From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from SA9PR02CU001.outbound.protection.outlook.com (mail-southcentralusazon11013055.outbound.protection.outlook.com [40.93.196.55]) (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 8F8C925785D for ; Tue, 16 Dec 2025 20:23:35 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=fail smtp.client-ip=40.93.196.55 ARC-Seal:i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1765916617; cv=fail; b=kBBDJimnym38ECz5pIZ1JjwYx5oj183CBFZ7nAJe8JBGKBq2Rgr7s6GTePfw7GNb0P3u9yBdl5SKkE75KN9FYgwU0OSNE1jazcUBZvyZkjnYOchq+zMv5cbQilgJKnFgLkoO9/JpNWo7fXyJsRRVjdcVA/tDvCbAcjLJdROMmEU= ARC-Message-Signature:i=2; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1765916617; c=relaxed/simple; bh=sT1PpC8jYnaK3msSXu/rNfEItc5W+jLoMXOWk3gnGQw=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: Content-Type:MIME-Version; b=FTwKR3EuDOxdq7SJ3B93vH2j0OAVf+s/IsgHCPKxEw3cx9SsuOiH/Lxq9KAqUJX03JYYoV7dnSWF1B5xfpIyUmhkuhfOkhz62b+lz21Nw7Ot6FfQZi6RIEzSDx9vqBxYmvTI3vR1rmVaBXhV6zDhoePeXD2sykMoqx+almkGSmY= 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=lDJl+h8F; arc=fail smtp.client-ip=40.93.196.55 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="lDJl+h8F" ARC-Seal: i=1; a=rsa-sha256; s=arcselector10001; d=microsoft.com; cv=none; b=oy03olduKqHv4eHILNmkdeKB9Tkdyz2XTiHStoJ8CVsU/Z6VI/si58ceQDfJgO1d3m41XahulvGGgPIs3jvKfxgi6s6O/0ENEloueAMYc6sAyrysvFE8RFwLuLTVBljK7pVDyNcx40FsXorP16iYTdy2C2+MxWlUwHDhqmQSXWfbG2ml3s3n2XD5R0UYhuKS+tFZT4E860DlkUpphN0sFaFK5YkrmseWeakghhLAOheeyABKRU98xz3lcAw8EBxKUqbsl+GdF/M/t0dVfc5NMf+p8eJ8FlJmSzspPJxpiHEpFEpiKeiLCU9wXvK5rhvCGjKRcldw9trIi7HBbwgZ+g== 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=RyEPIhSAEnPOBlO/16l3yr+VcwLidvnTP2ptqmdTGg4=; b=LeKjFheLC1EUzSes3VQuwLIy2u+8A7wYqc8fPnvNPWgdAGgJC9PY/oSU6XZ79ckpajZMht5AuhDazFvs/pg8HmJh0SJQ5qmYja/NjPm6AEz22uRBko25LMxCycu9sksHOb3llyzaD7FdalynllkqX3qCuORNKd6Qezz+DOXCartuxnzO7n1NZ8+xBId89+Z0TJxLC19h9Q8nELv/wVyAggOYZee9kCCYSCJ5VKAmctj9xVvuvDE1mEWuXYJdqcw2yUuxTBTR5D2cTToP4norhit4Mjn2SpRxRTAxe6fhluxYeaLwCDBXtJFKqCSjeO+7tURHne8QV4yWunZgTPDsCw== 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=RyEPIhSAEnPOBlO/16l3yr+VcwLidvnTP2ptqmdTGg4=; b=lDJl+h8FxFm3hWqlRNWXOTfD4FKqUuminGZjQRVLECztzHkBj5pE5c9JfMQZd/gxJSLzBVfFns7y2RagMhJbJoruoEoJTveeCWetbYcwYwAbGcSK8sgt7LHtpyPjPnwCiP48If3AJLkqHv9EujvpOiR2AuC9ha8dOi329MNVvq4Uho20YlmEFR1Lcoh1rJekmPaU062tTG/JP2hNNidYFG80v3L4hsn+rWvhPhwhQlUgr6erFM0/IXDHKWexW2RVJPe1FZBc2n7dUS3Papw1K+rvmw/Q0OkupR8wAdBp/oRgD/isExgqYL3njor0YTFAlYpJDe5i+pnHnKEjM/6MbA== Authentication-Results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=nvidia.com; Received: from DS7PR12MB9473.namprd12.prod.outlook.com (2603:10b6:8:252::5) by MW3PR12MB4458.namprd12.prod.outlook.com (2603:10b6:303:5d::10) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.9412.13; Tue, 16 Dec 2025 20:23:30 +0000 Received: from DS7PR12MB9473.namprd12.prod.outlook.com ([fe80::5189:ecec:d84a:133a]) by DS7PR12MB9473.namprd12.prod.outlook.com ([fe80::5189:ecec:d84a:133a%5]) with mapi id 15.20.9412.011; Tue, 16 Dec 2025 20:23:30 +0000 From: Zi Yan To: Vlastimil Babka Cc: Andrew Morton , Suren Baghdasaryan , Michal Hocko , Brendan Jackman , Johannes Weiner , David Rientjes , David Hildenbrand , Lorenzo Stoakes , "Liam R. Howlett" , Mike Rapoport , Joshua Hahn , Pedro Falcato , linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH RFC 1/2] mm, page_alloc, thp: prevent reclaim for __GFP_THISNODE THP allocations Date: Tue, 16 Dec 2025 15:23:27 -0500 X-Mailer: MailMate (2.0r6290) Message-ID: <41F074C5-DB6E-4D44-859C-AE366EA068B7@nvidia.com> In-Reply-To: <20251216-thp-thisnode-tweak-v1-1-0e499d13d2eb@suse.cz> References: <20251216-thp-thisnode-tweak-v1-0-0e499d13d2eb@suse.cz> <20251216-thp-thisnode-tweak-v1-1-0e499d13d2eb@suse.cz> Content-Type: text/plain X-ClientProxiedBy: BLAPR03CA0002.namprd03.prod.outlook.com (2603:10b6:208:32b::7) To DS7PR12MB9473.namprd12.prod.outlook.com (2603:10b6:8:252::5) 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: DS7PR12MB9473:EE_|MW3PR12MB4458:EE_ X-MS-Office365-Filtering-Correlation-Id: ea6b32d9-e9f3-4c53-89ab-08de3ce0f9f9 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0;ARA:13230040|366016|7416014|376014|1800799024|7053199007; X-Microsoft-Antispam-Message-Info: =?us-ascii?Q?BQ5sz9xnLYPQ5FPi7ruO9xVfKs+Izgud8XCU1GgwddkhWW+sHSloufhHXVTH?= =?us-ascii?Q?ryjYa/6F6cb9x89MU7wmg7IyaUc8DmVajUeM2QPQu0v1Y5uEGlTi04jZDWpG?= =?us-ascii?Q?D0cSybDRSwO2wzZ5Dei5tbaUj1AqLIotS+QIJN4c4vkFP/sXNeHAoCXL63cN?= =?us-ascii?Q?+rf4QgjrT93OcvPMlX8Fg6qOjxtmE6tDGYRYn9fis62lESQWXwn/m5gmTvJ2?= =?us-ascii?Q?eTRsU8I5y3egqeivNFCmA78avotCKp7dXNqRxq9nh5fQysLJX+z6Z57KzKiR?= =?us-ascii?Q?8b7ZUvULju8DcUoC0TaI4D6Dt8VovMMdc/Zf4Uug5sc2yqmBV4pgELBVtTQt?= =?us-ascii?Q?E+dDIx6UzfWGSJC9DUkjOqCydx7MMhGL9i1miUqxa+mcRRRjPyLxUB9E0mVD?= =?us-ascii?Q?DCjd607RaomBQ77aWFIfGFp05KSV+XtENM0T0UIot9nblgr5h/+dzvpmgKyg?= =?us-ascii?Q?ucj3uL4HVRP1C/6CUjznjMJv9rEiGf4D7CSu9bE7hKmsfcA8VhNGLlfPCc/4?= =?us-ascii?Q?EwLeTdoiiadyyR+/gXBW+9B0MjNZ9XV8y8OwGDEbVRcLWCh1e/3k1G8ngvri?= =?us-ascii?Q?Tk3revYpFToF/FQm0xxPvyFd8ShzWIZ1zanQNjpumSC8jCNdCEne+9p9jOBK?= =?us-ascii?Q?R22GGQhIpjPUMwBnxBPuaMjpUra+quMDnNBdB972Z4eff6GGJvraxrSgOqCT?= =?us-ascii?Q?yN37PEe+nImz2IBSjBQntN+7xJxV24/kuFAVieRBh8NTKbVsYhZjMczaB7hm?= =?us-ascii?Q?cMzM0ZcczfnuxaW6q5RsnqW8w4GVp/vpJl24EOrp7pAmsJ/6r6svlmfE/IPz?= =?us-ascii?Q?IiMdmaGySpL+f9seIQ8UllQ2Z1bIdEFezcEqRPRDqI3GRXQNp22s7DO8I2pH?= =?us-ascii?Q?aFSOjYcITzJdN84L6bVEQofC4xu6m9Bi6GxLnG5K8N8/kVzKxEhwj93b4kyy?