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 02D2BD1CA15 for ; Tue, 5 Nov 2024 03:01:41 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id 38F4A6B00A5; Mon, 4 Nov 2024 22:01:41 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 33F5D6B00A6; Mon, 4 Nov 2024 22:01:41 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 207446B00A7; Mon, 4 Nov 2024 22:01:41 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0012.hostedemail.com [216.40.44.12]) by kanga.kvack.org (Postfix) with ESMTP id EF8576B00A5 for ; Mon, 4 Nov 2024 22:01:40 -0500 (EST) Received: from smtpin25.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay06.hostedemail.com (Postfix) with ESMTP id 99E5CACA30 for ; Tue, 5 Nov 2024 03:01:40 +0000 (UTC) X-FDA: 82750539942.25.5A24792 Received: from mail-qv1-f46.google.com (mail-qv1-f46.google.com [209.85.219.46]) by imf02.hostedemail.com (Postfix) with ESMTP id 1BB4C8001E for ; Tue, 5 Nov 2024 03:00:36 +0000 (UTC) Authentication-Results: imf02.hostedemail.com; dkim=pass header.d=gmail.com header.s=20230601 header.b=RXoXYw7R; spf=pass (imf02.hostedemail.com: domain of laoar.shao@gmail.com designates 209.85.219.46 as permitted sender) smtp.mailfrom=laoar.shao@gmail.com; dmarc=pass (policy=none) header.from=gmail.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1730775563; 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=EuHFSUgmsz7WG7mvI/d8v4tpmdXYVDKyT1zJ7B6g1JA=; b=HGXA1MBj6fsMusOLMNK/TPoWnpPkKW47USuY+3JfpqaopVIch5QTblLfMbsBFoCOvPtrDk XIIbsnrLuYnqr5Z9ykO+E+oIwkm0O2KrB9auiYzQTMd18I5Qcnc5Wt5Z3R36ctUu3hkmTJ XqK89EiSr1PGZ+q6JtTtbM+XBH1EBCY= ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1730775563; a=rsa-sha256; cv=none; b=jgyPXD9Rbz+SSnMvx0Xxd2sebW17deMeodDrX/STGnwNF/3Lo8NO92oLOy1I6ouaGXXh+V xQ6IEW5+Q3ZHj9mpzsSQjyP+1EXFUBhQFeWDywc2b0BIlRCw+tHIEaz5Mcznw0u6pX6EAL XWXqaKv9lXAlialYyGaKklVeC4f8Q+4= ARC-Authentication-Results: i=1; imf02.hostedemail.com; dkim=pass header.d=gmail.com header.s=20230601 header.b=RXoXYw7R; spf=pass (imf02.hostedemail.com: domain of laoar.shao@gmail.com designates 209.85.219.46 as permitted sender) smtp.mailfrom=laoar.shao@gmail.com; dmarc=pass (policy=none) header.from=gmail.com Received: by mail-qv1-f46.google.com with SMTP id 6a1803df08f44-6d36f7cf765so20645076d6.0 for ; Mon, 04 Nov 2024 19:01:38 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1730775698; x=1731380498; darn=kvack.org; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:from:to:cc:subject:date :message-id:reply-to; bh=EuHFSUgmsz7WG7mvI/d8v4tpmdXYVDKyT1zJ7B6g1JA=; b=RXoXYw7RQkooOCwazg2Lf5sR0cS1e3K6pVl6qOEli4pckrkU0Fe6CwlWeOcCstR/Cm HN63Bs6F8tE/U0hxIRoJr6WwOqNMEBrOr1Lo1BW/TTCjAuIiWFozfYZM0ozce1iQBCkv 8y0aCNfE6Z0YvKjC5C+f4XKzkiEf9Zx6U40twdxpJ3aM1mK3yBf45YTqiZIZ8UuUyfBz DTBm6PHmWwcoz1INp2BjRfO/CaYv18KFk2e+bSpBMPOSyeiLjI+kE4wkccVt6xCRveOm hKI4MpO6P9Sl63ceDxy4G3WSLaRMil4u/cRXZ+8O1MS1qhqBeGvdTlP4pf5WitTy/UUa