Received-SPF: pass (google.com: domain of linux-kernel+bounces-157267-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
From: Zi Yan <ziy@nvidia.com>
To: David Hildenbrand <david@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>, Lance Yang <ioworker0@gmail.com>,
 akpm@linux-foundation.org, maskray@google.com, ryan.roberts@arm.com,
 21cnbao@gmail.com, mhocko@suse.com, fengwei.yin@intel.com,
 zokeefe@google.com, shy828301@gmail.com, xiehuan09@gmail.com,
 wangkefeng.wang@huawei.com, songmuchun@bytedance.com, peterx@redhat.com,
 minchan@kernel.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 1/1] mm/vmscan: avoid split PMD-mapped THP during
 shrink_folio_list()
Date: Wed, 24 Apr 2024 11:57:37 -0400
Message-ID: <333181FE-C2E6-4859-AF6A-0667A1F42037@nvidia.com>
In-Reply-To: <1a0ca018-8ad3-42b0-b98a-8e6b6862fc7a@redhat.com>
References: <20240422055213.60231-1-ioworker0@gmail.com>
 <ZiiHSwG_bnLJbwfb@casper.infradead.org>
 <1a0ca018-8ad3-42b0-b98a-8e6b6862fc7a@redhat.com>
Content-Type: multipart/signed;
 boundary="=_MailMate_A1F0F6A5-B89E-4673-BBC2-D899DA1157F2_=";
 micalg=pgp-sha512; protocol="application/pgp-signature"
Precedence: bulk
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0:
	=?us-ascii?Q?+aLBVLpcXuwQ0Eyec3KqcykslQdcZ1ivYLL05xaz3/+2lqZsZK6jiKvk2TvP?=
 =?us-ascii?Q?Q6xHWT50aDXiJAqSrvgInWmhAh8ByWmbho+s+e67b4IuYLqNM6D8JUlr26Mm?=
 =?us-ascii?Q?+kujnaB8FbRnBHsz9uKqULAajwfTTht2uKywD98NpWr8U+DJ23N0djfXpEhK?=
 =?us-ascii?Q?1sDtbcGtnyZv9U0XbxJLofUOTKCd8s7otfea5+GTcdImyAXmzX2LvJMRH5kh?=
 =?us-ascii?Q?SJJId9H9iBoPrcXrhNk3ZU7Dk0tH7lxWvtOj8r8Zu5eZ+eiDUNWaC1eHZsJm?=
 =?us-ascii?Q?mpJvslTJbc07GZ+fGoWzQu+ojPLdD+vlH/JFH3w2UuJIQpr5IKGwYdA84ouA?=
 =?us-ascii?Q?YwW9sUdy+BEGcezXYVPV9sq3SakW46F47FCocJEZ9p0SKifbkrZcf/7QLV78?=
 =?us-ascii?Q?B2hsO01pgS+xDE8ehRsln1t4xzuDBNM5SI+kDP4LKXgrPKB0l2YzGUVamQJH?=
 =?us-ascii?Q?w4JhPyrfl653sLCz5+cN5yLm2KHKs22qBKo1TdItKmySjJI7E1R13q+3QEib?=
 =?us-ascii?Q?KOW1B/8kG4BqjYluyRUlVsotnjmcR9OdcY4EYY4l9dlR8HtTOmKnhFASLqkH?=
 =?us-ascii?Q?1KaBDAOgmJmxm6UZtyln991OXdLzXcfk3isZ1NOubQNefCifWqldlbiPVLZ4?=
 =?us-ascii?Q?+GkSFZQsxxM4ZZ1gyauYla6aI0U9UzpSIObWz+ih4TvgWHn+EwuLZmZQy1Kg?=
 =?us-ascii?Q?jGP+Ez5ojxKOIkrLiXRBI276ZvEoe7+ZIP5mUE/BmytslE5amqnGUGo3HlUu?=
 =?us-ascii?Q?AFlHCr3aBMV43Ow7X3N+63GXQOoJbqe6ktPJVgspIjCw9MFgfzvEW0ztlE9F?=
 =?us-ascii?Q?9W8yL18nE/4pY4GbZkreuZ1jeZ/806ZdpR6dkapjvPd+x0tp4sdxYMkqoWy7?=
 =?us-ascii?Q?dxVSluvjlEOUrvR2wnaMTdEKTp/KMfrpy6YBGaSIy5muLu5BVWXaEqI0y0NN?=
 =?us-ascii?Q?EbmnH9npd80s2vBVPa9M54flxsZBLCrTJ/T2ETAYx+5Ru0z2wGCOfjx6E1XT?=
 =?us-ascii?Q?7cg8Tm8uAAUDIXw3mQiBpgbPYgsPNpHQojz/rFgXrddzbD6pZRAtxa2C537i?=
 =?us-ascii?Q?RPJYmySiNjfqHJkhIXErHjpQmtnTxLdOmyC2G1818iynacVuAuPrhjiENl4a?=
 =?us-ascii?Q?n3YaUuSQqffdK/8/x+pWYguRZIB/eH0QIYveJ2kMhViM1QHIge/5GW7LWG4I?=
 =?us-ascii?Q?+usCK04nAkoP3zzc4U86M3BaJI7i7QyC8Z17iLKMrJSX9YXQKo16eQXHIcv7?=
 =?us-ascii?Q?CfCZBKU8/bkpF9tIEnot0u8udSSLJYXqekca4fNJODwFzy1uMWrwjCJ0EFBQ?=
 =?us-ascii?Q?P37Vj58wlTOvq8BxUntyWXB6ufkSVASTuoZrS2MCm1A1CCvJJnWXnPh5t75E?=
 =?us-ascii?Q?126Wt1UEB+/zd4kpOTXokdTiU4lkC4js/AtKqdYd4j9TB0QWi4oD7CxWzbf6?=
 =?us-ascii?Q?Uymv9cs562/ur9Wc0jt6u8Ebdauc8yhCTkWjPU9VHcV2TcTHosGoJaNF0AX+?=
 =?us-ascii?Q?i5+Hgm2JIZgl+iDOt8agSz2PGMQBw/t3fOdamKTFQTMod2P4HxM93Ihu1bPH?=
 =?us-ascii?Q?mYma41QKALyPqt3ZV3E=3D?=
X-OriginatorOrg: Nvidia.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 9762596a-15f9-4ee7-074e-08dc647744cc
X-MS-Exchange-CrossTenant-AuthSource: DS7PR12MB5744.namprd12.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 24 Apr 2024 15:57:39.9154
 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 43083d15-7273-40c1-b7db-39efd9ccc17a
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: J0/E93HIP2E+yVgytJdiOq50XYsSlufiUohhUMQgnkXsY++pBn6dw8XJxs8wEY1B
X-MS-Exchange-Transport-CrossTenantHeadersStamped: MN2PR12MB4344

