From nobody Sun Oct 5 00:16:08 2025 Received: from NAM02-SN1-obe.outbound.protection.outlook.com (mail-sn1nam02on2089.outbound.protection.outlook.com [40.107.96.89]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 414EC2E5B0C for ; Tue, 12 Aug 2025 02:41:22 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=fail smtp.client-ip=40.107.96.89 ARC-Seal: i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754966484; cv=fail; b=kfJIEO2p6K4K1gHk9KzYomGk/IOBfoB2v0ij3bE1uQby13g7vmpyMkDJQ216kbo/uXlDNj7xxEbc6kpyUdM3Gn69OVHkgiMTSKiwU6/9oIrzADu4Ep9E1j/a9vwNEgK5pE20fITbUs3MgAiFqXRzpr8XWhUcntV2SFpgaiGjLYg= ARC-Message-Signature: i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754966484; c=relaxed/simple; bh=x1QkON2uZInQ8pqCNuXX17OV/c1IKdMmbuxJVi0v5AM=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: Content-Type:MIME-Version; b=E/GR9oWAaSZ9cEyp7Czt7wBFq6FPdTLU8W38GVUSe0+L0OJdCOd2wSaK7HP33W39vlnJIZ735ps3kVGR9pobm53mk0l/1zkvfrNAjiy+hkxhf07dgB1lZX0cAjSaoS71U/lK8xeRTbS3MNUSTrjohfuFOSnnUP5+KpZu+wfNdGw= ARC-Authentication-Results: i=2; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=nvidia.com; spf=fail smtp.mailfrom=nvidia.com; dkim=pass (2048-bit key) header.d=Nvidia.com header.i=@Nvidia.com header.b=WHrq9H5V; arc=fail smtp.client-ip=40.107.96.89 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=nvidia.com Authentication-Results: smtp.subspace.kernel.org; spf=fail smtp.mailfrom=nvidia.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=Nvidia.com header.i=@Nvidia.com header.b="WHrq9H5V" ARC-Seal: i=1; a=rsa-sha256; s=arcselector10001; d=microsoft.com; cv=none; b=gK2UsZCcBaif6ODZNgMID4AqwXvHT1izjr9zDDt+a5V5QwPng8msgUBP3qS4FxHv3sI4h+2t5Owo1u34E/o77LqyVzPX8HBmtt2aTBJv42OtS1LjY1zWrA6COnuQ63PA4/01zzWwcwjFRayaKGDfiBVEd+tIx92FONufvVMxhIWo0w1OiuU05pfQMlQvcIyoxLsvtkXZB9Fo68pDXmoTfZpU2lLO9Tx5HRkCO5YHy0Oppgsq85cstddKNdVoWxfcJ22nv6lRQ7/2KjODvTVwNuvu4QF3dPNOUhj6ki1YKn0uUB0HQbLYw5X+puICLbkZ6rRYo7Oil57XdpoYj2aYnw== 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=dcIeQ6mGEnh1gATpgNs0IJRozX0C5y8a8f0HsOJvhtk=; b=e0zUMklZr6M3vXj7ZtqnMpJYePP+aIgXHcejzMLNb6l0h/NHuK+PRQ0eEBcQMoOhwnPNhFo2HNdymeMzAE0VX3SIe2rk09KHhC9QHowr/5y+XryftLHwdXg9XyYN0tGG6pPqffE4IVaTsUL6IEJU0UCe67grKN/nAWFwfa4snhmXsRgD0LF8qPfRUC/8gjMtwve5qMKTMMRt8CIDT/tBMgDyZer1mE9HYGN5WuRvCc9As3wV9lqwAIVLpN9G7H+mq7/5CnU5pjBedZtGymlcV8pySZoP0LdhORmo3HMiGTK514ciGMiDCnyA3Gg6FGtcJr3oS/by3zzTRu0oYtyW8A== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=nvidia.com; dmarc=pass action=none header.from=nvidia.com; dkim=pass header.d=nvidia.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=Nvidia.com; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=dcIeQ6mGEnh1gATpgNs0IJRozX0C5y8a8f0HsOJvhtk=; b=WHrq9H5VUXcU9x26xuVK5OvqiNvSAyJAfJqfZ0ATCm2TRE+28I/mw8K3CTj/ecuc6Q5Y31ruwHPjy2MSsPrFqvCsW91ovvu5HzqPZO8yE6iePouS2cFrIZBd3ZyWQ1hUgSTjH9+9uC8EOZWl3g7BHQfQCDYsjRTnHq6s/V6Gs8tMscasoSbJBV7kgbi1nnfKvcObM08WgTqyOJhatBgWjvJ27/fTm28kFcFqKqbSXSS3UGXJPAWTEKz6/IOPwdWVNqEpLLpnALHdkPRgy0jI6sp8IlCkh3fWz95FQjtHUIW9PJQDMn2hA+e6kKGpFkmGuYDUnIbUmp86Zy8KUoAFuw== Authentication-Results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=nvidia.com; Received: from PH8PR12MB7277.namprd12.prod.outlook.com (2603:10b6:510:223::13) by MN0PR12MB6317.namprd12.prod.outlook.com (2603:10b6:208:3c2::12) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.9009.13; Tue, 12 Aug 2025 02:41:17 +0000 Received: from PH8PR12MB7277.namprd12.prod.outlook.com ([fe80::3a4:70ea:ff05:1251]) by PH8PR12MB7277.namprd12.prod.outlook.com ([fe80::3a4:70ea:ff05:1251%5]) with mapi id 15.20.9009.018; Tue, 12 Aug 2025 02:41:17 +0000 From: Balbir Singh To: dri-devel@lists.freedesktop.