Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
References: <20230531044018.17893-1-mpenttil@redhat.com>
 <87wn0pt13j.fsf@yhuang6-desk2.ccr.corp.intel.com>
User-agent: mu4e 1.8.10; emacs 28.2
From:   Alistair Popple <apopple@nvidia.com>
To:     "Huang, Ying" <ying.huang@intel.com>
Cc:     mpenttil@redhat.com, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org, John Hubbard <jhubbard@nvidia.com>,
        Ralph Campbell <rcampbell@nvidia.com>
Subject: Re: [PATCH v2] mm/migrate_device: Try to handle swapcache pages
Date:   Thu, 01 Jun 2023 22:25:35 +1000
In-reply-to: <87wn0pt13j.fsf@yhuang6-desk2.ccr.corp.intel.com>
Message-ID: <87wn0npdnc.fsf@nvidia.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0: =?utf-8?B?andtV3BvNGxIS2hSV1BDbjNxTzRCVkR0NktJZ3I3bHdHanJ3V0dINXdBOS9M?=
 =?utf-8?B?MXdOWEN2bXNHKzNCTmt4ZU1laGhieTZNUmdlQkdDWi9DUlVFNjAyRjdQa0wz?=
 =?utf-8?B?RHNBV0tKRzZVVjFqMnlSQkR6YWVaRDc5c3NMV1hSbnZ6UmR6VE5RSTkrcE1F?=
 =?utf-8?B?ekxOTC9lNTdUd0JzWHMwdWx2dmZRNUJIY3FXNk5lZkdDTEpJWXRwclJRVHFa?=
 =?utf-8?B?ZU83RFRFOEhrcXB4MG1YUDZnVjJFc1Rwd2c3ZW9ENFU1QmdSSGZVR0IxbjJL?=
 =?utf-8?B?MnN0OU1ya3dHRFE1UTNseXRVME9sbVV2R1ZUL2F4akR2bEttZG96NzEzdml4?=
 =?utf-8?B?ZjBkVklOTXhsbXV3WXFGSjVaZ09kZnUwYWR5NE1UMUdQSFoxdnkwd3pmUmoy?=
 =?utf-8?B?UU1nVU9nMGY1VTZ1eGRDdzRUR0xxUlBKOE05NEt6RGN3TzBTcFpTYkwzWHJ0?=
 =?utf-8?B?NlFjamI0RHA2UElHbGJ2djZuT0RCbC9TTHk2Q0NwOUJISWN4MG1ObHpheE9I?=
 =?utf-8?B?STlpMmFtSURBUWlhUUZrcE9OU2NwblZjR2p1Yk80ejdVbURTbkM0WnhnZERK?=
 =?utf-8?B?UWpwaUVoa2MyWTJiVFBkTmVNOGp2Z2NDa1hCcEVJbm01emZVVmtnWU0zd0d6?=
 =?utf-8?B?Nm41OEIzK25welFjcW1YYmY4NjR5MXE2NTJNTndQSExvaVo3Z2FQM1VOUEps?=
 =?utf-8?B?cnNVWk9MYUZKYVFkdGlmcllrZitOcERJVTh5aHhpajVtUmpaQUpmbVUwbHBS?=
 =?utf-8?B?bnhYZWNWKytRWUMvQ2ZqWTFxUTk4TDMvRy9PdlAwSHNkSkczM2xvTkptb25G?=
 =?utf-8?B?ZGFKNHY2MDMzWHJEUWVwWWlwQ2dGSzFwMkxPVjJ6RkplMTY4aENJTFNPZ09i?=
 =?utf-8?B?Vk5rWU1wNnZJWStibnJRdnNEbStxR1RWVEhEa3ZzV3I1RFVPelB2enBsc3Vn?=
 =?utf-8?B?WE50VFhnUjk0ZTlYTHFOSmd5MzF5M0VmK1pPenBSMzZSQ0VXWnh0WityUHov?=
 =?utf-8?B?Mk01WmFqTHdQR0tSM2pFdmwxa244MmQyckhQRGdGazJEenEyWnMvSTdldHVx?=
 =?utf-8?B?RHR4SFFCaGprbU5Tb1ptSDJTdWNvcDc3WWpRNFN1OW80K0ZtZ3l2cWhoMWJj?=
 =?utf-8?B?THdIWk01K3d1ZnpaVzlYWDczbnp5K1JRUFdmWlY1RUFqWVJHSk41R3IzRCtO?=
 =?utf-8?B?bWx2UndMNXk0blBVMHk0VE5YQ0FSUG1kNWc0akErY1pCSWNWTEFodU1ZNVBQ?=
 =?utf-8?B?RVc1eEdVTjl3L3Q4RGMwcGd5ZkdqYWx4NFBFdUdiZDFHSnNOM29zVCtVcWV2?=
 =?utf-8?B?ekVIYzlWb0Q3ZHpsNmYrM0hmUjdVU25GSURXMGttM05oRnZLS1YxWk1rU3pw?=
 =?utf-8?B?U1NieWRBRTdreXIwMmRidzZIR2pLdDgrejhyYzBVRU1mZC83bW5pL1F3Q3FV?=
 =?utf-8?B?UWFlTVFNUWxWV2RFV3djY3loakFHT1hJVzFPcWtjc0ZGYk96Rkl3Z3dMRHNv?=
 =?utf-8?B?TmtnbGltOFJuaUlHd3lRZlB3QUlnK1RjZ29DdGZmc3pRdnlGVWZBNGUvUUhk?=
 =?utf-8?B?Z2NOOTBVYnJlbDZUZ2d2NWd3b1NxcW5sYk9rQTJrMFVlMzZCQzQ4bW9BSUNM?=
 =?utf-8?B?czNCbWUwMVUxTlRvRG43OWhwSWgxdW8xaDFPY25wV0dteDdwMU9WM1JSSVZh?=
 =?utf-8?B?Y3RhMHoyT2VCRmpsaUx3NmxTK0dZVW80V0VuUkRiTmNzbDA1VE1veGtzaUhq?=
 =?utf-8?B?dWI1VGRsd2MyMi8yQmdUZmR2RkkyaUlZSDJma2tSQzlvd1QrU3RwTzlSMUN2?=
 =?utf-8?B?Rm5QWk5yZ3FlZklkbUkvNnRVcFBXd2FCbzZlOWZEWHI0UWZQSkllUFd3TmZ3?=
 =?utf-8?B?eko3T25seXdIRm54TDFqeWVEQkJKeDZiakd0Z05sZWEzMnpDenkwR1pwejNY?=
 =?utf-8?B?YUZRZGdWWGNUVXdTYjlGMnJDN2RPNERPUXovWjV2L2llMENmanFvN0ZSTE1H?=
 =?utf-8?B?aXJBZCtyV3ExdjMyaytaVjJ1dEVNbHBrdnFIWW1FOVRkQXNiaFc3dTlHZ0dI?=
 =?utf-8?B?aGxNSzBFZHJmNlVUTmJzY2twS2xxZEhnME1uMmZUUkRld0xUbDQ4VG5UQ05p?=
 =?utf-8?Q?YAGZZWxnfGpB0w4Qeu5h1sOKo?=
X-OriginatorOrg: Nvidia.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 945e0ea7-7eb0-406e-35fa-08db629c258f
X-MS-Exchange-CrossTenant-AuthSource: BYAPR12MB3176.namprd12.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 01 Jun 2023 12:31:38.9947
 (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: jzc4akFt//a+xVtpY6YwbWow2eMVL3ifjzJpZH+poRHYyubis/Cl0wxD+nnqFyy82de4C/h37Zn/TzH6Z4fs9A==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: MW3PR12MB4555
X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,DKIMWL_WL_HIGH,
        DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,FORGED_SPF_HELO,
        RCVD_IN_DNSWL_NONE,RCVD_IN_MSPIKE_H2,SPF_HELO_PASS,SPF_NONE,
        T_SCC_BODY_TEXT_LINE autolearn=no autolearn_force=no version=3.4.6
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on
        lindbergh.monkeyblade.net
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org


