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=-8.5 required=3.0 tests=INCLUDES_PATCH, MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_SANE_1 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 86244FA372A for ; Thu, 17 Oct 2019 10:01:11 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 419E52064B for ; Thu, 17 Oct 2019 10:01:11 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 419E52064B Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=kernel.org Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id CF6E48E0005; Thu, 17 Oct 2019 06:01:10 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id CA7058E0003; Thu, 17 Oct 2019 06:01:10 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id BBCD38E0005; Thu, 17 Oct 2019 06:01:10 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0180.hostedemail.com [216.40.44.180]) by kanga.kvack.org (Postfix) with ESMTP id 9B6618E0003 for ; Thu, 17 Oct 2019 06:01:10 -0400 (EDT) Received: from smtpin19.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with SMTP id 3592F180AE816 for ; Thu, 17 Oct 2019 10:01:10 +0000 (UTC) X-FDA: 76052833500.19.fang65_44525b3437a51 X-HE-Tag: fang65_44525b3437a51 X-Filterd-Recvd-Size: 6281 Received: from mx1.suse.de (mx2.suse.de [195.135.220.15]) by imf23.hostedemail.com (Postfix) with ESMTP for ; Thu, 17 Oct 2019 10:01:09 +0000 (UTC) X-Virus-Scanned: by amavisd-new at test-mx.suse.de Received: from relay2.suse.de (unknown [195.135.220.254]) by mx1.suse.de (Postfix) with ESMTP id D10C7B77F; Thu, 17 Oct 2019 10:01:07 +0000 (UTC) Date: Thu, 17 Oct 2019 12:01:06 +0200 From: Michal Hocko To: Qian Cai , Naoya Horiguchi Cc: "linux-kernel@vger.kernel.org" , "linux-mm@kvack.org" , David Hildenbrand , Mike Kravetz Subject: Re: memory offline infinite loop after soft offline Message-ID: <20191017100106.GF24485@dhcp22.suse.cz> References: <1570829564.5937.36.camel@lca.pw> <20191014083914.GA317@dhcp22.suse.cz> <20191017093410.GA19973@hori.linux.bs1.fc.nec.co.jp> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <20191017093410.GA19973@hori.linux.bs1.fc.nec.co.jp> User-Agent: Mutt/1.10.1 (2018-07-13) Content-Transfer-Encoding: quoted-printable X-Bogosity: Ham, tests=bogofilter, spamicity=0.000270, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On Thu 17-10-19 09:34:10, Naoya Horiguchi wrote: > On Mon, Oct 14, 2019 at 10:39:14AM +0200, Michal Hocko wrote: [...] > > diff --git a/mm/page_isolation.c b/mm/page_isolation.c > > index 89c19c0feadb..5fb3fee16fde 100644 > > --- a/mm/page_isolation.c > > +++ b/mm/page_isolation.c > > @@ -274,7 +274,7 @@ __test_page_isolated_in_pageblock(unsigned long p= fn, unsigned long end_pfn, > > * simple way to verify that as VM_BUG_ON(), though. > > */ > > pfn +=3D 1 << page_order(page); > > - else if (skip_hwpoisoned_pages && PageHWPoison(page)) > > + else if (skip_hwpoisoned_pages && PageHWPoison(compound_head(page)= )) > > /* A HWPoisoned page cannot be also PageBuddy */ > > pfn++; > > else >=20 > This fix looks good to me. The original code only addresses hwpoisoned = 4kB-page, > we seem to have this issue since the following commit, Thanks a lot for double checking Naoya! =20 > commit b023f46813cde6e3b8a8c24f432ff9c1fd8e9a64 > Author: Wen Congyang > Date: Tue Dec 11 16:00:45 2012 -0800 > =20 > memory-hotplug: skip HWPoisoned page when offlining pages >=20 > and extension of LTP coverage finally discovered this. Qian, could you give the patch some testing? --- >From 441a9515dcdb29bb0ca39ff995632907d959032f Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Thu, 17 Oct 2019 11:49:15 +0200 Subject: [PATCH] hugetlb, memory_hotplug: fix HWPoisoned tail pages