From mboxrd@z Thu Jan 1 00:00:00 1970 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 kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5197DC2D0CD for ; Mon, 19 May 2025 14:39:49 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id E38C06B00C1; Mon, 19 May 2025 10:39:43 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id DE8006B00C2; Mon, 19 May 2025 10:39:43 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id C3C7C6B00D4; Mon, 19 May 2025 10:39:43 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0015.hostedemail.com [216.40.44.15]) by kanga.kvack.org (Postfix) with ESMTP id 9DCA66B00C1 for ; Mon, 19 May 2025 10:39:43 -0400 (EDT) Received: from smtpin12.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay04.hostedemail.com (Postfix) with ESMTP id 030BA1A05E9 for ; Mon, 19 May 2025 14:39:47 +0000 (UTC) X-FDA: 83459916456.12.57129E4 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) by imf24.hostedemail.com (Postfix) with ESMTP id 92557180017 for ; Mon, 19 May 2025 14:39:45 +0000 (UTC) Authentication-Results: imf24.hostedemail.com; dkim=pass header.d=redhat.com header.s=mimecast20190719 header.b=Jy9D0lPg; spf=pass (imf24.hostedemail.com: domain of david@redhat.com designates 170.10.133.124 as permitted sender) smtp.mailfrom=david@redhat.com; dmarc=pass (policy=quarantine) header.from=redhat.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1747665585; h=from:from:sender:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references:dkim-signature; bh=fc1XxKKHBxhSvqtmcE/i6SWy2w5XcBTE8wKwLKGtMi4=; b=6e4MBeCPkdlYD+uqqZAGVrUSFQR/65iGW5HWUMppu1UODseYUzFLRkTZ+iXMz00ywhehfR hN5NBXEQnfDXJDIuIN/QGb7vfwUIg15EkoupqIYXcREYG3r77fCZr0/UlgU/hkKanmIPGZ NaZGEDx+KJpdZEezrq+orOsSqM/gVMM= ARC-Authentication-Results: i=1; imf24.hostedemail.com; dkim=pass header.d=redhat.com header.s=mimecast20190719 header.b=Jy9D0lPg; spf=pass (imf24.hostedemail.com: domain of david@redhat.com designates 170.10.133.124 as permitted sender) smtp.mailfrom=david@redhat.com; dmarc=pass (policy=quarantine) header.from=redhat.com ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1747665585; a=rsa-sha256; cv=none; b=daGu50xBrESqTiUpCM0hCACIFV5sBpkm5Gqt2YzfjyjWG22OWZ1vJDy2EUbR7WVOIZhTo7 ivlyo6SopaxtrxfKJxxEvT1kylIePgJY8qdjCyYNrGq2jL1CXwkRrZhq0x6lN6ZILP/QCx 5+k2Wzmc45/jCMpKqOTn50NrpMRXCJM= DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1747665584; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references:autocrypt:autocrypt; bh=fc1XxKKHBxhSvqtmcE/i6SWy2w5XcBTE8wKwLKGtMi4=; b=Jy9D0lPg+WLg/uJLTux5wxtKPoGqouJhy3ZAi1EeXRVgKdYUGXf0VVagcJXZT9Z/GWZPTB 3XJboR7TSLxDV7NZOVOvnQ6Jftq5hCl0gGb2sUMzib0N9cyeEN58gQ/7H1KPdoV/a4l3vO czIV7KU61ZkqiTpUy3iovJtr7epxv4U= Received: from