf-8?Q?Cobnp7/izY330Oi8=3D?= X-MS-Exchange-CrossTenant-Network-Message-Id: cb9ff43f-f500-49cf-c8ec-08de6f02302a X-MS-Exchange-CrossTenant-AuthSource: CH3PR11MB8706.namprd11.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 18 Feb 2026 15:27:12.3476 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 46c98d88-e344-4ed4-8496-4ed7712e255d X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: qCLtLEE976LJ+3feJ4uBKXRUGqDncJVvIIvYVzS44fr+f8h6GtcrOgUFNmFhfmkrXgIoJnTUQpiX4HWLTfvfZA== X-MS-Exchange-Transport-CrossTenantHeadersStamped: PH8PR11MB6612 X-OriginatorOrg: intel.com X-BeenThere: igt-dev@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Development mailing list for IGT GPU Tools List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: igt-dev-bounces@lists.freedesktop.org Sender: "igt-dev" --------------SB0wE4cPtzh0TDYjV2be5aXr Content-Type: text/plain; charset="UTF-8"; format=flowed Content-Transfer-Encoding: 8bit Hi Vinay On 12-02-2026 07:11, Belgaumkar, Vinay wrote: > > > On 1/30/2026 1:53 AM, Soham Purkait wrote: >> During GPUTOP engine initialization, the opened card fd is required to >> query engines and obtain configurations for PMU busyness calculation, but >> leaving it open causes the process to appear as a DRM client, which >> conflicts with KMS test requirements. To avoid this, card_fd is closed >> immediately after engine discovery, and engine queries are refactored to >> remove the lib/xe/* dependency in favor of direct ioctl calls, simplifying >> the implementation and reducing igt test library dependencies. >> >> v1: >> - Initialize pmu_device_obj to null. >> v2: >> - Remove dependency on lib/xe/* libraries. (Kamil) >> >> Fixes: c8106465683f ("tools/gputop/xe_gputop: Add gputop support for xe specific devices") >> Signed-off-by: Soham Purkait >> --- >> tools/gputop.src/xe_gputop.c | 100 ++++++++++++++++++++++++++++------- >> tools/gputop.src/xe_gputop.h | 3 +- >> 2 files changed, 81 insertions(+), 22 deletions(-) >> >> diff --git a/tools/gputop.src/xe_gputop.c b/tools/gputop.src/xe_gputop.c >> index bb2caa6ea..7793cd62f 100644 >> --- a/tools/gputop.src/xe_gputop.c >> +++ b/tools/gputop.src/xe_gputop.c >> @@ -1,8 +1,10 @@ >> // SPDX-License-Identifier: MIT >> /* >> - * Copyright © 2025 Intel Corporation >> + * Copyright © 2025-2026 Intel Corporation >> */ >> >> +#include >> + >> #include "xe_gputop.h" >> >> #define engine_ptr(pmu_device, n) (&(pmu_device)->engine + (n)) >> @@ -92,6 +94,7 @@ void xe_gputop_init(void *ptr, int index, >> >> obj = ((struct xe_gputop *)ptr) + index; >> obj->card = card; >> + obj->pmu_device_obj = NULL; >> } >> >> static int pmu_format_shift(int xe, const char *name) >> @@ -121,20 +124,54 @@ static int engine_cmp(const void *__a, const void *__b) >> return a->drm_xe_engine.engine_instance - b->drm_xe_engine.engine_instance; >> } >> >> +static int engine_query(int fd, struct drm_xe_query_engines **engine_q) >> +{ >> + struct drm_xe_device_query q; >> + int ret = 0, num_eng; >> + >> + memset(&q, 0, sizeof(q)); >> + q.query = DRM_XE_DEVICE_QUERY_ENGINES; >> + >> + ret = ioctl(fd, DRM_IOCTL_XE_DEVICE_QUERY, &q); >> + if (ret != 0) { >> + if ((errno != ENOSPC && errno != E2BIG) || q.size == 0) > why are we ignoring ENOSPC and E2BIG? I have rectified this check. Seems like the driver is not returning any of these during size check. >> + return -1; >> + } >> + if (q.size == 0) > we can check q.size just once instead of twice? Sure. >> + return 0; >> + >> + num_eng = (int)((q.size - sizeof(struct drm_xe_query_engines)) / >> + sizeof(struct drm_xe_engine_class_instance)); >> + >> + *engine_q = (struct drm_xe_query_engines *)calloc(1, q.size); >> + if (!*engine_q) >> + return -1; >> + >> + q.data = (uintptr_t)(*engine_q); > better to use to_user_pointer from the lib? For igt "tools" it has been asked to not use any library function from lib as these are meant to be used only with IGT tests. >> + >> + ret = ioctl(fd, DRM_IOCTL_XE_DEVICE_QUERY, &q); >> + if (ret != 0) { >> + free(*engine_q); >> + return -1; >> + } >> + >> + return num_eng; >> +} >> + >> void *xe_populate_engines(const void *obj, int index) >> { >> struct xe_gputop *ptr = ((struct xe_gputop *)obj) + index; >> struct igt_device_card *card = ptr->card; >> uint64_t engine_active_config, engine_total_config; >> uint64_t engine_class, engine_instance, gt_shift; >> - struct drm_xe_engine_class_instance *hwe; >> + struct drm_xe_query_engines *engine_q; >> struct xe_pmu_device *engines; >> char device[30]; >> - int ret = 0; >> - int card_fd; >> + int ret = 0, num_eng; >> + int card_fd = -1; >> >> if (!card || !strlen(card->card) || !strlen(card->render)) >> - return NULL; >> + goto err; >> >> if (strlen(card->card)) { >> card_fd = igt_open_card(card); >> @@ -142,39 +179,51 @@ void *xe_populate_engines(const void *obj, int index) >> card_fd = igt_open_render(card); >> } else { >> fprintf(stderr, "Failed to detect device!