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=-9.5 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,URIBL_BLOCKED,USER_AGENT_GIT 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 AC8E0C43331 for ; Fri, 27 Mar 2020 17:06:27 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 55BDE2073B for ; Fri, 27 Mar 2020 17:06:27 +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="Q9Z9XrgC" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 55BDE2073B 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 921E26B0071; Fri, 27 Mar 2020 13:06:09 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 7EA766B0078; Fri, 27 Mar 2020 13:06:09 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 355676B0072; Fri, 27 Mar 2020 13:06:09 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0052.hostedemail.com [216.40.44.52]) by kanga.kvack.org (Postfix) with ESMTP id E4F966B0070 for ; Fri, 27 Mar 2020 13:06:08 -0400 (EDT) Received: from smtpin11.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay02.hostedemail.com (Postfix) with ESMTP id CA2B13CE1 for ; Fri, 27 Mar 2020 17:06:08 +0000 (UTC) X-FDA: 76641770016.11.pets69_6fcf5c0d2a10f X-HE-Tag: pets69_6fcf5c0d2a10f X-Filterd-Recvd-Size: 12587 Received: from mail-lf1-f68.google.com (mail-lf1-f68.google.com [209.85.167.68]) by imf32.hostedemail.com (Postfix) with ESMTP for ; Fri, 27 Mar 2020 17:06:08 +0000 (UTC) Received: by mail-lf1-f68.google.com with SMTP id z23so8487120lfh.8 for ; Fri, 27 Mar 2020 10:06:07 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=shutemov-name.20150623.gappssmtp.com; s=20150623; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=5kQB28Q1srv5R/NFD9NiHhGkcte0ot9iIq76f8JehPk=; b=Q9Z9XrgC1agbGS6iVd1s2BTilTO0EpLN+sb0E1/T4r7u8Tvb/ifohsIOQxVZiio+VQ SOg+W0ADbWLo87YzursLjwt4RcWg7zpPiDeBYSN+wtK1dxLrdrN0FtDa8DCML9ASlM27 bsw0DhpdD0wElolWk9Z3X/JGQAkfKOAmpLQ4orKmxSwYv/1DkIGC6Q2T/V6IlhrfUpMk nj5Qf/2D2DeyTiKHJnw83cEmIl+yl7R+3GyT9i+laFYnxUy6D5kzUFPjLzN2qJU6GmXw 9g/+0SmJgLcq7O+XfJ7D3/0ifej4g5dwkW2frwZ04PGgV1XfGtkMwtUQbYujGNL05CDE 6psw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=5kQB28Q1srv5R/NFD9NiHhGkcte0ot9iIq76f8JehPk=; b=N/ECi14/L1A76l1pdPzQpN6lvOrBsLf9USVI8Nidm5STwVDVIKt/EB4159AAKAjbNc hLXkB/4WfHqImNC9bWpJFDf9DicODeWe00DzhAviaW2OPbKVYbBqgu6RgfBC2uMZvHT0 HBYntYv2VppSLivyeT8YGcLIhN6brs4Un5cVDOSdNa3PH+kS5iekVbi4l/CNANe6ZcIL c7XY+pOxWXZAuOAwA15dfXRGK+G3LcvdGem28dSs/MoyipgocsH512dBySoI0viJez+m RnaVJ00qeyXFBO5dr39if5wrgDcLNptDdo81Mlt6FYxTv29vH0FfTVCR40ugh81F4dhb PsIw== X-Gm-Message-State: AGi0PuaSPWcN6cIZBZaoxVGMRfDUtHTgZCMI35WmYyGq5qh8T3mByNp0 LJFVZwnkWiAQfhjufi2V+9SirG7hmMw= X-Google-Smtp-Source: APiQypJTg/7Jr+2YbWa9fzssusziciWK2L5tL//ePdoqgZeZK4AL4xepP8yE4nyjxIGV8HzMS9IywA== X-Received: by 2002:ac2:4316:: with SMTP id l22mr185811lfh.150.1585328766284; Fri, 27 Mar 2020 10:06:06 -0700 (PDT) Received: from box.localdomain ([86.57.175.117]) by smtp.gmail.com with ESMTPSA id x128sm3240696lff.67.2020.03.27.10.06.04 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 27 Mar 2020 10:06:05 -0700 (PDT) From: "Kirill A. Shutemov" X-Google-Original-From: "Kirill A. Shutemov" Received: by box.localdomain (Postfix, from