mail-wr1-f71.google.com (mail-wr1-f71.google.com [209.85.221.71]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-684-PBkj8KnoNmOJL7HsD48HPQ-1; Mon, 19 May 2025 10:39:41 -0400 X-MC-Unique: PBkj8KnoNmOJL7HsD48HPQ-1 X-Mimecast-MFC-AGG-ID: PBkj8KnoNmOJL7HsD48HPQ_1747665580 Received: by mail-wr1-f71.google.com with SMTP id ffacd0b85a97d-3a36c58ef62so834843f8f.1 for ; Mon, 19 May 2025 07:39:41 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1747665580; x=1748270380; h=content-transfer-encoding:in-reply-to:organization:autocrypt :content-language:from:references:cc:to:subject:user-agent :mime-version:date:message-id:x-gm-message-state:from:to:cc:subject :date:message-id:reply-to; bh=fc1XxKKHBxhSvqtmcE/i6SWy2w5XcBTE8wKwLKGtMi4=; b=MzKuZ6HG538MS9xj5ZBEpsrMXR4n9whFDwftw1lGWepnH/GRflSWfvPk/allnWfWPU 6IX0o+buRNSO6Z2DmPZP6W/vP5ODlPUCMS9MfQi9EqSf/T7QMh0KvAAMemAlrQpm7ZHQ /IAKgNft0QRnhkrt2N/EF0cIl0NeWjUvH/bdXimF777Gr+Qy22l+BjYexuyWc/+fpltF p9ITpa5l/N9B/kkj+SReJ5WcPBCDsuZA+bP3z8jay/50bXMMfsKdpTaq2Nlyl4SHsYme wtKEnUnfqfedYjPC4/wh8C1r64ItLe8h6kZxM/74Qo8lmteOXD0ED0/8RKV2sd6cje21 SPOg== X-Forwarded-Encrypted: i=1; AJvYcCUsP/te+7nqPlq85xBd3db7gXrp8ZOxlXPKevMaWDDKYCNZta5vw0kPFjiHUJzJeJFeHDV6BWBssw==@kvack.org X-Gm-Message-State: AOJu0YysOBfuy3xjT6IQGdG2Npkar9nOy7ODsAK4DEndmtEi3QtY4CBn o0JPUzPDee40BTZnQBdadjblr1CXLFf7cl4sanUdrXAp5NYWdbbYVzrivUQ+nt4Pl3JtKOVK4dH Whdgxqgwbi8Ngw3NseMzxQNXRvHGx56JyX6+bxfxkFxsrzpBsDc1J X-Gm-Gg: ASbGncuzOD0z259z0XKQ5CNO2L3MstZcl9SMsIrY8SydIqmTO9riP2VP3LN7vtlp/ij LX8TvgsUchelbnf6olW5cetcoP/hZx1Wpk1+LwvMt+7A3HWMfSyf1H1A9dmWge3qsQGsf8o54nR j3lZarQxKMKtKQda3YvSKp9GVw2MLkm2UTNRzOrxUWUgPZ+G/t+4Mzz04jeLEOsy3dBXSefUkCL fq9E0OE7LbeRrATuLicbndMELQu93HK7t0WhV1OsQv+nmWYDkMMsjN/ha9KSWUlJ/NFfUpA9F9j 6n8c7bWe87i05bqfaoU6qSeNXExqxQySpTtp0ZABZvYSS+ht3vczbSOlJHyUEMjIfz5I9jCD919 5gFbe1/B3+0XEUqQlj+zjRx1ELbMltgkwRGw3C/U= X-Received: by 2002:a05:6000:3111:b0:3a3:75fa:a0f6 with SMTP id ffacd0b85a97d-3a375faa340mr1893631f8f.50.1747665580206; Mon, 19 May 2025 07:39:40 -0700 (PDT) X-Google-Smtp-Source: AGHT+IEGbb+jxiMxPn0A7K1u6APHY87dcAiZht4gZaL68Kpwpnj/fmNBc4lbu//e0WBBsOYAYluLMg== X-Received: by 2002:a05:6000:3111:b0:3a3:75fa:a0f6 with SMTP id ffacd0b85a97d-3a375faa340mr1893594f8f.50.1747665579708; Mon, 19 May 2025 07:39:39 -0700 (PDT) Received: from ?IPV6:2003:d8:2f3c:3a00:5662:26b3:3e5d:438e? (p200300d82f3c3a00566226b33e5d438e.dip0.t-ipconnect.de. [2003:d8:2f3c:3a00:5662:26b3:3e5d:438e]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-3a35ca8cf66sm12807961f8f.87.2025.05.19.07.39.38 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Mon, 19 May 2025 07:39:39 -0700 (PDT) Message-ID: <85fd0566-30c8-4e1e-8ce9-5b5bb4f1fa84@redhat.com> Date: Mon, 19 May 2025 16:39:38 +0200 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH v1 1/2] mm/memory_hotplug: PG_offline_skippable for offlining memory blocks with PageOffline pages To: Zi Yan Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org, virtualization@lists.linux.dev, "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , =?UTF-8?Q?Eugenio_P=C3=A9rez?