From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from BL2PR02CU003.outbound.protection.outlook.com (mail-eastusazon11011025.outbound.protection.outlook.com [52.101.52.25]) (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 955822DB7B8; Thu, 13 Aug 2026 02:14:52 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=fail smtp.client-ip=52.101.52.25 ARC-Seal:i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786587294; cv=fail; b=RofOGKrBHbYMrFS7t9jXsnc9Lia684qc+SdYuVCHToW+uhGglkrYGJ/T8jxTScB3PzbPRzyUasdJjBtZOm9I5FnKzN7dBB3IAU6gMgK+u23JX0SClFME8RxYslAwAU5Fmb0iA6gjAn4SY8N0oNCyO9EyUzkkk6PEYJAJeO9QBxw= ARC-Message-Signature:i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786587294; c=relaxed/simple; bh=8Kae5pocnkaT65y8TcMPt7RVKk1RQeozCi16mvbv6Ns=; h=Content-Type:Date:Message-Id:Subject:Cc:To:From:References: In-Reply-To:MIME-Version; b=nEvWohOQoaM+/UAkbIx3UIybe3RcJiygWx4c24f/sjDI9QDNdPqA1r8RkuqVpNoYR1VEMoIhZPOav05EjqrHukuravXp3dYQe4tWGaOHGZmAWzOvzU2RAB883kTFysUYqyXvzqqBlpjxUsOH3G7dv54/kgMEro93r28GwjtTRho= 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=bA7Fb6g4; arc=fail smtp.client-ip=52.101.52.25 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="bA7Fb6g4" ARC-Seal: i=1; a=rsa-sha256; s=arcselector10001; d=microsoft.com; cv=none; b=NtY0Sa5BOr/WkJtixQVQpK/LFodxVCZQ5bKhhfygPUe5Mxf4/dCRwP1e/kUwoouVWbjpknXIA7qdT8Uvbbx4HWrhKiiyrtuyx4/pl77NqcS+PapHEwIW9UjxvCUybSzCpoKuXWymdTJB9Ob6X9NrQcdzuOjA+UE/24jUzIfyj7mm3CXX4k9F8jNsiEOMrugyMTrX7/YzXH1tMnml0+LAlYYTCZZKSb08dS9IDdzQjvdyoJGLEDE7dLHfUg9MD5F13G2yJ8VuW2Vh/L2Np8dKhrCeIeXMqI97nIuXoNyjNPVe+ixi/RANCfgc/U8j0T/EWWUCbsIr9gmmwZkYBiU7cg== 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=9PyQB4wCg5aCi5g77Qi+BZ314qUF0gDklMTrqzEnfZU=; b=cCqxyOH3k7vpoU08Yd3ewFcBpVeoj7Vwbdq6XW35mCOkTD5vc820idPJkUOftunmbgrB7AQ2rgemLD0gjC/1uHWmXcm7mep9hKRNJUk0hVC3DseeN5CE/Kuv14scFg9ujcCsa74BHVp3qcX/mD5TUhg2KGMNOAzud3g73LDgFlKxAZRe1+6qZFAnoG6fTvWPnAC4mgIDN33Fmhim8/TUtyCEuQmPJkQqbQjRHRMAXBMo5qQJVwiEVXXxde9YwBprNPt6fKXDuSRTT9LEMCS3bVmExosG5m68BmgUmWfPwCnO4HALgrB/mignMJkVv1sLMXHvkUW4e2Obg9lCnQxOqQ== 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=9PyQB4wCg5aCi5g77Qi+BZ314qUF0gDklMTrqzEnfZU=; b=bA7Fb6g49bTblcxc5mGfFu13P1Plwi/gUjAHutOTENiuIB3FVxDyI0RUYKmUYkJbbX6z3pbEgcJIkzhRRa3RwGFyim+dU2tiyCmhHRXFuAfvdLDqVTleiM2Hv6id9jhMQWFTi5y+YZCQtkjm1hZproXG2qkcPlcJ50CjQRQvJWWxnHKOFnWaNddJJQgXZCCRZnBBwy5W/QcbazZ1ADysJXIjniiUsrPLBS4nf8afkjYjPYT2LAryyqsmrnsBEpcCIy2tjHShAAbyAXV7v6yJCemFKvrnQ5iuY8myR0taVAvgOn+HDN0IY3sUyt9qz/tXAwXUXyy05mwhfjS4lolC/w== Authentication-Results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=nvidia.com; Received: from IA0PR12MB8374.namprd12.prod.outlook.com (2603:10b6:208:40e::7) by IA1PR12MB6210.namprd12.prod.outlook.com (2603:10b6:208:3e6::14) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.21.315.12; Thu, 13 Aug 2026 02:14:47 +0000 Received: from IA0PR12MB8374.namprd12.prod.outlook.com ([fe80::d85f:4c87:ae84:3f16]) by IA0PR12MB8374.namprd12.prod.outlook.com ([fe80::d85f:4c87:ae84:3f16%5]) with mapi id 15.21.0315.014; Thu, 13 Aug 2026 02:14:47 +0000 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=UTF-8 Date: Wed, 12 Aug 2026 22:14:46 -0400 Message-Id: Subject: Re: [PATCH v4 5/7] mm/khugepaged: Refactor the PTE state checks into a helper Cc: "Andrew Morton" , "Lorenzo Stoakes" , "Baolin Wang" , "Liam R. Howlett" , "Ryan Roberts" , "Dev Jain" , "Barry Song" , "Lance Yang" , "Usama Arif" , "Vlastimil Babka" , "Mike Rapoport" , "Suren Baghdasaryan" , "Michal Hocko" , "Jonathan Corbet" , "Shuah Khan" To: "David Hildenbrand (Arm)" , "Nico Pache (Red Hat)" , , , From: "Zi Yan" X-Mailer: aerc 0.21.0 References: <20260811-khugepaged_pte_refactor-v4-0-ddac39d61c4a@linux.dev> <20260811-khugepaged_pte_refactor-v4-5-ddac39d61c4a@linux.dev> <753cb38f-66b4-4bd0-8b53-7480efe30ec3@kernel.org> In-Reply-To: <753cb38f-66b4-4bd0-8b53-7480efe30ec3@kernel.org> X-ClientProxiedBy: MN2PR20CA0060.namprd20.prod.outlook.com (2603:10b6:208:235::29) To IA0PR12MB8374.namprd12.prod.outlook.com (2603:10b6:208:40e::7) 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: IA0PR12MB8374:EE_|IA1PR12MB6210:EE_ X-MS-Office365-Filtering-Correlation-Id: cf5022a7-6f18-46e1-2a08-08def8e0a5f2 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0;ARA:13230040|366016|1800799024|23010399003|376014|7416014|18002099003|22082099003|56012099006|4143699003|11063799006|6133799003|10067099003; X-Microsoft-Antispam-Message-Info: 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 X-Forefront-Antispam-Report: CIP:255.255.255.255;CTRY:;LANG:en;SCL:1;SRV:;IPV:NLI;SFV:NSPM;H:IA0PR12MB8374.namprd12.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230040)(366016)(1800799024)(23010399003)(376014)(7416014)(18002099003)(22082099003)(56012099006)(4143699003)(11063799006)(6133799003)(10067099003);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?utf-8?B?clhvVG1VNDgwZXkvTHcwbExvUzJrTjh3MDlITnRHSlloQjZ0TG9JWHkzWkRF?= =?utf-8?B?Wm05bklMdStYcnA3WWJ5UFovaEk3RWhUZGlDa1FTUmYvR2dFR1RxN1R3U1Zi?= =?utf-8?B?RzByN3FKWjBlOVYxZUZwQ0haSC9iNnRlN3pEdWFyNGx1Qy8rZU80VGFOODN0?