= =?us-ascii?Q?26IBHLPt7N2o+b6WhQ7dDLc5vf6GDwFAhmOWAiX0uCosP6TpncRZo8m8jzHg?= =?us-ascii?Q?w8DgEikLvJyYausZ2mEMc8/JuVgdVlGx8Qs4Rn3MAs+Ilksyc76B6KiPlxA+?= =?us-ascii?Q?96hxVvAJXceyWrWMTHIADDYHk3yjmAvF/WZyu9x5K6cE6FG28RgMILaounk7?= =?us-ascii?Q?r65R6uvMYPHoBRNhzH1p9LHuvnw+/qV4/Dmy2XkGQc5pYaXku97eUJnfyE4x?= =?us-ascii?Q?fTzdmIa9DGqU/AWR7XHHJMvAvXxQaAyhOD6/NXeCFEp4FDt9otugSPqkS+iR?= =?us-ascii?Q?ttZcDcYVBQXRFfXplwmmJhmruZqWxMa5vVBETmZ7mpg5nV+TyEQ9Mr+P+WBJ?= =?us-ascii?Q?mc5OtCgF/QA+bz+/W72sWnocOQOLoQocssXidU+V1VWbs1OhBGEuZaLGNp4P?= =?us-ascii?Q?H94jq3T0LmDVwMiCGnZ4S9zZQ23cLjsR+zjj9vsTQAbDrKMzANNJCic/b51V?= =?us-ascii?Q?OTsUv/CpXW3t/6rSTOGEkCzKbFPclZcL?= X-Forefront-Antispam-Report: CIP:255.255.255.255;CTRY:;LANG:en;SCL:1;SRV:;IPV:NLI;SFV:NSPM;H:DS7PR12MB9473.namprd12.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230040)(366016)(7416014)(376014)(1800799024)(7053199007);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?cz8GxCDn9G22Vl1pZ2F+AblL44vADqVjIlcOfZ5hwUQ6S5mvBKWktbWtdtjY?= =?us-ascii?Q?23AuNRkKHSroiypfBMHKBHvfpuDG0j30kbNleV0KVQ1BBPmKm/IddIcN3a3K?= =?us-ascii?Q?57wgbe0J2THBA45pElA6dtIlwH2E4kjdz5ChgV4ktr9H5xbahKGNmcm/AmP/?= =?us-ascii?Q?+EgHOHmro0kEqwo5ITKJ9qm/6bnfB75jdRY0S9He5FchhbsBRYvwqoggJbA2?= =?us-ascii?Q?alcDaaDgj09BUh8q/GlC6cbe9OEhHVngymdsfX1GeJTV0+F2ZjDfYJ4SwAkU?= =?us-ascii?Q?QfSGKMVZpViz1L4l1W2EbYZtXr7UyiKFqjPbf4muSV8zw5FU/j6YgDb3gldX?= =?us-ascii?Q?P1t9PnlDy+jsNbbfe3cwLD2Atarjb7vLo4kuvZns90LXtkVlDaYIr0fR/5/J?= =?us-ascii?Q?sPF/fMWbLnNoz9hUgCkiccy2nlNr5/Il1X3pSq1mTtOM+hAxQyrtH1YZ+eGq?= =?us-ascii?Q?MGkpsMzVszKmWF/i09iD5G0oBI0jXrW4Cgy42L8fHo3sVCynw6UTYYDv9oC6?= =?us-ascii?Q?loPoiruwyPJI/lyczUwZvOK5tnCTBVvGL9j1oaxdkPOECTojIYZD+yExYqCL?= =?us-ascii?Q?F5FJhldxWLvqLo5IamPS7kB+uDHNm/Qn7PtRKlEugg807zeaCZVrk09u3xbC?= =?us-ascii?Q?dcS61leK6jVTnnT4d1t8OVFBh1rWSBGOJeLJL8aUQLHbSAQ3QbvJGaqQm69d?= =?us-ascii?Q?vwfhgIh3Rv+AOVNVRm0iWVa+QOop00yCybtUN+dyr2cjIY5XtGFB7YwoouIV?= =?us-ascii?Q?RmQ5RhSlypuFw5w+9j5XVGtAy5K3Io6sGSY3TOUf0jK//cUT6+cr1Jcnk1HN?= =?us-ascii?Q?0B4tfTDeWF7hLctGT5WqnxLYyTq9C7tkc4hy6CswXJbe6yI/ndi7o40KTZqU?= =?us-ascii?Q?+oequVIXpi1lRgKar7qeA7EZVka5ruNUqIyctbQBgcatV4jSOEUYjS1FLHaL?= =?us-ascii?Q?RVKshy8qFZjv2T6LnhTV6dUxmdDw/dwmVCqHXryrOlU1gKmti3C6k/mNg1Ch?= =?us-ascii?Q?EdnEoxH6BMeMs1CwvtYTQNUKVFc6s9VrJG9boI0YCmmw0rrfuaI+gfygKqbE?= =?us-ascii?Q?geI/9AN16QKyTdT3eUFiUg7RpVSorySZQDgfH8rCTafd6T1sq/m88y0x0ZJM?= =?us-ascii?Q?QduqUaYduR8hbqY1KxrHkBPxsQHHK984O3Js0we3aoavgz7Tw0xCMc4s/C14?= =?us-ascii?Q?q72/DMxh9tNAx1YaYWxN2cDXKWp99uxay1maC5gGsvgKaa+D4StRX2+hML0H?= =?us-ascii?Q?RfJhejvgAGTeGQVlNfBypsZWSvAEMk1En8HzJ7z1Kpr4qTd9SIM50BzOEAiI?= =?us-ascii?Q?bR6arcVYWpLZJOT3wq7T/QRpE/GqbL5sTO3u3h3KxBj8nUc8c5cnSwWp4Ztq?= =?us-ascii?Q?fK27SVERPTdtXQZdMP6MjSjvXcgy20Wcco1oYXApWVBjNisOjT6ZiEo0y1Xt?