\n"); >> - return NULL; >> + goto err; >> } >> - xe_device_get(card_fd); >> + >> + num_eng = engine_query(card_fd, &engine_q); >> + if (num_eng <= 0) { >> + fprintf(stderr, "Engine query failed!\n"); >> + goto err; >> + } >> + >> engines = malloc(sizeof(struct xe_pmu_device) + >> - xe_number_engines(card_fd) * sizeof(struct xe_engine)); >> + num_eng * sizeof(struct xe_engine)); >> if (!engines) >> - return NULL; >> + goto err; >> >> memset(engines, 0, sizeof(struct xe_pmu_device) + >> - xe_number_engines(card_fd) * sizeof(struct xe_engine)); >> + num_eng * sizeof(struct xe_engine)); >> >> engines->num_engines = 0; >> gt_shift = pmu_format_shift(card_fd, "gt"); >> engine_class = pmu_format_shift(card_fd, "engine_class"); >> engine_instance = pmu_format_shift(card_fd, "engine_instance"); >> xe_perf_device(card_fd, device, sizeof(device)); >> + close(card_fd); >> + card_fd = -1; >> + >> engines->device = strdup(device); >> ret = perf_event_config(device, "engine-active-ticks", &engine_active_config); >> if (ret < 0) >> - return NULL; >> + goto err; >> >> ret = perf_event_config(device, "engine-total-ticks", &engine_total_config); >> if (ret < 0) >> - return NULL; >> + goto err; >> >> - xe_for_each_engine(card_fd, hwe) { >> + while (engines->num_engines < engine_q->num_engines) { >> uint64_t param_config; >> struct xe_engine *engine; >> + struct drm_xe_engine_class_instance hwe; >> + >> + hwe = engine_q->engines[engines->num_engines].instance; >> >> engine = engine_ptr(engines, engines->num_engines); >> - param_config = (uint64_t)hwe->gt_id << gt_shift | hwe->engine_class << engine_class >> - | hwe->engine_instance << engine_instance; >> - engine->drm_xe_engine = *hwe; >> + param_config = (uint64_t)hwe.gt_id << gt_shift | hwe.engine_class << engine_class >> + | hwe.engine_instance << engine_instance; >> + engine->drm_xe_engine = hwe; >> engine->engine_active_ticks.config = engine_active_config | param_config; >> engine->engine_total_ticks.config = engine_total_config | param_config; >> >> @@ -185,7 +234,7 @@ void *xe_populate_engines(const void *obj, int index) >> } >> >> ret = asprintf(&engine->display_name, "GT:%u %s/%u", >> - hwe->gt_id, >> + hwe.gt_id, >> class_display_name(engine->drm_xe_engine.engine_class), >> engine->drm_xe_engine.engine_instance); >> >> @@ -197,9 +246,11 @@ void *xe_populate_engines(const void *obj, int index) >> engines->num_engines++; >> } >> >> + free(engine_q); >> + >> if (!ret) { > > Not part of this patch, but looks like ret is overloaded here for > multiple things. Usually ret = 0 is a good thing? Should use num_bytes > or something if we expect asprintf above to return a positive value. > We should fix this as a separate patch. > Sure. Thanks, Soham > Thanks, > > Vinay. > >> errno = ret; >> - return NULL; >> + goto err; >> } >> >> qsort(engine_ptr(engines, 0), engines->num_engines, >> @@ -208,6 +259,15 @@ void *xe_populate_engines(const void *obj, int index) >> ptr->pmu_device_obj = engines; >> >> return engines; >> + >> +err: >> + if (card_fd >= 0) >> + close(card_fd); >> + if (engines) >> + free(engines); >> + if (engine_q) >> + free(engine_q); >> + return NULL; >> } >> >> static uint64_t pmu_read_multi(int fd, unsigned int num, uint64_t *val) >> @@ -263,11 +323,11 @@ int xe_pmu_init(const void *obj, int index) >> fd = _open_pmu(type, &engines->num_counters, &engine->engine_active_ticks, >> &engines->fd); >> if (fd < 0) >> - return -1; >> + return fd; >> fd = _open_pmu(type, &engines->num_counters, &engine->engine_total_ticks, >> &engines->fd); >> if (fd < 0) >> - return -1; >> + return fd; >> } >> return 0; >> } >> diff --git a/tools/gputop.src/xe_gputop.h b/tools/gputop.src/xe_gputop.h >> index 1e3856071..dcd82de7d 100644 >> --- a/tools/gputop.src/xe_gputop.h >> +++ b/tools/gputop.src/xe_gputop.h >> @@ -7,11 +7,10 @@ >> #define __XE_GPUTOP_H__ >> >> #include >> +#include >> >> -#include "igt_device_scan.h" >> #include "igt_perf.h" >> #include "utils.h" >> -#include "xe/xe_query.h" >> >> struct xe_pmu_pair { >> uint64_t cur; --------------SB0wE4cPtzh0TDYjV2be5aXr Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: 8bit