userid 1000) id 84E70100D2C; Fri, 27 Mar 2020 20:06:07 +0300 (+03) To: akpm@linux-foundation.org, Andrea Arcangeli Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org, "Kirill A. Shutemov" Subject: [PATCH 7/7] khugepaged: Introduce 'max_ptes_shared' tunable Date: Fri, 27 Mar 2020 20:06:01 +0300 Message-Id: <20200327170601.18563-8-kirill.shutemov@linux.intel.com> X-Mailer: git-send-email 2.26.0 In-Reply-To: <20200327170601.18563-1-kirill.shutemov@linux.intel.com> References: <20200327170601.18563-1-kirill.shutemov@linux.intel.com> MIME-Version: 1.0 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: ``max_ptes_shared`` speicies how many pages can be shared across multiple processes. Exeeding the number woul block the collapse:: /sys/kernel/mm/transparent_hugepage/khugepaged/max_ptes_shared A higher value may increase memory footprint for some workloads. By default, at least half of pages has to be not shared. Signed-off-by: Kirill A. Shutemov --- Documentation/admin-guide/mm/transhuge.rst | 7 ++ mm/khugepaged.c | 52 ++++++++++++-- tools/testing/selftests/vm/khugepaged.c | 83 ++++++++++++++++++++++ 3 files changed, 138 insertions(+), 4 deletions(-) diff --git a/Documentation/admin-guide/mm/transhuge.rst b/Documentation/a= dmin-guide/mm/transhuge.rst index bd5714547cee..d16e4f2bb70f 100644 --- a/Documentation/admin-guide/mm/transhuge.rst +++ b/Documentation/admin-guide/mm/transhuge.rst @@ -220,6 +220,13 @@ memory. A lower value can prevent THPs from being collapsed, resulting fewer pages being collapsed into THPs, and lower memory access performance. =20 +``max_ptes_shared`` speicies how many pages can be shared across multipl= e +processes. Exeeding the number woul block the collapse:: + + /sys/kernel/mm/transparent_hugepage/khugepaged/max_ptes_shared + +A higher value may increase memory footprint for some workloads. + Boot parameter =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 diff --git a/mm/khugepaged.c b/mm/khugepaged.c index c8c2c463095c..8e728a602491 100644 --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -28,6 +28,8 @@ enum scan_result { SCAN_SUCCEED, SCAN_PMD_NULL, SCAN_EXCEED_NONE_PTE, + SCAN_EXCEED_SWAP_PTE, + SCAN_EXCEED_SHARED_PTE, SCAN_PTE_NON_PRESENT, SCAN_PAGE_RO, SCAN_LACK_REFERENCED_PAGE, @@ -46,7 +48,6 @@ enum scan_result { SCAN_DEL_PAGE_LRU, SCAN_ALLOC_HUGE_PAGE_FAIL, SCAN_CGROUP_CHARGE_FAIL, - SCAN_EXCEED_SWAP_PTE, SCAN_TRUNCATED, SCAN_PAGE_HAS_PRIVATE, }; @@ -71,6 +72,7 @@ static DECLARE_WAIT_QUEUE_HEAD(khugepaged_wait); */ static unsigned int khugepaged_max_ptes_none __read_mostly; static unsigned int khugepaged_max_ptes_swap __read_mostly; +static unsigned int khugepaged_max_ptes_shared __read_mostly; =20 #define MM_SLOTS_HASH_BITS 10 static __read_mostly DEFINE_HASHTABLE(mm_slots_hash, MM_SLOTS_HASH_BITS)= ; @@ -290,15 +292,43 @@ static struct kobj_attribute khugepaged_max_ptes_sw= ap_attr =3D __ATTR(max_ptes_swap, 0644, khugepaged_max_ptes_swap_show, khugepaged_max_ptes_swap_store); =20 +static ssize_t khugepaged_max_ptes_shared_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + return sprintf(buf, "%u\n", khugepaged_max_ptes_shared); +} + +static