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Balbir Singh , Andrew Morton , David Hildenbrand , Zi Yan , Joshua Hahn , Rakie Kim , Byungchul Park , Gregory Price , Ying Huang , Alistair Popple , Oscar Salvador , Lorenzo Stoakes , Baolin Wang , "Liam R. Howlett" , Nico Pache , Ryan Roberts , Dev Jain , Barry Song , Lyude Paul , Danilo Krummrich , David Airlie , Simona Vetter , Ralph Campbell , =?UTF-8?q?Mika=20Penttil=C3=A4?= , Matthew Brost , Francois Dugast Subject: [v3 10/11] gpu/drm/nouveau: add THP migration support Date: Tue, 12 Aug 2025 12:40:35 +1000 Message-ID: <20250812024036.690064-11-balbirs@nvidia.com> X-Mailer: git-send-email 2.50.1 In-Reply-To: <20250812024036.690064-1-balbirs@nvidia.com> References: <20250812024036.690064-1-balbirs@nvidia.com> Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: SY6PR01CA0001.ausprd01.prod.outlook.com (2603:10c6:10:e8::6) To PH8PR12MB7277.namprd12.prod.outlook.com (2603:10b6:510:223::13) Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: PH8PR12MB7277:EE_|MN0PR12MB6317:EE_ X-MS-Office365-Filtering-Correlation-Id: 52293748-a871-44f0-b7c5-08ddd949b657 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0;ARA:13230040|376014|7416014|366016|1800799024; X-Microsoft-Antispam-Message-Info: =?utf-8?B?cDNGZXVpMFpERkhuSDBDSG13TEE3aEJTTklsMnV2a2x4akNzdkVuYUtRU0xn?= =?utf-8?B?bUI2ZHF2YXYzWUx6OUlNZGZvNWxyL1hFZzJxcjZJNzdjUG1Dd1pvdExiYzV3?= =?utf-8?B?emF6R280NzBsaENEUGtiRzZuZUFyWm05dG1DRWpxOW83K1FMVFpDVm9LTDVo?= =?utf-8?B?NUpmTGFpZnhnSk1ieis4WHJ6WWtxUG0va0lZN0REdHorWXBJdzQ3SG1OakN3?= =?utf-8?B?aFZSWGNlLzdRdU1vRDhEMVVHL1Y3SG81cy9yNElDWnJxYzJxc3FvWFVWcFlS?= =?utf-8?B?YkdPRHlZZ0hEbEdmS3hnWWdxZVFIanJuV2FYQ0pLLytjVDNaa0ZWMU9hejBL?= =?utf-8?B?a2VYcDY4OTF6L2hjSUdLUi9UL2RCemNMVCtZOHNHOEp3ZFQ1M3JIZkpJN0Z4?= =?utf-8?B?SGc4WGdCN0xlcjliTjdWMzNldEhqZFp2d0pmYjFvdllIVUM4RGNOdDdGb2V3?= =?utf-8?B?Ykxwa2IwNnZOakVkbVZ5T2dDdThNRCtwdWhHd0o1UkpTb1hoVFFtUVdsTWlz?= =?utf-8?B?OHhRRHJjcDZDY2FaeE84QkU1T3FINUE4NEU0K2VNNmxSbFV0RzJmVUQwM3Nx?= =?utf-8?B?bDhTSE1VamlYcEJGdU1aSVBjRGxoWnVvZU5xby8rdjE2SERNV2ZjcEh0VHVI?= =?utf-8?B?MkxuWmIyMk0vWmJleTNEdGJwT1YreEk5OXhhOStXREV2WUl2MiszYTV1ODJ6?= =?utf-8?B?YVRKTTRPSnlqVWJjSlY5aXZyUm9BcE5ZM3dnWStYamtmRWtQWFRaQlp5M3N4?= =?utf-8?B?eGMzMWhGVnVkbzFDbEJaL3lIdTl2aGN5QXVZN0E1Y01KdWRQMndLdjJad3g4?= =?utf-8?B?ekdWWWFrTHNmcU1aK21WY0cwcTJZNzRnUUttaXJzTWlaV3A1c0RsYzliRGk5?= =?utf-8?B?T3ZTbjhIQ2tLYkx4K2diUXVWLzEzL0p0b0J2RktjaVJKUUFHaXREdE96aEdY?= =?utf-8?B?K2tCbDFOejMzL2l6b1YxYTM2ODVwenM3TTRDejAya0s0d0U0ejBab0dkTlU5?= =?utf-8?B?NnZEeTJTZWZETy92a0ZSOVVhWSsvWlEvT214VitMdHNpWG1zVVJPa2VhRFJO?= =?utf-8?B?OXZHbTVRMCtSN3Bhd01OYlVWeWlHa2hGdS8rbzFMS3R2VzV3L2YvN3ptc3da?= =?utf-8?B?dy9vMEZRNVlVVHl0aWJUNDNUalpRcW1kamJNU1lPdE8rT21adzZlUHRkSFZa?= =?utf-8?B?ZlVwRTVieEtmeUQ0Mm9MbGMxd3VZYStqZ1pGZTZhWVhhcUxCQXJTZDVyeXlk?= =?utf-8?B?U1RDRjlIK1Vmc1NoL2tIMlI0TjQzUC9SZFp1eXN6TUF0ZXZJN3lxWGY1TkZM?= =?utf-8?B?bW1PaFNUWHZ2Z0RJcmVPcUh5c2FFSnFKdWppczNIUHpFVFdlZmUvQ3gvd3N0?= =?utf-8?B?NWpnSHYzcktPSnZZWnZ0WmhSaW5zeTZ5WmkyblZFM1J3d2dYK1U2bm1EdG9Y?= =?utf-8?B?RGNBbWtQQ3ByemZhYms0Q2pmV3RoNjhydC84MEJ0RGhUblVudXV1N1p3clJB?= =?utf-8?B?cTgyT3lVcWZzNjVrV0FvTEdxblkwZVhGdVVqSm1ERGRGTEp4NmFGcklBeHcz?= =?utf-8?B?dXA1bGpOVmEzZnQzRlZxaVl0Rm5BUEU2am1aRUhXakVHZEwwaDBXRFh3Z0cw?= =?utf-8?B?N2JpNDh5VjBwOGdWQ0t6eUEva1dXbHRUaGErbFVQQmhTWDlqZ2NvcVpOTk9F?= =?utf-8?B?cWRsN2NJZjdQK2l2NElrYjJJZ1dDZ1orRGZMdVdSa2YwMDY4dHRZRmZZcWRm?= =?utf-8?B?VzBSSFFZRFQzVy9TZFVkdTFwRXQ4YS9wRmYza05jWUxDekNmYjl4b0xQTFRB?= =?utf-8?B?MERsZnR2UlExSHl5SkQ2QnJJdWdhUFowbFFuN096TVl4aGxKN3NKeFRVcWZW?