From nobody Fri Feb 13 12:45:13 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 072F3CE7A94 for ; Mon, 25 Sep 2023 00:41:06 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230285AbjIYAlJ (ORCPT ); Sun, 24 Sep 2023 20:41:09 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59156 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230435AbjIYAlD (ORCPT ); Sun, 24 Sep 2023 20:41:03 -0400 Received: from mx0b-00069f02.pphosted.com (mx0b-00069f02.pphosted.com [205.220.177.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 0EDC2EE for ; Sun, 24 Sep 2023 17:40:55 -0700 (PDT) Received: from pps.filterd (m0246632.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMkBQE019937; Mon, 25 Sep 2023 00:40:14 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=E/x7mTYtq0AK43XWU8sgezCXfY3I6f53AD910fUr/7c=; b=bl59Dt8bk7PR84Rbqcv/sdCrimgMbMy3nWo8M72IV6ARW7ipkqtYVxlwjwtWyjzdrAvm GvaXAyCcwRnUqA3E2eIipSCwVeGIl1Zw1Swl+KWT6KO6G1ZLrxIMQ3gOT+Az9tU6JLjb xeeUhzeixQwSUo9Txkp0Bk6hb7NujkbyYfV0hnccwl6Nolyyct32p/Pa7kO+J0obqa5P QUOE0M5dwClVpcwDCZRhVImzMu38jsP6RoPQV9KlIqhC0U5K0qS0L+41mNVqj5V8FL77 9ebOSmmYffZcEY8Iy6kXSOndOemhGdAVkmCDI+ELj0j4SNMxq4HRJNY+KHJI0MQEAlZG 5w== Received: from phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta03.appoci.oracle.com [138.1.37.129]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9qmuae0r-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:14 +0000 Received: from pps.filterd (phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMIAuk034921; Mon, 25 Sep 2023 00:40:13 GMT Received: from nam10-bn7-obe.outbound.protection.outlook.com (mail-bn7nam10lp2103.outbound.protection.outlook.com [104.47.70.103]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf4184g-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:12 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=ckwdNHi+noSAp654ILw1G5ARO9n35tmokHSNp1GVhBQptzExWMizkuqKNrqfURs9WsdywjZe9cl19x6f72qhCIx/5jS3iKuffRvWWf6v1MNqxwEYEM7fpHzFeeAsg7BTPDrZ4qssp87VsRm73tDBtEAS0lBbOU/FI4m/AZlNC4JFaVyomAPVgn6NK6dJR0l8DYYuDwfaFDAo0w6hzPLAlfu0w6myhXFlHmH+csafdkEt+sCo0tvjjUc4SCimhZ6O9kabX6uwKddoxjBgkvfLOtQCDVIkByQ5/CTqGB0GVPttW4aOBwOP4qyqZzmMpPy7U+CqKRv3z3XVMp+D2uDr9A== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=E/x7mTYtq0AK43XWU8sgezCXfY3I6f53AD910fUr/7c=; b=hjz3em6me0I3VketzZRqYp0veMDMs60f+oUOUdQ4R9kZGeAxLV8yddsf+CeRt+IjmuX21sUx7uh5IR/JeCnv9KO4Ajn1PrxZbP2+zPcYwbpD4vAtCU6uxZxjAS8BrNo+Sld0h6i/T1m1XwSEg8PS/kwTBvE9s+bOaJ0+sNLe9dfPV2Ywhlxir3/iyAoPmC+8WtSn3XS4/yvnuwlbOFopxel/SQb+GueZr1Agf0r8bXTwIrvWdqguKT9L2+TwU3cjkz8hslYkskl4LQ4vNpGu1CmNiqXJMYxeL1UpTI9Ug0k/s/eu/FUoD+R3zB+2nxFAGARjGP7zZfDH1GisWVOszQ== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=E/x7mTYtq0AK43XWU8sgezCXfY3I6f53AD910fUr/7c=; b=cItdVJDbwHnl8rTvGlhc0gMUDjDFHSnIS3OwsAyLq6o+pFeGJ/UN0nZCKrHRqwn5Ok2gJQyjqyrmbltWEHWTiQmcDJtKNrOhSONamEUKOcEWB55Qb3IJwSblUJinzEVATOtYwX7VMLNbRwdFMEmbJ8znMpyg7dCF6sivoVyufPY= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by LV3PR10MB7819.namprd10.prod.outlook.com (2603:10b6:408:1b0::5) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:09 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:09 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz , James Houghton Subject: [PATCH v5 1/8] hugetlb: optimize update_and_free_pages_bulk to avoid lock cycles Date: Sun, 24 Sep 2023 17:39:45 -0700 Message-ID: <20230925003953.142620-2-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4P222CA0002.NAMP222.PROD.OUTLOOK.COM (2603:10b6:303:114::7) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|LV3PR10MB7819:EE_ X-MS-Office365-Filtering-Correlation-Id: 04c8fb6c-2fb6-4baf-11b6-08dbbd5ff8a6 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(396003)(366004)(39860400002)(136003)(346002)(376002)(230922051799003)(186009)(1800799009)(451199024)(2906002)(478600001)(6486002)(36756003)(6666004)(26005)(1076003)(38100700002)(83380400001)(6506007)(6512007)(2616005)(86362001)(41300700001)(4326008)(8676002)(8936002)(316002)(44832011)(5660300002)(66946007)(7416002)(15650500001)(66556008)(54906003)(66476007);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?/uFzmrdAFWSTH0uTAF8xy8XKBkuLa0+QrZf0mcw1PgrtBQ4DQCeKV2cPCDtI?= =?us-ascii?Q?MGCNDy2lZcG/3y9OpdPJ3ZRCS+kJ3cSvHwL4GQnZ/uaHOdfCWmRhhtCuT4Vv?= =?us-ascii?Q?hn4Idsj5ABgBBS8ciH6gD82xYVxNObAyMSR/BmwtkNVOp8rbh++Nv72BwZb4?= =?us-ascii?Q?LPp9Z2ys7qGzr+kJLZafPVlpg3kdf0xUlMzs0jhrbnJBmhaCARCZgddYCB+u?= =?us-ascii?Q?0xSmHhzJR0J8gIU0S5S1+e4+/GHGpV8OCLWcrOgtzueIEj9jKAVK8O0O4j8Q?= =?us-ascii?Q?8JhctC6F1U4vdnCYJq+1+aV722QY6+aKGE1ZOCKHNkowKDxdZ5aT/ZMrB/CO?= =?us-ascii?Q?JykRtWpyutL513gLgo24Z7Ivzk7j7DG6hPCANixTnKPnPXqR0HZ7d7gQnBEw?= =?us-ascii?Q?GzHR2QsX1JREe2Gdcn7xBDUXV+47UkTQ34nONa2q5BsppY871u8Ju48dNwf2?= =?us-ascii?Q?Ha7u9mPYoaPwkw7qopqnWDvSc4J7EcjZ/U81Z0zjyXr6KgGPYp+o9hqz+v46?= =?us-ascii?Q?UglRsSATEHMn5bncy1gwAmOtag2JsP1ORmAmleGu32QVD7kyEqkhQoLvm9RI?= =?us-ascii?Q?nx8XUuj39NlKskI34MFeoybMe8fvLILV+GbLsrTcld4pMBca5JCUPX1Bi23z?= =?us-ascii?Q?/t/R5Jg3b2F8kWCnZXTp7hfC3MsHUbYICwacpnW47XNZUDio7wlsjx4yHfbP?= =?us-ascii?Q?TVyn6T8qmSKe4YR3c/ZVi66HtaTMMR6qSvzUpd2r/dZlKxXT2ujJ4RIYfajd?= =?us-ascii?Q?IDYD0zjdw/4cK1dJuVy9RAmX3ZZKF0mY7W4EnV8Oo65iksEa73vT1513Jug3?= =?us-ascii?Q?rkx+0KO5uu0arN60FPfnQ8gs4nCHOzsu6udbzEQeD+5HLVqgBpnhEeXHEoej?= =?us-ascii?Q?HrtwlrACp+1xRcVfCr9cZptYTDcdaYxVbEOV6tlU+JuXqIXY5zrBFf1zihrm?= =?us-ascii?Q?Y6syGIPIoPIfi9S0bLCtVwiVdpId9cNRcBg6VQ/X7x6DWGiMykqiRBZm3/6w?= =?us-ascii?Q?jUCwjGdydEjNNQROA0Mm/jP+WGrrGWodSxiG1rAKOV43G5tw8hAapNOhvTFP?= =?us-ascii?Q?55XDvu+mAOOMbS6nI4tqzFSk5OFr16bxyG9El6Anaeip7RPM+ZYIi9yg8X1F?= =?us-ascii?Q?IugHO46OOAsY6TxQkDYf74THx50V8xGbay/sA9Y5aznRplhqH4lMFWcuRIpz?= =?us-ascii?Q?xvj2AImTuEh08Sl3Z2RTgW7hN6fooDh48WLvZF+ZgpuiungOC+NZk4pMd3KH?= =?us-ascii?Q?4C8ujqcMEwLsrhEn6R7CYAp43MiqpD3Wccj9rL0MsfLSYz093ivAEuS/Mmqz?= =?us-ascii?Q?tG+8lWh9cx92+SdHLe/EEHD8znt5KqxSHO1oK/UZWxZghOdwdM/zVD6cXbsx?= =?us-ascii?Q?kJ4hnDQOs+d3OSqm6jfI0RtHJuH9jjS3D9L+9jwYadzgrmFflwktX4wkEJyl?= =?us-ascii?Q?B7DZeI7xqapIKNWswV8bBuFm24YQLGOiDoUmBx2bo+1J9QOtPzTSxB6J5LKx?= =?us-ascii?Q?yFJoFYbo+heBr1NawXO35pJgJgVwzbT55bch7n1pjXGY6Rg0+8/8I9pSj0dJ?= =?us-ascii?Q?lh2PYp1fjjLbCVcFSTGgm8Cp/7gYsWFj6LYoA+EQ?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?uZlzhsX7mip/PGLvUTALTOQE0lu96aWRyWnWvdgjyZk3uPeXghaeflmJkPHI?= =?us-ascii?Q?E9ukX+8tt2Am5NjB8K35jhcduUxkySsg3HJ9whjwaeFABRH1k04yarvRAiWd?= =?us-ascii?Q?dqL4qzuJY8SqmO2RJ2wAYFv8iwuyeciFqq4FgGCnruRoNTSouodTkThwNU75?= =?us-ascii?Q?S9KL9UxjN5nEaqnv6B++xeaQxulnYyulRtaZTqfcaaa0BcIUvgrko9JVtmmQ?= =?us-ascii?Q?L/Keh1fFRZVSvJBdDQ8jKKPi51Y57m0icq0FVSFXF3jGLCzAx4kAD8Db3ypR?= =?us-ascii?Q?aqy49Q08bXaLj2YmTXogHbFTBz2MFj32yV1UomKj3bRhF0dm+mDWvxcK8MPs?= =?us-ascii?Q?+vtCanWzITfZvFoVvm1eFQCgvx+DHjKyApV/gzVK7mJDnU1ApYviyypGkXqr?= =?us-ascii?Q?GVQGSIoh55BpHvT1TI4qKflT6OluJDsblgRkdjWBMKopmavL2eL6h0qSW5nn?= =?us-ascii?Q?gharHZ4A0d/gfV4I41bh/AAJW0UcN4RT9qJIXcn1FVH9hSOsAj759eDuys/R?= =?us-ascii?Q?wAkb4cYv9rLNlzE1bND4hSISr/zlZB8jJkv0LnKL0987trEXUX43gxFj2fK4?= =?us-ascii?Q?pBJPLKMDGyZwpxGavL3kTAqlMjnY0ayVPzOhef08PD6iXGBuuDDAOq3nJY+q?= =?us-ascii?Q?sZ2akec1lemMImC70NfuHvkSt2HwJC0wqD3r7K8NBGxdxfgOPi9PpWagCRIu?= =?us-ascii?Q?PKxzo1uQnFOdCGkXGQT3Xfirhj170OYJyG/dvoAlWuNYhWFg3a2zEHZHox8U?= =?us-ascii?Q?Ngf5y7oU7AZuTAORWGEERCWu5kUR7g6qKT8dKXoQ0d1XjJEW/dJR3sf4jLr/?= =?us-ascii?Q?7hjbxMCeRKraP4q55CXhb51T788W9PzzwnGPFKwWhyHirpS34Jer0GlhP7kD?= =?us-ascii?Q?KIi4MY9ecADgqKXGBUzqBqMpAVnZefmxrb8xjrxpagsozxGgd2OGs7C+nvpB?= =?us-ascii?Q?+O4cxkBhuYZwBkoPckII3lSDWE6nc5GvrMMcONdQWWtW68hxfO+0GSxqJ1H1?= =?us-ascii?Q?jopIqtx3enU2tGxLNN1J3SC2LrLIZJqtr4eGLY6mKP+VRRp7oXNGzdqW+Pwc?= =?us-ascii?Q?AwEYw+iABbH78kVcIzoGFAw9Mebr8ObxlnWOaHWMK8CfJx8M9OkAMs2GPs+N?= =?us-ascii?Q?6ZGW7kvB2IGz?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: 04c8fb6c-2fb6-4baf-11b6-08dbbd5ff8a6 X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:09.6130 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: CBqM7Env+7t0Rs2aI8D6vJngdq6xQ1CSC/w98u2ruNHXZLjB8YWI+a5r41NfhXkLubV2Hig3S0KqOqfbMz+AQw== X-MS-Exchange-Transport-CrossTenantHeadersStamped: LV3PR10MB7819 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 bulkscore=0 mlxscore=0 mlxlogscore=999 suspectscore=0 phishscore=0 adultscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-GUID: NEeBp3iSSrbJTj79mHiI8-W2S2ERi9_O X-Proofpoint-ORIG-GUID: NEeBp3iSSrbJTj79mHiI8-W2S2ERi9_O Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" update_and_free_pages_bulk is designed to free a list of hugetlb pages back to their associated lower level allocators. This may require allocating vmemmmap pages associated with each hugetlb page. The hugetlb page destructor must be changed before pages are freed to lower level allocators. However, the destructor must be changed under the hugetlb lock. This means there is potentially one lock cycle per page. Minimize the number of lock cycles in update_and_free_pages_bulk by: 1) allocating necessary vmemmap for all hugetlb pages on the list 2) take hugetlb lock and clear destructor for all pages on the list 3) free all pages on list back to low level allocators Signed-off-by: Mike Kravetz Reviewed-by: Muchun Song Acked-by: James Houghton --- mm/hugetlb.c | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index de220e3ff8be..47159b9de633 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1837,7 +1837,46 @@ static void update_and_free_hugetlb_folio(struct hst= ate *h, struct folio *folio, static void update_and_free_pages_bulk(struct hstate *h, struct list_head = *list) { struct folio *folio, *t_folio; + bool clear_dtor =3D false; =20 + /* + * First allocate required vmemmmap (if necessary) for all folios on + * list. If vmemmap can not be allocated, we can not free folio to + * lower level allocator, so add back as hugetlb surplus page. + * add_hugetlb_folio() removes the page from THIS list. + * Use clear_dtor to note if vmemmap was successfully allocated for + * ANY page on the list. + */ + list_for_each_entry_safe(folio, t_folio, list, lru) { + if (folio_test_hugetlb_vmemmap_optimized(folio)) { + if (hugetlb_vmemmap_restore(h, &folio->page)) { + spin_lock_irq(&hugetlb_lock); + add_hugetlb_folio(h, folio, true); + spin_unlock_irq(&hugetlb_lock); + } else + clear_dtor =3D true; + } + } + + /* + * If vmemmmap allocation was performed on any folio above, take lock + * to clear destructor of all folios on list. This avoids the need to + * lock/unlock for each individual folio. + * The assumption is vmemmap allocation was performed on all or none + * of the folios on the list. This is true expect in VERY rare cases. + */ + if (clear_dtor) { + spin_lock_irq(&hugetlb_lock); + list_for_each_entry(folio, list, lru) + __clear_hugetlb_destructor(h, folio); + spin_unlock_irq(&hugetlb_lock); + } + + /* + * Free folios back to low level allocators. vmemmap and destructors + * were taken care of above, so update_and_free_hugetlb_folio will + * not need to take hugetlb lock. + */ list_for_each_entry_safe(folio, t_folio, list, lru) { update_and_free_hugetlb_folio(h, folio, false); cond_resched(); --=20 2.41.0 From nobody Fri Feb 13 12:45:13 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id A7723CE7A8B for ; Mon, 25 Sep 2023 00:41:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230170AbjIYAlO (ORCPT ); Sun, 24 Sep 2023 20:41:14 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59246 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230435AbjIYAlJ (ORCPT ); Sun, 24 Sep 2023 20:41:09 -0400 Received: from mx0b-00069f02.pphosted.com (mx0b-00069f02.pphosted.com [205.220.177.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 46D12116 for ; Sun, 24 Sep 2023 17:41:01 -0700 (PDT) Received: from pps.filterd (m0246632.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMkBQF019937; Mon, 25 Sep 2023 00:40:19 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=x/ssleWl8yTL8a0XIJeSZc6h/9ChBMuzWbujLkXEhts=; b=MRJ0g+Os7KBcJZ1V4DLMCVgCyK3c+iLSfXcUJb0WoDxnPUfBrp5tfuqegCGV+MZnhlt5 pIrxoofenKm3sRUkreRnPzAo4dKLLKupd7MHeg6f+oRWxwpRPjIcfjWtEDDXYy/VmiNs porfm/R3lqldW0OsEddnQcF0Mc23jjaXLlcVhRszvZqtsSZkbxpaZUO6i+zs1Hy7hzxP BPf9pZTozqGMhvehjMA4QYsq5bgyMlApbAkSN3L0aX5QKrY/468UcWHGAFYn/9Ls9M6M +ojwgwQWaxO8q4AnJGuHvRwdekBPCLD5yqEtxufwP/WBH31q4UcEIrLLigPLNHar/iby OQ== Received: from phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta01.appoci.oracle.com [138.1.114.2]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9qmuae0t-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:18 +0000 Received: from pps.filterd (phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38ONQiWu030565; Mon, 25 Sep 2023 00:40:17 GMT Received: from nam10-bn7-obe.outbound.protection.outlook.com (mail-bn7nam10lp2100.outbound.protection.outlook.com [104.47.70.100]) by phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf3kx04-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:17 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=DfDKtJhoAS+wGqMp5FYEmGv4prdFKBHVDFhPYvwjGkHtNqNm61Ymrma9M0Tits10hVU5LHzylfV8MFBg/gTGOyFKU/fIKd9hHIlLINx2mFoiZ2BIMdi7xccoJCb6mTZF8T+La3oL/HtFyFDidNaJo9AOlqtvM28GYumPBUbZ2mRPRPA/l7JHaSmf1pdHX9o2rIhC/PO0pSMNi0tmq0KowV4/GZchEZDa1yZEOEJzw8GYolCk1E9qZj3BhUOmyrifG3+CERDvHh0yUCw/lkePKfYQXlMvkRzPXv/pkrG7d/UiPlvol0rPmNfrFLlx0322mJwgZjw3CkVFduMV9Ev0ww== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=x/ssleWl8yTL8a0XIJeSZc6h/9ChBMuzWbujLkXEhts=; b=GVW1j81uM8LQ1K2Xxuu4e0gxzKoYf8LNKg0QYt2b4XKLTyGwev1ctzorWyCoGzDr5jyz31W/ItOUHNJio4D48jy17mxc8e6PNumJIW9lEyTNnGhOE2TGLj4VpCDWCF4CXBQJz9mt/9hGTxKAUNn5SZknAaL8xAhfvED90Af2rdq9Sr/n4wErEU1rrCZ6zm2oYnqf6ujmsTlQRX5yiDPEAbYBpTfZnydKyThxScFVQTnHN5X3+36oV1VQhSN0r3SgNaT4Uk0mZPm2BopLlLIu+0Shq6NnXevZHhZKqPzrhcsK+4P079TfEzWWol+i0bIdcGgASfpn9xPPzLyUR4npcA== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=x/ssleWl8yTL8a0XIJeSZc6h/9ChBMuzWbujLkXEhts=; b=NPCRh8P1Igas5LfaCVPZhUv9TZBpv2ko6XavACGK5S0RTh+rtnoXcm8PZZuryMjQ04PUMQqsRsQTQAC8rzrLyuhbHVWxl4J1jgumSFkYGyBjOoLQ9h4x8XdLldRz6J6oP37xS11IpEPTnBU9dyvHT1DTYj2JogDY4bpLvx73D40= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by LV3PR10MB7819.namprd10.prod.outlook.com (2603:10b6:408:1b0::5) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:15 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:15 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 2/8] hugetlb: restructure pool allocations Date: Sun, 24 Sep 2023 17:39:46 -0700 Message-ID: <20230925003953.142620-3-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4PR04CA0087.namprd04.prod.outlook.com (2603:10b6:303:6b::32) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|LV3PR10MB7819:EE_ X-MS-Office365-Filtering-Correlation-Id: ee005a7d-433f-4834-ed26-08dbbd5ffbd2 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(396003)(366004)(39860400002)(136003)(346002)(376002)(230922051799003)(186009)(1800799009)(451199024)(2906002)(478600001)(6486002)(36756003)(6666004)(26005)(1076003)(107886003)(38100700002)(83380400001)(6506007)(6512007)(2616005)(86362001)(41300700001)(4326008)(8676002)(8936002)(316002)(44832011)(5660300002)(30864003)(66946007)(7416002)(66556008)(54906003)(66476007);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?3U/pcRCzSG9KTbHbXGh+Tx4+DFDEXlHQqAPeiFp5wZ0XxFIgNhixYHoZbPYg?