= =?us-ascii?Q?wv9uJqSGxkCHC1yTQJTCENbHTBxIruBjpI+fhScEzPFvtNbRc1W6ude69FvO?= =?us-ascii?Q?k+mFCvhgvAOMJiOyIq4CmIucAgKQ5NMc0W30BdOAtKhUNPgv79bTI207Xw4J?= =?us-ascii?Q?eQVGD3nO9hrzT9M60z95hHSl1NrjNhZkE9Qbz8QP?= X-OriginatorOrg: Nvidia.com X-MS-Exchange-CrossTenant-Network-Message-Id: ea6b32d9-e9f3-4c53-89ab-08de3ce0f9f9 X-MS-Exchange-CrossTenant-AuthSource: DS7PR12MB9473.namprd12.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 16 Dec 2025 20:23:29.9167 (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: UAybyD9qWCPca1xslb/vMxU3GBkn70ua5gOagEpMdXdbFkuCP8NNIfnNC4+TiwHf X-MS-Exchange-Transport-CrossTenantHeadersStamped: MW3PR12MB4458 On 16 Dec 2025, at 10:54, Vlastimil Babka wrote: > Since commit cc638f329ef6 ("mm, thp: tweak reclaim/compaction effort of > local-only and all-node allocations"), THP page fault allocations have > settled on the following scheme (from the commit log): > > 1. local node only THP allocation with no reclaim, just compaction. > 2. for madvised VMA's or when synchronous compaction is enabled always - THP > allocation from any node with effort determined by global defrag setting > and VMA madvise > 3. fallback to base pages on any node > > Recent customer reports however revealed we have a gap in step 1 above. > What we have seen is excessive reclaim due to THP page faults on a NUMA > node that's close to its high watermark, while other nodes have plenty > of free memory. > > The problem with step 1 is that it promises no reclaim after the > compaction attempt, however reclaim is only avoided for certain > compaction outcomes (deferred, or skipped due to insufficient free base > pages), and not e.g. when compaction is actually performed but fails (we > did see compact_fail vmstat counter increasing). > > THP page faults can therefore exhibit a zone_reclaim_mode-like behavior, > which is not the intention. > > Thus add a check for __GFP_THISNODE that corresponds to this exact > situation and prevents continuing with reclaim/compaction once the > initial compaction attempt isn't successful in allocating the page. > > Note that commit cc638f329ef6 has not introduced this over-reclaim > possibility; it appears to exist in some form since commit 2f0799a0ffc0 > ("mm, thp: restore node-local hugepage allocations"). Followup commits > b39d0ee2632d ("mm, page_alloc: avoid expensive reclaim when compaction > may not succeed") and cc638f329ef6 have moved in the right direction, > but left the abovementioned gap. > > Fixes: 2f0799a0ffc0 ("mm, thp: restore node-local hugepage allocations") > Signed-off-by: Vlastimil Babka > --- > mm/page_alloc.c | 14 ++++++++++++++ > 1 file changed, 14 insertions(+) > LGTM. Acked-by: Zi Yan Best Regards, Yan, Zi