= =?utf-8?B?YWQrdmtMWkVLOHRLYmxqOGY1K3BjQ09kMUl3M2ZvOFdRRlB1NTFrZnUzUTJ3?= =?utf-8?Q?58DOc5sGZB8=3D?= X-Forefront-Antispam-Report: CIP:255.255.255.255;CTRY:;LANG:en;SCL:1;SRV:;IPV:NLI;SFV:NSPM;H:PH8PR12MB7277.namprd12.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230040)(376014)(7416014)(366016)(1800799024);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?utf-8?B?clNlRWcvWHVFY3hmVnk3ZkMyMlVpbEJPdjJJMHdCdlZCNnhDcUFSYmd4SlF5?= =?utf-8?B?K3FMbUk4dmtITmJqUTViaGNrdEFGNWpUT2tIOGRuSSszenlFVnhid0w2VlJt?= =?utf-8?B?d1pEVHRWNURrSXBhejlFMDMzUUJldmdGYlpCRTFPdUhUVDl3UkJRSDRQWU4v?= =?utf-8?B?ajM2NHByWUdTZGhYU2lsZExGcHp6ZkVxNVNrcEFBSWxkYkhLUjd0eXdsTGFP?= =?utf-8?B?RWgralM0VXJ0RmJLRzJkdWdEbXYyVldtVUIySzRRcjJoM1libWxOMWMybDc4?= =?utf-8?B?Q3k5b3BjRGR0WUlrVmJIT29pN0VIa3dPUHRFVThWdi9sUWdTVjFHQlZ3Y0lK?= =?utf-8?B?bDZ1MFlrY0hCTXUvaWtqYkRlUk52bkpiWk8wQ3cvakVxbUY4THpwUXUyQ3l1?= =?utf-8?B?ZmovNVUrZkljU2hsTmxBRzJyd0huMitTbVVjcEFQWmI5K2ppdjlwLzJraTNp?= =?utf-8?B?SU9ablNPTUFSejhNVDh5a0VzVGRqUjBrMEo4L0k0Wk1iaGR0NWY0ajhReWVE?= =?utf-8?B?YVlrMEh6Q09iQUEydGRpeDRJNG52RlFOL2VqUDNmS0JaQUJ3MnFwRlNJeHht?= =?utf-8?B?K3N3ejd0WlExVklNZG5WL1BOMmNXbmN1cXQ3VzdrMEtTTkpKZTJlOHVBbHM1?= =?utf-8?B?cmtYRnpCUHRoamkyVHkxU25oaTY3bXovKzE0dFY5NTQyZDdpUmhSOERkbUpy?= =?utf-8?B?WGFSSkF0cDlWQlUwaERBVjdFTHZGaDFyektmaGNyMkxvMkZmK1FtLzRuSTF4?= =?utf-8?B?Y1JsMklFUkpuMlZoTHFzdEZGTWRUdFFJb2hrc256U214SXczd3ZWVWRFaUNS?= =?utf-8?B?b0tuUWpGckRad0pYNXBoMjN6SnRhVUZFZE5qdzN5UndtOGJ0SUFjUU5JY0FK?= =?utf-8?B?RWd2U1JFL3l0eHBoRDhhc3Z2M0ViallhVFk5dURzSEVzMGJ1VGtaM2tUZnB4?= =?utf-8?B?OUJ4Q1ZSMEYyNGJEcDV0Y05Pb2tlZG5XYi92YzRQRTJkTjZMaDJ5ZTQxbHd2?= =?utf-8?B?K0NteW9Sam9yb0lNN3NDOVd4d0ZrNGlXK1d5a01rODdiMTNXWGNkL0dkQlpS?= =?utf-8?B?TW9YUERVTlo5Um9mTGhlbmZBTXpkN3VaUDl3T3RVdndoUzRqM3ZjRHYzTmhm?= =?utf-8?B?VEVrWTZFN2pScm5yR0J2UnovbmZ5bFNueldudFFYQ05RUTBZZWkrY3Mra2FU?= =?utf-8?B?Wi9YUGJ4ajQ0S3JHaDN2L0NOOU9VejNSaVZOSnl0OWV6cml0cXhGYm01Tmpz?= =?utf-8?B?TEVYZGpVVklHSjJTcWZzYjdSa2cva1JDSi9vbHRYWUpPMFJlZWhNZ1JJRGxR?= =?utf-8?B?RE1WS3ZNZmsrd2o1NFhQK1VYeHlTS2hIZ1JXQ2FwZmpvaFo4QU5lK3lQd3Uv?= =?utf-8?B?azNUa1VQeHFCS3NVZmVIMjg3SWNBckxPWC8xczRQT0lpek1hVDdqaXNmTkhV?= =?utf-8?B?cnlmQmp3bmlmaUZZeC9DZlF1dkl4SnFwa3E4bHlEWTFHblphNFhYM0hwcktl?= =?utf-8?B?czVOUmhNMHRNaTRoNCtCL2pYdEtrMUZsV1JpcWg1UlRPWGdNbDZSRzRuM01o?= =?utf-8?B?RHVqVWpHVVJ3SGhBMGZhYVFZZjcvMVlqWCtPT28yR1VFcjFCaGZOT2lCc3NK?= =?utf-8?B?QVdEa1ZEU2NkWVl3ZzdJVWxtc3R5WGlpN2dNSXpHUmVVYnRhRC83RUlCR0F5?= =?utf-8?B?dWVjU1hKMzlwMCtYZ3VjQno0QWpFYmczUHZFTjhVMm52SWI3NHgrZ0ozWk1G?= =?utf-8?B?UmNDVTN3UnQrazM5dUZ1WFUydkN2YkFvLy9yUCtnMnRMYUFmaXExNmFwOG5M?= =?utf-8?B?eGNVSUNHc0tFemJXZ0pSMGpBek9FYkRDdlZZa0JDbmVlNWtXTW5RNVV1a3JX?= =?utf-8?B?U2NnT2luaVFRbjcvaUJjd0trUUQ5bnpWZm5raXpqeXZDc3NJRG5sNHZ4TFEv?= =?utf-8?B?d2VucFF2dVFZeFZDdkdqMEFtbDltNEVQM3BMckl0aFFCdDRFSUVNMWVVeTJi?= =?utf-8?B?RXBVWlI2VVI4TC8za3VnbjltMjZ6TTZFTHVWcFRoWVdaeXBYem1hMGdVZmlz?= =?utf-8?B?MXRBWVE5MUpmalBWWVV3bmpRMWZyaGZ5K0xBYzd1TG55WitXV29pakhPcVdV?= =?utf-8?Q?XX44gzXINmfkxVSyjdeIurWlC?= X-OriginatorOrg: Nvidia.com X-MS-Exchange-CrossTenant-Network-Message-Id: 52293748-a871-44f0-b7c5-08ddd949b657 X-MS-Exchange-CrossTenant-AuthSource: PH8PR12MB7277.namprd12.