Hi Vinay

On 12-02-2026 07:11, Belgaumkar, Vinay wrote:


On 1/30/2026 1:53 AM, Soham Purkait wrote:
During GPUTOP engine initialization, the opened card fd is required to
query engines and obtain configurations for PMU busyness calculation, but
leaving it open causes the process to appear as a DRM client, which
conflicts with KMS test requirements. To avoid this, card_fd is closed
immediately after engine discovery, and engine queries are refactored to
remove the lib/xe/* dependency in favor of direct ioctl calls, simplifying
the implementation and reducing igt test library dependencies.

v1:
 - Initialize pmu_device_obj to null.
v2:
 - Remove dependency on lib/xe/* libraries. (Kamil)

Fixes: c8106465683f ("tools/gputop/xe_gputop: Add gputop support for xe specific devices")
Signed-off-by: Soham Purkait <soham.purkait@intel.com>
---
 tools/gputop.src/xe_gputop.c | 100 ++++++++++++++++++++++++++++-------
 tools/gputop.src/xe_gputop.h |   3 +-
 2 files changed, 81 insertions(+), 22 deletions(-)

diff --git a/tools/gputop.src/xe_gputop.c b/tools/gputop.src/xe_gputop.c
index bb2caa6ea..7793cd62f 100644
--- a/tools/gputop.src/xe_gputop.c
+++ b/tools/gputop.src/xe_gputop.c
@@ -1,8 +1,10 @@
 // SPDX-License-Identifier: MIT
 /*
- * Copyright © 2025 Intel Corporation
+ * Copyright © 2025-2026 Intel Corporation
  */
 