= =?us-ascii?Q?/4ZQgtSb9W9xiOhh/o2mMTHcSo1NiFaXjyaJSivmYPKeeII0nvDEi0wEN8k2?= =?us-ascii?Q?CEA9yvh3bHLv0sM6SyuBQFXJg3fZDjPxZq0xXYL3PtOdkyoxCwmNlI6mICgy?= =?us-ascii?Q?2AQW7nWfgURMXe9sMkjgydtw1kyWmfQBjLn+lClO1XrRtPJVuLfhDBcxhkdV?= =?us-ascii?Q?OBlc4YZuX9jDdfEI1QvNCeQJA9B0ICSncIAkm9vBwkZMI6aVg2AcKUH1eUTP?= =?us-ascii?Q?G23WIBQe1buuHgJC6hkPPSIt3XdUR0ec5pASMPs7hLvBNSCLfFN8bioBChXf?= =?us-ascii?Q?et0hZ5Nk9Zq8gCH/P66AJlDq69oi4ivRmwK4H5HenQzy80ftHAlQHAL462UX?= =?us-ascii?Q?f1IyiFJTf29te58Wov6HQTfNcnj9xbiGcMT9EsTBSAXDs0hTmHXmN7GODyBA?= =?us-ascii?Q?IW/B6mzat1+H/1vOUoPA/3As5tBK9I/ZlsirMm2ljH3s2sFySwS7oexCs5Y2?= =?us-ascii?Q?XkLDviyFgxyIGrOJt2f+TL6W2jobIUXm/7xXmp96KD/cfayd7NRAKblA88HC?= =?us-ascii?Q?9tJDx3R1hg4o7AdPtZELFVixvXxWLWQ+ZjhGxOcOhxcS0cSChSq9uSn7QhhE?= =?us-ascii?Q?7Aux0Sz8eMGro+W+tGwvaoZ/AWY9ka3ScuZ9FZha788rPDD8nv9xibVP9DFN?= =?us-ascii?Q?290XIuMmCbXVeVMTqoVOQmh0KC06ZmiS0C0zBG/N6RGVU9lPtQhgm9GY12uU?= =?us-ascii?Q?YR/kcYeyPBGnemySMsRJFgeVrMkFrPfmFDl4+D4ot0GSkIWDl1ST0hEuBMwl?= =?us-ascii?Q?3OWdLKUw1AxDNhIdQHvw8JSS5I4xq2jFOhjvMpPLn8B/C9Tw6l0t4knivT7b?= =?us-ascii?Q?vIUEkTUrH0Ds0XPC0sg3KZotfw+fHHgkUOiuvizKAbq++yXQQFL+h5GFA1BW?= =?us-ascii?Q?KwpfRstIM4jpx0bpExr4Np/zlX+sns0qIqv04gkUqM0jqLwJnjoHLpXr8TMg?= =?us-ascii?Q?nxtWftIoPBH+tWhjNkdvqRKPH8Po1AsXZRxoS/rQ2Lx9zxMnCg0IlqMrYRmW?= =?us-ascii?Q?CkNreHIqG15v/2K5WLxTqwq/jcJnPoSUbqmYTZ1g/dFDPZ1hFMC73HImkgFi?= =?us-ascii?Q?4KRzN7bSRA34kUbWr99m+AwgBkr7l5jv8P+IpZFHaoyLXKig7CFSFBnwHV8I?= =?us-ascii?Q?WaPfI+23kfmaO8Ipb4cXT9KP3sKczAVmIinl7s3TzwiOPb1ytEIS66vhYK1i?= =?us-ascii?Q?8s/qWV18oUac7nyNskl/vKNQwkBQxH83Y8jPhiVb4oe73C0UiGxgz+s6epJh?= =?us-ascii?Q?61BUAr/Zja+lWZRmufXkwScOxERoHzvQJ1R+EPYcGalIci/R1fAJUBzRLr/8?= =?us-ascii?Q?notYTfjAp6amZ7o9TqZ1rcvoAxgIc63kMlhphA438iPgNiEVAUZZcO0XIt6b?= =?us-ascii?Q?J2fMe3v8P/4VcNEtEA8Abs1RLcGT1maaaxxAqUuX4puWWN2mBGs+5jZsdtX9?= =?us-ascii?Q?Z1NENkrM6yUPSyjUNz6w6HiWMYuBmRe4P4mCMCGzVbSwwAeEgD7kmn2tZl6O?= =?us-ascii?Q?vyfc4050cRoDag70fQiaQ4hLiHG8FwGF6DtTUVGI?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?E59jRxOWLkj8nxhIelt/YEE4f2SNYKv255OYEJ6xG7INaribjAscLYT9quiA?= =?us-ascii?Q?cFEVGyRU6a62uG8nG3YGpZUZ95Ryk9HW/OD4RDDYW7T6q/exWcrezyoOASF0?= =?us-ascii?Q?W+tVufuzmLo/DCDByV2Fqc2laX4gzkNYP7Z8MwGUVkK92wXWfCCzfb1wgbK7?= =?us-ascii?Q?/8BqfrwZIobyrHJ8JQYxTdviDSEVezRDyjhQxm+bFtJh2W/Y+A2RpnYXmBW6?= =?us-ascii?Q?7RJRUWymZERX+hZqIVr9if68Q9IU4LFNKi1nER1kqAYNRr1nbMEsDVF7IR6r?= =?us-ascii?Q?j/O5sMHy+UgIIywV6eSZfmwyd0stX3p9u/GYRytYy6Z8ziojgf7JgPblKVHK?= =?us-ascii?Q?ePXxkGGo30Jwahz6/9++zXla4K5wxv9EgT/FdPV3s/t+6o0pLcMlzkTeItss?= =?us-ascii?Q?7FDuC9+9ZTnBKDYLUGGVixz+LOW9DTSf2lE0GXpbQ+Iwl3WqvEzkp46naVmb?= =?us-ascii?Q?Y6+XHQmkzE1USW7ID6etso/h9r2ySEWk1pecmpArVIzvualILLe0yiDYFyzv?= =?us-ascii?Q?IjolQ52CK8nj4kXH9xG5i7flURwOVVagXW5UwswOSeApMI8kDk7vnDcKPSJ7?= =?us-ascii?Q?TKE5rqbnybPoh4e9VHKbotrHezl88XFhDIjpRlkugrcmzNfey4WIHGI+3yUz?= =?us-ascii?Q?o0Lc8ggMk3YECgOuuAu0l/h9hc6lNxT0LFYNn/lX12JffuyCZN6HwKecdOCe?= =?us-ascii?Q?eWFN3vVyYeru612KCe2dUQBbn2xPmLKaNuoVKyiUDwRvZd9OJwmFetqPtVPR?= =?us-ascii?Q?7lVYJaTeoAi+0/kQazMcH02KW1OBQyjlEANLvmB7vsOziNqXTMLIyR5kj9e2?= =?us-ascii?Q?TxoQtr6e2EyNa8pBm3oOUyEG/0ouKw49nqXD8Lh/jeJwxLDMmSgtBt3qNwXS?= =?us-ascii?Q?CWKXCHULpZBm7aIyOrQcodpm7Yl1JO1TNkvO7zTpM9HRMXl1oYSjeWqbFaCC?= =?us-ascii?Q?ERCiv/ybYLtwH4UwiEIgc464xItp/aZMQWYsdghKF0B8Spz2iocSV+slNYcM?= =?us-ascii?Q?JhYPWmhI7dc0f+Hn6MzNzJlhYV0vhxS8mpHwI+eN69sgm1p0t7saomCmotGD?= =?us-ascii?Q?wcIU7F+y9wRZp4IXfpDsKlq2WTlW90OeS+BTov6ptwCbgWn2RHJOMgO5tqVE?= =?us-ascii?Q?XOWINROMx+92?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: ee005a7d-433f-4834-ed26-08dbbd5ffbd2 X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:14.9167 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: mgZu+a33F1ieo4oNSddTax5LyoEqMGJ0TchvVu63zY86o5kY4LB1sDK6SivfQ+kucb77nqWGSQzkrokal+4/QA== X-MS-Exchange-Transport-CrossTenantHeadersStamped: LV3PR10MB7819 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 suspectscore=0 phishscore=0 mlxscore=0 malwarescore=0 mlxlogscore=999 bulkscore=0 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-GUID: 47kf-UyU0VW8jJYZpEUPyBeCy5cuSIAe X-Proofpoint-ORIG-GUID: 47kf-UyU0VW8jJYZpEUPyBeCy5cuSIAe Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Allocation of a hugetlb page for the hugetlb pool is done by the routine alloc_pool_huge_page. This routine will allocate contiguous pages from a low level allocator, prep the pages for usage as a hugetlb page and then add the resulting hugetlb page to the pool. In the 'prep' stage, optional vmemmap optimization is done. For performance reasons we want to perform vmemmap optimization on multiple hugetlb pages at once. To do this, restructure the hugetlb pool allocation code such that vmemmap optimization can be isolated and later batched. The code to allocate hugetlb pages from bootmem was also modified to allow batching. No functional changes, only code restructure. Signed-off-by: Mike Kravetz Reviewed-by: Muchun Song --- mm/hugetlb.c | 179 ++++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 140 insertions(+), 39 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 47159b9de633..64f50f3844fc 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1970,16 +1970,21 @@ static void __prep_account_new_huge_page(struct hst= ate *h, int nid) h->nr_huge_pages_node[nid]++; } =20 -static void __prep_new_hugetlb_folio(struct hstate *h, struct folio *folio) +static void init_new_hugetlb_folio(struct hstate *h, struct folio *folio) { folio_set_hugetlb(folio); - hugetlb_vmemmap_optimize(h, &folio->page); INIT_LIST_HEAD(&folio->lru); hugetlb_set_folio_subpool(folio, NULL); set_hugetlb_cgroup(folio, NULL); set_hugetlb_cgroup_rsvd(folio, NULL); } =20 +static void __prep_new_hugetlb_folio(struct hstate *h, struct folio *folio) +{ + init_new_hugetlb_folio(h, folio); + hugetlb_vmemmap_optimize(h, &folio->page); +} + static void prep_new_hugetlb_folio(struct hstate *h, struct folio *folio, = int nid) { __prep_new_hugetlb_folio(h, folio); @@ -2190,16 +2195,9 @@ static struct folio *alloc_buddy_hugetlb_folio(struc= t hstate *h, return page_folio(page); } =20 -/* - * Common helper to allocate a fresh hugetlb page. All specific allocators - * should use this function to get new hugetlb pages - * - * Note that returned page is 'frozen': ref count of head page and all ta= il - * pages is zero. - */ -static struct folio *alloc_fresh_hugetlb_folio(struct hstate *h, - gfp_t gfp_mask, int nid, nodemask_t *nmask, - nodemask_t *node_alloc_noretry) +static struct folio *__alloc_fresh_hugetlb_folio(struct hstate *h, + gfp_t gfp_mask, int nid, nodemask_t *nmask, + nodemask_t *node_alloc_noretry) { struct folio *folio; bool retry =3D false; @@ -2212,6 +2210,7 @@ static struct folio *alloc_fresh_hugetlb_folio(struct= hstate *h, nid, nmask, node_alloc_noretry); if (!folio) return NULL; + if (hstate_is_gigantic(h)) { if (!prep_compound_gigantic_folio(folio, huge_page_order(h))) { /* @@ -2226,32 +2225,80 @@ static struct folio *alloc_fresh_hugetlb_folio(stru= ct hstate *h, return NULL; } } - prep_new_hugetlb_folio(h, folio, folio_nid(folio)); =20 return folio; } =20 +static struct folio *only_alloc_fresh_hugetlb_folio(struct hstate *h, + gfp_t gfp_mask, int nid, nodemask_t *nmask, + nodemask_t *node_alloc_noretry) +{ + struct folio *folio; + + folio =3D __alloc_fresh_hugetlb_folio(h, gfp_mask, nid, nmask, + node_alloc_noretry); + if (folio) + init_new_hugetlb_folio(h, folio); + return folio; +} + /* - * Allocates a fresh page to the hugetlb allocator pool in the node interl= eaved - * manner. + * Common helper to allocate a fresh hugetlb page. All specific allocators + * should use this function to get new hugetlb pages + * + * Note that returned page is 'frozen': ref count of head page and all ta= il + * pages is zero. */ -static int alloc_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowe= d, - nodemask_t *node_alloc_noretry) +static struct folio *alloc_fresh_hugetlb_folio(struct hstate *h, + gfp_t gfp_mask, int nid, nodemask_t *nmask, + nodemask_t *node_alloc_noretry) { struct folio *folio; - int nr_nodes, node; + + folio =3D __alloc_fresh_hugetlb_folio(h, gfp_mask, nid, nmask, + node_alloc_noretry); + if (!folio) + return NULL; + + prep_new_hugetlb_folio(h, folio, folio_nid(folio)); + return folio; +} + +static void prep_and_add_allocated_folios(struct hstate *h, + struct list_head *folio_list) +{ + struct folio *folio, *tmp_f; + + /* Add all new pool pages to free lists in one lock cycle */ + spin_lock_irq(&hugetlb_lock); + list_for_each_entry_safe(folio, tmp_f, folio_list, lru) { + __prep_account_new_huge_page(h, folio_nid(folio)); + enqueue_hugetlb_folio(h, folio); + } + spin_unlock_irq(&hugetlb_lock); +} + +/* + * Allocates a fresh hugetlb page in a node interleaved manner. The page + * will later be added to the appropriate hugetlb pool. + */ +static struct folio *alloc_pool_huge_folio(struct hstate *h, + nodemask_t *nodes_allowed, + nodemask_t *node_alloc_noretry) +{ gfp_t gfp_mask =3D htlb_alloc_mask(h) | __GFP_THISNODE; + int nr_nodes, node; =20 for_each_node_mask_to_alloc(h, nr_nodes, node, nodes_allowed) { - folio =3D alloc_fresh_hugetlb_folio(h, gfp_mask, node, + struct folio *folio; + + folio =3D only_alloc_fresh_hugetlb_folio(h, gfp_mask, node, nodes_allowed, node_alloc_noretry); - if (folio) { - free_huge_folio(folio); /* free it into the hugepage allocator */ - return 1; - } + if (folio) + return folio; } =20 - return 0; + return NULL; } =20 /* @@ -3264,25 +3311,35 @@ static void __init hugetlb_folio_init_vmemmap(struc= t folio *folio, */ static void __init gather_bootmem_prealloc(void) { + LIST_HEAD(folio_list); struct huge_bootmem_page *m; + struct hstate *h, *prev_h =3D NULL; =20 list_for_each_entry(m, &huge_boot_pages, list) { struct page *page =3D virt_to_page(m); struct folio *folio =3D (void *)page; - struct hstate *h =3D m->hstate; + + h =3D m->hstate; + /* + * It is possible to have multiple huge page sizes (hstates) + * in this list. If so, process each size separately. + */ + if (h !=3D prev_h && prev_h !=3D NULL) + prep_and_add_allocated_folios(prev_h, &folio_list); + prev_h =3D h; =20 VM_BUG_ON(!hstate_is_gigantic(h)); WARN_ON(folio_ref_count(folio) !=3D 1); =20 hugetlb_folio_init_vmemmap(folio, h, HUGETLB_VMEMMAP_RESERVE_PAGES); - prep_new_hugetlb_folio(h, folio, folio_nid(folio)); + __prep_new_hugetlb_folio(h, folio); /* If HVO fails, initialize all tail struct pages */ if (!HPageVmemmapOptimized(&folio->page)) hugetlb_folio_init_tail_vmemmap(folio, HUGETLB_VMEMMAP_RESERVE_PAGES, pages_per_huge_page(h)); - free_huge_folio(folio); /* add to the hugepage allocator */ + list_add(&folio->lru, &folio_list); =20 /* * We need to restore the 'stolen' pages to totalram_pages @@ -3292,6 +3349,8 @@ static void __init gather_bootmem_prealloc(void) adjust_managed_page_count(page, pages_per_huge_page(h)); cond_resched(); } + + prep_and_add_allocated_folios(h, &folio_list); } =20 static void __init hugetlb_hstate_alloc_pages_onenode(struct hstate *h, in= t nid) @@ -3325,9 +3384,22 @@ static void __init hugetlb_hstate_alloc_pages_onenod= e(struct hstate *h, int nid) h->max_huge_pages_node[nid] =3D i; } =20 +/* + * NOTE: this routine is called in different contexts for gigantic and + * non-gigantic pages. + * - For gigantic pages, this is called early in the boot process and + * pages are allocated from memblock allocated or something similar. + * Gigantic pages are actually added to pools later with the routine + * gather_bootmem_prealloc. + * - For non-gigantic pages, this is called later in the boot process after + * all of mm is up and functional. Pages are allocated from buddy and + * then added to hugetlb pools. + */ static void __init hugetlb_hstate_alloc_pages(struct hstate *h) { unsigned long i; + struct folio *folio; + LIST_HEAD(folio_list); nodemask_t *node_alloc_noretry; bool node_specific_alloc =3D false; =20 @@ -3369,14 +3441,25 @@ static void __init hugetlb_hstate_alloc_pages(struc= t hstate *h) =20 for (i =3D 0; i < h->max_huge_pages; ++i) { if (hstate_is_gigantic(h)) { + /* + * gigantic pages not added to list as they are not + * added to pools now. + */ if (!alloc_bootmem_huge_page(h, NUMA_NO_NODE)) break; - } else if (!alloc_pool_huge_page(h, - &node_states[N_MEMORY], - node_alloc_noretry)) - break; + } else { + folio =3D alloc_pool_huge_folio(h, &node_states[N_MEMORY], + node_alloc_noretry); + if (!folio) + break; + list_add(&folio->lru, &folio_list); + } cond_resched(); } + + /* list will be empty if hstate_is_gigantic */ + prep_and_add_allocated_folios(h, &folio_list); + if (i < h->max_huge_pages) { char buf[32]; =20 @@ -3510,7 +3593,9 @@ static int adjust_pool_surplus(struct hstate *h, node= mask_t *nodes_allowed, static int set_max_huge_pages(struct hstate *h, unsigned long count, int n= id, nodemask_t *nodes_allowed) { - unsigned long min_count, ret; + unsigned long min_count; + unsigned long allocated; + struct folio *folio; LIST_HEAD(page_list); NODEMASK_ALLOC(nodemask_t, node_alloc_noretry, GFP_KERNEL); =20 @@ -3587,7 +3672,8 @@ static int set_max_huge_pages(struct hstate *h, unsig= ned long count, int nid, break; } =20 - while (count > persistent_huge_pages(h)) { + allocated =3D 0; + while (count > (persistent_huge_pages(h) + allocated)) { /* * If this allocation races such that we no longer need the * page, free_huge_folio will handle it by freeing the page @@ -3598,15 +3684,32 @@ static int set_max_huge_pages(struct hstate *h, uns= igned long count, int nid, /* yield cpu to avoid soft lockup */ cond_resched(); =20 - ret =3D alloc_pool_huge_page(h, nodes_allowed, + folio =3D alloc_pool_huge_folio(h, nodes_allowed, node_alloc_noretry); - spin_lock_irq(&hugetlb_lock); - if (!ret) + if (!folio) { + prep_and_add_allocated_folios(h, &page_list); + spin_lock_irq(&hugetlb_lock); goto out; + } + + list_add(&folio->lru, &page_list); + allocated++; =20 /* Bail for signals. Probably ctrl-c from user */ - if (signal_pending(current)) + if (signal_pending(current)) { + prep_and_add_allocated_folios(h, &page_list); + spin_lock_irq(&hugetlb_lock); goto out; + } + + spin_lock_irq(&hugetlb_lock); + } + + /* Add allocated pages to the pool */ + if (!list_empty(&page_list)) { + spin_unlock_irq(&hugetlb_lock); + prep_and_add_allocated_folios(h, &page_list); + spin_lock_irq(&hugetlb_lock); } =20 /* @@ -3632,8 +3735,6 @@ static int set_max_huge_pages(struct hstate *h, unsig= ned long count, int nid, * Collect pages to be removed on list without dropping lock */ while (min_count < persistent_huge_pages(h)) { - struct folio *folio; - folio =3D remove_pool_hugetlb_folio(h, nodes_allowed, 0); if (!folio) break; --=20 2.41.0 From nobody Fri Feb 13 12:45:13 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 85DD8CE7A91 for ; Mon, 25 Sep 2023 00:41:48 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231497AbjIYAlw (ORCPT ); Sun, 24 Sep 2023 20:41:52 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55550 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231494AbjIYAl3 (ORCPT ); Sun, 24 Sep 2023 20:41:29 -0400 Received: from mx0a-00069f02.pphosted.com (mx0a-00069f02.pphosted.com [205.220.165.