prope= rly MIME-Version: 1.0 Content-Type: text/plain; charset=3DUTF-8 Content-Transfer-Encoding: 8bit Qian Cai has noticed that hwpoisoned hugetlb pages prevent memory offlining from making a forward progress. He has nailed down the issue to be __test_page_isolated_in_pageblock always returning EBUSY because of soft offlined page: [=A0=A0101.665160][ T8885] pfn =3D 77501, end_pfn =3D 78000 [=A0=A0101.665245][ T8885] page:c00c000001dd4040 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 [=A0=A0101.665329][ T8885] flags: 0x3fffc000000000() [=A0=A0101.665391][ T8885] raw: 003fffc000000000 0000000000000000 fffffff= f01dd0500 0000000000000000 [=A0=A0101.665498][ T8885] raw: 0000000000000000 0000000000000000 0000000= 0ffffffff 0000000000000000 [=A0=A0101.665588][ T8885] page dumped because: soft_offline [=A0=A0101.665639][ T8885] page_owner tracks the page as freed [=A0=A0101.665697][ T8885] page last allocated via order 5, migratetype M= ovable, gfp_mask 0x346cca(GFP_HIGHUSER_MOVABLE|__GFP_NOWARN|__GFP_RETRY_MAYFAIL|__GFP_COMP= |__GFP_ THISNODE) [=A0=A0101.665924][ T8885]=A0=A0prep_new_page+0x3c0/0x440 [=A0=A0101.665962][ T8885]=A0=A0get_page_from_freelist+0x2568/0x2bb0 [=A0=A0101.666059][ T8885]=A0=A0__alloc_pages_nodemask+0x1b4/0x670 [=A0=A0101.666115][ T8885]=A0=A0alloc_fresh_huge_page+0x244/0x6e0 [=A0=A0101.666183][ T8885]=A0=A0alloc_migrate_huge_page+0x30/0x70 [=A0=A0101.666254][ T8885]=A0=A0alloc_new_node_page+0xc4/0x380 [=A0=A0101.666325][ T8885]=A0=A0migrate_pages+0x3b4/0x19e0 [=A0=A0101.666375][ T8885]=A0=A0do_move_pages_to_node.isra.29.part.30+0x4= 4/0xa0 [=A0=A0101.666464][ T8885]=A0=A0kernel_move_pages+0x498/0xfc0 [=A0=A0101.666520][ T8885]=A0=A0sys_move_pages+0x28/0x40 [=A0=A0101.666643][ T8885]=A0=A0system_call+0x5c/0x68 [=A0=A0101.666665][ T8885] page last free stack trace: [=A0=A0101.666704][ T8885]=A0=A0__free_pages_ok+0xa4c/0xd40 [=A0=A0101.666773][ T8885]=A0=A0update_and_free_page+0x2dc/0x5b0 [=A0=A0101.666821][ T8885]=A0=A0free_huge_page+0x2dc/0x740 [=A0=A0101.666875][ T8885]=A0=A0__put_compound_page+0x64/0xc0 [=A0=A0101.666926][ T8885]=A0=A0putback_active_hugepage+0x228/0x390 [=A0=A0101.666990][ T8885]=A0=A0migrate_pages+0xa78/0x19e0 [=A0=A0101.667048][ T8885]=A0=A0soft_offline_page+0x314/0x1050 [=A0=A0101.667117][ T8885]=A0=A0sys_madvise+0x1068/0x1080 [=A0=A0101.667185][ T8885]=A0=A0system_call+0x5c/0x68 The reason is that __test_page_isolated_in_pageblock doesn't recognize hugetlb tail pages as the HWPoison bit is not transferred from the head page. Pfn walker then doesn't recognize those pages and so EBUSY is returned up the call chain. The proper fix would be to handle HWPoison throughout the huge page but considering there is a WIP to rework that code considerably let's go with a simple and easily backportable workaround and simply check the the head of a compound page for the HWPoison flag. Reported-and-analyzed-by: Qian Cai Fixes: b023f46813cd ("memory-hotplug: skip HWPoisoned page when offlining= pages") Cc: stable Signed-off-by: Michal Hocko --- mm/page_isolation.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/page_isolation.c b/mm/page_isolation.c index 89c19c0feadb..5fb3fee16fde 100644 --- a/mm/page_isolation.c +++ b/mm/page_isolation.c @@ -274,7 +274,7 @@ __test_page_isolated_in_pageblock(unsigned long pfn, = unsigned long end_pfn, * simple way to verify that as VM_BUG_ON(), though. */ pfn +=3D 1 << page_order(page); - else if (skip_hwpoisoned_pages && PageHWPoison(page)) + else if (skip_hwpoisoned_pages && PageHWPoison(compound_head(page))) /* A HWPoisoned page cannot be also PageBuddy */ pfn++; else --=20 2.20.1 --=20 Michal Hocko SUSE Labs