= , Andrew Morton , Oscar Salvador , Vlastimil Babka , Suren Baghdasaryan , Michal Hocko , Brendan Jackman , Johannes Weiner , "Matthew Wilcox (Oracle)" References: <20250514111544.1012399-1-david@redhat.com> <20250514111544.1012399-2-david@redhat.com> <7E5BD96D-971F-4AFC-AC09-503310BE8E68@nvidia.com> <0F0BB46A-F7C2-40BE-A045-C65525956D52@nvidia.com> From: David Hildenbrand Autocrypt: addr=david@redhat.com; keydata= xsFNBFXLn5EBEAC+zYvAFJxCBY9Tr1xZgcESmxVNI/0ffzE/ZQOiHJl6mGkmA1R7/uUpiCjJ dBrn+lhhOYjjNefFQou6478faXE6o2AhmebqT4KiQoUQFV4R7y1KMEKoSyy8hQaK1umALTdL QZLQMzNE74ap+GDK0wnacPQFpcG1AE9RMq3aeErY5tujekBS32jfC/7AnH7I0v1v1TbbK3Gp XNeiN4QroO+5qaSr0ID2sz5jtBLRb15RMre27E1ImpaIv2Jw8NJgW0k/D1RyKCwaTsgRdwuK Kx/Y91XuSBdz0uOyU/S8kM1+ag0wvsGlpBVxRR/xw/E8M7TEwuCZQArqqTCmkG6HGcXFT0V9 PXFNNgV5jXMQRwU0O/ztJIQqsE5LsUomE//bLwzj9IVsaQpKDqW6TAPjcdBDPLHvriq7kGjt WhVhdl0qEYB8lkBEU7V2Yb+SYhmhpDrti9Fq1EsmhiHSkxJcGREoMK/63r9WLZYI3+4W2rAc UucZa4OT27U5ZISjNg3Ev0rxU5UH2/pT4wJCfxwocmqaRr6UYmrtZmND89X0KigoFD/XSeVv jwBRNjPAubK9/k5NoRrYqztM9W6sJqrH8+UWZ1Idd/DdmogJh0gNC0+N42Za9yBRURfIdKSb B3JfpUqcWwE7vUaYrHG1nw54pLUoPG6sAA7Mehl3nd4pZUALHwARAQABzSREYXZpZCBIaWxk ZW5icmFuZCA8ZGF2aWRAcmVkaGF0LmNvbT7CwZgEEwEIAEICGwMGCwkIBwMCBhUIAgkKCwQW AgMBAh4BAheAAhkBFiEEG9nKrXNcTDpGDfzKTd4Q9wD/g1oFAl8Ox4kFCRKpKXgACgkQTd4Q 9wD/g1oHcA//a6Tj7SBNjFNM1iNhWUo1lxAja0lpSodSnB2g4FCZ4R61SBR4l/psBL73xktp rDHrx4aSpwkRP6Epu6mLvhlfjmkRG4OynJ5HG1gfv7RJJfnUdUM1z5kdS8JBrOhMJS2c/gPf wv1TGRq2XdMPnfY2o0CxRqpcLkx4vBODvJGl2mQyJF/gPepdDfcT8/PY9BJ7FL6Hrq1gnAo4 3Iv9qV0JiT2wmZciNyYQhmA1V6dyTRiQ4YAc31zOo2IM+xisPzeSHgw3ONY/XhYvfZ9r7W1l pNQdc2G+o4Di9NPFHQQhDw3YTRR1opJaTlRDzxYxzU6ZnUUBghxt9cwUWTpfCktkMZiPSDGd KgQBjnweV2jw9UOTxjb4LXqDjmSNkjDdQUOU69jGMUXgihvo4zhYcMX8F5gWdRtMR7DzW/YE BgVcyxNkMIXoY1aYj6npHYiNQesQlqjU6azjbH70/SXKM5tNRplgW8TNprMDuntdvV9wNkFs 9TyM02V5aWxFfI42+aivc4KEw69SE9KXwC7FSf5wXzuTot97N9Phj/Z3+jx443jo2NR34XgF 89cct7wJMjOF7bBefo0fPPZQuIma0Zym71cP61OP/i11ahNye6HGKfxGCOcs5wW9kRQEk8P9 M/k2wt3mt/fCQnuP/mWutNPt95w9wSsUyATLmtNrwccz63XOwU0EVcufkQEQAOfX3n0g0fZz Bgm/S2zF/kxQKCEKP8ID+Vz8sy2GpDvveBq4H2Y34XWsT1zLJdvqPI4af4ZSMxuerWjXbVWb T6d4odQIG0fKx4F8NccDqbgHeZRNajXeeJ3R7gAzvWvQNLz4piHrO/B4tf8svmRBL0ZB5P5A 2uhdwLU3NZuK22zpNn4is87BPWF8HhY0L5fafgDMOqnf4guJVJPYNPhUFzXUbPqOKOkL8ojk CXxkOFHAbjstSK5Ca3fKquY3rdX3DNo+EL7FvAiw1mUtS+5GeYE+RMnDCsVFm/C7kY8c2d0G NWkB9pJM5+mnIoFNxy7YBcldYATVeOHoY4LyaUWNnAvFYWp08dHWfZo9WCiJMuTfgtH9tc75 