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 713A0CC1 for ; Sun, 24 Sep 2023 17:41:14 -0700 (PDT) Received: from pps.filterd (m0246627.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OLnhAH028065; Mon, 25 Sep 2023 00:40:24 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=OP+HI4Dx3GzIthduXhchAaGMojbE9QNz8Df1OYn0zTs=; b=Zq57aqcvxjb+HAt5LLm2lXlf7drbgDbAJKcykAVUrZf6K0VUoTxpCxnRqUOs7q/+dW7Y 7VnPiO/RU580e6gFgJwiTKJxr80EPxarus+oYCSORIINy7WEoFo5tE/QgN9EXfmJ9ejv 3xF7Rpb9CenAvpJnxi6yTa59Fz4jXkUkxgenmIJzdXIFZGTLqEr91uI7bMIiOvC0hvkR qSP+SQSJGgFrgtA3eZNV5Ni8+zADfdiWveudCf2Nh/jAg2v0faSai/W84SiK9Yaore7i kqV8BU/aQ8oY+CDS2Nmickzo2ORFQay2+i5am2fb6auI1gVHNcOsZBZkdUIrOoJYAswO bg== Received: from phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta03.appoci.oracle.com [138.1.37.129]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9pm22gbr-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:23 +0000 Received: from pps.filterd (phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38P08QcP036610; Mon, 25 Sep 2023 00:40:23 GMT Received: from nam12-bn8-obe.outbound.protection.outlook.com (mail-bn8nam12lp2172.outbound.protection.outlook.com [104.47.55.172]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf4189e-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:22 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=e9AtgfQb0LCBWql3DtZ4yjTOM0+ZNUiCg2Lqlm5v1DNwDwSZHu5lqHC/00dOB0kwXdyrC4gjF78YLN7TlVFtkqa3Ra9ux/oaposBpPUXjL06oTmvYfaKES2L+vR77Pfk+wk5SfB6FnES7LCaqqBkt2skZbHLh0T+7PYB99zXNBVEv/F78dNVObYprnmhYQbqxoW0RAG+ikx2ctBLpfw/2lA7DK+xVhv6bHUyFQP0ReqZCNKn2QfVNAydjrudjyEAOQoaXj14kFkLC7sPcdzf2KKd1dla2ThKAWnSiyqJoT5sGOzPaI0Vp3gUik7GfhLnpoz3vxLsf6IG3ASgXmC2bQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=OP+HI4Dx3GzIthduXhchAaGMojbE9QNz8Df1OYn0zTs=; b=XOytx3JT8+usVebdL1rJpkddmUbwsFE1Hue2cyDWek1w5K+gqdH9SOOnamhWmJzrx9cjvjeJEWIcbIFJLlB/CCX0lfD1HH3PUdW5MkUSdTbZeHqNIMxK20TKPnP+rmCOJ71YoYA/+PaZ/2063ogxyrVuKqw0Bzc9CGsnYvQ8Hcitjfy/055RvLWch6Tl3yIL3VkSNN+LdazC7gAI+HUGTEk16AIyDaqJP5cCiXgIb1Zd7H/PY/s1iUjKXmLDHXxkQAaioMFeDXtNYQFoId7sO2Y2LmCCG/tWabSgsyolNSmK4EQwzfShwiOAEO9GWiaje9aRBFapb9KjiFl3I5vAZg== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=OP+HI4Dx3GzIthduXhchAaGMojbE9QNz8Df1OYn0zTs=; b=LQEoN7VxJH0MtcCAnJPUvyknZm0vUFNly/GG/xrxNEGaFeOGjkNVRUnElsmsgImGERwEjtzGgussV2T2ahZbFGIguWBv/Ft43GpQLLC/xQphJhIteSrpwM+Pr6TqoSb5xYyFO6WTtuuOxY8I2YVIN+mAlJUQSOB3HMpxonX8wMc= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by DS0PR10MB7174.namprd10.prod.outlook.com (2603:10b6:8:df::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:19 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:18 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 3/8] hugetlb: perform vmemmap optimization on a list of pages Date: Sun, 24 Sep 2023 17:39:47 -0700 Message-ID: <20230925003953.142620-4-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4PR03CA0213.namprd03.prod.outlook.com (2603:10b6:303:b9::8) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|DS0PR10MB7174:EE_ X-MS-Office365-Filtering-Correlation-Id: 6246bc1b-2423-4a98-7ac5-08dbbd5ffe24 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(376002)(346002)(39860400002)(366004)(396003)(136003)(230922051799003)(186009)(1800799009)(451199024)(6666004)(478600001)(6486002)(107886003)(2616005)(1076003)(26005)(83380400001)(36756003)(86362001)(38100700002)(5660300002)(6506007)(6512007)(66556008)(66476007)(54906003)(66946007)(316002)(7416002)(2906002)(44832011)(41300700001)(4326008)(8676002)(8936002);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?xzuWm6g4P9gJCcq0IjWjNymgCMafHMS/T5jshGvV4+I0HsrAcBuZIzNk1nMF?= =?us-ascii?Q?3fasB9GRikG/aYFgzlfk0TT3/vBj07p6MENqelf48eJ3JqH5WVB2cv4CSTZ8?= =?us-ascii?Q?VFemgZ2hhIqbHU2HmlT2voo9hZAFBul+ukm7Fw8pORBec1TdDl/vYIp4urFX?= =?us-ascii?Q?J4++GVl43GWov+Qc0mjTi+64jXi6+H+JFzhLeC1dIGtiSJTqbVec/XxxK801?= =?us-ascii?Q?CtGce15W0XuBJgwrOFZuABa5zuxz2yb6wy1LBjQbwK8WgUABHr45/lLRQYlo?= =?us-ascii?Q?w4Gf20c0T1HMnRc1Iy7H043Zn6t6Fjt7ctpmSKlIadEJzCFzBFWh7CgNEwXc?= =?us-ascii?Q?mVsfWtJza4qC4jO4BFxIsCI2qfXs3ra6WS5+ruXQ13al7Xb67XZ9WDkg5EnI?= =?us-ascii?Q?p/DvPftlv7Gj1fyfgTEJT3FnAjwirWY246YUP81jrlEyys1w475Lf+I6AoGt?= =?us-ascii?Q?9oj45Ki0fAD7bhwQF4yQNg+wf/tpkDRWm+HeIKgSJQG8awSegB7OVe7KdIAQ?= =?us-ascii?Q?NF5E88+WMxMHMhJsLd7ZNVw2JXWsCY7WEmYmNY1KEm515m0KalSXKdLIW8aV?= =?us-ascii?Q?Che6Scj0tdoazYq7QzYd1IsNDmVqugT++dDgQORjHGnGavMqrdIjyvDhofTF?= =?us-ascii?Q?rq8W2+69VcrFBwK4XyGuZzU9UlvHo3ARDquOExCk/QUdunG5H7BwdAiafQ5x?= =?us-ascii?Q?437oQ/2r8U7pbl+WfDOq1R485TYwO6OwAEBt3UkapdWNI4VoDMkaBZDe/ZOX?= =?us-ascii?Q?IiOYf5eVtVOR6P4vwkuASQRhpYzxhmfMbp/JmOEkS8vFyJukLeet5l6yCm1P?= =?us-ascii?Q?kDkjly5LfwIF+gVEAwV+jHgbHBzFRYxavIX78FBVC+l8zL7MfTNxovsDM6PW?= =?us-ascii?Q?z4uTIKfQieAiNFiDG8+FYnbvqhGWsfInjyLNGcXywIIv9Orwm4cDYmNCSPMT?= =?us-ascii?Q?nkZ0wV1+/PqMK8ApBCxgy7SrecoS9nrjQX3lFHHD4bQ5CcRcbY/5R23lMAwf?= =?us-ascii?Q?bNX63Hqo7VB0mfvEjJ7839x4TtB6U+dcSRppfFEFLJY59/ha7+jJ4/6YY9d9?= =?us-ascii?Q?vgHXDt0gs2f8UgUgH1A8jzahuEnazSl0ORssBOinueNp0fe51MDmv7X8UwAV?= =?us-ascii?Q?g3tmCtJ7Qs5FK3fx4J9if+5qeTNuQp/iH1T8MIexXuDPb+A3mS/hmaZNKR36?= =?us-ascii?Q?KqyK/DmLRwYrhlaj6N3CeHPYCopqsrXHyQdxrt7ClziEjFrIwGVklpQHsOIc?= =?us-ascii?Q?8GZUzZdnrQjHTZ1JGXUdi789Ow+9X3M+xt+4T/DGf619vGVkmfL5Dj2eCYTN?= =?us-ascii?Q?eC6hmY+xSYYwFZlTavYhHGTQlJIHvr0wrgM7ykS7d1TBv9uUrPLXJ3dkG+PK?= =?us-ascii?Q?FQEkC9vea4aYYvST4xi3Q08SW/sUAqlkHj6XBfwP4W7gIXJzhv4zerdVQOSM?= =?us-ascii?Q?OEhWMTVT3iqwFJI4GEb5LMu/9mT2FWCcfHlF2sTFdjn7gH1yAUm5dmlAA0dS?= =?us-ascii?Q?HNPm4dx0sQtANHxCZtWWAl3PA3pRpFJSjuhCgkPx4erZ9e02JUj+xQpxdnho?= =?us-ascii?Q?+mG/8p2H5V4fMEScTmVko+cVVBhqF7O/QttF2GF0?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?vzhDeaXjmQapCNu59SMJ0POiAwZdCzkhuJXMf5D+u5B3O+/w1OtQEmqUbQ7r?= =?us-ascii?Q?oKi4SU38icTrVeXwtXQIm1S77riK2O78+xDO9yorHjcLDILoYolJ0pxEhMzP?= =?us-ascii?Q?K7y/ZTxbg3aAE9VtQdxuegv7RQHvThpiAIA4g78Cc66Y2w+0PricFb4SpEKl?= =?us-ascii?Q?8XVcrLXiGx/D0b5kqXwOE4JeHz0dVzIEsqZw88AEisXlLJSHwK4p7Gk3DYfi?= =?us-ascii?Q?3JZe2ZYCYyR1mZUtaZcBdAOVq6purCOFsjpX4emZOffrwYO9HL7UPs58PyCZ?= =?us-ascii?Q?tKb4LHGNtuXRLL5XaHuI78pufvKGkPTuqN+yJwM6lIk1t0Rc8z1nhPz9VDbg?= =?us-ascii?Q?YbGVAz/6tDsxH65xMvaa6UK0p+M4s9PDs/ZQVd9Kom87sjq+p6wWpRgeFfFv?= =?us-ascii?Q?/ZWtlStH5nc93SJ0pUh0f3PdKrslsM6aonOXOLF0Fd2QtPuo9fJHJHhqgJ3H?= =?us-ascii?Q?7BjSIWtsZvAwrutZU7Cz5HHuTcabEPdcGRLgZ9x5FnLj6EaZCcSCqbFvemLE?= =?us-ascii?Q?eAJKcvAuZFCX94O73efi6G5fxwn3UN0e4g2V9CG+Uf4sA+mVqtGNq30yijvh?= =?us-ascii?Q?XD9cNj8MBEgQFxYzxk/AWeCg5FeRpmamlRxOhYf8TWByoUZTXwQfI2XceHy4?= =?us-ascii?Q?J0Q4LvaOnek7yT96GTkrm5Aq/sG/YgcsYYMaS5t5yl+U6vOXUZNiHONiO5gc?= =?us-ascii?Q?ASD7YoEEiwPmYfC1nJ1GHGT7O5PMSIWzEMT80Rp4OjFXAEaijc5KcCL6JaxX?= =?us-ascii?Q?qTCqRQmTRsk9ldW4xr7aaZ2q03pbVSFccrTxMeVrh0BUDrP4XS9bJzLLRj1U?= =?us-ascii?Q?5PHE1GS5wL7abLViHYWh/pWN+fx6U86Ng6YMFnRRgGz7dp2I+3OSyeZZF/Fe?= =?us-ascii?Q?l83rs+PidG/LqP7znWmfQMQfuuqzuKx+vF2YL7s5xzrkL/2hhvMqYZJugWCc?= =?us-ascii?Q?6o1lzW2rA91HDaP2tR+icRKA2zSrON5Y+vPc5pk1Q79XVOm/FZle3P0c9AME?= =?us-ascii?Q?7Z964TkgAH3+Nym+iSrM0McQfO+319zfQGAgZrV4OFE4N/OTrBBJB1ubtZka?= =?us-ascii?Q?jYLYgqujhd/B1z5tRpeofuZGw2KiDy/3Q2k83lJ6tAATodcEC1a20jvsBAr9?= =?us-ascii?Q?Lg+utQ2jMP6c?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: 6246bc1b-2423-4a98-7ac5-08dbbd5ffe24 X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:18.7828 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: Nny5f5EbPAH42zElviylXdVDTjlj7cyzKkLshO2aq9oU19pTeNgFUKhhQZOUmlUcKwoomtURqtJrUCVdpjU5Mw== X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS0PR10MB7174 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 bulkscore=0 mlxscore=0 mlxlogscore=999 suspectscore=0 phishscore=0 adultscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-ORIG-GUID: dE9xvf161eH_uNoglxSDdfY0ecc_L7N3 X-Proofpoint-GUID: dE9xvf161eH_uNoglxSDdfY0ecc_L7N3 Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" When adding hugetlb pages to the pool, we first create a list of the allocated pages before adding to the pool. Pass this list of pages to a new routine hugetlb_vmemmap_optimize_folios() for vmemmap optimization. Due to significant differences in vmemmmap initialization for bootmem allocated hugetlb pages, a new routine prep_and_add_bootmem_folios is created. We also modify the routine vmemmap_should_optimize() to check for pages that are already optimized. There are code paths that might request vmemmap optimization twice and we want to make sure this is not attempted. Signed-off-by: Mike Kravetz Reviewed-by: Muchun Song --- mm/hugetlb.c | 42 ++++++++++++++++++++++++++++++++++-------- mm/hugetlb_vmemmap.c | 11 +++++++++++ mm/hugetlb_vmemmap.h | 5 +++++ 3 files changed, 50 insertions(+), 8 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 64f50f3844fc..da0ebd370b5f 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -2269,6 +2269,9 @@ static void prep_and_add_allocated_folios(struct hsta= te *h, { struct folio *folio, *tmp_f; =20 + /* Send list for bulk vmemmap optimization processing */ + hugetlb_vmemmap_optimize_folios(h, folio_list); + /* Add all new pool pages to free lists in one lock cycle */ spin_lock_irq(&hugetlb_lock); list_for_each_entry_safe(folio, tmp_f, folio_list, lru) { @@ -3305,6 +3308,34 @@ static void __init hugetlb_folio_init_vmemmap(struct= folio *folio, prep_compound_head((struct page *)folio, huge_page_order(h)); } =20 +static void __init prep_and_add_bootmem_folios(struct hstate *h, + struct list_head *folio_list) +{ + struct folio *folio, *tmp_f; + + /* Send list for bulk vmemmap optimization processing */ + hugetlb_vmemmap_optimize_folios(h, folio_list); + + /* Add all new pool pages to free lists in one lock cycle */ + spin_lock_irq(&hugetlb_lock); + list_for_each_entry_safe(folio, tmp_f, folio_list, lru) { + if (!folio_test_hugetlb_vmemmap_optimized(folio)) { + /* + * If HVO fails, initialize all tail struct pages + * We do not worry about potential long lock hold + * time as this is early in boot and there should + * be no contention. + */ + hugetlb_folio_init_tail_vmemmap(folio, + HUGETLB_VMEMMAP_RESERVE_PAGES, + pages_per_huge_page(h)); + } + __prep_account_new_huge_page(h, folio_nid(folio)); + enqueue_hugetlb_folio(h, folio); + } + spin_unlock_irq(&hugetlb_lock); +} + /* * Put bootmem huge pages into the standard lists after mem_map is up. * Note: This only applies to gigantic (order > MAX_ORDER) pages. @@ -3325,7 +3356,7 @@ static void __init gather_bootmem_prealloc(void) * in this list. If so, process each size separately. */ if (h !=3D prev_h && prev_h !=3D NULL) - prep_and_add_allocated_folios(prev_h, &folio_list); + prep_and_add_bootmem_folios(prev_h, &folio_list); prev_h =3D h; =20 VM_BUG_ON(!hstate_is_gigantic(h)); @@ -3333,12 +3364,7 @@ static void __init gather_bootmem_prealloc(void) =20 hugetlb_folio_init_vmemmap(folio, h, HUGETLB_VMEMMAP_RESERVE_PAGES); - __prep_new_hugetlb_folio(h, folio); - /* If HVO fails, initialize all tail struct pages */ - if (!HPageVmemmapOptimized(&folio->page)) - hugetlb_folio_init_tail_vmemmap(folio, - HUGETLB_VMEMMAP_RESERVE_PAGES, - pages_per_huge_page(h)); + init_new_hugetlb_folio(h, folio); list_add(&folio->lru, &folio_list); =20 /* @@ -3350,7 +3376,7 @@ static void __init gather_bootmem_prealloc(void) cond_resched(); } =20 - prep_and_add_allocated_folios(h, &folio_list); + prep_and_add_bootmem_folios(h, &folio_list); } =20 static void __init hugetlb_hstate_alloc_pages_onenode(struct hstate *h, in= t nid) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 76682d1d79a7..4558b814ffab 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -483,6 +483,9 @@ int hugetlb_vmemmap_restore(const struct hstate *h, str= uct page *head) /* Return true iff a HugeTLB whose vmemmap should and can be optimized. */ static bool vmemmap_should_optimize(const struct hstate *h, const struct p= age *head) { + if (HPageVmemmapOptimized((struct page *)head)) + return false; + if (!READ_ONCE(vmemmap_optimize_enabled)) return false; =20 @@ -572,6 +575,14 @@ void hugetlb_vmemmap_optimize(const struct hstate *h, = struct page *head) SetHPageVmemmapOptimized(head); } =20 +void hugetlb_vmemmap_optimize_folios(struct hstate *h, struct list_head *f= olio_list) +{ + struct folio *folio; + + list_for_each_entry(folio, folio_list, lru) + hugetlb_vmemmap_optimize(h, &folio->page); +} + static struct ctl_table hugetlb_vmemmap_sysctls[] =3D { { .procname =3D "hugetlb_optimize_vmemmap", diff --git a/mm/hugetlb_vmemmap.h b/mm/hugetlb_vmemmap.h index 4573899855d7..c512e388dbb4 100644 --- a/mm/hugetlb_vmemmap.h +++ b/mm/hugetlb_vmemmap.h @@ -20,6 +20,7 @@ #ifdef CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP int hugetlb_vmemmap_restore(const struct hstate *h, struct page *head); void hugetlb_vmemmap_optimize(const struct hstate *h, struct page *head); +void hugetlb_vmemmap_optimize_folios(struct hstate *h, struct list_head *f= olio_list); =20 static inline unsigned int hugetlb_vmemmap_size(const struct hstate *h) { @@ -48,6 +49,10 @@ static inline void hugetlb_vmemmap_optimize(const struct= hstate *h, struct page { } =20 +static inline void hugetlb_vmemmap_optimize_folios(struct hstate *h, struc= t list_head *folio_list) +{ +} + static inline unsigned int hugetlb_vmemmap_optimizable_size(const struct h= state *h) { return 0; --=20 2.41.0 From nobody Fri Feb 13 12:45:13 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B9064CE7A95 for ; Mon, 25 Sep 2023 00:41:45 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230360AbjIYAlt (ORCPT ); Sun, 24 Sep 2023 20:41:49 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55366 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231499AbjIYAla (ORCPT ); Sun, 24 Sep 2023 20:41:30 -0400 Received: from mx0a-00069f02.pphosted.com (mx0a-00069f02.pphosted.com [205.220.165.