From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from BN1PR04CU002.outbound.protection.outlook.com (mail-eastus2azon11010064.outbound.protection.outlook.com [52.101.56.64]) (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 0990A4749F0 for ; Thu, 27 Aug 2026 15:38:57 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=fail smtp.client-ip=52.101.56.64 ARC-Seal:i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787845139; cv=fail; b=ByvexL9TvOsfGtAiQMTer7lAgYIMYAEXnCeOGpgHyTH+RsvqFYUuBxIZq5xkOpTeOeubCAbirKHiaPdWx7gYlcqmM4X2/vbFj+wYeMSEjzTaJZFSSb1lLF9cwx0Ur8TjeJ6d+O4fAJ1NlAfZtNcL5gWGIVpE0ET6Eig3xavsnPQ= ARC-Message-Signature:i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787845139; c=relaxed/simple; bh=/cs0bNXvO59NGRpRYXJkqxZj4FSGEBvddo6M+44+hyA=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: Content-Type:MIME-Version; b=IZ68FhumqJgQyD6VA1/iPMSsvmI4Kp/NdNByuuWpqrGE4ZaoN0gqJrJPxOIY2HASZtO9z2y3K0TVYFrLZyOetGAv46HDeRtVBUNr0a3xKfuN0/Cdoa/iw5YcRAQbpeMzeCNxDqLQGF/TwPA03llqoCWCqGxF5cM5JyRKevpnWbA= 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=L6p32YtR; arc=fail smtp.client-ip=52.101.56.64 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="L6p32YtR" ARC-Seal: i=1; a=rsa-sha256; s=arcselector10001; d=microsoft.com; cv=none; b=vadX95kZVOrIL48SSlzPd34XDgoWuevmOM0kTRYujRsncervx9tnnlBjCnybsjcYtt48vuyZ7VDN2IvTAkDrDwKVbU4VR7gcrV/P40PpOFrmUZUqGrmHM6PStV/9msVe+vEo/D3Q46AP6GjoRaRwbhNMmWBY3TAfeCDWbDXUqNGnTIBls1BJ0kzrn5SMBB7ktb+MhgVrZZJ6Rv0vzyi6Mir5IeE/w8svl2QFKRTIqvzN8VyLHsfS6DMsfsyDuQMLSARHHdD2VfUI5TkeF7Us/jFBfyepZsbqIUdBRlPhcP/baldx9i+3+HKTG+m1QSncy1aLsTdZN8ceQ3Gc59nM2w== 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=GFatwpLJ3wuT3UO7t4Mt4XxYFCr/8qeij9gI8Q78dBE=; b=JzaHumqJOzm8fQPaEDsZ5p9ExKexNMpo9oE9SC4fv5VsdE9+PugFPgXSQVhauylnIm+vBe3v3dIybjQaGZhLQt1NvyD6iOIP8AziK/k4m4ORMk/hZd/PIK0y3wyDgPcqKpwJTo4GRntM8o8yO/RzjyAYQHc0bhum/ITPkLaYSflnD+AYWwd1Ol2aLFVyqlS3i65kR/I3EFmSMpZmo83/stz5/8sUM26Paa75VM9f/O4EqOgHYEEW335iTsAruYDTOJAUumTc+yo3qDbVliBZEMZPM1lP0C9ffgMY19Hpo+OGC63Xj3a7dOZPK35JTKM0JFcXnktJiFPgjxZ0nKfTIQ== 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=GFatwpLJ3wuT3UO7t4Mt4XxYFCr/8qeij9gI8Q78dBE=; b=L6p32YtRTzpKUfOqL/omHsMClt/3LmmO7xwR+DP39fNwY0xnXSDyWxZp7adYP3RlDlzkLwpVRlL9xLbPV66bpG7a/jPUblD5BnHmOpf54GbSbbJPC9MfIMd1CpLogiCtUgq8SGixASdqnsxbg2cjj6b/nbMeuH973Xhk0HqvsR0/4mB6pdnoANeiYzfEUoEuDJK7H1TatRtw/4Wm+1AdnADcGXkQJhwQWxjGynkNGRAd9FuswZYBnSL5xzWkUNXmZF6uMcmN907KM6iVAYa7DenVGraz6RD9YUPutn9dbZcOQVzIcx8qU43MmugaBMOu266T1eD9RHGGRIABnWk17w== 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 DS5PPF23E22D637.namprd12.prod.outlook.com (2603:10b6:f:fc00::647) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.21.360.10; Thu, 27 Aug 2026 15:38:50 +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 15.21.0360.008; Thu, 27 Aug 2026 15:38:47 +0000 From: Zi Yan To: "David Hildenbrand (Arm)" Cc: Kiryl Shutsemau , akpm@linux-foundation.org, ljs@kernel.org, hannes@cmpxchg.org, usama.arif@linux.dev, lance.yang@linux.dev, hughd@google.com, baolin.wang@linux.alibaba.com, baohua@kernel.org, liam@infradead.org, nico.pache@linux.dev, dev.jain@arm.com, ryan.roberts@arm.com, balbirs@nvidia.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org, "Kiryl Shutsemau (Meta)" Subject: Re: [PATCH 1/5] mm/huge_memory: do not touch frozen folios in deferred_split_isolate() Date: Thu, 27 Aug 2026 11:38:45 -0400 X-Mailer: MailMate (3.0r7024) Message-ID: <50FDE99F-3E71-47B8-BD88-05AD771C4BDF@nvidia.com> In-Reply-To: References: <20260826162101.1314941-1-kirill@shutemov.name> <20260826162101.1314941-2-kirill@shutemov.name> Content-Type: text/plain Content-Transfer-Encoding: quoted-printable X-MS-Reactions: disallow X-ClientProxiedBy: BN0PR03CA0008.namprd03.prod.outlook.com (2603:10b6:408:e6::13) 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_|DS5PPF23E22D637:EE_ X-MS-Office365-Filtering-Correlation-Id: 5e626c73-189d-44a0-2df8-08df0451490e X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0;ARA:13230040|7416014|376014|1800799024|366016|23010399003|10067099003|56012099006|5023799004|11063799006|4143699003|18002099003|22082099003; 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)(7416014)(376014)(1800799024)(366016)(23010399003)(10067099003)(56012099006)(5023799004)(11063799006)(4143699003)(18002099003)(22082099003);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?PfY+hJGVmvkW9mBQaVl9jxs+6XG1EqxW7ZoTCLKnMc4V1dj8pUrcbKk/YixF?= =?us-ascii?Q?eGDBOUyhvAaU9BU7e+10P5cQZVK+2EFyOG1JM6bfX3furXfU5WjMmosBskxb?= =?us-ascii?Q?aCKyn1u4f8IDQOgLBkP9cw+iz4IZDArwJxPQGIrGsYvtfHhdVt9khRse8N0l?= =?us-ascii?Q?qAu/tl137vdLbCMFJGfZx0Sfsav2YD6ZGRjM2WI5UKTPNZInpoVF4lMb1V86?= =?us-ascii?Q?7BBKTj4SGKC5NPmgJpj/GbEnIvcMlcDoAP5uJ99+2yKlJuwHJk3S97V/TYLV?= =?us-ascii?Q?gOVI2ycv71eI4iWdB4SE5eNmYqskdfbmi52U6D6kTiraiMaFuS1HdR+2nMwo?= =?us-ascii?Q?3NnqWRCxye5LPtNfRDkSQxk5sEv3fW7LgWJvWXK+QSDmbcJahRIdQLMLnKWu?= =?us-ascii?Q?Q/gkYenF+0Y4B+CJr+pTh8idREHH2ZlAswas+25zS8GcpyMRo67b2hLzub0q?= =?us-ascii?Q?LBGG3EW+59+mbfeWUb7pSftApLDaLExuqRvBjVTkjPJxeYps01VMLA9jrf/d?= =?us-ascii?Q?MYD3nl0go66RFyvYX8FCByQFtcdafk2mZw6JsVm3hl30zBYW0MTH9xwfuCQj?= =?us-ascii?Q?nzneKV/MaPccG+ftgULuE48LBSs0LrokSzd2CPkMO6fbfhg2nJxVxZ3BkeA7?= =?us-ascii?Q?ZYuPlPRUjdJXO8TQVO/w65lmcHU+ifgYJRick1vmnCVTwEKLY3rhr4fEX/PV?