= =?utf-8?B?S1FiczNXdDFpT1o1OEVFSVZwSkJoU2pXa2hlZmZYcjBJN0diZ2E2Y0tiNFRW?= =?utf-8?B?MkZaY1VYMmR3ZzlWR3YzY0NzaWdOMHZIZ3NLN1RCclBhbS90TTlWaWJaWUFW?= =?utf-8?B?Tlc3WWdYTFBaM3FOZTJ2SkdsdlBCZzhLUlFyTjRDTXprV2hXY1ZPL0RvZmhQ?= =?utf-8?B?SHhXNEo1c3V6eWJFWmRENTZ6cGtrdkJpRGJyQzFPckoyRG44VGNma256Ymdw?= =?utf-8?B?S0xnMFdaWmcrcnY2UGJ5NDNPbGF1a3dXMmFJVnp0M0lod0pSdE4zck9XUFRL?= =?utf-8?B?NDZoY0hrNHBZVU1hcHovaG5uNzZGd3h6c2liM04vSGhEcFNmdkRCQkhBcUhZ?= =?utf-8?B?Vy92dENFWmxoVGp1aEQxckx0Qy9FMHZSandFNTM1NHQ4SlAwWUdOQ0p6Z1F6?= =?utf-8?B?c21rUWZWRHNHS3M2SmduTmVmSWgrdStGWUVtT1JlQkRjZ2JXTUo1dk5tTnFF?= =?utf-8?B?em9QY1BiZm1FVDgyTEM2V0g0WTdsUEQ5ZzlvVWxHQzdBZENiRllJbS9xRVlN?= =?utf-8?B?ZExGSVBVbkRVaVFvZGpGci9OYjQ1RGZpQWwra2M0NDQ0RElUNTBKNXhQMGlQ?= =?utf-8?B?ZCtJZHlzMmJTc2RBbENTQTY2dkJ4cCtUSHBVRVFuVEdnM2podnhqK0dKR3o2?= =?utf-8?B?N3Rhcy9FNDk4YmxGOHU2ZXBNTVlKclpTd2RvSnYrVXhHMGVtQ0JXYlJrNkVF?= =?utf-8?B?emlvYmdXMnZNOG1rdVBtQXg4eUxDWFhpRXQ3OGFUdHp6UHNMaDB4RVVBN3Zr?= =?utf-8?B?a3FxWWpGNmsyaEtnQXlwZkhKM3VteWZLOWJqSUd3TWpwOXlzb1VzUzNJRVNT?= =?utf-8?B?bVhzMjVFU01BQVJCNC80MENsVXJzVlV5UGRkT0svNG5OM0VBdGYxSUZEUDFW?= =?utf-8?B?cEcxM05UTWNyYzgzNVBTT2dVT2hma1R0K0MyN1VucWJQY3Y3dWtidDU5OTZ2?= =?utf-8?B?YUFtdHFYMlZNdkVJdzBvdXl3Q1RnRHR1WlNyRzVhUkZhTWJGQzJRUnpVVlZF?= =?utf-8?B?WlF3YnNub0wrcGY2UE9meDdMQWhjbHVNMm1hSnRSeFRTajhMeCtCU3VRTXFp?= =?utf-8?B?M0NwbktuZGZlNTBqOGx3aTRwd1pqK1QvenB1dUdEYXYrUFpYVlhDamtkL25Z?= =?utf-8?B?ZDlNYkg4QVY5SHZTbDE2cTFGSFJLM3ZXZWNmVG5HZFV2Ymc2MVIzL0RFYVJ0?= =?utf-8?B?aGhQQitqb2hPSFI0VjJhbXI3V2JqRUovZHNqR0UrM2R1eHZRVjVIUHZzeUpM?= =?utf-8?B?T21CSk93UjhPTW4wZjFjV2NVTk1seEFoQmZGNFlkM3FLdmdWVUwxT2FjNVl6?= =?utf-8?B?cnlBeUI0RC9CZGZPN05xdm1sblpUL3lobW8wWFdCRlEzRjlsSUlmbFJEd2M1?= =?utf-8?B?QTZtNWRSUlY0dnRSRDZLVERZUkl3cXFOc25ZcUluREVQZmFkSk5HNG81RlJS?= =?utf-8?B?dVRzbUM3U245ZXEzK3VIS2VGRHlPbkU1cmRXU0QrdUJlM09kTVAzRk1hd01x?= =?utf-8?B?ZjdYY3ltWEFWVGpKMC9pajcrMC9IblJEWlZud2cycW8xM3FxK3U0VWx2azJH?= =?utf-8?B?UW8vd0dQVHhlcXZ6T3NJRjVNVGYxRm5QemI3OGtGeEFKcGNOZ1JzNFlZRTdC?= =?utf-8?B?UHNUdytia1VtOUxmdDJLNXpuRGZQeEJDZ3VPWE5HUVFxSS9UR1JDbzdGakRP?= =?utf-8?B?NXVNWHpGREJaa3J4ci9HMy9pcVZXVTRkODdhRGNoU3hESUt4R0oxL0dtNzRK?= =?utf-8?B?eUM0WlRiVmVnUXltUGJQL3VhdndkVHZkNDJub1ZWMU1Jcmw5KzZJbXJTaWZj?= =?utf-8?B?SnB3ekRVMGVId3ZjdGhZWXc1WnJ2VjFFYTY5elJWdDZ3SStoeHhOSmxaWGox?= =?utf-8?B?SWVlZk13azlIQzRuTlJVaHFiekR5cndBWHBHREVuZENqaCszekhld3NnTlVR?= =?utf-8?B?VzdtckZkNmRnN1Q4SGtKSlU1VWljVUJsaGJ1MDBGbWRKT2NZTExoSmpaNU4v?= =?utf-8?B?VEpzWmZLVWpHRWEvRUYvNzBKVVFIT2w0NktIb0o1T3BKN1c1U0FwQkpRbEll?= =?utf-8?B?NDNkNkl6TWNGMGJUaTk2dkpSNTJxcDQ2Z0p6d2dYV3pTSDFLQi9kY0s0Sndu?