7QanMVdPt6fDK8UUXIBLQ2TWr/sQKE9xtFuEmoQGlE1l6bGaDnnMLcYu+Asp3kDT0w4zYGsx 5r6XQVRH4+5N6eHZiaeYtFOujp5n+pjBaQK7wUUjDilPQ5QMzIuCL4YjVoylWiBNknvQWBXS lQCWmavOT9sttGQXdPCC5ynI+1ymZC1ORZKANLnRAb0NH/UCzcsstw2TAkFnMEbo9Zu9w7Kv AxBQXWeXhJI9XQssfrf4Gusdqx8nPEpfOqCtbbwJMATbHyqLt7/oz/5deGuwxgb65pWIzufa N7eop7uh+6bezi+rugUI+w6DABEBAAHCwXwEGAEIACYCGwwWIQQb2cqtc1xMOkYN/MpN3hD3 AP+DWgUCXw7HsgUJEqkpoQAKCRBN3hD3AP+DWrrpD/4qS3dyVRxDcDHIlmguXjC1Q5tZTwNB boaBTPHSy/Nksu0eY7x6HfQJ3xajVH32Ms6t1trDQmPx2iP5+7iDsb7OKAb5eOS8h+BEBDeq 3ecsQDv0fFJOA9ag5O3LLNk+3x3q7e0uo06XMaY7UHS341ozXUUI7wC7iKfoUTv03iO9El5f XpNMx/YrIMduZ2+nd9Di7o5+KIwlb2mAB9sTNHdMrXesX8eBL6T9b+MZJk+mZuPxKNVfEQMQ a5SxUEADIPQTPNvBewdeI80yeOCrN+Zzwy/Mrx9EPeu59Y5vSJOx/z6OUImD/GhX7Xvkt3kq Er5KTrJz3++B6SH9pum9PuoE/k+nntJkNMmQpR4MCBaV/J9gIOPGodDKnjdng+mXliF3Ptu6 3oxc2RCyGzTlxyMwuc2U5Q7KtUNTdDe8T0uE+9b8BLMVQDDfJjqY0VVqSUwImzTDLX9S4g/8 kC4HRcclk8hpyhY2jKGluZO0awwTIMgVEzmTyBphDg/Gx7dZU1Xf8HFuE+UZ5UDHDTnwgv7E th6RC9+WrhDNspZ9fJjKWRbveQgUFCpe1sa77LAw+XFrKmBHXp9ZVIe90RMe2tRL06BGiRZr jPrnvUsUUsjRoRNJjKKA/REq+sAnhkNPPZ/NNMjaZ5b8Tovi8C0tmxiCHaQYqj7G2rgnT0kt WNyWQQ== Organization: Red Hat In-Reply-To: <0F0BB46A-F7C2-40BE-A045-C65525956D52@nvidia.com> X-Mimecast-Spam-Score: 0 X-Mimecast-MFC-PROC-ID: KvwH9jENz5Ky8MZOGlya1ceHpiZ7ocuxS_hvKGyR0cY_1747665580 X-Mimecast-Originator: redhat.com Content-Language: en-US Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Rspamd-Server: rspam10 X-Stat-Signature: 13ui387fddstm4jytihn8q4wzpyxbb7c X-Rspamd-Queue-Id: 92557180017 X-Rspam-User: X-HE-Tag: 1747665585-106406 X-HE-Meta: U2FsdGVkX1+yOHc7ehfao/l6IyNHD8iR97DTwres8d0G1+aLcRgBgfJJ4Mu1GE/8IQcKR+UZbgQb80et9dJq2jW+GXUtv/wRNQeoIMqKmhUXeXlPlXoO8INBDEGoWOVh2cqmbd6VDQRD3xCm5vOMDcwS4yZkF0ChXorcBPoUBu8/8wiWVvLMQUdwHSYZxGR5CN2B0TQnrr7xZFb48wrhovY7iIs3Fq1e4yTIunP9/sNGTc/UdrJ4bt1E7jbgsY6U4rGnOY0jnCXT+882HQ179UWgddSkgDfwNozmojKmN+wZg31bLKXfNITtvsB4dSyqtRpLyOrF+s0QSDL9K/6N1a9fk8dw8Jz6kjJQ7Za1ZtE8mNGNEPsxmiAFBubjn0IFVWQEe3Ld/3TpuRGf+GL0QuuDZ9ufh4cn64CKtywBldiHS1PsxrLMDI4zA/Ffjd/uMxxIg6memTUh/VUlYOACE/M5sd/AFLDgbBjYZaYbsqZ2D7pTWHGT4QTRyOeBX5d9ofYfvi8CaH2WrRmqyTwRrtzgjYYQqp02DgOuDmsJosmbz1D638IDwMyPtoN5nF6MMmXU9bGki52zM6vJ/JQdtdylUN+881iVFI+udO4ogDW1FnHb6mXYk8KNTNk3aZd1nMi12CJw6haPRShFSw1eZNr7qPevVYlNENtU0vRY6zmp06qwck/C0PcHdBObHGJwmkuXCo8HgtscXi6utusj8tJbk+6/3hZKQMPvTczP6EovdIDZ1b0UsQ4wu4/mXsUjOHmPECeBnW53hfGdrjS5knYsWNSSN42Jgxl++gZxISwNhJ3Tf9YfyYQWexO/RaBbVyyJChNjH1wnfYloRMuWwI+huJpHg/piBWFUHMC6phPqAp8s/zemJAb/VAs4RvfjmE2Sk969GyWcVGqnt8wBiM1u19465MLzWZA9CxE2FFvOeRnQx9bzbfUkx8uXmMQtjk80MlBlvgvoeWVlxs6 2FZ0u6cj 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 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: List-Subscribe: List-Unsubscribe: >>> Some thoughts after reading the