ssize_t khugepaged_max_ptes_shared_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + int err; + unsigned long max_ptes_shared; + + err =3D kstrtoul(buf, 10, &max_ptes_shared); + if (err || max_ptes_shared > HPAGE_PMD_NR-1) + return -EINVAL; + + khugepaged_max_ptes_shared =3D max_ptes_shared; + + return count; +} + +static struct kobj_attribute khugepaged_max_ptes_shared_attr =3D + __ATTR(max_ptes_shared, 0644, khugepaged_max_ptes_shared_show, + khugepaged_max_ptes_shared_store); + static struct attribute *khugepaged_attr[] =3D { &khugepaged_defrag_attr.attr, &khugepaged_max_ptes_none_attr.attr, + &khugepaged_max_ptes_swap_attr.attr, + &khugepaged_max_ptes_shared_attr.attr, &pages_to_scan_attr.attr, &pages_collapsed_attr.attr, &full_scans_attr.attr, &scan_sleep_millisecs_attr.attr, &alloc_sleep_millisecs_attr.attr, - &khugepaged_max_ptes_swap_attr.attr, NULL, }; =20 @@ -360,6 +390,7 @@ int __init khugepaged_init(void) khugepaged_pages_to_scan =3D HPAGE_PMD_NR * 8; khugepaged_max_ptes_none =3D HPAGE_PMD_NR - 1; khugepaged_max_ptes_swap =3D HPAGE_PMD_NR / 8; + khugepaged_max_ptes_shared =3D HPAGE_PMD_NR / 2; =20 return 0; } @@ -546,7 +577,7 @@ static int __collapse_huge_page_isolate(struct vm_are= a_struct *vma, { struct page *page =3D NULL; pte_t *_pte; - int none_or_zero =3D 0, result =3D 0, referenced =3D 0; + int none_or_zero =3D 0, shared =3D 0, result =3D 0, referenced =3D 0; bool writable =3D false; LIST_HEAD(compound_pagelist); =20 @@ -575,6 +606,12 @@ static int __collapse_huge_page_isolate(struct vm_ar= ea_struct *vma, =20 VM_BUG_ON_PAGE(!PageAnon(page), page); =20 + if (page_mapcount(page) > 1 && + ++shared > khugepaged_max_ptes_shared) { + result =3D SCAN_EXCEED_SHARED_PTE; + goto out; + } + if (PageCompound(page)) { struct page *p; page =3D compound_head(page); @@ -1160,7 +1197,8 @@ static int khugepaged_scan_pmd(struct mm_struct *mm= , { pmd_t *pmd; pte_t *pte, *_pte; - int ret =3D 0, none_or_zero =3D 0, result =3D 0, referenced =3D 0; + int ret =3D 0, result =3D 0, referenced =3D 0; + int none_or_zero =3D 0, shared =3D 0; struct page *page =3D NULL; unsigned long _address; spinlock_t *ptl; @@ -1210,6 +1248,12 @@ static int khugepaged_scan_pmd(struct mm_struct *m= m, goto out_unmap; } =20 + if (page_mapcount(page) > 1 && + ++shared > khugepaged_max_ptes_shared) { + result =3D SCAN_EXCEED_SHARED_PTE; + goto out_unmap; + } + page =3D compound_head(page); =20 /* diff --git a/tools/testing/selftests/vm/khugepaged.c b/tools/testing/self= tests/vm/khugepaged.c index 193bde6a1534..3a98d5b2d6d8 100644 --- a/tools/testing/selftests/vm/khugepaged.c +++ b/tools/testing/selftests/vm/khugepaged.c @@ -77,6 +77,7 @@ struct khugepaged_settings { unsigned int scan_sleep_millisecs; unsigned int max_ptes_none; unsigned int max_ptes_swap; + unsigned int max_ptes_shared; unsigned long pages_to_scan; }; =20 @@ -276,6 +277,7 @@ static void write_settings(struct settings *settings) khugepaged->scan_sleep_millisecs); write_num("khugepaged/max_ptes_none", khugepaged->max_ptes_none); write_num("khugepaged/max_ptes_swap", khugepaged->max_ptes_swap); + write_num("khugepaged/max_ptes_shared", khugepaged->max_ptes_shared); write_num("khugepaged/pages_to_scan", khugepaged->pages_to_scan); } =20 @@ -312,6 +314,7 @@ static void save_settings(void) read_num("khugepaged/scan_sleep_millisecs"), .max_ptes_none =3D read_num("khugepaged/max_ptes_none"), .max_ptes_swap =3D read_num("khugepaged/max_ptes_swap"), + .max_ptes_shared =3D read_num("khugepaged/max_ptes_shared"), .pages_to_scan =3D read_num("khugepaged/pages_to_scan"), }; success("OK"); @@ -786,12 +789,90 @@ static void collapse_fork_compound(void) fail("Fail"); fill_memory(p, 0, page_size); =20 + write_num("khugepaged/max_ptes_shared", hpage_pmd_nr - 1); if (wait_for_scan("Collapse PTE table full of compound pages in child"= , p)) fail("Timeout"); else if (check_huge(p)) success("OK"); else fail("Fail"); + write_num("khugepaged/max_ptes_shared", + default_settings.khugepaged.max_ptes_shared); + + validate_memory(p, 0, hpage_pmd_size); + munmap(p, hpage_pmd_size); + exit(exit_status); + } + + wait(&wstatus); + exit_status +=3D WEXITSTATUS(wstatus); + + printf("Check if parent still has huge page..."); + if (check_huge(p)) + success("OK"); + else + fail("Fail"); + validate_memory(p, 0, hpage_pmd_size); + munmap(p, hpage_pmd_size); +} + +static void collapse_max_ptes_shared() +{ + int max_ptes_shared =3D read_num("khugepaged/max_ptes_shared"); + int wstatus; + void *p; + + p =3D alloc_mapping(); + + printf("Allocate huge page..."); + madvise(p, hpage_pmd_size, MADV_HUGEPAGE); + fill_memory(p, 0, hpage_pmd_size); + if (check_huge(p)) + success("OK"); + else + fail("Fail"); + + printf("Share huge page over fork()..."); + if (!fork()) { + /* Do not touch settings on child exit */ + skip_settings_restore =3D true; + exit_status =3D 0; + + if (check_huge(p)) + success("OK"); + else + fail("Fail"); + + printf("Trigger CoW in %d of %d...", + hpage_pmd_nr - max_ptes_shared - 1, hpage_pmd_nr); + fill_memory(p, 0, (hpage_pmd_nr - max_ptes_shared - 1) * page_size); + if (!check_huge(p)) + success("OK"); + else + fail("Fail"); + + if (wait_for_scan("Do not collapse with max_ptes_shared exeeded", p)) + fail("Timeout"); + else if (!check_huge(p)) + success("OK"); + else + fail("Fail"); + + printf("Trigger CoW in %d of %d...", + hpage_pmd_nr - max_ptes_shared, hpage_pmd_nr); + fill_memory(p, 0, (hpage_pmd_nr - max_ptes_shared) * page_size); + if (!check_huge(p)) + success("OK"); + else + fail("Fail"); + + + if (wait_for_scan("Collapse with max_ptes_shared PTEs shared", p)) + fail("Timeout"); + else if (check_huge(p)) + success("OK"); + else + fail("Fail"); =20 validate_memory(p, 0, hpage_pmd_size); munmap(p, hpage_pmd_size); @@ -820,6 +901,7 @@ int main(void) =20 default_settings.khugepaged.max_ptes_none =3D hpage_pmd_nr - 1; default_settings.khugepaged.max_ptes_swap =3D hpage_pmd_nr / 8; + default_settings.khugepaged.max_ptes_shared =3D hpage_pmd_nr / 2; default_settings.khugepaged.pages_to_scan =3D hpage_pmd_nr * 8; =20 save_settings(); @@ -836,6 +918,7 @@ int main(void) collapse_full_of_compound(); collapse_fork(); collapse_fork_compound(); + collapse_max_ptes_shared(); =20 restore_settings(0); } --=20 2.26.0