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7C5FDCCA for ; Sun, 24 Sep 2023 17:41:15 -0700 (PDT) Received: from pps.filterd (m0246629.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OKqRiP013276; Mon, 25 Sep 2023 00:40:26 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=RRqnSj9V2V9kQEWWcxwgMfag+uHNl9i5WXsEUMojxeE=; b=wtvFBpcP7X+6b76qtOKdZt38nhZV684R5yUlpkX51gjUTJwU7cWTGL53QjaI0XiOSjUK +uzUq5YAbISYgIRU14NfKAGQ8mJOkuIP/guJLNZV9PWfs4BlsOC5kR7okVOKm5a/7ahe acAVo3WTbjWBnBK4YljZDjNX+og80nvocjbIbLIpaieCLo9Yr2tMVGDt2Yn+YHd0SyMk qOTWYtjx9fEA8orUd5qYuvQWbbSORaoENUKKO9UjMEi69oV1s4dy/1HbvYMvlD1vLY7m hkyC8YBoSI7c3I9VWjJhkhh7npctcoFkJLkTjnp1JTRUL5AdXY0Xc7V646UL2rJO5RUz bA== Received: from iadpaimrmta02.imrmtpd1.prodappiadaev1.oraclevcn.com (iadpaimrmta02.appoci.oracle.com [147.154.18.20]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9qwbadqb-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:26 +0000 Received: from pps.filterd (iadpaimrmta02.imrmtpd1.prodappiadaev1.oraclevcn.com [127.0.0.1]) by iadpaimrmta02.imrmtpd1.prodappiadaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38ON0wxH030950; Mon, 25 Sep 2023 00:40:25 GMT Received: from nam12-bn8-obe.outbound.protection.outlook.com (mail-bn8nam12lp2169.outbound.protection.outlook.com [104.47.55.169]) by iadpaimrmta02.imrmtpd1.prodappiadaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf9usgb-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:24 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=ZhKuE1OpGeKbr3tTK7lrTkBEvHzzmVOnp4eb5vF/3QBh9/c0TBKh+W0o5h5hzwS/t+i4NSzi01V8xBm07rxHIBpDgMWbdM3gAQmPPUP/NS/1MoBMb6BF0ybyHcc0ijNGsXVo4bErCz1qLffphl8DSFO7BqwYC4wdgdO1G67oBYCEpjv9Sp1xWX+Zm4B7sk/yGU76Q7RoytS2RSPe1q1f+vNM1aFrKyf7wR9BsTchMuoMUdEmBGwuFkSOtAxzksvf9Bzs6fH1UvaFcLrPj9Zz/Yk74cenmju4jrXiCAEX1KDd5GlEL6Twr+PMCNQ0gof7z2P2ceoH8jzRHGkllagsxg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=RRqnSj9V2V9kQEWWcxwgMfag+uHNl9i5WXsEUMojxeE=; b=FgbIVfKYBpE+vCxE7lWhwjlsohd1TYXgg3FZYIKdvc7wNvE9vQRoa/TiawOlx/zDXy6Djjtm+cfqAzh3Gt9Uo3bGjdaGkEQeLUx+K1zXZRVUz9h9rgBRu4bCgOkKPfq95rm1cbfoh0wzJVRSOZP/teVNgsJ3AUYnmeIDXOJ/cjR3EKJmOwinDJSk0izmuPr3cirbyaIOWOqepAtbw8Wzo38IYmKcg1Zs94h8ND/Zee5HMclyWUzJ6Vjj2RXj1I8Hh10m13PPehbNL3lA5W1+AJo1w6YmihnYaIxRQ0Za1vovEANmpD9/YdHYct5E0AZQHLxLkVH6ireVjGUmTcnyRg== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=RRqnSj9V2V9kQEWWcxwgMfag+uHNl9i5WXsEUMojxeE=; b=mAc5Gxuns37/uGSbKerPv0LLUqU8v1VGe99JPmcyr7GmKXQ4fvrJzLMWleWgKoNDlv7wOsOzaDFWH4ISA+DKIC5tKnJsOh472gi9b83RWxpkL5moyVU+zW1Xv5vhgqx+kKK+UpFStMKblMogE/bpJOZuEgVJjsrPJK8A+6/s3dg= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by DS0PR10MB7174.namprd10.prod.outlook.com (2603:10b6:8:df::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:22 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:22 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 4/8] hugetlb: perform vmemmap restoration on a list of pages Date: Sun, 24 Sep 2023 17:39:48 -0700 Message-ID: <20230925003953.142620-5-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4PR03CA0346.namprd03.prod.outlook.com (2603:10b6:303:dc::21) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|DS0PR10MB7174:EE_ X-MS-Office365-Filtering-Correlation-Id: 9f83b1df-a9fb-4670-b7d3-08dbbd600086 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(376002)(346002)(39860400002)(366004)(396003)(136003)(230922051799003)(186009)(1800799009)(451199024)(6666004)(478600001)(6486002)(107886003)(2616005)(1076003)(26005)(83380400001)(36756003)(86362001)(38100700002)(5660300002)(6506007)(6512007)(66556008)(66476007)(54906003)(66946007)(316002)(7416002)(2906002)(44832011)(41300700001)(4326008)(8676002)(8936002);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?uFjQq2jS7U2YH8zJM5Ssw1feLmf5lBhPeM6ySlWLFJNu/akuQPH46bQAPOM+?= =?us-ascii?Q?A9xb7cGe+9FKHHIHLA0zLJ9VW2o5Hl26dYsZKbcbf2s1VdneEFuAMf5iyTMB?= =?us-ascii?Q?5syROlxYVI51vRJATtpNfRNoEts0FG3tOjH251LR8tTl580hdBybRC2QPuz3?= =?us-ascii?Q?BM4RBZXrHjLucAgiyoVYcReD64qU3o0q3tNtOkhUtwBK0mTEQDOnBqwwBD5+?= =?us-ascii?Q?7e7IaU7LX/SQGjlI4RUx3iV/iBvYD4XG5V02P5PBW4uB3vDwqpkMawCK0oYF?= =?us-ascii?Q?fffct/QIo8gfXEI/TOTyzULOggg6VnOSM3DqE+8ya5KhOC3Qscj5vwsYonkU?= =?us-ascii?Q?Kj2M9cVjyrOBx8ZzlxXQ7LpIqHp4puiZTsSDngwD/GHa0Pnf6Fri5PUWd79m?= =?us-ascii?Q?UzaSlrikczrYkyb96N+WZKvI4KQsgKe/eOujjITW73Fg3z7Zpq8RFATfLc6K?= =?us-ascii?Q?Wi5IP5Wtb2vjlxS2fVAohY7Pp0vQRLj5S6Sn8XczVWLOL09P+lxekL7rGL8z?= =?us-ascii?Q?TPcud0I3Vf4vvlQkFpzSQ2yfQAXLgEEXSdpVT7MAdNdJHZ3OtEaTXeCRZDEe?= =?us-ascii?Q?+eCVcXK3ZfkE0J3ZfVi1FUfb9iVWk8GbSPyF2cU/ZRD8poRvnI6GfdBDP9vz?= =?us-ascii?Q?6bJY1NHePzv9qNYAORKNH99GdSHHCx2bIeEk/i7+2ErUvYzAdLikDcprIkED?= =?us-ascii?Q?U+6DUz/U8IjlTlJzsWkxWO6zFCgpsCsUF9CJ0F0x4Sx/ZsQnAZQosSjCvujR?= =?us-ascii?Q?H4NryeNOlS4KI+eG0uQwro+elLnUsx+09V0hB6iogr7b+Embm1KVEIk2rCox?= =?us-ascii?Q?NnJCyYn1nmAffATmQnxRIJ1XPKTcDLnfZXoLjhF5LZC9Ldh31gXKvFWQh1q4?= =?us-ascii?Q?+xKjSjq7dyv03xKuw7S14W6WJQERmmSPy1XgYDS/1xaOGYiL9y7ZO1m1Em24?= =?us-ascii?Q?lyv4xN3sAmJcH8sbEjoLIG4dSvj3GhymmQsxy/mDK4pFBtPkMEC4uCkMkITC?= =?us-ascii?Q?NR/bQ7eE1Sit/PP9YlmlKy+BW+rRYdoubde5vF0aCGPMzplT5h5gVzYNAfXJ?= =?us-ascii?Q?3glGsP9j37Fbw1d74XFCozr4AI1jdv54mFhYRYssXQGyDGhaSYlzT9mjeS3p?= =?us-ascii?Q?gz5PTOrg/8ZAsqr7iVVKV3Pjde0liYxXy2TLqgXO9G86yEESg2fZXEWpHnta?= =?us-ascii?Q?Ry45t21nv1BhubQ6PGK+3TphLvQrWPzEpgbSFeQDAeOIShg4dJadw+zqijuh?= =?us-ascii?Q?JQmoO131WTBmyorKTQLtP/j+ZSwphGTLqTvzcTi4dYjvFd4LNYfag5Jf0QAu?= =?us-ascii?Q?RY7ELaihJgfkEkUVvj+Bf8MWpjT6TiHldb4DfRuN7nQrIeFRFn1SpsHgw59H?= =?us-ascii?Q?eRjCL/DJDzzEi4+0cCX+02zZS3GYZPHbZoubl6nDQM+6AiHvZaGygl93l/NI?= =?us-ascii?Q?CwJ2YMhMBZPhuEv88Ay6q2hK2mJYJ3geOJdPBGrPlGLZleWt9MMWP5vAjSWr?= =?us-ascii?Q?JNwSuRh83cgmFhmw0qUz51KYEd54r0RxopgZBtbCc0RkF77Ld1/H4AWFqVQy?= =?us-ascii?Q?rP/XR6NLgd868suow0ldfyXeYEpJuFJikVQ1Dk8k?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?/gnYhUnuXgNiVlt1bj9WPxHb136/wpVHv/5uxoRVdmrqitxGlhgfDkR3K4Sc?= =?us-ascii?Q?k0T2SEIqUu8t3TrWT/OzfXk+XfAwVQaLiAwxCNjfJETmItSctw1kYjoxedLE?= =?us-ascii?Q?qrEYumJy8rdCQdrQVBOj0axGEiQjsM+oFMx9Th/bnp2ZLUckZefmi6b7pFPI?= =?us-ascii?Q?LkYQWp4cNxntGGgUv2LWaeahxGTw8cdl2IAIZ2kgP7uorqmEkmP/U2oQt913?= =?us-ascii?Q?m6jM95TzICTeoR2y+yZ8Y6xcow5q01touW7+2ZAAySM64ZR79RCf2sCEujc4?= =?us-ascii?Q?Jxl6SU1iGw900OeQqqKpIzDyfxMKyKBhAmyUqXluwKotcQcqNqy4XzCHHDE+?= =?us-ascii?Q?vZa7CdUxH3I0HszcWgPzt9dlujY+0WGy+2zAzQvNBz0gf6j6cC+30vMnP8jg?= =?us-ascii?Q?rG21V7tZ5H3ZsvGKtBMdiUOGBg0vHIfj18xXXVmbGAtb5z/L/pVMriddRmea?= =?us-ascii?Q?BMhRQz21wTcr3c4kBRlMBgsmOyWJXvqqLzpnKFyJXZ/eo27lcBUGWi1RIpAL?= =?us-ascii?Q?XB0YjGcBIPLR4oc7bexuAttSjwGFGAKPzeEHBGUEgy3iMgyySo3In8QRaVe+?= =?us-ascii?Q?INEKvuSVn+jhIxOcvPvReoXbOCs6zC7Z7JMY0nS55Zz5d6X7S+61RIwsgQ1g?= =?us-ascii?Q?6VQ5DecHOQo3PWrwUQNg/6/NEgmbMd6RPZxiGNJw0oQbMESEK8a3WSLKqrBT?= =?us-ascii?Q?b+ZtAwNZKYRpeosl6z/1Qq+CU/l+STa30CipBjfZYP6NZOliob4aeN3QaI52?= =?us-ascii?Q?gROpNMh7eWihcblAwwFOXcmf2VPP9Ycsxd44jd1Z3kFOKOfCPgztJzYtVnGO?= =?us-ascii?Q?Z4B0Pjy5DzUD5Q/7Wg1xyqsYk3LfKSMq5OCHeF7fLZ2fLKguEvomw8dO1xkC?= =?us-ascii?Q?hzbJYAesVpFJ4OlKXiVWMjvv9wYyoXs9AOwHMLOEXlL6N1LnJ2JsEzdhB+WQ?= =?us-ascii?Q?a2YNTJDxwQhcLNuJjpoHVNJUe+/IizNMfvxem+vp7bRctxL7Qx2sm3jmcSYY?= =?us-ascii?Q?Rj29NPvCau7bZhQcmXTPvabdmJtEvVY18zRS8z4r9QdavNLAnx4Il1XVJtqm?= =?us-ascii?Q?N1w+LbSWZc2xAEyVLNj1xXBXW0btq7I/E2wRBQECx9JSSSLlWH3hpWSCHw9J?= =?us-ascii?Q?b/3BllfWq2zc?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: 9f83b1df-a9fb-4670-b7d3-08dbbd600086 X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:22.7395 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: Dqm59EcUGTHSW0AKAIxjrin7LwuGtlEng0r7wmiW7BTQuhdE1cuC/HOLXsTOyKhD3KrfXn+tZWHqJ5PHilaiWQ== X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS0PR10MB7174 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 adultscore=0 spamscore=0 mlxscore=0 malwarescore=0 suspectscore=0 bulkscore=0 mlxlogscore=999 phishscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-GUID: Eb5O6Y-A3RZ-PuDzE7UVSIBLQgGwsH7h X-Proofpoint-ORIG-GUID: Eb5O6Y-A3RZ-PuDzE7UVSIBLQgGwsH7h Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The routine update_and_free_pages_bulk already performs vmemmap restoration on the list of hugetlb pages in a separate step. In preparation for more functionality to be added in this step, create a new routine hugetlb_vmemmap_restore_folios() that will restore vmemmap for a list of folios. This new routine must provide sufficient feedback about errors and actual restoration performed so that update_and_free_pages_bulk can perform optimally. Special care must be taken when encountering an error from hugetlb_vmemmap_restore_folios. We want to continue making as much forward progress as possible. A new routine bulk_vmemmap_restore_error handles this specific situation. Signed-off-by: Mike Kravetz --- mm/hugetlb.c | 98 +++++++++++++++++++++++++++++++------------- mm/hugetlb_vmemmap.c | 38 +++++++++++++++++ mm/hugetlb_vmemmap.h | 10 +++++ 3 files changed, 118 insertions(+), 28 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index da0ebd370b5f..53df35fbc3f2 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1834,50 +1834,92 @@ static void update_and_free_hugetlb_folio(struct hs= tate *h, struct folio *folio, schedule_work(&free_hpage_work); } =20 -static void update_and_free_pages_bulk(struct hstate *h, struct list_head = *list) +static void bulk_vmemmap_restore_error(struct hstate *h, + struct list_head *folio_list, + struct list_head *non_hvo_folios) { struct folio *folio, *t_folio; - bool clear_dtor =3D false; =20 - /* - * First allocate required vmemmmap (if necessary) for all folios on - * list. If vmemmap can not be allocated, we can not free folio to - * lower level allocator, so add back as hugetlb surplus page. - * add_hugetlb_folio() removes the page from THIS list. - * Use clear_dtor to note if vmemmap was successfully allocated for - * ANY page on the list. - */ - list_for_each_entry_safe(folio, t_folio, list, lru) { - if (folio_test_hugetlb_vmemmap_optimized(folio)) { + if (!list_empty(non_hvo_folios)) { + /* + * Free any restored hugetlb pages so that restore of the + * entire list can be retried. + * The idea is that in the common case of ENOMEM errors freeing + * hugetlb pages with vmemmap we will free up memory so that we + * can allocate vmemmap for more hugetlb pages. + */ + list_for_each_entry_safe(folio, t_folio, non_hvo_folios, lru) { + list_del(&folio->lru); + spin_lock_irq(&hugetlb_lock); + __clear_hugetlb_destructor(h, folio); + spin_unlock_irq(&hugetlb_lock); + update_and_free_hugetlb_folio(h, folio, false); + cond_resched(); + } + } else { + /* + * In the case where there are no folios which can be + * immediately freed, we loop through the list trying to restore + * vmemmap individually in the hope that someone elsewhere may + * have done something to cause success (such as freeing some + * memory). If unable to restore a hugetlb page, the hugetlb + * page is made a surplus page and removed from the list. + * If are able to restore vmemmap and free one hugetlb page, we + * quit processing the list to retry the bulk operation. + */ + list_for_each_entry_safe(folio, t_folio, folio_list, lru) if (hugetlb_vmemmap_restore(h, &folio->page)) { spin_lock_irq(&hugetlb_lock); add_hugetlb_folio(h, folio, true); spin_unlock_irq(&hugetlb_lock); - } else - clear_dtor =3D true; - } + } else { + list_del(&folio->lru); + spin_lock_irq(&hugetlb_lock); + __clear_hugetlb_destructor(h, folio); + spin_unlock_irq(&hugetlb_lock); + update_and_free_hugetlb_folio(h, folio, false); + cond_resched(); + break; + } } +} + +static void update_and_free_pages_bulk(struct hstate *h, + struct list_head *folio_list) +{ + long ret; + struct folio *folio, *t_folio; + LIST_HEAD(non_hvo_folios); =20 /* - * If vmemmmap allocation was performed on any folio above, take lock - * to clear destructor of all folios on list. This avoids the need to - * lock/unlock for each individual folio. - * The assumption is vmemmap allocation was performed on all or none - * of the folios on the list. This is true expect in VERY rare cases. + * First allocate required vmemmmap (if necessary) for all folios. + * Carefully handle errors and free up any available hugetlb pages + * in an effort to make forward progress. */ - if (clear_dtor) { +retry: + ret =3D hugetlb_vmemmap_restore_folios(h, folio_list, &non_hvo_folios); + if (ret < 0) { + bulk_vmemmap_restore_error(h, folio_list, &non_hvo_folios); + goto retry; + } + + /* + * At this point, list should be empty, ret should be >=3D 0 and there + * should only be pages on the non_hvo_folios list. + * Do note that the non_hvo_folios list could be empty. + * Without HVO enabled, ret will be 0 and there is no need to call + * __clear_hugetlb_destructor as this was done previously. + */ + VM_WARN_ON(!list_empty(folio_list)); + VM_WARN_ON(ret < 0); + if (!list_empty(&non_hvo_folios) && ret) { spin_lock_irq(&hugetlb_lock); - list_for_each_entry(folio, list, lru) + list_for_each_entry(folio, &non_hvo_folios, lru) __clear_hugetlb_destructor(h, folio); spin_unlock_irq(&hugetlb_lock); } =20 - /* - * Free folios back to low level allocators. vmemmap and destructors - * were taken care of above, so update_and_free_hugetlb_folio will - * not need to take hugetlb lock. - */ - list_for_each_entry_safe(folio, t_folio, list, lru) { + list_for_each_entry_safe(folio, t_folio, &non_hvo_folios, lru) { update_and_free_hugetlb_folio(h, folio, false); cond_resched(); } diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 4558b814ffab..77f44b81ff01 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -480,6 +480,44 @@ int hugetlb_vmemmap_restore(const struct hstate *h, st= ruct page *head) return ret; } =20 +/** + * hugetlb_vmemmap_restore_folios - restore vmemmap for every folio on the= list. + * @h: hstate. + * @folio_list: list of folios. + * @non_hvo_folios: Output list of folios for which vmemmap exists. + * + * Return: number of folios for which vmemmap was restored, or an error co= de + * if an error was encountered restoring vmemmap for a folio. + * Folios that have vmemmap are moved to the non_hvo_folios + * list. Processing of entries stops when the first error is + * encountered. The folio that experienced the error and all + * non-processed