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 12 Aug 2025 02:41:17.3117 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 43083d15-7273-40c1-b7db-39efd9ccc17a X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: xKahMXWmA94MSg0kGLo4QDDPnyEhZ3Zka9CjDBjMtNtiL87f/oHrUgp5/dFL1HQuTxsWin5qf7XKGfhXnRlk1Q== X-MS-Exchange-Transport-CrossTenantHeadersStamped: MN0PR12MB6317 Change the code to add support for MIGRATE_VMA_SELECT_COMPOUND and appropriately handling page sizes in the migrate/evict code paths. Cc: Andrew Morton Cc: David Hildenbrand Cc: Zi Yan Cc: Joshua Hahn Cc: Rakie Kim Cc: Byungchul Park Cc: Gregory Price Cc: Ying Huang Cc: Alistair Popple Cc: Oscar Salvador Cc: Lorenzo Stoakes Cc: Baolin Wang Cc: "Liam R. Howlett" Cc: Nico Pache Cc: Ryan Roberts Cc: Dev Jain Cc: Barry Song Cc: Lyude Paul Cc: Danilo Krummrich Cc: David Airlie Cc: Simona Vetter Cc: Ralph Campbell Cc: Mika Penttil=C3=A4 Cc: Matthew Brost Cc: Francois Dugast Signed-off-by: Balbir Singh --- drivers/gpu/drm/nouveau/nouveau_dmem.c | 253 ++++++++++++++++++------- drivers/gpu/drm/nouveau/nouveau_svm.c | 6 +- drivers/gpu/drm/nouveau/nouveau_svm.h | 3 +- 3 files changed, 186 insertions(+), 76 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_dmem.c b/drivers/gpu/drm/nouve= au/nouveau_dmem.c index ca4932a150e3..408c1adf6f20 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dmem.c +++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c @@ -83,9 +83,15 @@ struct nouveau_dmem { struct list_head chunks; struct mutex mutex; struct page *free_pages; + struct folio *free_folios; spinlock_t lock; }; =20 +struct nouveau_dmem_dma_info { + dma_addr_t dma_addr; + size_t size; +}; + static struct nouveau_dmem_chunk *nouveau_page_to_chunk(struct page *page) { return container_of(page_pgmap(page), struct nouveau_dmem_chunk, @@ -112,10 +118,16 @@ static void nouveau_dmem_page_free(struct page *page) { struct nouveau_dmem_chunk *chunk =3D nouveau_page_to_chunk(page); struct nouveau_dmem *dmem =3D chunk->drm->dmem; + struct folio *folio =3D page_folio(page); =20 spin_lock(&dmem->lock); - page->zone_device_data =3D dmem->free_pages; - dmem->free_pages =3D page; + if (folio_order(folio)) { + folio_set_zone_device_data(folio, dmem->free_folios); + dmem->free_folios =3D folio; + } else { + page->zone_device_data =3D dmem->free_pages; + dmem->free_pages =3D page; + } =20 WARN_ON(!chunk->callocated); chunk->callocated--; @@ -139,20 +151,28 @@ static void nouveau_dmem_fence_done(struct nouveau_fe= nce **fence) } } =20 -static int nouveau_dmem_copy_one(struct nouveau_drm *drm, struct page *spa= ge, - struct page *dpage, dma_addr_t *dma_addr) +static int nouveau_dmem_copy_folio(struct nouveau_drm *drm, + struct folio *sfolio, struct folio *dfolio, + struct nouveau_dmem_dma_info *dma_info) { struct device *dev =3D drm->dev->dev; + struct page *dpage =3D folio_page(dfolio, 0); + struct page *spage =3D folio_page(sfolio, 0); =20 - lock_page(dpage); + folio_lock(dfolio); =20 - *dma_addr =3D dma_map_page(dev, dpage, 0, PAGE_SIZE, DMA_BIDIRECTIONAL); - if (dma_mapping_error(dev, *dma_addr)) + dma_info->dma_addr =3D dma_map_page(dev, dpage, 0, page_size(dpage), + DMA_BIDIRECTIONAL); + dma_info->size =3D page_size(dpage); + if (dma_mapping_error(dev, dma_info->dma_addr)) return -EIO; =20 - if (drm->dmem->migrate.copy_func(drm, 1, NOUVEAU_APER_HOST, *dma_addr, - NOUVEAU_APER_VRAM, nouveau_dmem_page_addr(spage))) { - dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); + if (drm->dmem->migrate.copy_func(drm, folio_nr_pages(sfolio), + NOUVEAU_APER_HOST, dma_info->dma_addr, + NOUVEAU_APER_VRAM, + nouveau_dmem_page_addr(spage))) { + dma_unmap_page(dev, dma_info->dma_addr, page_size(dpage), + DMA_BIDIRECTIONAL); return -EIO; } =20 @@ -165,21 +185,47 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct = vm_fault *vmf) struct nouveau_dmem *dmem =3D drm->dmem; struct nouveau_fence *fence; struct nouveau_svmm *svmm; - struct page *spage, *dpage; - unsigned long src =3D 0, dst =3D 0; - dma_addr_t