"Huang, Ying" <ying.huang@intel.com> writes:

> mpenttil@redhat.com writes:
>
>> From: Mika Penttil=C3=A4 <mpenttil@redhat.com>
>>
>> Migrating file pages and swapcache pages into device memory is not suppo=
rted.
>> The decision is done based on page_mapping(). For now, swapcache pages a=
re not migrated.
>>
>> Things can however be improved, for swapcache pages. Try to get rid of t=
he swap cache,
>> and if successful, go ahead as with other anonymous pages.
>>
>> Cc: Alistair Popple <apopple@nvidia.com>
>> Cc: John Hubbard <jhubbard@nvidia.com>
>> Cc: Ralph Campbell <rcampbell@nvidia.com>
>> Cc: Huang Ying <ying.huang@intel.com>
>> Signed-off-by: Mika Penttil=C3=A4 <mpenttil@redhat.com>
>> ---
>>
>> v2:
>>   - use folio_test_anon() (Huang, Ying)
>>
>> =20
>>  mm/migrate_device.c | 24 ++++++++++++++++++------
>>  1 file changed, 18 insertions(+), 6 deletions(-)
>>
>> diff --git a/mm/migrate_device.c b/mm/migrate_device.c
>> index d30c9de60b0d..829bbc526758 100644
>> --- a/mm/migrate_device.c
>> +++ b/mm/migrate_device.c
>> @@ -747,13 +747,25 @@ static void __migrate_device_pages(unsigned long *=
src_pfns,
>> =20
>>  		if (is_device_private_page(newpage) ||
>>  		    is_device_coherent_page(newpage)) {
>> -			/*
>> -			 * For now only support anonymous memory migrating to
>> -			 * device private or coherent memory.
>> -			 */
>> +
>>  			if (mapping) {
>> -				src_pfns[i] &=3D ~MIGRATE_PFN_MIGRATE;
>> -				continue;
>> +				struct folio *folio;
>> +
>> +				folio =3D page_folio(page);
>> +
>> +				/*
>> +				 * For now only support anonymous memory migrating to
>> +				 * device private or coherent memory.
>> +				 *
>> +				 * Try to get rid of swap cache if possible.
>
> I think we can delete the following 2 lines of comments.  They add
> nothing except what code says already.

They are also a bit confusing, because a private file-backed mapping is
treated as anonymous so folio_test_anon() won't filter those
out.=20

I will test this tomorrow but the change looks good so please add:

Reviewed-by: Alistair Popple <apopple@nvidia.com>

> Otherwise looks good to me.
>
> Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
>
>> +				 * page is here file or swapcache page, could be shmem also
>> +				 * folio_test_anon() filters out file and shmem
>> +				 *
>> +				 */
>> +				if (!folio_test_anon(folio) || !folio_free_swap(folio)) {
>> +					src_pfns[i] &=3D ~MIGRATE_PFN_MIGRATE;
>> +					continue;
>> +				}
>>  			}
>>  		} else if (is_zone_device_page(newpage)) {
>>  			/*