+#include <sys/ioctl.h>
+
 #include "xe_gputop.h"
 
 #define engine_ptr(pmu_device, n) (&(pmu_device)->engine + (n))
@@ -92,6 +94,7 @@ void xe_gputop_init(void *ptr, int index,
 
 	obj = ((struct xe_gputop *)ptr) + index;
 	obj->card = card;
+	obj->pmu_device_obj = NULL;
 }
 
 static int pmu_format_shift(int xe, const char *name)
@@ -121,20 +124,54 @@ static int engine_cmp(const void *__a, const void *__b)
 		return a->drm_xe_engine.engine_instance - b->drm_xe_engine.engine_instance;
 }
 
+static int engine_query(int fd, struct drm_xe_query_engines **engine_q)
+{
+	struct drm_xe_device_query q;
+	int ret = 0, num_eng;
+
+	memset(&q, 0, sizeof(q));
+	q.query = DRM_XE_DEVICE_QUERY_ENGINES;
+
+	ret = ioctl(fd, DRM_IOCTL_XE_DEVICE_QUERY, &q);
+	if (ret != 0) {
+		if ((errno != ENOSPC && errno != E2BIG) || q.size == 0)
why are we ignoring ENOSPC and E2BIG? 
I have rectified this check. Seems like the driver is not returning any of these during size check. 
+			return -1;
+	}
+	if (q.size == 0)
we can check q.size just once instead of twice? 
Sure. 
+		return 0;
+
+	num_eng = (int)((q.size - sizeof(struct drm_xe_query_engines)) /
+			sizeof(struct drm_xe_engine_class_instance));
+
+	*engine_q = (struct drm_xe_query_engines *)calloc(1, q.size);
+	if (!*engine_q)
+		return -1;
+
+	q.data = (uintptr_t)(*engine_q);
better to use to_user_pointer from the lib? 
For igt "tools" it has been asked to not use any library function from lib as these are meant to be used only with IGT tests. 
+
+	ret = ioctl(fd, DRM_IOCTL_XE_DEVICE_QUERY, &q);
+	if (ret != 0) {
+		free(*engine_q);
+		return -1;
+	}
+
+	return num_eng;
+}
+
 void *xe_populate_engines(const void *obj, int index)
 {
 	struct xe_gputop *ptr = ((struct xe_gputop *)obj) + index;
 	struct igt_device_card *card = ptr->card;
 	uint64_t engine_active_config, engine_total_config;
 	uint64_t engine_class, engine_instance, gt_shift;
-	struct drm_xe_engine_class_instance *hwe;
+	struct drm_xe_query_engines *engine_q;
 	struct xe_pmu_device *engines;
 	char device[30];
-	int ret = 0;
-	int card_fd;
+	int ret = 0, num_eng;
+	int card_fd = -1;
 
 	if (!card || !strlen(card->card) || !strlen(card->render))
-		return NULL;
+		goto err;
 
 	if (strlen(card->card)) {
 		card_fd = igt_open_card(card);
@@ -142,39 +179,51 @@ void *xe_populate_engines(const void *obj, int index)
 		card_fd = igt_open_render(card);
 	} else {
 		fprintf(stderr, "Failed to detect device!\n");
-		return NULL;
+		goto err;
 	}
-	xe_device_get(card_fd);
+
+	num_eng = engine_query(card_fd, &engine_q);
+	if (num_eng <= 0) {
+		fprintf(stderr, "Engine query failed!\n");
+		goto err;
+	}
+
 	engines = malloc(sizeof(struct xe_pmu_device) +
-			 xe_number_engines(card_fd) * sizeof(struct xe_engine));
+			 num_eng * sizeof(struct xe_engine));
 	if (!engines)
-		return NULL;
+		goto err;
 
 	memset(engines, 0, sizeof(struct xe_pmu_device) +
-	       xe_number_engines(card_fd) * sizeof(struct xe_engine));
+	       num_eng * sizeof(struct xe_engine));
 
 	engines->num_engines = 0;
 	gt_shift = pmu_format_shift(card_fd, "gt");
 	engine_class = pmu_format_shift(card_fd, "engine_class");
 	engine_instance = pmu_format_shift(card_fd, "engine_instance");
 	xe_perf_device(card_fd, device, sizeof(device));
+	close(card_fd);
+	card_fd = -1;
+
 	engines->device = strdup(device);
 	ret = perf_event_config(device, "engine-active-ticks", &engine_active_config);
 	if (ret < 0)
-		return NULL;
+		goto err;
 
 	ret = perf_event_config(device, "engine-total-ticks", &engine_total_config);
 	if (ret < 0)
-		return NULL;
+		goto err;
 