= =?utf-8?Q?86JBDZR9AKzS5B9P0W?= X-OriginatorOrg: Nvidia.com X-MS-Exchange-CrossTenant-Network-Message-Id: cf5022a7-6f18-46e1-2a08-08def8e0a5f2 X-MS-Exchange-CrossTenant-AuthSource: IA0PR12MB8374.namprd12.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 13 Aug 2026 02:14:47.5243 (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: 9GLLQvi4mGlgb8WvYRpNaQvuAkRvsaJ95It/gQArYf4ER/RbI6gnLRKsN/KbYn7M X-MS-Exchange-Transport-CrossTenantHeadersStamped: IA1PR12MB6210 On Wed Aug 12, 2026 at 5:51 AM EDT, David Hildenbrand (Arm) wrote: >> Huh, that looks odd. >>=20 >> That should just be a VM_WARN_ON_FOLIO(true, ..) or sth like that. >>=20 >> But in collapse_scan_pmd() that warning never existed? So this raises ey= ebrows. >>=20 >> [...] >>=20 >> I'll play with it to see if we can do better and will reply here later. > > > Okay, I think below is what we should do. > > There is one behavioral change: we now longer trace the last folio, which= is the right > thing to do IMHO. > > And I think there is one fix we should pull out and evaluate first: A zer= opage with a > uffd-wp marker is not checked properly, IIUC. > > > diff --git a/mm/khugepaged.c b/mm/khugepaged.c > index 5a06e3942e889..8a223659fce2d 100644 > --- a/mm/khugepaged.c > +++ b/mm/khugepaged.c > @@ -119,6 +119,18 @@ struct collapse_control { > DECLARE_BITMAP(mthp_present_ptes, MAX_PTRS_PER_PTE); > }; > =20 > +struct collapse_anon_pte_check_ctx { > + struct collapse_control *cc; > + struct vm_area_struct *vma; > + unsigned int order; > + int none_or_zero_ptes; > + int nonpresent_ptes; > + int shared_ptes; > + unsigned int max_ptes_none; > + int max_ptes_swap; > + unsigned int max_ptes_shared; > +}; > + > /** > * struct khugepaged_scan - cursor for scanning > * @mm_head: the head of the mm list to scan > @@ -696,74 +708,104 @@ static void count_collapse_event(unsigned int orde= r, enum vm_event_item vm_event > count_mthp_stat(order, mthp_event); > } > =20 > +static enum scan_result collapse_anon_pte_check(pte_t pteval, > + unsigned long addr, struct collapse_anon_pte_check_ctx *ctx, > + struct folio **foliop) > +{ > + *foliop =3D NULL; > + > + /* > + * Don't collapse if any of the small PTEs are armed with uffd > + * write protection. Marking the new huge pmd as write protected > + * could bring userfault messages that fall outside of the > + * registered range. > + */ > + if ((pte_present(pteval) && pte_uffd(pteval)) || > + (!pte_present(pteval) && pte_swp_uffd_any(pteval))) > + return SCAN_PTE_UFFD; > + > + if (pte_none_or_zero(pteval)) { > + if (++ctx->none_or_zero_ptes > ctx->max_ptes_none) { > + count_collapse_event(ctx->order, THP_SCAN_EXCEED_NONE_PTE, > + MTHP_STAT_COLLAPSE_EXCEED_NONE); > + return SCAN_EXCEED_NONE_PTE; > + } > + return SCAN_SUCCEED; > + } > + if (!pte_present(pteval)) { > + if (ctx->max_ptes_swap < 0) > + return SCAN_PTE_NON_PRESENT; > + if (++ctx->nonpresent_ptes > ctx->max_ptes_swap) { So max_ptes_swap is actually max_ptes_nonpresent. But due to khugepaged's max_ptes_swap config name, we just keep the variable and related function names that way? > + count_collapse_event(ctx->order, THP_SCAN_EXCEED_SWAP_PTE, > + MTHP_STAT_COLLAPSE_EXCEED_SWAP); > + return SCAN_EXCEED_SWAP_PTE; > + } > + return SCAN_SUCCEED; > + } > + > + *foliop =3D vm_normal_folio(ctx->vma, addr, pteval); > + if (unlikely(!*foliop) || unlikely(folio_is_zone_device(*foliop))) > + return SCAN_PAGE_NULL; > + > + /* > + * If the vma has the VM_DROPPABLE flag, the collapse will > + * preserve the lazyfree property without needing to skip. > + */ > + if (ctx->cc->is_khugepaged && !(ctx->vma->vm_flags & VM_DROPPABLE) && > + folio_test_lazyfree(*foliop) && !pte_dirty(pteval)) > + return SCAN_PAGE_LAZYFREE; > + > + if (!folio_test_anon(*foliop)) > + return SCAN_PAGE_ANON; > + > + if (folio_maybe_mapped_shared(*foliop)) { > + /* > + * TODO: Support shared pages without leading to further > + * mTHP collapses. Currently bringing in new pages via > + * shared may cause a future higher order collapse on a > + * rescan of the same range. > + */ > + if (++ctx->shared_ptes > ctx->max_ptes_shared) { > + count_collapse_event(ctx->order, THP_SCAN_EXCEED_SHARED_PTE, > + MTHP_STAT_COLLAPSE_EXCEED_SHARED); > + return SCAN_EXCEED_SHARED_PTE; > + } > + } > + > + return SCAN_SUCCEED; > +} > + > static enum scan_result __collapse_huge_page_isolate(struct vm_area_stru= ct *vma, > unsigned long start_addr, pte_t *pte, struct collapse_control *cc, > unsigned int order, struct list_head *compound_pagelist) > { > + result =3D collapse_anon_pte_check(pteval, addr, &ctx, &folio); > + if (result !=3D SCAN_SUCCEED) { > + VM_WARN_ON_ONCE(result =3D=3D SCAN_PAGE_ANON); > goto out; > } > + if (!folio) > + continue; SCAN_SUCCEED + folio !=3D NULL means to proceed with the pte, while SCAN_SUCCEED + folio =3D=3D NULL means to skip the pte. collapse_anon_pte_check() probably needs to document this? Yes, it is straightforward at the call sites, since without a folio the following code cannot be executed. Or you think that is self-documented. --=20 Best Regards, Yan, Zi