dma_addr =3D 0; + struct page *dpage; vm_fault_t ret =3D 0; struct migrate_vma args =3D { .vma =3D vmf->vma, - .start =3D vmf->address, - .end =3D vmf->address + PAGE_SIZE, - .src =3D &src, - .dst =3D &dst, .pgmap_owner =3D drm->dev, .fault_page =3D vmf->page, - .flags =3D MIGRATE_VMA_SELECT_DEVICE_PRIVATE, + .flags =3D MIGRATE_VMA_SELECT_DEVICE_PRIVATE | + MIGRATE_VMA_SELECT_COMPOUND, + .src =3D NULL, + .dst =3D NULL, }; + unsigned int order, nr; + struct folio *sfolio, *dfolio; + struct nouveau_dmem_dma_info dma_info; + + sfolio =3D page_folio(vmf->page); + order =3D folio_order(sfolio); + nr =3D 1 << order; =20 + /* + * Handle partial unmap faults, where the folio is large, but + * the pmd is split. + */ + if (vmf->pte) { + order =3D 0; + nr =3D 1; + } + + if (order) + args.flags |=3D MIGRATE_VMA_SELECT_COMPOUND; + + args.start =3D ALIGN_DOWN(vmf->address, (1 << (PAGE_SHIFT + order))); + args.vma =3D vmf->vma; + args.end =3D args.start + (PAGE_SIZE << order); + args.src =3D kcalloc(nr, sizeof(*args.src), GFP_KERNEL); + args.dst =3D kcalloc(nr, sizeof(*args.dst), GFP_KERNEL); + + if (!args.src || !args.dst) { + ret =3D VM_FAULT_OOM; + goto err; + } /* * FIXME what we really want is to find some heuristic to migrate more * than just one page on CPU fault. When such fault happens it is very @@ -190,20 +236,26 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct = vm_fault *vmf) if (!args.cpages) return 0; =20 - spage =3D migrate_pfn_to_page(src); - if (!spage || !(src & MIGRATE_PFN_MIGRATE)) - goto done; - - dpage =3D alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vmf->vma, vmf->addres= s); - if (!dpage) + if (order) + dpage =3D folio_page(vma_alloc_folio(GFP_HIGHUSER | __GFP_ZERO, + order, vmf->vma, vmf->address), 0); + else + dpage =3D alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vmf->vma, + vmf->address); + if (!dpage) { + ret =3D VM_FAULT_OOM; goto done; + } =20 - dst =3D migrate_pfn(page_to_pfn(dpage)); + args.dst[0] =3D migrate_pfn(page_to_pfn(dpage)); + if (order) + args.dst[0] |=3D MIGRATE_PFN_COMPOUND; + dfolio =3D page_folio(dpage); =20 - svmm =3D spage->zone_device_data; + svmm =3D folio_zone_device_data(sfolio); mutex_lock(&svmm->mutex); nouveau_svmm_invalidate(svmm, args.start, args.end); - ret =3D nouveau_dmem_copy_one(drm, spage, dpage, &dma_addr); + ret =3D nouveau_dmem_copy_folio(drm, sfolio, dfolio, &dma_info); mutex_unlock(&svmm->mutex); if (ret) { ret =3D VM_FAULT_SIGBUS; @@ -213,19 +265,33 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct = vm_fault *vmf) nouveau_fence_new(&fence, dmem->migrate.chan); migrate_vma_pages(&args); nouveau_dmem_fence_done(&fence); - dma_unmap_page(drm->dev->dev, dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); + dma_unmap_page(drm->dev->dev, dma_info.dma_addr, PAGE_SIZE, + DMA_BIDIRECTIONAL); done: migrate_vma_finalize(&args); +err: + kfree(args.src); + kfree(args.dst); return ret; } =20 +static void nouveau_dmem_folio_split(struct folio *head, struct folio *tai= l) +{ + if (tail =3D=3D NULL) + return; + tail->pgmap =3D head->pgmap; + folio_set_zone_device_data(tail, folio_zone_device_data(head)); +} + static const struct dev_pagemap_ops nouveau_dmem_pagemap_ops =3D { .page_free =3D nouveau_dmem_page_free, .migrate_to_ram =3D nouveau_dmem_migrate_to_ram, + .folio_split =3D nouveau_dmem_folio_split, }; =20 static int -nouveau_dmem_chunk_alloc(struct nouveau_drm *drm, struct page **ppage) +nouveau_dmem_chunk_alloc(struct nouveau_drm *drm, struct page **ppage, + bool is_large) { struct nouveau_dmem_chunk *chunk; struct resource *res; @@ -274,16 +340,21 @@ nouveau_dmem_chunk_alloc(struct nouveau_drm *drm, str= uct page **ppage) pfn_first =3D chunk->pagemap.range.start >> PAGE_SHIFT; page =3D pfn_to_page(pfn_first); spin_lock(&drm->dmem->lock); - for (i =3D 0; i < DMEM_CHUNK_NPAGES - 1; ++i, ++page) { - page->zone_device_data =3D drm->dmem->free_pages; - drm->dmem->free_pages =3D page; + + if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE) || !is_large) { + for (i =3D 0; i < DMEM_CHUNK_NPAGES - 1; ++i, ++page) { + page->zone_device_data =3D drm->dmem->free_pages; + drm->dmem->free_pages =3D page; + } } + *ppage =3D page; chunk->callocated++; spin_unlock(&drm->dmem->lock); =20 - NV_INFO(drm, "DMEM: registered %ldMB of device memory\n", - DMEM_CHUNK_SIZE >> 20); + NV_INFO(drm, "DMEM: registered %ldMB of %sdevice memory %lx %lx\n", + DMEM_CHUNK_SIZE >> 20, is_large ? "THP " : "", pfn_first, + nouveau_dmem_page_addr(page)); =20 return 0; =20 @@ -298,27 +369,37 @@ nouveau_dmem_chunk_alloc(struct nouveau_drm *drm, str= uct page **ppage) } =20 static struct page * -nouveau_dmem_page_alloc_locked(struct nouveau_drm *drm) +nouveau_dmem_page_alloc_locked(struct nouveau_drm *drm, bool is_large) { struct nouveau_dmem_chunk *chunk; struct page *page =3D NULL; + struct folio *folio =3D NULL; int ret; + unsigned int order =3D 0; =20 spin_lock(&drm->dmem->lock); - if (drm->dmem->free_pages) { + if (is_large && drm->dmem->free_folios) { + folio =3D drm->dmem->free_folios; + drm->dmem->free_folios =3D folio_zone_device_data(folio); + chunk =3D nouveau_page_to_chunk(page); + chunk->callocated++; + spin_unlock(&drm->dmem->lock); + order =3D DMEM_CHUNK_NPAGES; + } else if (!is_large && drm->dmem->free_pages) { page =3D drm->dmem->free_pages; drm->dmem->free_pages =3D page->zone_device_data; chunk =3D nouveau_page_to_chunk(page); chunk->callocated++; spin_unlock(&drm->dmem->lock); + folio =3D page_folio(page); } else { spin_unlock(&drm->dmem->lock); - ret =3D nouveau_dmem_chunk_alloc(drm, &page); + ret =3D nouveau_dmem_chunk_alloc(drm, &page, is_large); if (ret) return NULL; } =20 - zone_device_page_init(page); + zone_device_folio_init(folio, order); return page; } =20 @@ -369,12 +450,12 @@ nouveau_dmem_evict_chunk(struct nouveau_dmem_chunk *c= hunk) { unsigned long i, npages =3D range_len(&chunk->pagemap.range) >> PAGE_SHIF= T; unsigned long *src_pfns, *dst_pfns; - dma_addr_t *dma_addrs; + struct nouveau_dmem_dma_info *dma_info; struct nouveau_fence *fence; =20 src_pfns =3D kvcalloc(npages, sizeof(*src_pfns), GFP_KERNEL | __GFP_NOFAI= L); dst_pfns =3D kvcalloc(npages, sizeof(*dst_pfns), GFP_KERNEL | __GFP_NOFAI= L); - dma_addrs =3D kvcalloc(npages, sizeof(*dma_addrs), GFP_KERNEL | __GFP_NOF= AIL); + dma_info =3D kvcalloc(npages, sizeof(*dma_info), GFP_KERNEL | __GFP_NOFAI= L); =20 migrate_device_range(src_pfns, chunk->pagemap.range.start >> PAGE_SHIFT, npages); @@ -382,17 +463,28 @@ nouveau_dmem_evict_chunk(struct nouveau_dmem_chunk *c= hunk) for (i =3D 0; i < npages; i++) { if (src_pfns[i] & MIGRATE_PFN_MIGRATE) { struct page *dpage; + struct folio *folio =3D page_folio( + migrate_pfn_to_page(src_pfns[i])); + unsigned int order =3D folio_order(folio); + + if (src_pfns[i] & MIGRATE_PFN_COMPOUND) { + dpage =3D folio_page( + folio_alloc( + GFP_HIGHUSER_MOVABLE, order), 0); + } else { + /* + * _GFP_NOFAIL because the GPU is going away and there + * is nothing sensible we can do if we can't copy the + * data back. + */ + dpage =3D alloc_page(GFP_HIGHUSER | __GFP_NOFAIL); + } =20 - /* - * _GFP_NOFAIL because the GPU is going away and there - * is nothing sensible we