From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-vn0-f46.google.com (mail-vn0-f46.google.com [209.85.216.46]) by kanga.kvack.org (Postfix) with ESMTP id 0BE886B0038 for ; Wed, 15 Apr 2015 03:27:43 -0400 (EDT) Received: by vnbf190 with SMTP id f190so11951403vnb.1 for ; Wed, 15 Apr 2015 00:27:42 -0700 (PDT) Received: from tyo202.gate.nec.co.jp (TYO202.gate.nec.co.jp. [210.143.35.52]) by mx.google.com with ESMTPS id y186si2304533oia.39.2015.04.15.00.27.41 for (version=TLSv1 cipher=RC4-SHA bits=128/128); Wed, 15 Apr 2015 00:27:42 -0700 (PDT) From: Naoya Horiguchi Subject: [PATCH] mm/memory-failure: call shake_page() when error hits thp tail page Date: Wed, 15 Apr 2015 07:25:46 +0000 Message-ID: <1429082714-26115-1-git-send-email-n-horiguchi@ah.jp.nec.com> Content-Language: ja-JP Content-Type: text/plain; charset="iso-2022-jp" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Sender: owner-linux-mm@kvack.org List-ID: To: Andrew Morton Cc: Dean Nelson , Andi Kleen , Andrea Arcangeli , Hidetoshi Seto , Jin Dongming , "linux-mm@kvack.org" , "linux-kernel@vger.kernel.org" Currently memory_failure() calls shake_page() to sweep pages out from pcpli= sts only when the victim page is 4kB LRU page or thp head page. But we should d= o this for a thp tail page too. Consider that a memory error hits a thp tail page whose head page is on a pcplist when memory_failure() runs. Then, the current kernel skips shake_pa= ges() part, so hwpoison_user_mappings() returns without calling split_huge_page()= nor try_to_unmap() because PageLRU of the thp head is still cleared due to the = skip of shake_page(). As a result, me_huge_page() runs for the thp, which is a broken behavior. This patch fixes this problem by calling shake_page() for thp tail case. Fixes: 385de35722c9 ("thp: allow a hwpoisoned head page to be put back to L= RU") Signed-off-by: Naoya Horiguchi Cc: stable@vger.kernel.org # v3.4+ --- mm/memory-failure.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git v4.0.orig/mm/memory-failure.c v4.0/mm/memory-failure.c index d487f8dc6d39..2cc1d578144b 100644 --- v4.0.orig/mm/memory-failure.c +++ v4.0/mm/memory-failure.c @@ -1141,10 +1141,10 @@ int memory_failure(unsigned long pfn, int trapno, i= nt flags) * The check (unnecessarily) ignores LRU pages being isolated and * walked by the page reclaim code, however that's not a big loss. */ - if (!PageHuge(p) && !PageTransTail(p)) { - if (!PageLRU(p)) - shake_page(p, 0); - if (!PageLRU(p)) { + if (!PageHuge(p)) { + if (!PageLRU(hpage)) + shake_page(hpage, 0); + if (!PageLRU(hpage)) { /* * shake_page could have turned it free. */ --=20 2.1.0 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org