related code: >>> >>> offline_pages() is a little convoluted, since it has two steps to make >>> sure a range of memory can be offlined: >>> 1. start_isolate_page_range() checks unmovable (maybe not migratable >>> is more precise) pages using has_unmovable_pages(), but leaves unmovable >>> PageOffline() page handling to the driver; >> >> Right, it's best-effort. For ZONE_MOVABLE we're skipping the checks completely. >> >> I was wondering for a longer time that -- with the isolate flag being a separate bit soon :) -- we could simply isolate the whole range, and then fail if we stumble over > > Talking about that, I will need your input on my change to move_pfn_range_to_zone() > in online_pages()[1]. Making MIGRATE_ISOLATE a separate bit means if you isolate > a pageblock without a migratetype, unisolating it will give an unpredictable > migratetype. Sorry for my late reply on that. I think, the rule should be that you initialize the migratetype once, before any isolation+un-isolation. So that should only require care when adding new pageblocks (memory hotplug). > > [1] https://lore.kernel.org/linux-mm/20250509200111.3372279-3-ziy@nvidia.com/ > >> an unmovable page during migration. That is, get rid of has_unmovable_pages() entirely. Un-doing the isolation would then properly preserve the migratetype -- no harm done? >> >> Certainly worth a look. What do you think about that? > > In principle, the method should work and simplifies the code. But it suffers more > penalty (pages are migrated) when an unmovable page is encountered after the > isolation, since before nothing will be migrated. To mitigate this, > we probably would want some retry mechanism. For example, register a callback > to each unmovable page and once the unmovable page is deallocated, > alloc_contig_range() can move forward progress. I was wondering if we could make the isolation code a bit simpler. For example, set all involved pageblocks isolated. If any is already isolated, we can easily back off. Once all are isolated, we could do the has_unmovable_pages() on the whole range, not a single pageblock. Then we could start migrating. I think the "issue" is, once we drop the zone lock, pages that are getting freed end up on the MIGRATE_ISOLATE list -- and I think we also must have moved the free pages already to the proper MIGRATE_ISOLATE list before we drop the zone lock. So the possible cleanups might be limited. > >> >>> 2. scan_movable_pages() and do_migrate_range() migrate pages and handle >>> different types of PageOffline() pages. >> >> Right, migrate what we can, skip these special once, and bail out on any others (at least in this patch, patch #2 restores the pre-virtio-mem behavior). >> >>> >>> It