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 X-Spam-Level: X-Spam-Status: No, score=-6.5 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_HELO_NONE,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id C8EF9C2BA2B for ; Fri, 10 Apr 2020 15:56:17 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 7C4FF20CC7 for ; Fri, 10 Apr 2020 15:56:17 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (2048-bit key) header.d=shutemov-name.20150623.gappssmtp.com header.i=@shutemov-name.20150623.gappssmtp.com header.b="t2Hk/GfO" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 7C4FF20CC7 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=shutemov.name Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 2241D8E0044; Fri, 10 Apr 2020 11:56:17 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 1D34B8E0003; Fri, 10 Apr 2020 11:56:17 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 0EA2F8E0044; Fri, 10 Apr 2020 11:56:17 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0194.hostedemail.com [216.40.44.194]) by kanga.kvack.org (Postfix) with ESMTP id E7D2F8E0003 for ; Fri, 10 Apr 2020 11:56:16 -0400 (EDT) Received: from smtpin09.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with ESMTP id 9FF08180AD811 for ; Fri, 10 Apr 2020 15:56:16 +0000 (UTC) X-FDA: 76692397152.09.cat35_19a49cccabe26 X-HE-Tag: cat35_19a49cccabe26 X-Filterd-Recvd-Size: 7677 Received: from mail-lf1-f65.google.com (mail-lf1-f65.google.com [209.85.167.65]) by imf06.hostedemail.com (Postfix) with ESMTP for ; Fri, 10 Apr 2020 15:56:16 +0000 (UTC) Received: by mail-lf1-f65.google.com with SMTP id f8so1629910lfe.12 for ; Fri, 10 Apr 2020 08:56:15 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=shutemov-name.20150623.gappssmtp.com; s=20150623; h=date:from:to:cc:subject:message-id:references:mime-version :content-disposition:content-transfer-encoding:in-reply-to; bh=5oXzzweU9+b1N5+cGj6pvbd+Y8KpBQDkz61NLnD/b7c=; b=t2Hk/GfOBLvuuJ342Lrin5iCn8jW7Hdl70kSAp5chiaxVe3jyF6APhVBUz0JD9wR9C csCdUGmml/oC1rstoIwryKphaM37BYtFVKrv+Uz5ib0rljmOqRayvK04I+Ux3bBdZVZU gidlwxUtSxfhwE5i6DHR013lTGAu27xnhQ61ditOHV5ZRlPksyJtrt/KC9KC72u32hLC kNju4qiF4rOQOszUwaVQAvaZHVMM6+RuNrCa79UvgBzMyJnW/j3sAZ75qtAnkjiJi0UK 86ipF3Ny/DWYDMNdjEAFMrqs73wIyoWWxB3kkB9oq5Xhs5ZNcQ/4IMxXsnqzS5Xf1218 gnjA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:date:from:to:cc:subject:message-id:references :mime-version:content-disposition:content-transfer-encoding :in-reply-to; bh=5oXzzweU9+b1N5+cGj6pvbd+Y8KpBQDkz61NLnD/b7c=; b=nPuQVbC2KBp7eTwF5zKdIzR/iVUUUPR7za2enHTrVeAiyNa40q2SdS4Ac2m5E+JRKD oVS5ddY7lbuFMYfUXAM7OU0ycF/QCmagz9gEXSrFrXSI9fkYSoD1YY1WkIHXIjw6HneZ 5R+/FJ2PkFRL6RO4NrIFo2ChRq+8xkXEmE/gKxKQuMVEVbf6r+ai77y4xGlWMSFQMn64 In9847lTRYhf7C1laPEDPr99Jm13tRqYgMMo6JlO/8KIN9VYV3bBkwzF9D8owrURhyYf 1K5dBuIrD4fhQQRz5eNshoRRiTYelqn5Cc2RY0dYi+prCNG1bTQ0hqq7g2JF0g+NHOms ulGQ== X-Gm-Message-State: AGi0PubzTu1jJJh31HSK9Cy2o41uHLY1HHxNk2AOCEzmMX9ar48tafE5 OQpT1f7NtVxy0NvgQkgBfCk3/eC+yLQ= X-Google-Smtp-Source: APiQypI9qvle8foeI1LCtXi2mzT1MhdY9shhQWtO1nZMume9pEjCI8OG2k/lFRxBm7w9FG5W1mkOMw== X-Received: by 2002:a05:6512:686:: with SMTP id t6mr3048293lfe.163.1586534174501; Fri, 10 Apr 2020 08:56:14 -0700 (PDT) Received: from box.localdomain ([86.57.175.117]) by smtp.gmail.com with ESMTPSA id w10sm1264769ljh.85.2020.04.10.08.56.13 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 10 Apr 2020 08:56:13 -0700 (PDT) Received: by box.localdomain (Postfix, from userid 1000) id DA8EE102110; Fri, 10 Apr 2020 18:56:13 +0300 (+03) Date: Fri, 10 Apr 2020 18:56:13 +0300 From: "Kirill A. Shutemov" To: Yang Shi Cc: akpm@linux-foundation.org, Andrea Arcangeli , Zi Yan , linux-mm@kvack.org, linux-kernel@vger.kernel.org, "Kirill A. Shutemov" Subject: Re: [PATCHv2 5/8] khugepaged: Allow to callapse a page shared across fork Message-ID: <20200410155613.cxqplrmwckezx3hl@box> References: <20200403112928.19742-1-kirill.shutemov@linux.intel.com> <20200403112928.19742-6-kirill.shutemov@linux.intel.com> <20200408131044.xzlheacvslrbwrja@box> <107630f5-bbde-3f78-23e9-6f6b3113d709@linux.alibaba.com> <61569ead-2bbd-6d18-c04c-3251218a3fcd@linux.alibaba.