fVZg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1730775698; x=1731380498; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=EuHFSUgmsz7WG7mvI/d8v4tpmdXYVDKyT1zJ7B6g1JA=; b=QZLRFxBlMEadWvQwOuwxjzE7F31JR+WOcCAkXPRRllsGNq47oHTcjgdW0n9zwCQ6Al JvSBW7VpBnFoKhugrDTkaULs1aCNMqNadDkhsoyExuXdnw/RQ8YLNraslbQbZBHi/8+U rIKemCeDBTKVHFZI1m3jJ4PfkWmjNELehPTQEONkYst/gGUxxZcbqU6mnlb9bLOU59js p5ojSOH+DQrkdLWUb3Lsqivp9zYAx3BLWTlgBSWFIAtXUmYXNmZhZZthpY32INBES7U6 R9dSuGV57KT5ZGU4X/sa63oKRro3LuABQX1VeM2iIydeGH+nOaHVh/WtnlcUZxc96vba tk8A== X-Forwarded-Encrypted: i=1; AJvYcCV5Nr6Yjb6qhOAi8FV9Aj6hVvUCN9SoJS1OlR0ot8izRz/OChHbUtFokkLcQKAG/apkBNmlilUrpQ==@kvack.org X-Gm-Message-State: AOJu0YyKdvFL2WEGqzVX+X8e8ws6iCMfCqN7ai3w3+WlvsuVDnd29gkb rnLOdF9UwAQoconYG1nS0GqZCPFQGgzsZaS/2oNyaSL7LdWTm9eD3smQhiZ3cGy1+NZWVf96Cqw Y3424y80Ul2eqB6Yo2Ji+K8fMznw= X-Google-Smtp-Source: AGHT+IFgjpEeUJ9o5NYbXi2YGrKNgYqShK4CP7xKAPW4g3FA2ykBQEfVMrkhK+y67JiMDlKg5ZYT7c9Tdj2KRq4SxlA= X-Received: by 2002:a05:6214:4a86:b0:6cb:e6fa:4960 with SMTP id 6a1803df08f44-6d35bff55dbmr219466406d6.0.1730775697849; Mon, 04 Nov 2024 19:01:37 -0800 (PST) MIME-Version: 1.0 References: <20241104143015.34684-1-laoar.shao@gmail.com> In-Reply-To: From: Yafang Shao Date: Tue, 5 Nov 2024 11:01:01 +0800 Message-ID: Subject: Re: [RFC PATCH] mm: Add large folio support for async readahead To: Matthew Wilcox Cc: akpm@linux-foundation.org, linux-mm@kvack.org Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Rspam-User: X-Rspamd-Server: rspam12 X-Rspamd-Queue-Id: 1BB4C8001E X-Stat-Signature: mcbibe4u3ea8khyrat5yedmzews6jfy5 X-HE-Tag: 1730775636-364575 X-HE-Meta: U2FsdGVkX18iccQTBrs3W/HLTdXmaYMJ7cUAawb6ggj1bKm6Sk8pTyr4nT7geokFzuqjYex+h21NmRZllFEKTpBQ9S0D/rmrkKD4V2YxJlGBhn5knNuLXWDX2pqvzw7PLb5EY7KwRRnd9e152spHlmq+S3rMrehA+YnbnZbvOu65+CGrKshrxi9ZGhLwobJ0TQn17W0vuVDmSN7P7/ocfIvfN7sk/J6aC1HhCWatZVKLbSt2hiRcrVAACgcbIAzZ8nkUQiLMqMH2mYKK+5W/XSJqkVhCkYTnnJnRx/UXRhcetwX2jbmEYDvJEH+7T0q4wethx9sDRumpLkgeP3L6wV/GtQYbwromgvPTHHCQ0hqk9CJaA6IY6SLI5NEmWT+l1Vj+nuiZ6lICM84Lyc+GhIZSXvSWqTtR66OY90RE8u81f+LuGXzYJ1tFW5bAdGe0FOka6XJESD/ujycE15aijvczI43CRwkJTOcOrJpk8UXhKTB4Gd9iXX40C0slp9ZLjmCuOPtx38GI4f5Ivv/5dSRCqyu5hkQmIgElJRRrokhXd7I6JEx7lWm2kcr9k6aDBHOpci8D87U0t3u9J2Io7s7WJZTjbCCcm707kBS/ZJLcjxLdvlqql6etumm48ZcOTBM/ZENTB81apudJJhTKzxbMZeOdtgsHoecaFFt9mQoNdRzVWS7RUeXGtAETZWP4Ff1BRwxhCy5dMP4Kv9e/Z9l3be1E8Y/yAwIFfBHowxJ+e8xOX81umGAau0HS+ut2Lzxw8EVc2vtXu5voN6egI5FtbKZKdTGGTljRoLbzTGetnW0pEWQIU6+zPvs4SpRAfUOTW1XWKYbTY/RudEVhNjQNOKu+wYKatlg8EC0uoY3HANDM2GNjzuyNxUXnnAS0hem7v/gsPcQCjJQ83PJ1NitJVFowkDRF1ukjQdPiFiWE5lkDqhMESd3PmQ4HW2d5RkpELgbBC084e7WVMEe QhoL7RBs +KilMHfUZiQNx8EzOdgLP91EKKVdur+TI4MJ9pTB/mbztFRQz+AdTj/cOUNkhVPLWRI9HdlgIjhWqfqFVUSVjWSUPwM+SIovvxl23uKvr0nh37x07ZS6d5luNgiCOtDUOdBjSnou4xgGeGEIHUYkT/+NGvV+BizpkFR2l4wmJUJA+0O0J412w1d4q4nJkhAOo8UjLFQqP5lhVPjqmxaEYhvH7WHAhgU14ehTm/oWaokvKwtjcFzcu5pxU/xl3rbvAtuu0teBclLtnS8FkQjvhO80Z+TzFN3CvhXdWmWUyPhE02z+HUnSUJPzyhct/pd+vpBLOC0j+0uAZypEXLn0yqZZDc/nFZ070eT6CKqsdv0TC54YtZv4M/m/OBdvRDrRhKWOpTOuvWoBqqOzeOmzCMKWh5cVv4Lg7Ujsq 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: On Tue, Nov 5, 2024 at 1:00=E2=80=AFAM Matthew Wilcox = wrote: > > On Mon, Nov 04, 2024 at 10:30:15PM +0800, Yafang Shao wrote: > > When testing large folio support with XFS on our servers, we observed t= hat > > only a few large folios are mapped when reading large files via mmap. T= his > > behavior occurs because large folio support is currently implemented on= ly > > for sync readahead, not for async readahead. Consequently, while the > > first filemap fault may map to a large folio, subsequent filemap faults= are > > mapped to regular folios. This can be verified with a simple test case,= as > > shown below: > > I awear I tested this and it worked at the time. Here's what's supposed > to happen: > > You touch the first byte of the mapping, and there's no page in the page > cache. So we go into the sync readahead path, and force it to read two > PMDs. From your email, I assume this is succeeding. The readahead > flag gets set on the second PMD so that when we get to 2MB, we go into > the 'time to do more readahead' path, ie the 'do_async_mmap_readahead' > function you patch below. > > Now, page_cache_async_ra() does this: > > unsigned int order =3D folio_order(folio); > ... > if (index =3D=3D expected) { > ra->start +=3D ra->size; > ra->size =3D get_next_ra_size(ra, max_pages); > ra->async_size =3D ra->size; > goto readit; > } > readit: > ractl->_index =3D ra->start; > page_cache_ra_order(ractl, ra, order); > > So it should already be doing readahead of PMD size. Can you dig into > what's going wrong, now that you understand that this is supposed to > work already? Upon further analysis, it appears that the behavior depends on the /sys/block/*/queue/read_ahead_kb setting. On my test server, this parameter is set to 128KB. For the current approach to work without modification, it would need to be 2MB-aligned. However, I believe MADV_HUGEPAGE behavior should not be dependent on the value of read_ahead_kb. It would be more robust if the kernel automatically aligned it to 2MB when MADV_HUGEPAGE is enabled. The following changes ensure compatibility with non-2MB-aligned read_ahead_kb values: diff --git a/mm/readahead.c b/mm/readahead.c index 3dc6c7a128dd..c024245497cc 100644 --- a/mm/readahead.c +++ b/mm/readahead.c @@ -641,9 +641,9 @@ void page_cache_async_ra(struct readahead_control *ract= l, expected =3D round_down(ra->start + ra->size - ra->async_size, 1UL << order); if (index =3D=3D expected) { - ra->start +=3D ra->size; - ra->size =3D get_next_ra_size(ra, max_pages); - ra->async_size =3D ra->size; + ra->start +=3D ALIGN(ra->size, HPAGE_PMD_NR); + ra->size =3D ALIGN(get_next_ra_size(ra, max_pages), HPAGE_P= MD_NR); + ra->async_size =3D ALIGN(ra->size, HPAGE_PMD_NR); goto readit; } In addition, adjustments may be needed in the sync readahead path to allow MADV_HUGEPAGE readahead sizes to be tunable based on read_ahead_kb. I will continue to analyze this and submit a revised version accordingly. -- Regards Yafang