can do if we can't copy the - * data back. - */ - dpage =3D alloc_page(GFP_HIGHUSER | __GFP_NOFAIL); dst_pfns[i] =3D migrate_pfn(page_to_pfn(dpage)); - nouveau_dmem_copy_one(chunk->drm, - migrate_pfn_to_page(src_pfns[i]), dpage, - &dma_addrs[i]); + nouveau_dmem_copy_folio(chunk->drm, + page_folio(migrate_pfn_to_page(src_pfns[i])), + page_folio(dpage), + &dma_info[i]); } } =20 @@ -403,8 +495,9 @@ nouveau_dmem_evict_chunk(struct nouveau_dmem_chunk *chu= nk) kvfree(src_pfns); kvfree(dst_pfns); for (i =3D 0; i < npages; i++) - dma_unmap_page(chunk->drm->dev->dev, dma_addrs[i], PAGE_SIZE, DMA_BIDIRE= CTIONAL); - kvfree(dma_addrs); + dma_unmap_page(chunk->drm->dev->dev, dma_info[i].dma_addr, + dma_info[i].size, DMA_BIDIRECTIONAL); + kvfree(dma_info); } =20 void @@ -607,31 +700,35 @@ nouveau_dmem_init(struct nouveau_drm *drm) =20 static unsigned long nouveau_dmem_migrate_copy_one(struct nouveau_drm *drm, struct nouveau_svmm *svmm, unsigned long src, - dma_addr_t *dma_addr, u64 *pfn) + struct nouveau_dmem_dma_info *dma_info, u64 *pfn) { struct device *dev =3D drm->dev->dev; struct page *dpage, *spage; unsigned long paddr; + bool is_large =3D false; =20 spage =3D migrate_pfn_to_page(src); if (!(src & MIGRATE_PFN_MIGRATE)) goto out; =20 - dpage =3D nouveau_dmem_page_alloc_locked(drm); + is_large =3D src & MIGRATE_PFN_COMPOUND; + dpage =3D nouveau_dmem_page_alloc_locked(drm, is_large); if (!dpage) goto out; =20 paddr =3D nouveau_dmem_page_addr(dpage); if (spage) { - *dma_addr =3D dma_map_page(dev, spage, 0, page_size(spage), + dma_info->dma_addr =3D dma_map_page(dev, spage, 0, page_size(spage), DMA_BIDIRECTIONAL); - if (dma_mapping_error(dev, *dma_addr)) + dma_info->size =3D page_size(spage); + if (dma_mapping_error(dev, dma_info->dma_addr)) goto out_free_page; - if (drm->dmem->migrate.copy_func(drm, 1, - NOUVEAU_APER_VRAM, paddr, NOUVEAU_APER_HOST, *dma_addr)) + if (drm->dmem->migrate.copy_func(drm, folio_nr_pages(page_folio(spage)), + NOUVEAU_APER_VRAM, paddr, NOUVEAU_APER_HOST, + dma_info->dma_addr)) goto out_dma_unmap; } else { - *dma_addr =3D DMA_MAPPING_ERROR; + dma_info->dma_addr =3D DMA_MAPPING_ERROR; if (drm->dmem->migrate.clear_func(drm, page_size(dpage), NOUVEAU_APER_VRAM, paddr)) goto out_free_page; @@ -645,7 +742,7 @@ static unsigned long nouveau_dmem_migrate_copy_one(stru= ct nouveau_drm *drm, return migrate_pfn(page_to_pfn(dpage)); =20 out_dma_unmap: - dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); + dma_unmap_page(dev, dma_info->dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); out_free_page: nouveau_dmem_page_free_locked(drm, dpage); out: @@ -655,27 +752,33 @@ static unsigned long nouveau_dmem_migrate_copy_one(st= ruct nouveau_drm *drm, =20 static void nouveau_dmem_migrate_chunk(struct nouveau_drm *drm, struct nouveau_svmm *svmm, struct migrate_vma *args, - dma_addr_t *dma_addrs, u64 *pfns) + struct nouveau_dmem_dma_info *dma_info, u64 *pfns) { struct nouveau_fence *fence; unsigned long addr =3D args->start, nr_dma =3D 0, i; + unsigned long order =3D 0; + + for (i =3D 0; addr < args->end; ) { + struct folio *folio; =20 - for (i =3D 0; addr < args->end; i++) { + folio =3D page_folio(migrate_pfn_to_page(args->dst[i])); + order =3D folio_order(folio); args->dst[i] =3D nouveau_dmem_migrate_copy_one(drm, svmm, - args->src[i], dma_addrs + nr_dma, pfns + i); - if (!dma_mapping_error(drm->dev->dev, dma_addrs[nr_dma])) + args->src[i], dma_info + nr_dma, pfns + i); + if (!dma_mapping_error(drm->dev->dev, dma_info[nr_dma].dma_addr)) nr_dma++; - addr +=3D PAGE_SIZE; + i +=3D 1 << order; + addr +=3D (1 << order) * PAGE_SIZE; } =20 nouveau_fence_new(&fence, drm->dmem->migrate.chan); migrate_vma_pages(args); nouveau_dmem_fence_done(&fence); - nouveau_pfns_map(svmm, args->vma->vm_mm, args->start, pfns, i); + nouveau_pfns_map(svmm, args->vma->vm_mm, args->start, pfns, i, order); =20 while (nr_dma--) { - dma_unmap_page(drm->dev->dev, dma_addrs[nr_dma], PAGE_SIZE, - DMA_BIDIRECTIONAL); + dma_unmap_page(drm->dev->dev, dma_info[nr_dma].dma_addr, + dma_info[nr_dma].size, DMA_BIDIRECTIONAL); } migrate_vma_finalize(args); } @@ -689,20 +792,24 @@ nouveau_dmem_migrate_vma(struct nouveau_drm *drm, { unsigned long npages =3D (end - start) >> PAGE_SHIFT; unsigned long max =3D min(SG_MAX_SINGLE_ALLOC, npages); - dma_addr_t *dma_addrs; struct migrate_vma args =3D { .vma =3D vma, .start =3D start, .pgmap_owner =3D drm->dev, - .flags =3D MIGRATE_VMA_SELECT_SYSTEM, + .flags =3D MIGRATE_VMA_SELECT_SYSTEM + | MIGRATE_VMA_SELECT_COMPOUND, }; unsigned long i; u64 *pfns; int ret =3D -ENOMEM; + struct nouveau_dmem_dma_info *dma_info; =20 if (drm->dmem =3D=3D NULL) return -ENODEV; =20 + if (IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) + max =3D max(HPAGE_PMD_NR, max); + args.src =3D kcalloc(max, sizeof(*args.src), GFP_KERNEL); if (!args.src) goto out; @@ -710,8 +817,8 @@ nouveau_dmem_migrate_vma(struct nouveau_drm *drm, if (!args.dst) goto out_free_src; =20 - dma_addrs =3D kmalloc_array(max, sizeof(*dma_addrs), GFP_KERNEL); - if (!dma_addrs) + dma_info =3D kmalloc_array(max, sizeof(*dma_info), GFP_KERNEL); + if (!dma_info) goto out_free_dst; =20 pfns =3D nouveau_pfns_alloc(max); @@ -729,7 +836,7 @@ nouveau_dmem_migrate_vma(struct nouveau_drm *drm, goto out_free_pfns; =20 if (args.cpages) - nouveau_dmem_migrate_chunk(drm, svmm, &args, dma_addrs, + nouveau_dmem_migrate_chunk(drm, svmm, &args, dma_info, pfns); args.start =3D args.end; } @@ -738,7 +845,7 @@ nouveau_dmem_migrate_vma(struct nouveau_drm *drm, out_free_pfns: nouveau_pfns_free(pfns); out_free_dma: - kfree(dma_addrs); + kfree(dma_info); out_free_dst: kfree(args.dst); out_free_src: diff --git a/drivers/gpu/drm/nouveau/nouveau_svm.c b/drivers/gpu/drm/nouvea= u/nouveau_svm.c index 6fa387da0637..b8a3378154d5 100644 --- a/drivers/gpu/drm/nouveau/nouveau_svm.c +++ b/drivers/gpu/drm/nouveau/nouveau_svm.c @@ -921,12 +921,14 @@ nouveau_pfns_free(u64 *pfns) =20 void nouveau_pfns_map(struct nouveau_svmm *svmm, struct mm_struct *mm, - unsigned long addr, u64 *pfns, unsigned long npages) + unsigned long addr, u64 *pfns, unsigned long npages, + unsigned int page_shift) { struct nouveau_pfnmap_args *args =3D nouveau_pfns_to_args(pfns); =20 args->p.addr =3D addr; - args->p.size =3D npages << PAGE_SHIFT; + args->p.size =3D npages << page_shift; + args->p.page =3D page_shift; =20 mutex_lock(&svmm->mutex); =20 diff --git a/drivers/gpu/drm/nouveau/nouveau_svm.h b/drivers/gpu/drm/nouvea= u/nouveau_svm.h index e7d63d7f0c2d..3fd78662f17e 100644 --- a/drivers/gpu/drm/nouveau/nouveau_svm.h +++ b/drivers/gpu/drm/nouveau/nouveau_svm.h @@ -33,7 +33,8 @@ void nouveau_svmm_invalidate(struct nouveau_svmm *svmm, u= 64 start, u64 limit); u64 *nouveau_pfns_alloc(unsigned long npages); void nouveau_pfns_free(u64 *pfns); void nouveau_pfns_map(struct nouveau_svmm *svmm, struct mm_struct *mm, - unsigned long addr, u64 *pfns, unsigned long npages); + unsigned long addr, u64 *pfns, unsigned long npages, + unsigned int page_shift); #else /* IS_ENABLED(CONFIG_DRM_NOUVEAU_SVM) */ static inline void nouveau_svm_init(struct nouveau_drm *drm) {} static inline void nouveau_svm_fini(struct nouveau_drm *drm) {} --=20 2.50.1