--=_MailMate_A1F0F6A5-B89E-4673-BBC2-D899DA1157F2_=
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On 24 Apr 2024, at 3:17, David Hildenbrand wrote:

> On 24.04.24 06:15, Matthew Wilcox wrote:
>> On Mon, Apr 22, 2024 at 01:52:13PM +0800, Lance Yang wrote:
>>> When the user no longer requires the pages, they would use
>>> madvise(MADV_FREE) to mark the pages as lazy free. IMO, they would no=
t
>>> typically rewrite to the given range.
>>>
>>> At present, PMD-mapped THPs that are marked as lazyfree during
>>> shrink_folio_list() are unconditionally split, which may be unnecessa=
ry.
>>> If the THP is clean, its PMD is also clean, and there are no unexpect=
ed
>>> references, then we can attempt to remove the PMD mapping from it. Th=
is
>>> change will improve the efficiency of memory reclamation in this case=
=2E
>>
>> Does this happen outside of benchmarks?  I'm really struggling to see
>> how we end up in this situation.  We have a clean THP without swap
>> backing, so it's full of zeroes, but for some reason we haven't used t=
he
>> shared huge zero page?  What is going on?
>
> It's not full of zeroes.
>
> User space called MADV_FREE on a PMD-mapped THP.
>
> During MADV_FREE, we mark the PTEs as clean, the folio as clean and sd =
"lazyfree" (no swap backend). If, during memory reclaim, we detect that (=
a) the folio is still clean (b) the PTEs are still clean and (c) there ar=
e no unexpected references (GUP), user space didn't re-write to that memo=
ry again, so we can just discard the memory "lazily".

It seems that try_to_unmap_one() does not support unmapping PMD-mapped fo=
lios.
Maybe adding that support instead of a special case handling?

--
Best Regards,
Yan, Zi

--=_MailMate_A1F0F6A5-B89E-4673-BBC2-D899DA1157F2_=
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename=signature.asc
Content-Type: application/pgp-signature; name=signature.asc

-----BEGIN PGP SIGNATURE-----
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=1eku
-----END PGP SIGNATURE-----

--=_MailMate_A1F0F6A5-B89E-4673-BBC2-D899DA1157F2_=--