= =?us-ascii?Q?XWfGY4mjn4uAqoYnvuFyd9S5YpNSoM4kzpdZU7EgyWUfRZozXZiD/TQlrPbl?= =?us-ascii?Q?j6lejPuStxiwbB4qdh4DCDikb3S9jyxHxHp54XgymPnkNTdk9tdHTnoo21cg?= =?us-ascii?Q?u3m37x2uolip8meuOLOkjWGC06MlLxlgGnW8C4T78ePDhPO11g1rjE91AFd2?= =?us-ascii?Q?bnf2BcQFuIvV79msHPz/leZlygzs5eKasm2bW8D7MwikFAAaRczT+yjDB3ZQ?= =?us-ascii?Q?IO+QCssJb3lx2dojnaRxdRupBAnkLGMKJLE+jGSP++q0aj3HIY8niiJ+sHig?= =?us-ascii?Q?4QofVRkWnFGx0tmvtyVf+DXMNbpu0EepehneG2QlB4NaAyRUUPulPe4iO8lX?= =?us-ascii?Q?VpySCFwpTMuL+ndBnm8UPl09Ekn0uAd3z1X2T2rOr8BQvrH0ewPAEhQlQ1jM?= =?us-ascii?Q?mth+3wjC9Urq9Dn/ZBzgZExmqm1OhD360rpax5XIYRVxTaQi/cXV/z2+IPZP?= =?us-ascii?Q?vSJGlfcQU0wQ8DjuGPWuN9uAhL1q3gKMcHGQn4UJP1kx4LRDad63lZg0U1jM?= =?us-ascii?Q?Q3+VO6C00QFAhVkLz39uixQTODZtzGHLLmjNwnCybyZXkmqb4sliLHXYiTSc?= =?us-ascii?Q?meD3b0GdLRWXfzmGYTBOQBSIRbYpwpUs/s4/LTww/IkShTSWIh/+YSmrd5HW?= =?us-ascii?Q?HanM4dv3YSnayZqTIifHwDV7LkToXV8O9rAsK+dcWL3McHpLtG5h3937rB67?= =?us-ascii?Q?nGIykltDJI/YyWHI5kLVAfHCj/hlWC+2rUukZE1tx6RVeORo1oagvnhdVdut?= =?us-ascii?Q?lwWKoWoyPp2TeJivr+u5UNUn44Ce+E9rFvKy34xWIIguSvShXiC+G2VQC3xN?= =?us-ascii?Q?yDuqhAr1gsuzdHzUtkg3XScPvDGQTk2qh3ZuPfW5yZXW1kA1H+XFrnU6680r?= =?us-ascii?Q?xnDFj5IuVbic14+AGvZOZ+KgF/kPOeiVCZEIkEm8mdZVxBhg7wIBZlRDkkQ4?= =?us-ascii?Q?GaYFKm+Pr8GkHe/dJ5w69HZ0XNJ3eb2ItYIHv0eiMrL84aV02uMF2SEjier1?= =?us-ascii?Q?gfsz9uHpOQ=3D=3D?= X-OriginatorOrg: Nvidia.com X-MS-Exchange-CrossTenant-Network-Message-Id: 5e626c73-189d-44a0-2df8-08df0451490e X-MS-Exchange-CrossTenant-AuthSource: IA0PR12MB8374.namprd12.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 27 Aug 2026 15:38:47.5913 (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: cIEDQ8wtVGr+29P1ZpolY0r9A74RuLTuLP+YauvXgcY9draP219EXjvj06JJe5iT X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS5PPF23E22D637 On 27 Aug 2026, at 11:23, David Hildenbrand (Arm) wrote: > On 8/26/26 18:20, Kiryl Shutsemau wrote: >> From: "Kiryl Shutsemau (Meta)" >> >> deferred_split_isolate() probes each queued folio with folio_try_get()= =2E >> folio_try_get() failure is treated as a lost race with folio_put(): cl= ear >> PG_partially_mapped, correct MTHP_STAT_NR_ANON_PARTIALLY_MAPPED, take >> the folio off the queue. >> >> The folio_put() race is the most common case for !folio_try_get(), but= >> it is not the only option. Another scenario is folio_ref_freeze(). >> >> A zero refcount in such cases does not mean the folio is going away. = It >> means "don't touch me" and current deferred_split_isolate() doesn't >> respect it. It can lead to unqueueing folios from the deferred list fo= r >> no reason: > > Yes. > >> >> CPU 0 CPU 1 >> --------------------------- ------------------------------ >> freeze a mapped folio deferred_split_scan() >> folio_ref_freeze() folio_try_get() fails >> folio_clear_partially_mapped() >> NR_ANON_PARTIALLY_MAPPED-- >> folio off the queue >> give up, put it back >> folio_ref_unfreeze() >> >> The folio is still partially mapped, but it is no longer a split candi= date. >> Nothing queues it again until part of it is unmapped once more. >> >> Skip the folio instead: whoever freezes the folio, owns it and owner i= s >> responsible for its fate. It also covers the folio_put() case: >> __folio_put() unqueues the folio via folio_unqueue_deferred_split(). >> >> Reported-by: Lance Yang >> Link: https://lore.kernel.org/all/20260824131224.73344-1-lance.yang@li= nux.dev/ > > Fixes? > >> Assisted-by: Claude-Code:claude-opus-5 >> Signed-off-by: Kiryl Shutsemau (Meta) >> --- >> mm/huge_memory.c | 19 ++++--------------- >> 1 file changed, 4 insertions(+), 15 deletions(-) >> >> diff --git a/mm/huge_memory.c b/mm/huge_memory.c >> index ced400f72d43..6281ed993243 100644 >> --- a/mm/huge_memory.c >> +++ b/mm/huge_memory.c >> @@ -4590,22 +4590,11 @@ static enum lru_status deferred_split_isolate(= struct list_head *item, >> struct folio *folio =3D container_of(item, struct folio, _deferred_l= ist); >> struct list_head *freeable =3D cb_arg; >> >> - if (folio_try_get(folio)) { >> - list_lru_isolate_move(lru, item, freeable); >> - return LRU_REMOVED; >> - } >> + /* Lost race to folio_put() or the folio is under folio_ref_freeze()= */ >> + if (!folio_try_get(folio)) >> + return LRU_SKIP; >> >> - /* >> - * We lost race with folio_put(). Read folio state before the >> - * isolate: folio_unqueue_deferred_split() checks list_empty() >> - * locklessly, so once removed the folio can be freed any time. >> - */ >> - if (folio_test_partially_mapped(folio)) { >> - folio_clear_partially_mapped(folio); >> - mod_mthp_stat(folio_order(folio), >> - MTHP_STAT_NR_ANON_PARTIALLY_MAPPED, -1); >> - } >> - list_lru_isolate(lru, item); >> + list_lru_isolate_move(lru, item, freeable); >> return LRU_REMOVED; >> } >> > > Who will clean up the stats that we used to clean up? That should be me= ntioned > in the commit log, otherwise it looks like some piece of the puzzle is = missing. It is kinda implied in the last paragraph in the commit log (copied below= ), but I agree that stating it explicitly is better. |> Skip the folio instead: whoever freezes the folio, owns it and owner i= s |> responsible for its fate. It also covers the folio_put() case: |> __folio_put() unqueues the folio via folio_unqueue_deferred_split(). Best Regards, Yan, Zi