-	xe_for_each_engine(card_fd, hwe) {
+	while (engines->num_engines < engine_q->num_engines) {
 		uint64_t  param_config;
 		struct xe_engine *engine;
+		struct drm_xe_engine_class_instance hwe;
+
+		hwe = engine_q->engines[engines->num_engines].instance;
 
 		engine = engine_ptr(engines, engines->num_engines);
-		param_config = (uint64_t)hwe->gt_id << gt_shift | hwe->engine_class << engine_class
-			| hwe->engine_instance << engine_instance;
-		engine->drm_xe_engine = *hwe;
+		param_config = (uint64_t)hwe.gt_id << gt_shift | hwe.engine_class << engine_class
+			| hwe.engine_instance << engine_instance;
+		engine->drm_xe_engine = hwe; 
 		engine->engine_active_ticks.config = engine_active_config | param_config;
 		engine->engine_total_ticks.config = engine_total_config | param_config;
 
@@ -185,7 +234,7 @@ void *xe_populate_engines(const void *obj, int index)
 		}
 
 		ret = asprintf(&engine->display_name, "GT:%u %s/%u",
-			       hwe->gt_id,
+			       hwe.gt_id,
 			       class_display_name(engine->drm_xe_engine.engine_class),
 			       engine->drm_xe_engine.engine_instance);
 
@@ -197,9 +246,11 @@ void *xe_populate_engines(const void *obj, int index)
 		engines->num_engines++;
 	}
 
+	free(engine_q);
+
 	if (!ret) {

Not part of this patch, but looks like ret is overloaded here for multiple things. Usually ret = 0 is a good thing? Should use num_bytes or something if we expect asprintf above to return a positive value. We should fix this as a separate patch.

Sure.

Thanks,
Soham

Thanks,

Vinay.

 		errno = ret;
-		return NULL;
+		goto err;
 	}
 
 	qsort(engine_ptr(engines, 0), engines->num_engines,
@@ -208,6 +259,15 @@ void *xe_populate_engines(const void *obj, int index)
 	ptr->pmu_device_obj = engines;
 
 	return engines;
+
+err:
+	if (card_fd >= 0)
+		close(card_fd);
+	if (engines)
+		free(engines);
+	if (engine_q)
+		free(engine_q);
+	return NULL;
 }
 
 static uint64_t pmu_read_multi(int fd, unsigned int num, uint64_t *val)
@@ -263,11 +323,11 @@ int xe_pmu_init(const void *obj, int index)
 		fd = _open_pmu(type, &engines->num_counters, &engine->engine_active_ticks,
 			       &engines->fd);
 		if (fd < 0)
-			return -1;
+			return fd;
 		fd = _open_pmu(type, &engines->num_counters, &engine->engine_total_ticks,
 			       &engines->fd);
 		if (fd < 0)
-			return -1;
+			return fd;
 	}
 	return 0;
 }
diff --git a/tools/gputop.src/xe_gputop.h b/tools/gputop.src/xe_gputop.h
index 1e3856071..dcd82de7d 100644
--- a/tools/gputop.src/xe_gputop.h
+++ b/tools/gputop.src/xe_gputop.h
@@ -7,11 +7,10 @@
 #define __XE_GPUTOP_H__
 
 #include <dirent.h>
+#include <xe_drm.h>
 
-#include "igt_device_scan.h"
 #include "igt_perf.h"
 #include "utils.h"
-#include "xe/xe_query.h"
 
 struct xe_pmu_pair {
 	uint64_t cur;
--------------SB0wE4cPtzh0TDYjV2be5aXr--