folios will remain on folio_list. + */ +long hugetlb_vmemmap_restore_folios(const struct hstate *h, + struct list_head *folio_list, + struct list_head *non_hvo_folios) +{ + struct folio *folio, *t_folio; + long restored =3D 0; + long ret =3D 0; + + list_for_each_entry_safe(folio, t_folio, folio_list, lru) { + if (folio_test_hugetlb_vmemmap_optimized(folio)) { + ret =3D hugetlb_vmemmap_restore(h, &folio->page); + if (ret) + break; + restored++; + } + + /* Add non-optimized folios to output list */ + list_move(&folio->lru, non_hvo_folios); + } + + if (!ret) + ret =3D restored; + return ret; +} + /* Return true iff a HugeTLB whose vmemmap should and can be optimized. */ static bool vmemmap_should_optimize(const struct hstate *h, const struct p= age *head) { diff --git a/mm/hugetlb_vmemmap.h b/mm/hugetlb_vmemmap.h index c512e388dbb4..0b7710f90e38 100644 --- a/mm/hugetlb_vmemmap.h +++ b/mm/hugetlb_vmemmap.h @@ -19,6 +19,9 @@ =20 #ifdef CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP int hugetlb_vmemmap_restore(const struct hstate *h, struct page *head); +long hugetlb_vmemmap_restore_folios(const struct hstate *h, + struct list_head *folio_list, + struct list_head *non_hvo_folios); void hugetlb_vmemmap_optimize(const struct hstate *h, struct page *head); void hugetlb_vmemmap_optimize_folios(struct hstate *h, struct list_head *f= olio_list); =20 @@ -45,6 +48,13 @@ static inline int hugetlb_vmemmap_restore(const struct h= state *h, struct page *h return 0; } =20 +static long hugetlb_vmemmap_restore_folios(const struct hstate *h, + struct list_head *folio_list, + struct list_head *non_hvo_folios) +{ + return 0; +} + static inline void hugetlb_vmemmap_optimize(const struct hstate *h, struct= page *head) { } --=20 2.41.0 From nobody Fri Feb 13 12:45:13 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id CFF69CE7A8B for ; Mon, 25 Sep 2023 00:41:15 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231421AbjIYAlT (ORCPT ); Sun, 24 Sep 2023 20:41:19 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47074 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231386AbjIYAlO (ORCPT ); Sun, 24 Sep 2023 20:41:14 -0400 Received: from mx0b-00069f02.pphosted.com (mx0b-00069f02.pphosted.com [205.220.177.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 51324FF for ; Sun, 24 Sep 2023 17:41:03 -0700 (PDT) Received: from pps.filterd (m0246630.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OLsBVB012416; Mon, 25 Sep 2023 00:40:29 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=o8COXuu6/TtJ5IDPv6UZqVnZDWI+bSmwncS3etnX9L8=; b=Yuew2zGAFlLf+GSdB6WUQc0eR1wB367C/h7u89fejbcki5JbIEAzy07125Kla/KcL7Lm NDZkc9qIUCEMUKgSCoScCDrBlzVRN1np1gjsnDtGHE9xqLOoa34hmwTlGmET6IY4IP92 DgWtG4uex2tRPJ+9+pyk/Q4D2Yj4c/0xtnS2xQ8umfAeZUa5ySvnUgJPj4FicYHIyPgJ 22iOXq9xWY8hhJsY3u5ZivHKgngc9plDcN1aicojg3CHicPnV2wfeer0n7fcuh4T+iGy 5xfwnwzUpCP1+4FIOPjK/wn1synDUe1v05pI1UQ8f69pkLAqXw9543BvaFrfVxauCbl4 kg== Received: from phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta01.appoci.oracle.com [138.1.114.2]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9pee2er8-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:29 +0000 Received: from pps.filterd (phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38ON16v4030612; Mon, 25 Sep 2023 00:40:28 GMT Received: from nam12-bn8-obe.outbound.protection.outlook.com (mail-bn8nam12lp2168.outbound.protection.outlook.com [104.47.55.168]) by phxpaimrmta01.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf3kx5m-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:28 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=n4klr3ScJFtCkvuCRgcYgXA+KZ/JPkZlQEHcXJd5Z4NzOSj2ynNEdRmuaKsLgYR7bvVN/ommf73H+30h3o8YKUS4Q/8PAa3dO6vIzM7Gb2lFjRmmuizXmnprilH9Ka/filizGQEj/7A8SraCybvOhhzFbCB7Of4LyB4w2jSOmLvmaOcT+sxxcFJrDxidRKF3aa3NKsSjqrfgh+RZyaxG2nCefk7jMW9AfY/SvwrDKjEmyHTNi8WO5f7DYmsIrhtqeKMAdZLjdyo1H+v6f64AcyVhIvvK4+DEFbDhOP2Uj0/JFTbo2HErMH+8oV9mWXETyrAQMEHX7pEWwlyYvaMYQw== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=o8COXuu6/TtJ5IDPv6UZqVnZDWI+bSmwncS3etnX9L8=; b=GUJpr7eU0RvQGfJvfomQuo6/LcmjFavwwX14V+rd8PzIRLEHgPqFs+3b/DTXpH33Zt0rIglUyzlDkAETpXowYzj6w1m5P2M9/R47hcagMjQfagrovVQXy129S8m59O2gtrGuKuhz80O5ZsOcXHCzLYCsfNgIHbj8ge8Hlgj/4SNPGiECipPnEJmzWIUa6XScp453T3WPuvcglk4DDxhcYOSWFjTc+432jJC7hAM8KsU4/g7miqVXVwIJc5CVqHy0oerWsfaZ2isigH6HWmcdt3zJzrH03hPitYuDCZKGsqgAYXbgpbM7Qd4H1F4m2d9zeJyJy3uVlLcotu0THV6Zrg== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=o8COXuu6/TtJ5IDPv6UZqVnZDWI+bSmwncS3etnX9L8=; b=zHczSrmxLbuzD11AkyzH94PsVcnCCbxCYtM5ZcYjnORh56lW67tbNIQqMCKZKzZSBbog5450jDJdCI7zAN5dToWvyJO9FS0VmkDO/Thw1LueYJF/2ZEenGvsDz0yqoJ2gXvKaDZntzk2g6mMKL/8Jag0i/nAfmDbnX4RBr8nkLs= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by DS0PR10MB7174.namprd10.prod.outlook.com (2603:10b6:8:df::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:25 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:25 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 5/8] hugetlb: batch freeing of vmemmap pages Date: Sun, 24 Sep 2023 17:39:49 -0700 Message-ID: <20230925003953.142620-6-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4PR03CA0241.namprd03.prod.outlook.com (2603:10b6:303:b4::6) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|DS0PR10MB7174:EE_ X-MS-Office365-Filtering-Correlation-Id: a086f848-ddd3-481e-7a84-08dbbd600244 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(376002)(346002)(39860400002)(366004)(396003)(136003)(230922051799003)(186009)(1800799009)(451199024)(6666004)(478600001)(6486002)(107886003)(2616005)(1076003)(26005)(83380400001)(36756003)(86362001)(38100700002)(5660300002)(6506007)(6512007)(66556008)(66476007)(54906003)(66946007)(316002)(7416002)(2906002)(44832011)(41300700001)(4326008)(8676002)(8936002);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?kdYfMeLYzhY9KOqx1pXc5m7EEfEYHEWPnA0frMueO5fN3iKMvXSFB2t6A23s?= =?us-ascii?Q?6GlVxSxuEgcCEclxAqH2k1dRNk5EUxYVZCgnHH8qARfR4ugNIxo+dRDM3P2f?= =?us-ascii?Q?eDWATOmsOXoTI4pxS1HttnvWLpYSw/TIP6G/s1dyu3E+filijddF4E0TOcHE?= =?us-ascii?Q?H5Uf0jeUvxgRLcXGYjHJ22rJzsZSDYmi518a6zD/t6m9orK++D8wMagtSqgm?= =?us-ascii?Q?aN/Ew4Ikgij1JKwXTbwXggwjhhhO5H/Ia/d49Sxxox9h/HyKA5Yb4FmbkamO?= =?us-ascii?Q?sBvMGW7FifjYXPut7DkUkuMkWZ8mjstrTvfjtFg92686VBDGHg9aUORBFbe2?= =?us-ascii?Q?8KKLKPJ67oOZfNEfNHoSvO7q/UGMIhZSkLSKzZLIl1/aXL8Dixh05UM6ilb9?= =?us-ascii?Q?h4NB0JkL4UJhkSbTT/KGZuZuRHm0g64qszQYZwK3m//eaZ08eklM6CPb+zww?= =?us-ascii?Q?yreV7drwnOC9D+cvGXDzGE756434SWmDhpbwlf/htWuI11EbzpcHcMH1X/S1?= =?us-ascii?Q?o/ZkjJsksD86bIbu1wXp/8Yjz7jgeYaTlzQVj1L3rOlkWF4OBGQeRHgKXdjz?= =?us-ascii?Q?RVS1u4qGtb/l5nyTUoTJ3svZkQ5JrZhc+RWSffy7QM+q95Ym0JmtxT30Vx7Q?= =?us-ascii?Q?sT3fz8RJL0YznnvN55bZ4qxtlkXkgRr2bscKPiirQzFKmCQmcDVYXLoKA3y3?= =?us-ascii?Q?jtJCNzmW43MYe6F5MQMUHyBi0ssL4ewEimPdyyDQ/EDkyfFoj1qiyserG3aM?= =?us-ascii?Q?PEIByj2TSuxzbkigjTWohTr0xthswWADqkmNaYfWojKkuRl25y+P9xyRlCtB?= =?us-ascii?Q?xwx0TsyGNvtyhApPZA2urDkRn2XVG9ZDNKB6CN4JFDVI+rI/nRoKVJIwBQDn?= =?us-ascii?Q?vSnPpbZtjN/Qo1j4CWdKM8Q/DZy0r5izh/8L98z1MC56uJ6AvqFQiHH2lHm5?= =?us-ascii?Q?sUfuv25ukN+PxzqUdQoGpIxtmTsYqbcu3qblaAIsGOsOFR2/qsDL8CCoTtXy?= =?us-ascii?Q?RFvbRatQsxoqVpBU4lL98V95XxTvCkESYylMPSMTe34lbnQqLb1lUCE0pdhz?= =?us-ascii?Q?yMIXC1zJXs1thlzrSCtg4yiOGFGt8/KNmbvswKiJT4agcGZmp6EAFZJ8bmX3?= =?us-ascii?Q?O+y66nOnV2pKSptZyjY2dQZgKTXn0S4VwdAqw1nQNGavpPc0EbRAsk7K7qAG?= =?us-ascii?Q?/iZrvq3JRuAx7WGX/Rl6pvqTxOxLOuivS+RVdDmd4upLm6/EOq5wogbieBI1?= =?us-ascii?Q?26ec4kX/I15HkiBWm4uWg29Rc0RS2mcCfFiCtjq+J8ZEAFFnw5cU1evZErBT?= =?us-ascii?Q?JvYfqls52CetV+P4Vnv0JRxXR62ASJQ1hHKg4dUmZZJJslkh74EqxIQJTrNx?= =?us-ascii?Q?9Xnu8QfNxG1Jp2sOuVA7Ag40gcmECA/77QkwM6V5l6GfhDBsrC3ljeLNUbSC?= =?us-ascii?Q?tp6PwmobaKx+y/DSyM0aByEyrBs4cyyLhV8WdgOanp+GMOR8eF+x+m0aqJKR?= =?us-ascii?Q?3sIiUtoiBNvh1Zv4WE+wpFmxLnfHhrMuMcdvFFOYO5Ix14OgbtMxwPhVLd5O?= =?us-ascii?Q?ctU6Qb2zuVk/tnBSMsWiDF75aSZ8yHPRzrXaTvYI?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?VtVkcdYxMg+3hLA/Ll2dGWsOVUmocMKZx3zxl/d8PAirh6Ix31SeyFnTFAu7?= =?us-ascii?Q?eCLqpgNdyIof59/EHAtRw9+yio9uFiEGAWCGKjIvPwsVVTDj4aE6z0SE/Lo2?= =?us-ascii?Q?Pb4/aEQ34FLE+dzz/+dP644oGJqci0ti5TIp01LgMMa4K/12B8+EJAvt9Llo?= =?us-ascii?Q?njoOR5fo8DuZGA0GQI16VNL1hjoKgg9uoNaY4a7s40XOc5BmRDqZSvdgG+9j?= =?us-ascii?Q?VjwWf5jsg9k9doLkIO7AkrzbponKpDvjvLevNeutoXiV34QeM/OjTtC9ycos?= =?us-ascii?Q?S14pWT++ckUd6o6buB/4+RQNAO1NjLmpzIz2wHtn567U1KCu0yOxXVZ5ECW4?= =?us-ascii?Q?MIW3/rqmfINJhnT3Iuf8IEW/NdM78XQ0HgOxjBV8EPPEfwx7kk5uQM7kOEPG?= =?us-ascii?Q?3Y8TeojersaU0orpoXpWDV6u+AUC4Qu53t0EE2/+WTwJemKf82vVhzWNn97R?= =?us-ascii?Q?aBd5T+riny1CD8/QPuRWZK+O+u0WXQ+gzfS4Ym4kup3ZFI6cdIxve6bS8bsR?= =?us-ascii?Q?P8wV+/rni8Nq8kmeYIoKsTvynEvmO+fV4AX41Y4qQFoNa5l3ysSyqLFKiWX1?= =?us-ascii?Q?2PYocR9f1LqyxUlgtrcjcvvMko9vh34NEC7xHohq9DEC3aPYfIB/pKgxpUIX?= =?us-ascii?Q?nSR8RmZra5PkqiRQ44BRdVQU/XjeZ1djCN35HIbVnZMXXn2pRKnzC+zrNHHt?= =?us-ascii?Q?dBkbrxgYapFImI0kvdiyOCiKFCUfccVRKs3feovEmR9SsCeoMFz48qrXX/rF?= =?us-ascii?Q?/kO88PsqaP5xECuxGZDKugjZs/8AtHWRonM14p93yqp9/VMQlp10WmteP/Om?= =?us-ascii?Q?nQSD9YWfGIz7L7KpyXyWiUd5BO4fY2O32I0Wl9Z0ullyavrDwji/mSqGxMpB?= =?us-ascii?Q?qWik/rYDrZxgaOCJIjjnc/M1g+0d6e0JBNqb4bCBUoqaHfam+qGVNdTCNNj1?= =?us-ascii?Q?tebf6Z0yKh/xBBu0PyfNTnEjVNuCybD4yz1sgLfQajIjUh6RVVymbe6smL2J?= =?us-ascii?Q?30ES30jpB98bfO94dn6szgykf7WVye1JrnnFIyeZL3eOc7BAp6vJBsxEs5Hf?= =?us-ascii?Q?gocsGRioHMx3cvpZWilvbz2Fz0EhTxDzuMmaOBzCFu8i0bL+o1piE3gWRcyV?= =?us-ascii?Q?SpwcEQafZIMd?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: a086f848-ddd3-481e-7a84-08dbbd600244 X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:25.6591 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: HQtMswecAE5UcXhpY3XkHarBN8f9JRva9EcwfFFYXTdl74A/RjVnUoEJsrTSpTSalfb9k2M4rmGZi/JIe4A81g== X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS0PR10MB7174 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 suspectscore=0 phishscore=0 mlxscore=0 malwarescore=0 mlxlogscore=999 bulkscore=0 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-ORIG-GUID: iqZ_EHcWbixoAeEcHiHuCI4XdJF8jeHx X-Proofpoint-GUID: iqZ_EHcWbixoAeEcHiHuCI4XdJF8jeHx Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Now that batching of hugetlb vmemmap optimization processing is possible, batch the freeing of vmemmap pages. When freeing vmemmap pages for a hugetlb page, we add them to a list that is freed after the entire batch has been processed. This enhances the ability to return contiguous ranges of memory to the low level allocators. Signed-off-by: Mike Kravetz Reviewed-by: Muchun Song --- mm/hugetlb_vmemmap.c | 82 ++++++++++++++++++++++++++++++-------------- 1 file changed, 56 insertions(+), 26 deletions(-) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 77f44b81ff01..4ac521e596db 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -251,7 +251,7 @@ static void vmemmap_remap_pte(pte_t *pte, unsigned long= addr, } =20 entry =3D mk_pte(walk->reuse_page, pgprot); - list_add_tail(&page->lru, walk->vmemmap_pages); + list_add(&page->lru, walk->vmemmap_pages); set_pte_at(&init_mm, addr, pte, entry); } =20 @@ -306,18 +306,20 @@ static void vmemmap_restore_pte(pte_t *pte, unsigned = long addr, * @end: end address of the vmemmap virtual address range that we want to * remap. * @reuse: reuse address. + * @vmemmap_pages: list to deposit vmemmap pages to be freed. It is calle= rs + * responsibility to free pages. * * Return: %0 on success, negative error code otherwise. */ static int vmemmap_remap_free(unsigned long start, unsigned long end, - unsigned long reuse) + unsigned long reuse, + struct list_head *vmemmap_pages) { int ret; - LIST_HEAD(vmemmap_pages); struct vmemmap_remap_walk walk =3D { .remap_pte =3D vmemmap_remap_pte, .reuse_addr =3D reuse, - .vmemmap_pages =3D &vmemmap_pages, + .vmemmap_pages =3D vmemmap_pages, }; int nid =3D page_to_nid((struct page *)reuse); gfp_t gfp_mask =3D GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN; @@ -334,7 +336,7 @@ static int vmemmap_remap_free(unsigned long start, unsi= gned long end, if (walk.reuse_page) { copy_page(page_to_virt(walk.reuse_page), (void *)walk.reuse_addr); - list_add(&walk.reuse_page->lru, &vmemmap_pages); + list_add(&walk.reuse_page->lru, vmemmap_pages); } =20 /* @@ -365,15 +367,13 @@ static int vmemmap_remap_free(unsigned long start, un= signed long end, walk =3D (struct vmemmap_remap_walk) { .remap_pte =3D vmemmap_restore_pte, .reuse_addr =3D reuse, - .vmemmap_pages =3D &vmemmap_pages, + .vmemmap_pages =3D vmemmap_pages, }; =20 vmemmap_remap_range(reuse, end, &walk); } mmap_read_unlock(&init_mm); =20 - free_vmemmap_page_list(&vmemmap_pages); - return ret; } =20 @@ -389,7 +389,7 @@ static int alloc_vmemmap_page_list(unsigned long start,= unsigned long end, page =3D alloc_pages_node(nid, gfp_mask, 0); if (!page) goto out; - list_add_tail(&page->lru, list); + list_add(&page->lru, list); } =20 return 0; @@ -577,24 +577,17 @@ static bool vmemmap_should_optimize(const struct hsta= te *h, const struct page *h return true; } =20 -/** - * hugetlb_vmemmap_optimize - optimize @head page's vmemmap pages. - * @h: struct hstate. - * @head: the head page whose vmemmap pages will be optimized. - * - * This function only tries to optimize @head's vmemmap pages and does not - * guarantee that the optimization will succeed after it returns. The call= er - * can use HPageVmemmapOptimized(@head) to detect if @head's vmemmap pages - * have been optimized. - */ -void hugetlb_vmemmap_optimize(const struct hstate *h, struct page *head) +static int __hugetlb_vmemmap_optimize(const struct hstate *h, + struct page *head, + struct list_head *vmemmap_pages) { + int ret =3D 0; unsigned long vmemmap_start =3D (unsigned long)head, vmemmap_end; unsigned long vmemmap_reuse; =20 VM_WARN_ON_ONCE(!PageHuge(head)); if (!vmemmap_should_optimize(h, head)) - return; + return ret; =20 static_branch_inc(&hugetlb_optimize_vmemmap_key); =20 @@ -604,21 +597,58 @@ void hugetlb_vmemmap_optimize(const struct hstate *h,= struct page *head) =20 /* * Remap the vmemmap virtual address range [@vmemmap_start, @vmemmap_end) - * to the page which @vmemmap_reuse is mapped to, then free the pages - * which the range [@vmemmap_start, @vmemmap_end] is mapped to. + * to the page which @vmemmap_reuse is mapped to. Add pages previously + * mapping the range to vmemmap_pages list so that they can be freed by + * the caller. */ - if (vmemmap_remap_free(vmemmap_start, vmemmap_end, vmemmap_reuse)) + ret =3D vmemmap_remap_free(vmemmap_start, vmemmap_end, vmemmap_reuse, vme= mmap_pages); + if (ret) static_branch_dec(&hugetlb_optimize_vmemmap_key); else SetHPageVmemmapOptimized(head); + + return ret; +} + +/** + * hugetlb_vmemmap_optimize - optimize @head page's vmemmap pages. + * @h: struct hstate. + * @head: the head page whose vmemmap pages will be optimized. + * + * This function only tries to optimize @head's vmemmap pages and does not + * guarantee that the optimization will succeed after it returns. The call= er + * can use HPageVmemmapOptimized(@head) to detect if @head's vmemmap pages + * have been optimized. + */ +void hugetlb_vmemmap_optimize(const struct hstate *h, struct page *head) +{ + LIST_HEAD(vmemmap_pages); + + __hugetlb_vmemmap_optimize(h, head, &vmemmap_pages); + free_vmemmap_page_list(&vmemmap_pages); } =20 void hugetlb_vmemmap_optimize_folios(struct hstate *h, struct list_head *f= olio_list) { struct folio *folio; + LIST_HEAD(vmemmap_pages); + + list_for_each_entry(folio, folio_list, lru) { + int ret =3D __hugetlb_vmemmap_optimize(h, &folio->page, + &vmemmap_pages); + + /* + * Pages to be freed may have been accumulated. If we + * encounter an ENOMEM, free what we have and try again. + */ + if (ret =3D=3D -ENOMEM && !list_empty(&vmemmap_pages)) { + free_vmemmap_page_list(&vmemmap_pages); + INIT_LIST_HEAD(&vmemmap_pages); + __hugetlb_vmemmap_optimize(h, &folio->page, &vmemmap_pages); + } + } =20 - list_for_each_entry(folio, folio_list, lru) - hugetlb_vmemmap_optimize(h, &folio->page); + free_vmemmap_page_list(&vmemmap_pages); } =20 static struct ctl_table hugetlb_vmemmap_sysctls[] =3D { --=20 2.41.0 From nobody Fri Feb 13 12:45:13 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 59E90CE7A8B for ; Mon, 25 Sep 2023 00:41:29 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231537AbjIYAlc (ORCPT ); Sun, 24 Sep 2023 20:41:32 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47118 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230418AbjIYAlQ (ORCPT ); Sun, 24 Sep 2023 20:41:16 -0400 Received: from mx0b-00069f02.pphosted.com (mx0b-00069f02.pphosted.com [205.220.177.