might make the logic cleaner if start_isolate_page_range() can >>> have a callback to allow driver to handle PageOffline() pages. >> >> We have to identify them repeadetly, so start_isolate_page_range() would not be sufficient. Also, callbacks are rather tricky for the case where we cannot really stabilize the page. > > During start_isolate_page_range(), all pageblocks are isolated, so > free pages cannot be used by anyone else, meaning no new PageOffline() > pages or any other unmovable pages, right? Yes, that is correct. Pages (incl. PageOffline) pages could get freed back to the buddy. But we don't want to store per-page callbacks either way. What would work is a new notifier chain (protected by RCU etc), that we can ask for each PageOffline page. Not sure I prefer that of the approach here that is significantly simpler -- and we do have the spare bit for PageOffline pages in the new memdec world (for PageOffline, we'll probably need 2/3 flags in the long run). > >> >>> >>> Hmm, Skippable PageOffline() pages are virtio-mem specific, I wonder >>> why offline_pages() takes care of them. Shouldn't virtio-mem driver >>> handle them? >>> I also realize that the two steps in offline_pages()> are very similar to alloc_contig_range() and virtio-mem is using >>> alloc_contig_range(). I wonder if the two steps in offline_pages() >>> can be abstracted out and shared with virtio-mem.Yes, offline_pages()> operates at memory section granularity, different from the granularity, >>> pageblock size, of alloc_contig_range() used in virtio-mem, but >>> both are trying to check unmovable pages and migrate movable pages. >> >> Not sure I get completely what you mean. virtio-mem really wants to use alloc_contig_range() to allocate ranges it wants to unplug, because it will fail fast and allows for smaller ranges. >> >> offline_pages() is primarily also about offlining the memory section, which virtio-mem must do in order to remove the Linux memory block. > > To clarify, I mean the steps of start_isolate_page_range(), scan_movable_pages(), > do_migrate_range(), dissolve_free_hugetlb_folios() and test_pages_isolated() in > offline_pages() is very similar to the steps of start_isolate_page_range(), > __alloc_contig_migrate_range(), replace_free_hugepage_folios(), > and test_pages_isolate() in alloc_contig_range(). So I wonder if a common > function routine can be shared. Ah, yes, I had the same idea a couple of years back, but never got to it. There are subtle differences (e.g., memory offlining keeps retrying and is allowed to dissolve free hugetlb folios), but these could likely be modified using a parameter (similar to how we already special-case offlining). Thanks! -- Cheers, David / dhildenb