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <61569ead-2bbd-6d18-c04c-3251218a3fcd@linux.alibaba.com> Content-Transfer-Encoding: quoted-printable 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: On Thu, Apr 09, 2020 at 05:03:00PM -0700, Yang Shi wrote: >=20 >=20 > On 4/8/20 11:51 AM, Yang Shi wrote: > >=20 > >=20 > > On 4/8/20 6:10 AM, Kirill A. Shutemov wrote: > > > On Mon, Apr 06, 2020 at 01:50:56PM -0700, Yang Shi wrote: > > > >=20 > > > > On 4/3/20 4:29 AM, Kirill A. Shutemov wrote: > > > > > The page can be included into collapse as long as it doesn't > > > > > have extra > > > > > pins (from GUP or otherwise). > > > > >=20 > > > > > Signed-off-by: Kirill A. Shutemov > > > > > --- > > > > > =A0=A0 mm/khugepaged.c | 25 ++++++++++++++----------- > > > > > =A0=A0 1 file changed, 14 insertions(+), 11 deletions(-) > > > > >=20 > > > > > diff --git a/mm/khugepaged.c b/mm/khugepaged.c > > > > > index 57ff287caf6b..1e7e6543ebca 100644 > > > > > --- a/mm/khugepaged.c > > > > > +++ b/mm/khugepaged.c > > > > > @@ -581,11 +581,18 @@ static int > > > > > __collapse_huge_page_isolate(struct vm_area_struct *vma, > > > > > =A0=A0=A0=A0=A0=A0=A0=A0=A0=A0 } > > > > > =A0=A0=A0=A0=A0=A0=A0=A0=A0=A0 /* > > > > > -=A0=A0=A0=A0=A0=A0=A0=A0 * cannot use mapcount: can't collapse= if there's a gup pin. > > > > > -=A0=A0=A0=A0=A0=A0=A0=A0 * The page must only be referenced by= the scanned process > > > > > -=A0=A0=A0=A0=A0=A0=A0=A0 * and page swap cache. > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * Check if the page has any GUP (or o= ther external) pins. > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * The page table that maps the page h= as been > > > > > already unlinked > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * from the page table tree and this p= rocess cannot get > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * additinal pin on the page. > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * New pins can come later if the page= is shared across fork, > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * but not for the this process. It is= fine. The > > > > > other process > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0 * cannot write to the page, only trig= ger CoW. > > > > > =A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0 */ > > > > > -=A0=A0=A0=A0=A0=A0=A0 if (page_count(page) !=3D 1 + PageSwapCa= che(page)) { > > > > > +=A0=A0=A0=A0=A0=A0=A0 if (total_mapcount(page) + PageSwapCache= (page) !=3D > > > > > +=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0 page_count(page)= ) { > > > > This check looks fine for base page, but what if the page is > > > > PTE-mapped THP? > > > > The following patch made this possible. > > > >=20 > > > > If it is PTE-mapped THP and the page is in swap cache, the > > > > refcount would be > > > > 512 + the number of PTE-mapped pages. > > > >=20 > > > > Shall we do the below change in the following patch? > > > >=20 > > > > extra_pins =3D PageSwapCache(page) ? nr_ccompound(page) - 1 : 0; > > > > if (total_mapcount(page) + PageSwapCache(page) !=3D page_count(pa= ge) - > > > > extra_pins) { > > > > ... > > > Looks like you're right. > > >=20 > > > It would be nice to have a test case to demonstrate the issue. > > >=20 > > > Is there any way to trigger moving the page to swap cache? I don't > > > see it > > > immediately. > >=20 > > It sounds not easy to trigger since it totally depends on timing, I'm > > wondering we may have to use MADV_PAGEOUT? Something below off the to= p > > of my head may trigger this? > >=20 > >=20 > > =A0=A0=A0 CPU=A0=A0=A0 =A0=A0 A=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0= =A0=A0=A0 =A0 =A0=A0=A0 =A0 =A0=A0=A0 =A0=A0=A0 CPU=A0=A0=A0 B =A0=A0=A0= =A0=A0=A0 =A0=A0=A0 > > =A0=A0=A0 =A0=A0=A0 CPU=A0=A0=A0 C > > In parent: > > MADV_HUGEPAGE > > page fault to fill with THP > > fork > > =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0= =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 In Child: > > MADV_NOHUGEPAGE > > MADV_DONTNEED (split pmd) > > MADV_PAGEOUT > > =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0= =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0= =A0=A0 -> > > add_to_swap > > =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0= =A0=A0=A0 khugepaged scan parent and try to > > collapse PTE-mapped > >=20 > > =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0= =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0= =A0=A0 -> > > try_to_unmap > >=20 > > When doing MADV_DONTNEED we need make sure head page is unmapped sinc= e > > MADV_PAGEOUT would call page_mapcount(page) to skip shared mapping. >=20 > This can't trigger it since MADV_PAGEOUT would isolate the page from lr= u so > khugepaged_scan_pmd() can't reach the refcount check at all. >=20 > If try_to_unmap() or pageout() fails, the page may stay in swap cache. = But I > didn't think of an easy way to make this happen. Okay, I'll fix it without a test case. --=20 Kirill A. Shutemov