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 96145193 for ; Sun, 24 Sep 2023 17:41:07 -0700 (PDT) Received: from pps.filterd (m0246631.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMgE6b008084; Mon, 25 Sep 2023 00:40:32 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=3vbONjsZvravfMZFWf/jFmc0JqQCVqxQwTo9QP2uYlc=; b=F+PJBwmzztUsekG8vqtAwbqjm85aQ3HlPAI7UjIA3NEcdfReVKXxQ0MghsOmgYudBHQM KBw4MPBGBuFUUc9fVlPrOMTEfnX3Nb/zXaMMZqhMOUNx+1lfTrNQuvrlZPxxj3SB/5e4 YQIeZGZE2TlOo9Jkikp8StY/vrPyqYQK3y7WXK4SLf16NgHRaZeC0w2/LNsXywX3c12+ YXTFfia1DInbctJbEkxqP5Q2FZ6eF+5bX21c+ZcdJTnJXMKYWLITzSKKI1Tq/e1S/FIR ylof5Wi4BniuJCDZW6e4fkHVXocAmUvV+EWbyLVuuc2l3n2eWRNFqYHHMJ9fUuBBCrtR 7w== Received: from phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta03.appoci.oracle.com [138.1.37.129]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9pt3jebu-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:32 +0000 Received: from pps.filterd (phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMdVUJ034972; Mon, 25 Sep 2023 00:40:31 GMT Received: from nam12-bn8-obe.outbound.protection.outlook.com (mail-bn8nam12lp2174.outbound.protection.outlook.com [104.47.55.174]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf418cq-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:31 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=B6NbKBBGo9dhIS3jE2z6o3KkaJNh2MaAynLWX+yOGt7uDIR3iCBL+zzXXlNyvADOBAgRZE1sGzlxwlFg5VrWss+il7NsNlKzGD+rT1f3swsqeD7j6CPLOH0SeYhGVJ944QTkT4E/3rNDOAzE5sGXX7ARF8Z3jXXxOIbYUOg9nDb+Lz+U7ozLQIaULg6771p4kacIHWxbFwAaCFBLxW4LnjS+qJXYHtqM6Q23Lbr1Z+agT/4O13mxF9ODiY4I2GVdjzWgmdh8HtByr+qzFX9kccv2DzhzVPliYQ8Co+01tKjt9KDhsMjRyzNGEwrr9954KFHfUIo/VHo53Iio6fkSYg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=3vbONjsZvravfMZFWf/jFmc0JqQCVqxQwTo9QP2uYlc=; b=EY6rrnmLxNuwEYdk0CnZ4iUN6HJHaTuh9fmGgu5HtZYUrVMBa5/MNDUqJuW0h/a/86NuioVXCE7JdK629jffrWXnQnpGbKJPOFCxCzLRmdgeIA3RsKmu3tp3c6AWIprHcqZOFLER+E/gj1CIVegxW3l8MeoMVoTyO/b2MYAHIQ2ZUc7ASOh0VxX+0rPyXA/zzjZI94q25XGP4SIabkbgHzZZW2sxAlsFot1UWo4qvIuZENtddomrM6raSeJ+BOOzaWHSXTb+izpqp/JuU2txFBEvu5CG1rVas26aV3dHm4hdwgOuQfb2Iqo3LUXwz1R5VrfJ+2WVdyVa/YVVRU76oQ== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=3vbONjsZvravfMZFWf/jFmc0JqQCVqxQwTo9QP2uYlc=; b=QLUXRsi2Gidkbk3HfEX8CN+VowLG44K8d1ep5PiOKfLXcDB/3Ah8AL6Pew9f0pSJVdv/MBqsMxt8DOboVdYRj6GfYjD2y7slGX3zN4fVRQ00wvdjBN7kYA9YEpuhVK/pOe06wvusGE7VqyPnGNEOKKQhgovnPwIF/CBDCBNi9Fo= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by DS0PR10MB7174.namprd10.prod.outlook.com (2603:10b6:8:df::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:28 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:28 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 6/8] hugetlb: batch PMD split for bulk vmemmap dedup Date: Sun, 24 Sep 2023 17:39:50 -0700 Message-ID: <20230925003953.142620-7-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4PR03CA0110.namprd03.prod.outlook.com (2603:10b6:303:b7::25) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|DS0PR10MB7174:EE_ X-MS-Office365-Filtering-Correlation-Id: 3e3a11d5-1b6e-4528-78bc-08dbbd6003fd X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(376002)(346002)(39860400002)(366004)(396003)(136003)(230922051799003)(186009)(1800799009)(451199024)(6666004)(478600001)(6486002)(107886003)(2616005)(1076003)(26005)(83380400001)(36756003)(86362001)(38100700002)(5660300002)(6506007)(6512007)(66556008)(66476007)(54906003)(66946007)(316002)(7416002)(2906002)(44832011)(41300700001)(4326008)(8676002)(8936002);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?9JT88nIvbWRpTRyc06e0Fwj8iWg7/5ALT9WD36k3MxHe//Kya4IHIxFmVUzP?= =?us-ascii?Q?W/pLQ3D7SBqD0Y1t94v+ZRvOvJs/ZHV/4ElXyAApqxfhF+CdHXbVOqFtVAPc?= =?us-ascii?Q?+BZwE4lYICa8YJMDRfiYzRmFvb3Z4BcQ55rSj22Mze+gAUPKiQYuM4IWnX4Z?= =?us-ascii?Q?Q0tTWdGaBEHq/wSTnX7iyhFfGN4zwxCiwu5VjsIz+5Qc9udJCgU5wYg1T/ih?= =?us-ascii?Q?VSY+MKtI//4oyZYfKXPPO/JwqlQe0Dj5H2GDZT1vv0xREbyRX7aiNtFOJRKj?= =?us-ascii?Q?Qb+SmP5Fu1naS92nfubgKo2xTBsQjvnKHmoAWAS/6srX91P+gm3FqBwR7liY?= =?us-ascii?Q?EhBPRzk1NPcZx0jRV8bnGHwskWBiuXWnAyju2sxRu4B497IQikgyRAaeUcLO?= =?us-ascii?Q?LsTzE3DhUYrZTrmhY76cRAtL7dvuXPhPkH2Ctc2xzlwLi0YQ0xh8TB4Fn/PY?= =?us-ascii?Q?WK13ZzStiCZ3vIoC27BfRN/udsJoBobiNLuXXXrzSaDvNFWmw1vFY3FCEzrs?= =?us-ascii?Q?i+FjF0tqIuw5SgVMZqpW2jjl0NTzAfCUHcAHOe4sE/WurAnznYMhZdunWzAj?= =?us-ascii?Q?t78TRRw5micTYOokv3EmiVSpdkDT3EN6yDICe93GRNmAgF4NJb8nE2oH5MFf?= =?us-ascii?Q?5YzQzOTvPA3fw89NMgyNAM5djFoNX7lTFP3B4wWejIkE/tmv1c6N05xexSNM?= =?us-ascii?Q?wVKB2rA7AZzU1BNIWiNa5YWrws/OWtumBVwWOUS/DYbea6038SJ8S8I3KZeg?= =?us-ascii?Q?MecF/tz3qKokWOOnCSfcWy00eL+bFfwzzfcgRiuXX2fsSbn8ePHh19+peN7g?= =?us-ascii?Q?8ch+UrJCHuvN6LG8SzVWp5Bqcjq5xr/sndCZ7NiGNu5pq2ElBUmWYQY2h9yS?= =?us-ascii?Q?BVJKOUZyWuA6mC1A006em9jIMSfiDzo7eKXhHjw1PJT+5o8rRAPM/A6IVtXh?= =?us-ascii?Q?+uhKrZGQukKwFPq3rwzSZ1YSuEJ5k+eubU8tNK7zSaqoHrACMP2cEcxRAtyO?= =?us-ascii?Q?vj7uh9ljC9MRXHqKyiYUetiw14NqqBxigJpDCIPoSXMuJgim9knW/lV+l0l3?= =?us-ascii?Q?/hC8rJ0OacH3bAzG1WX03xRAdBwLV9tlcg4VidlCICrOzD3SUu+IvSmumkcB?= =?us-ascii?Q?LzIMRzKmqsvJEFkJFhGgTWqofSC74uFsVX6VDhCVYZfg8od9LYt9YsryTiMA?= =?us-ascii?Q?/Co3O+ZroKad2N7sIHcHUuhR15lU5pqudJsCSXRmm3TmmHIS2btmX/P/Z4mT?= =?us-ascii?Q?K12jJFyrmkYLW0tgdlWdCJC7zrZNKWnSMXM4IGMsSBR2GenV3oXF+GNDLEhM?= =?us-ascii?Q?7z7RVTC/maZr4sVlS875JzedEO9OUn/Cn+aFjudMxdXlH+nMZDwow62zM3vI?= =?us-ascii?Q?8XtqoGIw381TiD3HSnPzrnLbGo+6P9FUWOIBLuDqCXheaiicbRh+yJmBnvT4?= =?us-ascii?Q?TyE7xeMPtiLcEAyKy5vJkHjW39HiB6jP86Np8/iicN3K2zWHJHsjoCvtxjFA?= =?us-ascii?Q?Cnsbm3mnnf4rmiUu7Ym0Z1k/aafJuer+x4A26vDik2Jv35n+iCrPqgP7O9fI?= =?us-ascii?Q?MnzsO9rO2qCNzpKgQ1F/tqlHmwb4FObTfgEaChOq?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?8J6jof/3Lm0Wrxd2KipEte5pJFTUyr8ZJ65apVORtS1PPWBQaDvzWR/j2va9?= =?us-ascii?Q?2+LbVKOAkN98QQITjsTaHi3C5vpbd9RzEZMze4WgDwo4I8Po95/0ypz70DxF?= =?us-ascii?Q?MhfkoZFyU/KLoJhQ585yUM1GjtVSyxyId5Adxxsd5ZpkwnxB6pQrw3A5NS2a?= =?us-ascii?Q?AVWN9V6QhtfvFR/nfh53B/7JQpISJj+ct1qAuUNs/4rCxQpriXJhlTh29XrT?= =?us-ascii?Q?S9WusKSCWuG+Busi88mrWwtIUjZZu8ENJV1BdKVId3O9/r4ZIvnrCEpsJ0QY?= =?us-ascii?Q?oIoSEMc1ige7V2cJPml+tVTJimL+R3vOSUnCx1uCbhZbRbWoIJ73AUNTkiCz?= =?us-ascii?Q?yAvDGTuAwLHy2HWWkmaBOeGqOPZVl3ORDxPYAFFAmnOl8LP8iFYMx56f24NE?= =?us-ascii?Q?mfCekGFf6P7kWnV3l9Hh1oKwmqwYTEFyv+sxb0HjafcAlgYpGOVGzFeKm7pB?= =?us-ascii?Q?B8j5/sSTn5PkyfC3kex59qRhDhTmFMEpqRXj46kEpW9Hlg90F9iFFHzFSEUb?= =?us-ascii?Q?EPtsjRDT6RBWnN+R3Bb2ZDVtwoxfm60gNVT1TU3ZYnIQfjF/kziAK7VHNpVB?= =?us-ascii?Q?Aw1shSYmsoM1UH1y7hL6JluK6cmZGgvbBKmiMPbaCdLFBZHHa/2HPUyGAA9E?= =?us-ascii?Q?p2qxLOFHlylovZ+chcLvxF3cKrq7s5t2Z7ovwuI8knXH/SBwV0tX/Xl5DRe7?= =?us-ascii?Q?KH3CStUER7Nueri1023MfyweIn1Pk/8xTTR3dSwFKANFjI+hxyjUekY1wlix?= =?us-ascii?Q?aRRrAugQU4GO3bmWpwnbT1+1D6AFCbbey3npWPtBmr+Lsd0OK/RdV08yw3Y1?= =?us-ascii?Q?wj3quH5mHPkJoZ9Q23KfHBnXa/RfUUXrUMx5HX1U9kTRHe6lLHJqjHhkXPUO?= =?us-ascii?Q?ndGagYZp/L3FgnfmTcwuZHDjRnhJn0SXxOBPTInK5z/hPWgcFCYlNo4ni/XK?= =?us-ascii?Q?N6sJmrJ8XKdpePET5CYP8Y99olujaTYKqIx5Ei7ikzmTqA4I0c+yJKCM26hq?= =?us-ascii?Q?jW22lLWgRPtDmCnuDj7jiD7Fxi3K3mdcridaU36bRum/fvpbu1gmV/gv3dH4?= =?us-ascii?Q?Rq7DXPNOCOe4qOsGSTzN3ChuGjsWkDMYATjj1qBEdFEL1x2LYKsht5DxKjtu?= =?us-ascii?Q?qOT6gWrrRg0u?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: 3e3a11d5-1b6e-4528-78bc-08dbbd6003fd X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:28.6340 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: hxSsbGAF37EVBvpTcUlKUJpRq+NldUGsGqC3YnRbt7ls61TU03hWguJOkoRYzzUGRG5Uri72mtIhehxLWgLg/A== X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS0PR10MB7174 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 bulkscore=0 mlxscore=0 mlxlogscore=999 suspectscore=0 phishscore=0 adultscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-ORIG-GUID: cNu9uOhmoybWbAmYCp6XNgJa1581flfM X-Proofpoint-GUID: cNu9uOhmoybWbAmYCp6XNgJa1581flfM Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Joao Martins In an effort to minimize amount of TLB flushes, batch all PMD splits belonging to a range of pages in order to perform only 1 (global) TLB flush. Add a flags field to the walker and pass whether it's a bulk allocation or just a single page to decide to remap. First value (VMEMMAP_SPLIT_NO_TLB_FLUSH) designates the request to not do the TLB flush when we split the PMD. Rebased and updated by Mike Kravetz Signed-off-by: Joao Martins Signed-off-by: Mike Kravetz Reviewed-by: Muchun Song --- mm/hugetlb_vmemmap.c | 92 ++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 88 insertions(+), 4 deletions(-) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 4ac521e596db..10739e4285d5 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -27,6 +27,8 @@ * @reuse_addr: the virtual address of the @reuse_page page. * @vmemmap_pages: the list head of the vmemmap pages that can be freed * or is mapped from. + * @flags: used to modify behavior in vmemmap page table walking + * operations. */ struct vmemmap_remap_walk { void (*remap_pte)(pte_t *pte, unsigned long addr, @@ -35,9 +37,13 @@ struct vmemmap_remap_walk { struct page *reuse_page; unsigned long reuse_addr; struct list_head *vmemmap_pages; + +/* Skip the TLB flush when we split the PMD */ +#define VMEMMAP_SPLIT_NO_TLB_FLUSH BIT(0) + unsigned long flags; }; =20 -static int split_vmemmap_huge_pmd(pmd_t *pmd, unsigned long start) +static int split_vmemmap_huge_pmd(pmd_t *pmd, unsigned long start, bool fl= ush) { pmd_t __pmd; int i; @@ -80,7 +86,8 @@ static int split_vmemmap_huge_pmd(pmd_t *pmd, unsigned lo= ng start) /* Make pte visible before pmd. See comment in pmd_install(). */ smp_wmb(); pmd_populate_kernel(&init_mm, pmd, pgtable); - flush_tlb_kernel_range(start, start + PMD_SIZE); + if (flush) + flush_tlb_kernel_range(start, start + PMD_SIZE); } else { pte_free_kernel(&init_mm, pgtable); } @@ -127,11 +134,20 @@ static int vmemmap_pmd_range(pud_t *pud, unsigned lon= g addr, do { int ret; =20 - ret =3D split_vmemmap_huge_pmd(pmd, addr & PMD_MASK); + ret =3D split_vmemmap_huge_pmd(pmd, addr & PMD_MASK, + !(walk->flags & VMEMMAP_SPLIT_NO_TLB_FLUSH)); if (ret) return ret; =20 next =3D pmd_addr_end(addr, end); + + /* + * We are only splitting, not remapping the hugetlb vmemmap + * pages. + */ + if (!walk->remap_pte) + continue; + vmemmap_pte_range(pmd, addr, next, walk); } while (pmd++, addr =3D next, addr !=3D end); =20 @@ -198,7 +214,8 @@ static int vmemmap_remap_range(unsigned long start, uns= igned long end, return ret; } while (pgd++, addr =3D next, addr !=3D end); =20 - flush_tlb_kernel_range(start, end); + if (walk->remap_pte) + flush_tlb_kernel_range(start, end); =20 return 0; } @@ -297,6 +314,36 @@ static void vmemmap_restore_pte(pte_t *pte, unsigned l= ong addr, set_pte_at(&init_mm, addr, pte, mk_pte(page, pgprot)); } =20 +/** + * vmemmap_remap_split - split the vmemmap virtual address range [@start, = @end) + * backing PMDs of the directmap into PTEs + * @start: start address of the vmemmap virtual address range that we = want + * to remap. + * @end: end address of the vmemmap virtual address range that we wa= nt to + * remap. + * @reuse: reuse address. + * + * Return: %0 on success, negative error code otherwise. + */ +static int vmemmap_remap_split(unsigned long start, unsigned long end, + unsigned long reuse) +{ + int ret; + struct vmemmap_remap_walk walk =3D { + .remap_pte =3D NULL, + .flags =3D VMEMMAP_SPLIT_NO_TLB_FLUSH, + }; + + /* See the comment in the vmemmap_remap_free(). */ + BUG_ON(start - reuse !