From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail202.messagelabs.com (mail202.messagelabs.com [216.82.254.227]) by kanga.kvack.org (Postfix) with SMTP id C23556B01DB for ; Mon, 15 Mar 2010 10:33:03 -0400 (EDT) Received: by pxi34 with SMTP id 34so1731382pxi.22 for ; Mon, 15 Mar 2010 07:33:02 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <20100315142124.GL18274@csn.ul.ie> References: <1268412087-13536-1-git-send-email-mel@csn.ul.ie> <1268412087-13536-3-git-send-email-mel@csn.ul.ie> <28c262361003141728g4aa40901hb040144c5a4aeeed@mail.gmail.com> <20100315143420.6ec3bdf9.kamezawa.hiroyu@jp.fujitsu.com> <20100315112829.GI18274@csn.ul.ie> <1268657329.1889.4.camel@barrios-desktop> <20100315142124.GL18274@csn.ul.ie> Date: Mon, 15 Mar 2010 23:33:01 +0900 Message-ID: <28c262361003150733j4d0c708eyfc525e1e3c5c1f7d@mail.gmail.com> Subject: Re: [PATCH 02/11] mm,migration: Do not try to migrate unmapped anonymous pages From: Minchan Kim Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Sender: owner-linux-mm@kvack.org To: Mel Gorman Cc: KAMEZAWA Hiroyuki , Andrew Morton , Andrea Arcangeli , Christoph Lameter , Adam Litke , Avi Kivity , David Rientjes , KOSAKI Motohiro , Rik van Riel , linux-kernel@vger.kernel.org, linux-mm@kvack.org List-ID: On Mon, Mar 15, 2010 at 11:21 PM, Mel Gorman wrote: > On Mon, Mar 15, 2010 at 09:48:49PM +0900, Minchan Kim wrote: >> On Mon, 2010-03-15 at 11:28 +0000, Mel Gorman wrote: >> > The use after free looks like >> > >> > 1. page_mapcount(page) was zero so anon_vma was no longer reliable >> > 2. rcu lock taken but the anon_vma at this point can already be garbag= e because the >> > =C2=A0 =C2=A0process exited >> > 3. call try_to_unmap, looks up tha anon_vma and locks it. This causes = problems >> > >> > I thought the race would be closed but there is still a very tiny wind= ow there all >> > right. The following alternative should close it. What do you think? >> > >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 if (PageAnon(page)) { >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 rcu_read_lock(); >> > >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 /* >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0* If the= page has no mappings any more, just bail. An >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0* unmapp= ed anon page is likely to be freed soon but worse, >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0* it's p= ossible its anon_vma disappeared between when >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0* the pa= ge was isolated and when we reached here while >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0* the RC= U lock was not held >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0*/ >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 if (!page_mapc= ount(page)) { >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = rcu_read_unlock(); >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 goto uncharge; >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 } >> > >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 rcu_locked =3D= 1; >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 anon_vma =3D p= age_anon_vma(page); >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 atomic_inc(&an= on_vma->external_refcount); >> > =C2=A0 =C2=A0 =C2=A0 =C2=A0 } >> > >> > The rcu_unlock label is not used here because the reference counts wer= e not taken in >> > the case where page_mapcount =3D=3D 0. >> > >> >> Please, repost above code with your use-after-free scenario comment. >> > > This will be the replacement patch so. > > =3D=3D=3D=3D CUT HERE =3D=3D=3D=3D > mm,migration: Do not try to migrate unmapped anonymous pages > > rmap_walk_anon() was triggering errors in memory compaction that look lik= e > use-after-free errors. The problem is that between the page being isolate= d > from the LRU and rcu_read_lock() being taken, the mapcount of the page > dropped to 0 and the anon_vma gets freed. This can happen during memory > compaction if pages being migrated belong to a process that exits before > migration completes. Hence, the use-after-free race looks like > > =C2=A01. Page isolated for migration > =C2=A02. Process exits > =C2=A03. page_mapcount(page) drops to zero so anon_vma was no longer reli= able > =C2=A04. unmap_and_move() takes the rcu_lock but the anon_vma is already = garbage > =C2=A04. call try_to_unmap, looks up tha anon_vma and "locks" it but the = lock > =C2=A0 =C2=A0is garbage. > > This patch checks the mapcount after the rcu lock is taken. If the > mapcount is zero, the anon_vma is assumed to be freed and no further > action is taken. > > Signed-off-by: Mel Gorman > Acked-by: Rik van Riel Reviewed-by: Minchan Kim --=20 Kind regards, Minchan Kim -- 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