From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752183AbcERHap (ORCPT ); Wed, 18 May 2016 03:30:45 -0400 Received: from mx2.suse.de ([195.135.220.15]:49644 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750870AbcERHan (ORCPT ); Wed, 18 May 2016 03:30:43 -0400 Subject: Re: [PATCH v1] mm: bad_page() checks bad_flags instead of page->flags for hwpoison page To: Naoya Horiguchi , Andrew Morton References: <1463470975-29972-1-git-send-email-n-horiguchi@ah.jp.nec.com> Cc: Mel Gorman , linux-mm@kvack.org, linux-kernel@vger.kernel.org, Naoya Horiguchi From: Vlastimil Babka Message-ID: <573C1A21.6030006@suse.cz> Date: Wed, 18 May 2016 09:30:41 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.7.2 MIME-Version: 1.0 In-Reply-To: <1463470975-29972-1-git-send-email-n-horiguchi@ah.jp.nec.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 05/17/2016 09:42 AM, Naoya Horiguchi wrote: > There's a race window between checking page->flags and unpoisoning, which > taints kernel with "BUG: Bad page state". That's overkill. It's safer to > use bad_flags to detect hwpoisoned page. > > Signed-off-by: Naoya Horiguchi > --- > mm/page_alloc.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git tmp/mm/page_alloc.c tmp_patched/mm/page_alloc.c > index 5b269bc..4e0fa37 100644 > --- tmp/mm/page_alloc.c > +++ tmp_patched/mm/page_alloc.c > @@ -522,8 +522,8 @@ static void bad_page(struct page *page, const char *reason, > static unsigned long nr_shown; > static unsigned long nr_unshown; > > - /* Don't complain about poisoned pages */ > - if (PageHWPoison(page)) { > + /* Don't complain about hwpoisoned pages */ > + if (bad_flags == __PG_HWPOISON) { This will wrongly return prematurely on !CONFIG_MEMORY_FAILURE where __PG_HWPOISON == 0 and bad_page() called for other reasons than bad flags? > page_mapcount_reset(page); /* remove PageBuddy */ > return; > } >