=3D PAGE_SIZE); + + mmap_read_lock(&init_mm); + ret =3D vmemmap_remap_range(reuse, end, &walk); + mmap_read_unlock(&init_mm); + + return ret; +} + /** * vmemmap_remap_free - remap the vmemmap virtual address range [@start, @= end) * to the page which @reuse is mapped to, then free vmemmap @@ -320,6 +367,7 @@ static int vmemmap_remap_free(unsigned long start, unsi= gned long end, .remap_pte =3D vmemmap_remap_pte, .reuse_addr =3D reuse, .vmemmap_pages =3D vmemmap_pages, + .flags =3D 0, }; int nid =3D page_to_nid((struct page *)reuse); gfp_t gfp_mask =3D GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN; @@ -368,6 +416,7 @@ static int vmemmap_remap_free(unsigned long start, unsi= gned long end, .remap_pte =3D vmemmap_restore_pte, .reuse_addr =3D reuse, .vmemmap_pages =3D vmemmap_pages, + .flags =3D 0, }; =20 vmemmap_remap_range(reuse, end, &walk); @@ -419,6 +468,7 @@ static int vmemmap_remap_alloc(unsigned long start, uns= igned long end, .remap_pte =3D vmemmap_restore_pte, .reuse_addr =3D reuse, .vmemmap_pages =3D &vmemmap_pages, + .flags =3D 0, }; =20 /* See the comment in the vmemmap_remap_free(). */ @@ -628,11 +678,45 @@ void hugetlb_vmemmap_optimize(const struct hstate *h,= struct page *head) free_vmemmap_page_list(&vmemmap_pages); } =20 +static int hugetlb_vmemmap_split(const struct hstate *h, struct page *head) +{ + unsigned long vmemmap_start =3D (unsigned long)head, vmemmap_end; + unsigned long vmemmap_reuse; + + if (!vmemmap_should_optimize(h, head)) + return 0; + + vmemmap_end =3D vmemmap_start + hugetlb_vmemmap_size(h); + vmemmap_reuse =3D vmemmap_start; + vmemmap_start +=3D HUGETLB_VMEMMAP_RESERVE_SIZE; + + /* + * Split PMDs on the vmemmap virtual address range [@vmemmap_start, + * @vmemmap_end] + */ + return vmemmap_remap_split(vmemmap_start, vmemmap_end, vmemmap_reuse); +} + void hugetlb_vmemmap_optimize_folios(struct hstate *h, struct list_head *f= olio_list) { struct folio *folio; LIST_HEAD(vmemmap_pages); =20 + list_for_each_entry(folio, folio_list, lru) { + int ret =3D hugetlb_vmemmap_split(h, &folio->page); + + /* + * Spliting the PMD requires allocating a page, thus lets fail + * early once we encounter the first OOM. No point in retrying + * as it can be dynamically done on remap with the memory + * we get back from the vmemmap deduplication. + */ + if (ret =3D=3D -ENOMEM) + break; + } + + flush_tlb_all(); + list_for_each_entry(folio, folio_list, lru) { int ret =3D __hugetlb_vmemmap_optimize(h, &folio->page, &vmemmap_pages); --=20 2.41.0 From nobody Fri Feb 13 12:45:14 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 6F74DCE7A94 for ; Mon, 25 Sep 2023 00:42:02 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231669AbjIYAmG (ORCPT ); Sun, 24 Sep 2023 20:42:06 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47160 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231409AbjIYAlu (ORCPT ); Sun, 24 Sep 2023 20:41:50 -0400 Received: from mx0a-00069f02.pphosted.com (mx0a-00069f02.pphosted.com [205.220.165.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DAB41112 for ; Sun, 24 Sep 2023 17:41:29 -0700 (PDT) Received: from pps.filterd (m0333521.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMddcC022306; Mon, 25 Sep 2023 00:40:35 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=Tsc70UZvKWYWI+spZQiVZ3u2z8ZAo9dlRTusEOzzs6g=; b=0mwvBOvbVCLPNB+9wmhrfUT9UjP9vzveaNxuSnpXuFoChUGMuCYr//UR9P6zj9vJYvDm uz2ToE+4rZFu6I/EaBNjHQ0J7SCl18c0S7CM61QRjeJYWCTU1jV1fwzzYn3jCrzu6VCS QMrUVVAumSb/52P5vuHwFLgQOmntswxKyQToiaFuNJlrztLAns3F6nexf9OW4FxUiHsR BSZY5cd3J34AsXbL210NyD4SdbfwiDdUZ/pYDTdOMIQ6bAfr09UdLSBVBc/1ALCxoZ6V XwPoEe09tQ9MlwyMwSOrYMgZiqupyw31u4Ly60/+ZTND+kbu6XECBqv6eGHzyFBqWZLG 7A== Received: from phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta03.appoci.oracle.com [138.1.37.129]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9pxbtg3d-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:35 +0000 Received: from pps.filterd (phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38OMXIhI034971; Mon, 25 Sep 2023 00:40:34 GMT Received: from nam12-bn8-obe.outbound.protection.outlook.com (mail-bn8nam12lp2170.outbound.protection.outlook.com [104.47.55.170]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf418dn-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:34 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=LGJ/Y0fX5Yp8Qde+tG4jxMiC2+8VJVYKF6EPIWOHI7vMfut4AXB5RPFLA5xtFTaD/gNQ+FSOfMBpztrvsqvsIBLWpa8dGzz/7po3JhgHsLFCIQHM+QG22z61sGMcibCE3vJqfMlIUHS2s43nlD1mBwAMRZlAUQgyH48jq1EYNhGmxX3ALCXyshCjERTSAevR5C0FfPuxmru9a1GxavGK56ow36rt/hqHWzW3FzFYB6BSSzrFcQmLuSf5JUA/C17MO2SQmV4MQ9OaRlvWd3yzU2rg9ipQr4SVwcbj195PwUIgRE0LfMpuMaviLKgN6Xd9n309GiRVEoI9tK2Q4jFz2Q== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=Tsc70UZvKWYWI+spZQiVZ3u2z8ZAo9dlRTusEOzzs6g=; b=Qy/kjnYZRh82QdLWPZ//ntOwHBRn/ImHJ1yapP74v+wS1x0cz2FXofNSN1r8YaDGjZOUqkJW/7gGZMRY5qB1AsZzdHLdjF6QNvnEFtNQBX+ncIKJtpdO+AuzF7UbWF0HjhVHibzRuJFJTR0aeJv1dYKVrsNhf+fA/zue2VHmiwzg3io0UspK36Ae+FzkhBedpEZe3mCT7PzCXVPAkrQs7wSOjR/1eCUvtCsTj0bBNWIcBM2NXP7UjNwFbfeADvHrTnGeYxhIiwjenXQBtvvWV4NfXnHUUox5OpLhvMmoGhc6/fP15IVhY3t+kf89C/uw8BcQCm3ir5imfqnBdtvVQA== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=Tsc70UZvKWYWI+spZQiVZ3u2z8ZAo9dlRTusEOzzs6g=; b=o6Kv24iAnTi3QMTS14vyhYxUytbpNvOlMVIexZfNFnXKvHa4yK8EPdLOddABq2iPow+QD2LmcHWUoiHqPiltVrICbCgmPY+4+pK+sQgXuBWNEVUQ7joh6fsgMV/I/sgSme+2ISnOYose2xlGjjEDZiRDKUlciyfac2fyT7N4Qt0= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by DS0PR10MB7174.namprd10.prod.outlook.com (2603:10b6:8:df::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:32 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:32 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 7/8] hugetlb: batch TLB flushes when freeing vmemmap Date: Sun, 24 Sep 2023 17:39:51 -0700 Message-ID: <20230925003953.142620-8-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4PR03CA0010.namprd03.prod.outlook.com (2603:10b6:303:8f::15) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|DS0PR10MB7174:EE_ X-MS-Office365-Filtering-Correlation-Id: fc6548a3-1225-4bee-9c79-08dbbd6005dc X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(376002)(346002)(39860400002)(366004)(396003)(136003)(230922051799003)(186009)(1800799009)(451199024)(6666004)(478600001)(6486002)(107886003)(2616005)(1076003)(26005)(83380400001)(36756003)(86362001)(38100700002)(5660300002)(6506007)(6512007)(66556008)(66476007)(54906003)(66946007)(316002)(7416002)(2906002)(44832011)(41300700001)(4326008)(8676002)(8936002);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?+FT0U+rXagdxaBpXvRH6A8viVamXer/15pJdzxBzmHNHxt3QIJlE9Ike35IK?= =?us-ascii?Q?XxhoYb95DjTsNFWXF+VR7Tpep1CvO6teZb8+6ubAbIXACXvAaG3WU4Aci5EM?= =?us-ascii?Q?Qurccik68pgVGi2basPcr7LoJ27W/sllxfTfknzGfkKNHx5mO/RDzjDB9uQu?= =?us-ascii?Q?0w/i2C8+7E29mHKG16qbt9iNobzZYMvj4B8vNyKwbnJNf/VZEE44BxpUn1Eh?= =?us-ascii?Q?mnLFLlpGwOGJxmqE5n5OZb1bOkg2SMiCXIcnuo1IVL36/AX4WKe3I/eqVX5K?= =?us-ascii?Q?/5jxKxmtFwuCnuUHRHy7f0OCZmrmf9ScgiqIsDMdb9UcM9X/enaVry3o6pKB?= =?us-ascii?Q?J7q5u8tJAnPWpzJMQoiwcACWbstlmVES6R35aQEmQqsKu70+qeBSmF7T4AR6?= =?us-ascii?Q?u5bNNC6wAvsa63NdFG7PXW3sgOhU4r0azn36czlveh2SkeLXh72H0nNBMYhI?= =?us-ascii?Q?wQjQQJCQKHcKZVORBYd0QnTZlkvLA2dAVVqBkgR8enqi5koJQAN2VeHg/Vhq?= =?us-ascii?Q?mjWYXOUxxrWUDqOsdoQHwTh9aB1OzkAXpYK+T5xcabVtWCZF91BaOWeg4Nkx?= =?us-ascii?Q?qB8tDiBfJNFYqb6nn6Ks41/9u/6jnJjx0YaBh0XaMU4b7EcJQs/SfHq+/+tW?= =?us-ascii?Q?X0FCmauPWoDi4dhUrqcn1nvALKxn3BvHkmo5UKCD+CfoeJjw4z6lKmJ79bSr?= =?us-ascii?Q?V+hWEV6KpMVyyr39r1AVcPF64F/ku8c+Nid8H1h2CIATPevXLiDGgwm5QDU6?= =?us-ascii?Q?xuknUQCQ1XKlomNOAHtxiBqssTvVf+0xU9fWKf7S3U5hOjCC1yI5fMveX0pv?= =?us-ascii?Q?6OQ4Ww8zUdAt7wQwn1PsrgTsODsGRtl3tfDCT8anMrFW5EwTTVCSBGTiEjJQ?= =?us-ascii?Q?+N5n5N398KBAgijWvg1OonjABaIW8zdOGJ6QWq5zESu3dw757MYa2RWghv3G?= =?us-ascii?Q?aL9qUXtmeSp3L1KiZOv1z4ulOhmPuQfrQsI+Gm96JAJUb6E2JkU+h1QaWHiM?= =?us-ascii?Q?hH1K7koYsW7Gp/Xy71OxHJXCSsoC6UHs6h4FYf0wIRx/hLVof5eNwdIfZfxA?= =?us-ascii?Q?v/skAYYkFiLDzGPTtfR0DmTpAztNRtyNW8JhZfVHLCcHJWnE9pIkKBW7Eecd?= =?us-ascii?Q?WtLPgnsdU5OlpwZzLhvYw7pCcInRSClYUvIfoGkREaAPoCLqaQBSGKKB9dsz?= =?us-ascii?Q?0BTQvbNXvPu8sW5WJqdGiRNoOY5dFZqS00d7zYqbUqGPYGk5rRAAuJl/I841?= =?us-ascii?Q?0k652D08y8eqT6mHQW89aCHjDJxKZl8wkwfjr572YUWz76GxoVOx89KipBk3?= =?us-ascii?Q?S3KKlbcDGL9mZA2GFJx/lW/1Htt9E/SehGzl3lGU+xJW4rUjYMFc5K643OUq?= =?us-ascii?Q?y9J30Iq/NnOpa8WrAxYSsriQcLQTP3eIKS3HEhJhzYukCZ5o6K2l2gyuPdyQ?= =?us-ascii?Q?KE3qRTBkbL5FLdkjuDDqotP/EOONYzFgTPGOa/xYaRRubQN1Au8DhMqzpDcZ?= =?us-ascii?Q?bBOlblzfNjhYrE0BsRpU1PJrhILs2BM4479dVYwzPouPRrJzIf4d2E2qvRZr?= =?us-ascii?Q?WwAiffsXY/ngzisLTuE+Yk6wdhVxe/87HPEs68TB?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?gLDkoY78XwDjxK71tkdPLK32EdcLSYlIIpk23bxTEISMenCn5YnQ5nUGOMQT?= =?us-ascii?Q?h8henrECRVftCe3fw5Vrljo4LPAI5cK8TveTMe4t+dSHtsl78yqn+XmSyMKM?= =?us-ascii?Q?fsuTAVutaLO/tAjQKYw/ubP6yGD0CrnO0UoByOF41Hv4AWvO4fVp96G5Zs/Z?= =?us-ascii?Q?QNpwBNDIDgtWtSQit7KL3Dij6fdbYRz1FCm5GyElzvPcphRcDl0b4Zv54Pwq?= =?us-ascii?Q?0VZuLLXCKyLyeQ0X6R22xoprNbdMtxTN7/d011rGqinTqK+gmo7S+oSooXIP?= =?us-ascii?Q?TaKQ5XtHfS/Z5x9KaSb1g5TxtTpugfYByds/OucB/p9jebf62McczU1f7PKa?= =?us-ascii?Q?LACny95UQX7QOar/8jZHPBoaYRrViF+tghXaB0BzWExq9l1t4W0VJxaevFXl?= =?us-ascii?Q?4vh4l1fZBhoRhRuFpbBw+BfuMmOw4XGLkknx9URXJeT1JV4BxzKYC6JTwVSW?= =?us-ascii?Q?1grdFpdLhMdzQQqIOdO1kDpT0c6oSXkpKKPCnEXcowYBadTKIO8u18YDKZHm?= =?us-ascii?Q?Oht5AZOkEZI89LrTo5P7lETJq1HVS2LtUwOYonBZm4c8+26kPRcSNR9XxZrR?= =?us-ascii?Q?CltRIZpOpyY8IsJrDPEH/2dKFgN2jLN0qJxfQnMVgiigWgWM/f6AcR6NTz4L?= =?us-ascii?Q?QSnh35HoBWTukNoaqnkT+2peEgl3cMbzmto8bZUyx2pcJ8ARm9eAKratwEY9?= =?us-ascii?Q?Zg0eOe3j5tR8uYXo2BaeUAet0hvspe0m3ZVQAi1t5XXEHAnG9ja0evDCq31J?= =?us-ascii?Q?9qRznKt1MX13P6TsdVvt5FDstp4GAR8TxKrnMiXYHDNucAN+37eR7WKNa9GJ?= =?us-ascii?Q?iiaJPSHIj/HzVnekkL9lyx9mmeycXDO4fmGoX/11z1YrjUBY4k0nWLzUKmSx?= =?us-ascii?Q?oWCe1RX3TZgeTFGaJbRJHUR9qbCw9qGBsr0rRE0Bcfn0SDo7XFoJNArz6uIQ?= =?us-ascii?Q?MeUKKGstMjEPwPslO6nb55/RYm2V6MF5dUj1QWIuGZ/k4GYocOTPPrlotcGK?= =?us-ascii?Q?tPQM0YASMq7GmPaRXcqO1tcx4KyQuiUQeEA3gLsZ8MK6FKgKXoe/Rk7MdRNO?= =?us-ascii?Q?9M+1Ea+mSthGJq94K6wdg0n4/1WzsCFg0JLX/2GmyIIa51TqhgFRfFe+Vy13?= =?us-ascii?Q?1oFqzE8ZN0F9?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: fc6548a3-1225-4bee-9c79-08dbbd6005dc X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:32.0203 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: UHZO/MVh+La6u6QEaBF0pt3o3Im2W0WynsbBFfThtwqrmDmPuDD00ChygjMO4204N9L+fnT3nyKndrW7HZM8sg== X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS0PR10MB7174 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 bulkscore=0 mlxscore=0 mlxlogscore=999 suspectscore=0 phishscore=0 adultscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-GUID: q1r9Iwf7IgyvQH6_qtZinAH-tKL6oW2F X-Proofpoint-ORIG-GUID: q1r9Iwf7IgyvQH6_qtZinAH-tKL6oW2F Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Joao Martins Now that a list of pages is deduplicated at once, the TLB flush can be batched for all vmemmap pages that got remapped. Expand the flags field value to pass whether to skip the TLB flush on remap of the PTE. The TLB flush is global as we don't have guarantees from caller that the set of folios is contiguous, or to add complexity in composing a list of kVAs to flush. Modified by Mike Kravetz to perform TLB flush on single folio if an error is encountered. Signed-off-by: Joao Martins Signed-off-by: Mike Kravetz Reviewed-by: Muchun Song --- mm/hugetlb_vmemmap.c | 49 ++++++++++++++++++++++++++++++++++---------- 1 file changed, 38 insertions(+), 11 deletions(-) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 10739e4285d5..9df350372046 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -40,6 +40,8 @@ struct vmemmap_remap_walk { =20 /* Skip the TLB flush when we split the PMD */ #define VMEMMAP_SPLIT_NO_TLB_FLUSH BIT(0) +/* Skip the TLB flush when we remap the PTE */ +#define VMEMMAP_REMAP_NO_TLB_FLUSH BIT(1) unsigned long flags; }; =20 @@ -214,7 +216,7 @@ static int vmemmap_remap_range(unsigned long start, uns= igned long end, return ret; } while (pgd++, addr =3D next, addr !=3D end); =20 - if (walk->remap_pte) + if (walk->remap_pte && !(walk->flags & VMEMMAP_REMAP_NO_TLB_FLUSH)) flush_tlb_kernel_range(start, end); =20 return 0; @@ -355,19 +357,21 @@ static int vmemmap_remap_split(unsigned long start, u= nsigned long end, * @reuse: reuse address. * @vmemmap_pages: list to deposit vmemmap pages to be freed. It is calle= rs * responsibility to free pages. + * @flags: modifications to vmemmap_remap_walk flags * * Return: %0 on success, negative error code otherwise. */ static int vmemmap_remap_free(unsigned long start, unsigned long end, unsigned long reuse, - struct list_head *vmemmap_pages) + struct list_head *vmemmap_pages, + unsigned long flags) { int ret; struct vmemmap_remap_walk walk =3D { .remap_pte =3D vmemmap_remap_pte, .reuse_addr =3D reuse, .vmemmap_pages =3D vmemmap_pages, - .flags =3D 0, + .flags =3D flags, }; int nid =3D page_to_nid((struct page *)reuse); gfp_t gfp_mask =3D GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN; @@ -629,7 +633,8 @@ static bool vmemmap_should_optimize(const struct hstate= *h, const struct page *h =20 static int __hugetlb_vmemmap_optimize(const struct hstate *h, struct page *head, - struct list_head *vmemmap_pages) + struct list_head *vmemmap_pages, + unsigned long flags) { int ret =3D 0; unsigned long vmemmap_start =3D (unsigned long)head, vmemmap_end; @@ -640,6 +645,18 @@ static int __hugetlb_vmemmap_optimize(const struct hst= ate *h, return ret; =20 static_branch_inc(&hugetlb_optimize_vmemmap_key); + /* + * Very Subtle + * If VMEMMAP_REMAP_NO_TLB_FLUSH is set, TLB flushing is not performed + * immediately after remapping. As a result, subsequent accesses + * and modifications to struct pages associated with the hugetlb + * page could be to the OLD struct pages. Set the vmemmap optimized + * flag here so that it is copied to the new head page. This keeps + * the old and new struct pages in sync. + * If there is an error during optimization, we will immediately FLUSH + * the TLB and clear the flag below. + */ + SetHPageVmemmapOptimized(head); =20 vmemmap_end =3D vmemmap_start + hugetlb_vmemmap_size(h); vmemmap_reuse =3D vmemmap_start; @@ -651,11 +668,12 @@ static int __hugetlb_vmemmap_optimize(const struct hs= tate *h, * mapping the range to vmemmap_pages list so that they can be freed by * the caller. */ - ret =3D vmemmap_remap_free(vmemmap_start, vmemmap_end, vmemmap_reuse, vme= mmap_pages); - if (ret) + ret =3D vmemmap_remap_free(vmemmap_start, vmemmap_end, vmemmap_reuse, + vmemmap_pages, flags); + if (ret) { static_branch_dec(&hugetlb_optimize_vmemmap_key); - else - SetHPageVmemmapOptimized(head); + ClearHPageVmemmapOptimized(head); + } =20 return ret; } @@ -674,7 +692,7 @@ void hugetlb_vmemmap_optimize(const struct hstate *h, s= truct page *head) { LIST_HEAD(vmemmap_pages); =20 - __hugetlb_vmemmap_optimize(h, head, &vmemmap_pages); + __hugetlb_vmemmap_optimize(h, head, &vmemmap_pages, 0); free_vmemmap_page_list(&vmemmap_pages); } =20 @@ -719,19 +737,28 @@ void hugetlb_vmemmap_optimize_folios(struct hstate *h= , struct list_head *folio_l =20 list_for_each_entry(folio, folio_list, lru) { int ret =3D __hugetlb_vmemmap_optimize(h, &folio->page, - &vmemmap_pages); + &vmemmap_pages, + VMEMMAP_REMAP_NO_TLB_FLUSH); =20 /* * Pages to be freed may have been accumulated. If we * encounter an ENOMEM, free what we have and try again. + * This can occur in the case that both spliting fails + * halfway and head page allocation also failed. In this + * case __hugetlb_vmemmap_optimize() would free memory + * allowing more vmemmap remaps to occur. */ if (ret =3D=3D -ENOMEM && !list_empty(&vmemmap_pages)) { + flush_tlb_all(); free_vmemmap_page_list(&vmemmap_pages); INIT_LIST_HEAD(&vmemmap_pages); - __hugetlb_vmemmap_optimize(h, &folio->page, &vmemmap_pages); + __hugetlb_vmemmap_optimize(h, &folio->page, + &vmemmap_pages, + VMEMMAP_REMAP_NO_TLB_FLUSH); } } =20 + flush_tlb_all(); free_vmemmap_page_list(&vmemmap_pages); } =20 --=20 2.41.0 From nobody Fri Feb 13 12:45:14 2026 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id CA911CE7A95 for ; Mon, 25 Sep 2023 00:41:41 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230435AbjIYAlp (ORCPT ); Sun, 24 Sep 2023 20:41:45 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47088 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231497AbjIYAla (ORCPT ); Sun, 24 Sep 2023 20:41:30 -0400 Received: from mx0b-00069f02.pphosted.com (mx0b-00069f02.pphosted.com [205.220.177.32]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id CDB70CC5 for ; Sun, 24 Sep 2023 17:41:14 -0700 (PDT) Received: from pps.filterd (m0246630.