Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Song Liu <songliubraving@fb.com>
To:     Andrew Morton <akpm@linux-foundation.org>
CC:     open list <linux-kernel@vger.kernel.org>,
        "linux-mm@kvack.org" <linux-mm@kvack.org>,
        "matthew.wilcox@oracle.com" <matthew.wilcox@oracle.com>,
        Kernel Team <Kernel-team@fb.com>,
        "william.kucharski@oracle.com" <william.kucharski@oracle.com>,
        "kirill.shutemov@linux.intel.com" <kirill.shutemov@linux.intel.com>,
        "Johannes Weiner" <hannes@cmpxchg.org>,
        Hugh Dickins <hughd@google.com>
Subject: Re: [PATCH v3] mm,thp: recheck each page before collapsing file THP
Thread-Topic: [PATCH v3] mm,thp: recheck each page before collapsing file THP
Thread-Index: AQHVhd5xLYpzi8cJoUynxajagY1vhadhKiEAgABE9QA=
Date:   Sat, 19 Oct 2019 05:24:00 +0000
Message-ID: <9DC29F5B-1DF5-408F-BEDF-FD1FBAAB1361@fb.com>
References: <20191018180345.4188310-1-songliubraving@fb.com>
 <20191018181712.91dd9e9f9941642300e1b8d9@linux-foundation.org>
In-Reply-To: <20191018181712.91dd9e9f9941642300e1b8d9@linux-foundation.org>
Accept-Language: en-US
Content-Language: en-US
received-spf: None (protection.outlook.com: fb.com does not designate
 permitted sender hosts)
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1BC2B7C527BB184C8A960CE6306B5F95@namprd15.prod.outlook.com>
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-Network-Message-Id: 31a989e3-2e79-4b08-c8ab-08d754548cb6
X-MS-Exchange-CrossTenant-originalarrivaltime: 19 Oct 2019 05:24:00.9230
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 8ae927fe-1255-47a7-a2af-5f3a069daaa2
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: kd3nGSiLxO1Mzh9QqU1L5wz6T4d2Vnpt99znTkZp01hyThkrk0ARQWz66P9XTYKAWn/oqtQP46owLhWZBRp5zQ==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: MWHPR15MB1584
X-OriginatorOrg: fb.com
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.95,1.0.8
 definitions=2019-10-19_01:2019-10-18,2019-10-19 signatures=0
X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 bulkscore=0
 mlxlogscore=999 priorityscore=1501 lowpriorityscore=0 malwarescore=0
 suspectscore=0 adultscore=0 phishscore=0 clxscore=1015 mlxscore=0
 spamscore=0 impostorscore=0 classifier=spam adjust=0 reason=mlx
 scancount=1 engine=8.12.0-1908290000 definitions=main-1910190051
X-FB-Internal: deliver
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org


> On Oct 18, 2019, at 6:17 PM, Andrew Morton <akpm@linux-foundation.org> wr=
ote:
>=20
> On Fri, 18 Oct 2019 11:03:45 -0700 Song Liu <songliubraving@fb.com> wrote=
:
>=20
>> In collapse_file(), after locking the page, it is necessary to recheck
>> that the page is up-to-date. Add PageUptodate() check for both shmem THP
>> and file THP.
>>=20
>> Current khugepaged should not try to collapse dirty file THP, because it
>> is limited to read only text. Add a PageDirty check and warning for file
>> THP. This is added after page_mapping() check, because if the page is
>> truncated, it might be dirty.
>=20
> When fixing a bug, please always fully describe the end-user visible
> effects of that bug.  This is vital information for people who are
> considering the fix for backporting.

The end user effect is, corruption in page cache. While grouping pages=20
into a huge page, the page cache mistakenly includes some pages that=20
are not uptodate, and considers the huge page is uptodate.=20

I am attaching updated commit log to the end.=20

>=20
> I'm suspecting that you've found a race condition which can trigger a
> VM_BUG_ON_PAGE(), which is rather serious.  But that was just a wild
> guess.  Please don't make us wildly guess :(
>=20
> The old code looked rather alarming:
>=20
> 			} else if (!PageUptodate(page)) {
> 				xas_unlock_irq(&xas);
> 				wait_on_page_locked(page);
> 				if (!trylock_page(page)) {
> 					result =3D SCAN_PAGE_LOCK;
> 					goto xa_unlocked;
> 				}
> 				get_page(page);
>=20
> We don't have a ref on that page.  After we've released the xarray lock
> we have no business playing with *page at all, correct?

Yeah, this piece is not just redundant, but also buggy. I am also=20
including some information about it.=20

Updated commit log:

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D 8< =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

In collapse_file(), for !is_shmem case, current check cannot guarantee=20
the locked page is up-to-date. Specifically, xas_unlock_irq() should not
be called before lock_page() and get_page(); and it is necessary to=20
recheck PageUptodate() after locking the page.=20

With this bug and CONFIG_READ_ONLY_THP_FOR_FS=3Dy, madvise(HUGE)'ed .text=20
may contain corrupted data. This is because khugepaged mistakenly=20
collapses some not up-to-date sub pages into a huge page, and assumes the=20
huge page is up-to-date. This will NOT corrupt data in the disk, because=20
the page is read-only and never written back. Fix this by properly=20
checking PageUptodate() after locking the page. This check replaces=20
"VM_BUG_ON_PAGE(!PageUptodate(page), page);".=20

Also, move PageDirty() check after locking the page. Current khugepaged=20
should not try to collapse dirty file THP, because it is limited to=20
read-only .text. Add a warning with the PageDirty() check as it should=20
not happen. This warning is added after page_mapping() check, because=20
if the page is truncated, it might be dirty.

Fixes: 99cb0dbd47a1 ("mm,thp: add read-only THP support for (non-shmem) FS"=
)
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: William Kucharski <william.kucharski@oracle.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Song Liu <songliubraving@fb.com>

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D 8< =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

Thanks,
Song