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 38OLsBVC012416; Mon, 25 Sep 2023 00:40:38 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : content-transfer-encoding : content-type : mime-version; s=corp-2023-03-30; bh=wop91Vn2thHKTczlqkauN3ZgszKMNCLF+DVzOTx6zso=; b=ep1qyqEVmrzBRT2+RbRmWKqmAwDbD1zK6SVv1Cyx5hp27Zozadjz4NnMSd0jDzgyOwIm z7/b0IHwKR2t5yvxDOMlmc9mwAWNpqGbQCyecgjCviDtfD8oMMdq9B8KTj/xv+WNo0e2 EneIQovAVqE9AUMtFj0vDxPmTompUync9x5uAwET24uPpkpY3ikU6hf0ieAsH75ffLQR rUjaxuuiTnnzXLa/rkahPQxoiEhNiA7Z1sprZizq4c/ns3p3e4Dh6dlUQ87nEfwuB83g b4Y0Ol4eLgtWQJCEyabpZa/BA5IibI432VIqvSWQ0dln/kexKQMGenmZlo/1BCqL/Ri8 fg== Received: from phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (phxpaimrmta03.appoci.oracle.com [138.1.37.129]) by mx0b-00069f02.pphosted.com (PPS) with ESMTPS id 3t9pee2era-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:38 +0000 Received: from pps.filterd (phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com [127.0.0.1]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (8.17.1.19/8.17.1.19) with ESMTP id 38ONoR1C034959; Mon, 25 Sep 2023 00:40:37 GMT Received: from nam12-bn8-obe.outbound.protection.outlook.com (mail-bn8nam12lp2174.outbound.protection.outlook.com [104.47.55.174]) by phxpaimrmta03.imrmtpd1.prodappphxaev1.oraclevcn.com (PPS) with ESMTPS id 3t9pf418em-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 25 Sep 2023 00:40:37 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=JBWfd9CL4AsBujLRTY4kcV5TUp4jdIs2yZGXOK+dtNLO4B1xn7E4Bls8mtY8VZTCxibSwgGFwd5yAq2h/lmd9qw/4110hF+8wNBtn0/E+rnmCtWy//EyVbmkFKT74ew/ZdNIuF+U/ViZHcznXmMumPMzrt2hfHfkChr8Dl8jPEt8hT+dBloPd8zOuQMNjPPdmH+ZfAck8mZwSsBWl17d0y8cDBmc2fiQXcANYq2zQ9cS17VG+L6B6h8jLWOcRsjOuBpqaiwGPat/HYFNtIfVCft5Ad2M2YFDUO6ebML6x020OZyyLwUQVcWR6Mu2W3JUIszgkm1P57bPqUznArP+JQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; 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=wop91Vn2thHKTczlqkauN3ZgszKMNCLF+DVzOTx6zso=; b=GXHR66F8K7zQVewYTa1fh71lfihDIay6xt8k6AgAZsSUE4sfy1eIQsk677+C2vgt6p9b9wmA+iWlupI5pLPr0u+SRBNW1II2lJMJmvb/04sz8kJ7fLCrd1I4LmDCS+n3tU1H9a+4nbR05T4XU1KKeBpQLFzUzwOo1Eklwe4dhz4W9qhxyQwxSoHdFiMkXvcjxuoX9h6rOkhP6N4HjV3Tif21nPIOWFcb0PInR1yGRfnXym8Au0C+voWAxLmqleYiLj/Z0MCW6iYKA4rW1BWpuifAxVgWLlbkHBHP3OXorqWgfkuIwn///DCEUAhXWruweoQMsoEQ7qRVtOTJYATIfg== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=oracle.com; dmarc=pass action=none header.from=oracle.com; dkim=pass header.d=oracle.com; arc=none DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.onmicrosoft.com; s=selector2-oracle-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=wop91Vn2thHKTczlqkauN3ZgszKMNCLF+DVzOTx6zso=; b=gIowThSl/plu+lqovuRlFR2YnBDhEQQu2BEDP/j7M6JKCvrdm69VCU3MglXOQ0nXe21kC61ymba5gYnTUAIVGQgWjBZTtUfiyxI58H2rMD3f2nAKSUDN/+yJ0l058WV1glcr1gKai4AKya8vobHO/FrvK4Pjs6A2ybfu8UVGvI0= Received: from BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) by DS0PR10MB7174.namprd10.prod.outlook.com (2603:10b6:8:df::11) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6813.28; Mon, 25 Sep 2023 00:40:34 +0000 Received: from BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054]) by BY5PR10MB4196.namprd10.prod.outlook.com ([fe80::c621:12ca:ba40:9054%5]) with mapi id 15.20.6813.027; Mon, 25 Sep 2023 00:40:34 +0000 From: Mike Kravetz To: linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Muchun Song , Joao Martins , Oscar Salvador , David Hildenbrand , Miaohe Lin , David Rientjes , Anshuman Khandual , Naoya Horiguchi , Barry Song <21cnbao@gmail.com>, Michal Hocko , Matthew Wilcox , Xiongchun Duan , Andrew Morton , Mike Kravetz Subject: [PATCH v5 8/8] hugetlb: batch TLB flushes when restoring vmemmap Date: Sun, 24 Sep 2023 17:39:52 -0700 Message-ID: <20230925003953.142620-9-mike.kravetz@oracle.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230925003953.142620-1-mike.kravetz@oracle.com> References: <20230925003953.142620-1-mike.kravetz@oracle.com> Content-Transfer-Encoding: quoted-printable X-ClientProxiedBy: MW4P223CA0024.NAMP223.PROD.OUTLOOK.COM (2603:10b6:303:80::29) To BY5PR10MB4196.namprd10.prod.outlook.com (2603:10b6:a03:20d::23) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: BY5PR10MB4196:EE_|DS0PR10MB7174:EE_ X-MS-Office365-Filtering-Correlation-Id: 850d5942-369b-42eb-e66d-08dbbd6007a3 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; 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:BY5PR10MB4196.namprd10.prod.outlook.com;PTR:;CAT:NONE;SFS:(13230031)(376002)(346002)(39860400002)(366004)(396003)(136003)(230922051799003)(186009)(1800799009)(451199024)(6666004)(478600001)(6486002)(107886003)(2616005)(1076003)(26005)(83380400001)(36756003)(86362001)(38100700002)(5660300002)(6506007)(6512007)(66556008)(66476007)(54906003)(66946007)(316002)(7416002)(2906002)(44832011)(41300700001)(4326008)(8676002)(8936002);DIR:OUT;SFP:1101; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?tDKvKEUfyS3vqsfjPMflm7+M/+Ms9NnLXCWz2jMAhavmz2C8sJNuujTtWU9o?= =?us-ascii?Q?1V/rtn9W+BjZqth8/QoPn75AYLQadsrGHizpRlSgT9MB8Ux3YhDB/hy1W4HR?= =?us-ascii?Q?HRPbEd1xKVDD7OTFuyYcptE6rXD14s7n6ksJdx1F0vS7BTG3hg7mitzxtCRD?= =?us-ascii?Q?1ZoJGfNobr4iAVNPl7AqlBzHmdQF5epco7MTR6YnFAzXJB5jj3AMgBD6fWqc?= =?us-ascii?Q?ne6za5zj/tCsNI2HFDlr2dDaFrT1dNzJKVgm4xB9OsxjnunrQRli9Nv6SBiy?= =?us-ascii?Q?FUB6+/CONno29ZOO/KqjKEFkD0PyYqily6hy2/aF8GKCLGrt//Z1bbXKi1c5?= =?us-ascii?Q?hP4mWC+1tWCs7okUjVOAc+jMVmqTMVC6/xA0e78DWHTZxFg5tOYqDvFXPuKB?= =?us-ascii?Q?97Jb4LyJVY2bcuUzL7PZF/+oTdTWUNidBN2qSxtgu7EcCzNKjxcF1hPQnrCh?= =?us-ascii?Q?D/YIHMG9CsIfJ3PJgbATugxIOzwpo9x66oupilQ5JboAYmbSy1IW/OtqGwAJ?= =?us-ascii?Q?cd+mHUPoYGTShR0l5p/dblLO7Bjx+XV+IbNPnK1rGI0pNOkYDD2I5taNWwVE?= =?us-ascii?Q?HP1h57+hCiddBf4xb2xHEQSgP6O7+9+TzxmDEoT+xxQBrgzUGV2UDHTeKCZL?= =?us-ascii?Q?/+1bbGSP61kDrKS9FQN+AARSXvhubhbpdgsyqEXvBGZ2Y7xQL+Rw/SlH6qk4?= =?us-ascii?Q?sCfy9fJmX1d07hTUYkJdjVL1WTUMDZaK94/iCVqihjiAcNgbjySSxdSkorpg?= =?us-ascii?Q?EIL48TO9CMADdsWezlUr5fzFqwTeAXSexXlFR0/Zo7utzB4uz8GRoMFSuZ3N?= =?us-ascii?Q?r9KS6ZQt0boFDtafUMgBWDVMP5NY0y43TXdViPliDeSVz+iYT24mTTV5X8V5?= =?us-ascii?Q?GV6LFI/jSJ/0cZxQbaSJpxmzAxYmXQZM8/VzrdxGpLhSVoYgXzl3asC54H9B?= =?us-ascii?Q?qRbLCqEecZv7vftHV0DnbRBc0n0PlZbFtu/aefYuk3iZqnx2T/pgZ3oMdCFG?= =?us-ascii?Q?L+QopBReYkl1K5Z3CW/Az1SLw2i3FjVOEPTYvLjOPVlOo2UfIO4LWwKvby5u?= =?us-ascii?Q?qPE7WmH3MBfVq0/WGKl4EkFUOPCdAqMgEZD4/TqjEtC4Te87wL/dhXsuwRZU?= =?us-ascii?Q?4x9ohkJbrSrW8uKGuE7iH6pWGDMrTzSlk3u0CDC0jDraJrO3f+lF0LfCBmJ1?= =?us-ascii?Q?wLYHNQnWvdns2vHJn98k/9NmpcUvLPbVpkR2rce6DVJyC5m0jxu7aflqusY7?= =?us-ascii?Q?1eLBJFNL53SJEYzdR9xXSzqUIpugFWZn9bMhnrlxz1VyGTofMynppzw15301?= =?us-ascii?Q?ips/c2GMZemqYqqp8LihCQWeDN1o0E+4BrBfWM6Mgzv5bbsTQSXv9zxOx3Sa?= =?us-ascii?Q?bv6grtr+LVtvYqJQ9iJI9mUWasNXQqGhzmluXrBC44NB/GH5Hl7Z1gqL0evy?= =?us-ascii?Q?xERr5VryNUdmT22Jqtwc/fa2JtR0QkNG9kNuKBwLFnijaevG6qaQ517bgUuh?= =?us-ascii?Q?+LM1BHeE1SpytQGI1d7voNc7E5cien3bNaPNyMLIGUa2ekUPzekmshJZXp+O?= =?us-ascii?Q?K5vZbxwPUAddQ54DDV6qSdqXDd1mCcEkN4Dvs9KS?= X-MS-Exchange-AntiSpam-ExternalHop-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-ExternalHop-MessageData-0: =?us-ascii?Q?YHyqp6+/8bjXPej6JaTFF8SsdO4nCxAGxBsgEbJkTDlTb4bcoQXGC0XpcQHr?= =?us-ascii?Q?KRSsmJvpYBhYtPB6cAOe3UhhbIDonpAaBP6di9NCZiSyRISqnNS0I7aRYhCE?= =?us-ascii?Q?y1X7xtSeZHIOnM/ooGIyVj20DljBptkgrLizz1iLutuTq+I6assVMnYY8LMC?= =?us-ascii?Q?okBTck2T58MlgePPQLxC/NTohGPMNfbXkllCuV+TRvEKm9i6Y28NPo8n1kCU?= =?us-ascii?Q?fY607gWl92AhsgVtSbjPZ/fez7ySzMInILf1nkdHvdIX5jcCSJWg9b6PJjyb?= =?us-ascii?Q?iiq2mBD4AyRkDCGJlV4BQwmNNn3fr2srCD/NMcOfS9AhfA/kNDC7GY/O5xfU?= =?us-ascii?Q?MayDM0Z5XARkx1bFW8Stn8ZZjoMnNsTaPobRhIDmUfw92QGD+u6xX+3HuWjd?= =?us-ascii?Q?wVG0D3nWR0jfAw9atSlHxlCGW2sEWZZddiM/FBaCSHvC09nDyMXnO34MsjTu?= =?us-ascii?Q?togDGPgEXesX6zcqBHXbDoXqQSCL1p+FnfSecyQdJhNf7SeCk13Bq4qvTwpx?= =?us-ascii?Q?dgdvY4TVo4YkauczzYelqA/sptOnpHCd/4Z8Cg7G5Rj2MiPlFBa+LLzyDwDr?= =?us-ascii?Q?wAdvU3Rfv10ERCz5XvgnE/wLKcarYtiVhH7m0PqV82FAfnjmx4jOVJ3Zx9XM?= =?us-ascii?Q?eqsCgYf6Mn6gWyHFSgT3LxVozyhhbWH1joSJ7vk7JVbjwrrVZnSV97DJi3iM?= =?us-ascii?Q?EfvqVEBQ5VyLJAHBUoRyJ6yyHrmWXNoO9a+mfAlUEW0Dfrh5yKSThObU37t1?= =?us-ascii?Q?rdVh3NsDqeifR2aJ28thhxPBVc3cRnX0f10RxNjgy5qkibA8VQOmtWAMgHJf?= =?us-ascii?Q?sc8sfmNcHQpCh7xfsCvIKFiaHxuhHuwXLZXZJuOXopGUwjfXk0MKnEF3+Szg?= =?us-ascii?Q?fgNe1tvjxHuqgzgOF/Th18oZ7l4LOW4RMMZYDNRPiLJjCGgVxEHscOpRsyjh?= =?us-ascii?Q?jj03NVPauyQ0ijI23mwigUcMktd8cP8zjszNLQCWzDui0SdTx5Agv1P/qWL0?= =?us-ascii?Q?FLAPVjZIoGp1PHDFRXGFzCQCD4DOqNia6i1Horzmui0I/cfu7gIcHt8gY+NT?= =?us-ascii?Q?Lxqi64S3RGKm3343l2bGLx82ZpVJF9NYMJkAdPGjVm1FfgyaNfIgLjFrSMVa?= =?us-ascii?Q?17RU6nZfdcXU?= X-OriginatorOrg: oracle.com X-MS-Exchange-CrossTenant-Network-Message-Id: 850d5942-369b-42eb-e66d-08dbbd6007a3 X-MS-Exchange-CrossTenant-AuthSource: BY5PR10MB4196.namprd10.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 25 Sep 2023 00:40:34.6878 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 4e2c6054-71cb-48f1-bd6c-3a9705aca71b X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: 3paTb7dxTRgo759+XdEXmKnAP2aiveNoFdqXOPU52e19++MtE64aHco9zRP3MyTub9YFp8Uki6hy9XYIOZ0rQg== X-MS-Exchange-Transport-CrossTenantHeadersStamped: DS0PR10MB7174 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.267,Aquarius:18.0.980,Hydra:6.0.619,FMLib:17.11.176.26 definitions=2023-09-24_21,2023-09-21_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 bulkscore=0 mlxscore=0 mlxlogscore=999 suspectscore=0 phishscore=0 adultscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2309180000 definitions=main-2309250000 X-Proofpoint-ORIG-GUID: pSbux-Z7blrrWnBynn8BvvvvUoxkzByW X-Proofpoint-GUID: pSbux-Z7blrrWnBynn8BvvvvUoxkzByW Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Update the internal hugetlb restore vmemmap code path such that TLB flushing can be batched. Use the existing mechanism of passing the VMEMMAP_REMAP_NO_TLB_FLUSH flag to indicate flushing should not be performed for individual pages. The routine hugetlb_vmemmap_restore_folios is the only user of this new mechanism, and it will perform a global flush after all vmemmap is restored. Signed-off-by: Joao Martins Signed-off-by: Mike Kravetz --- mm/hugetlb_vmemmap.c | 39 ++++++++++++++++++++++++--------------- 1 file changed, 24 insertions(+), 15 deletions(-) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 9df350372046..d2999c303031 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -461,18 +461,19 @@ static int alloc_vmemmap_page_list(unsigned long star= t, unsigned long end, * @end: end address of the vmemmap virtual address range that we want to * remap. * @reuse: reuse address. + * @flags: modifications to vmemmap_remap_walk flags * * Return: %0 on success, negative error code otherwise. */ static int vmemmap_remap_alloc(unsigned long start, unsigned long end, - unsigned long reuse) + unsigned long reuse, unsigned long flags) { LIST_HEAD(vmemmap_pages); struct vmemmap_remap_walk walk =3D { .remap_pte =3D vmemmap_restore_pte, .reuse_addr =3D reuse, .vmemmap_pages =3D &vmemmap_pages, - .flags =3D 0, + .flags =3D flags, }; =20 /* See the comment in the vmemmap_remap_free(). */ @@ -494,17 +495,7 @@ EXPORT_SYMBOL(hugetlb_optimize_vmemmap_key); static bool vmemmap_optimize_enabled =3D IS_ENABLED(CONFIG_HUGETLB_PAGE_OP= TIMIZE_VMEMMAP_DEFAULT_ON); core_param(hugetlb_free_vmemmap, vmemmap_optimize_enabled, bool, 0); =20 -/** - * hugetlb_vmemmap_restore - restore previously optimized (by - * hugetlb_vmemmap_optimize()) vmemmap pages which - * will be reallocated and remapped. - * @h: struct hstate. - * @head: the head page whose vmemmap pages will be restored. - * - * Return: %0 if @head's vmemmap pages have been reallocated and remapped, - * negative error code otherwise. - */ -int hugetlb_vmemmap_restore(const struct hstate *h, struct page *head) +static int __hugetlb_vmemmap_restore(const struct hstate *h, struct page *= head, unsigned long flags) { int ret; unsigned long vmemmap_start =3D (unsigned long)head, vmemmap_end; @@ -525,7 +516,7 @@ int hugetlb_vmemmap_restore(const struct hstate *h, str= uct page *head) * When a HugeTLB page is freed to the buddy allocator, previously * discarded vmemmap pages must be allocated and remapping. */ - ret =3D vmemmap_remap_alloc(vmemmap_start, vmemmap_end, vmemmap_reuse); + ret =3D vmemmap_remap_alloc(vmemmap_start, vmemmap_end, vmemmap_reuse, fl= ags); if (!ret) { ClearHPageVmemmapOptimized(head); static_branch_dec(&hugetlb_optimize_vmemmap_key); @@ -534,6 +525,21 @@ int hugetlb_vmemmap_restore(const struct hstate *h, st= ruct page *head) return ret; } =20 +/** + * hugetlb_vmemmap_restore - restore previously optimized (by + * hugetlb_vmemmap_optimize()) vmemmap pages which + * will be reallocated and remapped. + * @h: struct hstate. + * @head: the head page whose vmemmap pages will be restored. + * + * Return: %0 if @head's vmemmap pages have been reallocated and remapped, + * negative error code otherwise. + */ +int hugetlb_vmemmap_restore(const struct hstate *h, struct page *head) +{ + return __hugetlb_vmemmap_restore(h, head, 0); +} + /** * hugetlb_vmemmap_restore_folios - restore vmemmap for every folio on the= list. * @h: hstate. @@ -557,7 +563,8 @@ long hugetlb_vmemmap_restore_folios(const struct hstate= *h, =20 list_for_each_entry_safe(folio, t_folio, folio_list, lru) { if (folio_test_hugetlb_vmemmap_optimized(folio)) { - ret =3D hugetlb_vmemmap_restore(h, &folio->page); + ret =3D __hugetlb_vmemmap_restore(h, &folio->page, + VMEMMAP_REMAP_NO_TLB_FLUSH); if (ret) break; restored++; @@ -567,6 +574,8 @@ long hugetlb_vmemmap_restore_folios(const struct hstate= *h, list_move(&folio->lru, non_hvo_folios); } =20 + if (restored) + flush_tlb_all(); if (!ret) ret =3D restored; return ret; --=20 2.41.0