[PATCH] mm/migrate: fix shmem xarray update during migration

[PATCH] mm/migrate: fix shmem xarray update during migration

[Zi Yan]

Pagecache uses multi-index entries for large folio, so does shmem. Only
swap cache still stores multiple entries for a single large folio.
Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
fixed swap cache but got shmem wrong by storing multiple entries for
a large shmem folio. Fix it by storing a single entry for a shmem
folio.

Fixes: fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
Reported-by: Liu Shixin <liushixin2@huawei.com>
Closes: https://lore.kernel.org/all/28546fb4-5210-bf75-16d6-43e1f8646080@huawei.com/
Signed-off-by: Zi Yan <ziy@nvidia.com>
---
 mm/migrate.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/mm/migrate.c b/mm/migrate.c
index 365c6daa8d1b..9db26f5527a8 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -44,6 +44,7 @@
 #include <linux/sched/sysctl.h>
 #include <linux/memory-tiers.h>
 #include <linux/pagewalk.h>
+#include <linux/shmem_fs.h>
 
 #include <asm/tlbflush.h>
 
@@ -524,7 +525,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
 			folio_set_swapcache(newfolio);
 			newfolio->private = folio_get_private(folio);
 		}
-		entries = nr;
+		/* shmem uses high-order entry */
+		if (shmem_mapping(mapping))
+			entries = 1;
+		else
+			entries = nr;
 	} else {
 		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
 		entries = 1;
-- 
2.47.2

Re: [PATCH] mm/migrate: fix shmem xarray update during migration

[Matthew Wilcox]

On Fri, Feb 28, 2025 at 10:42:19AM -0500, Zi Yan wrote:
> @@ -524,7 +525,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
>  			folio_set_swapcache(newfolio);
>  			newfolio->private = folio_get_private(folio);
>  		}
> -		entries = nr;
> +		/* shmem uses high-order entry */
> +		if (shmem_mapping(mapping))

It's be cheaper to check folio_test_anon() here, right?

Also, how did this bug remain unnoticed for almost 4 years?
Our testing is terrible ;-(

> +			entries = 1;
> +		else
> +			entries = nr;
>  	} else {
>  		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
>  		entries = 1;
> -- 
> 2.47.2
> 

Re: [PATCH] mm/migrate: fix shmem xarray update during migration

[Shivank Garg]

On 2/28/2025 9:12 PM, Zi Yan wrote:
> Pagecache uses multi-index entries for large folio, so does shmem. Only
> swap cache still stores multiple entries for a single large folio.
> Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
> fixed swap cache but got shmem wrong by storing multiple entries for
> a large shmem folio. Fix it by storing a single entry for a shmem
> folio.
> 
> Fixes: fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
> Reported-by: Liu Shixin <liushixin2@huawei.com>
> Closes: https://lore.kernel.org/all/28546fb4-5210-bf75-16d6-43e1f8646080@huawei.com/
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> ---
>  mm/migrate.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/migrate.c b/mm/migrate.c
> index 365c6daa8d1b..9db26f5527a8 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -44,6 +44,7 @@
>  #include <linux/sched/sysctl.h>
>  #include <linux/memory-tiers.h>
>  #include <linux/pagewalk.h>
> +#include <linux/shmem_fs.h>
>  
>  #include <asm/tlbflush.h>
>  
> @@ -524,7 +525,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
>  			folio_set_swapcache(newfolio);
>  			newfolio->private = folio_get_private(folio);
>  		}
> -		entries = nr;
> +		/* shmem uses high-order entry */
> +		if (shmem_mapping(mapping))
> +			entries = 1;
> +		else
> +			entries = nr;

LGTM functionally.
As a minor style suggestion, can we consider using a ternary operator:

	entries = shmem_mapping(mapping) ? 1 : nr;  /* shmem uses high-order entry */

This looks cleaner to me.

Reviewed-by: Shivank Garg <shivankg@amd.com>

>  	} else {
>  		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
>  		entries = 1;

Re: [PATCH] mm/migrate: fix shmem xarray update during migration

[Zi Yan]

On 28 Feb 2025, at 11:35, Matthew Wilcox wrote:

> On Fri, Feb 28, 2025 at 10:42:19AM -0500, Zi Yan wrote:
>> @@ -524,7 +525,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
>>  			folio_set_swapcache(newfolio);
>>  			newfolio->private = folio_get_private(folio);
>>  		}
>> -		entries = nr;
>> +		/* shmem uses high-order entry */
>> +		if (shmem_mapping(mapping))
>
> It's be cheaper to check folio_test_anon() here, right?

Yes and it gets rid of the new include. Let me send v2.

>
> Also, how did this bug remain unnoticed for almost 4 years?
> Our testing is terrible ;-(

Probably not 4 years. Before fc346d0a70a1 (2023), shmem was right,
swap cache was wrong. After fc346d0a70a1, shmem is wrong, swap cache
is right. And before Baolin’s patch, shmem only has PMD size folio
to use multi-index entries. Maybe getting PMD size folio is
really impossible when compaction runs a lot?

>
>> +			entries = 1;
>> +		else
>> +			entries = nr;
>>  	} else {
>>  		VM_BUG_ON_FOLIO(folio_test_swapcache(folio), folio);
>>  		entries = 1;
>> -- 
>> 2.47.2
>>


Best Regards,
Yan, Zi

Re: [PATCH] mm/migrate: fix shmem xarray update during migration

[Zi Yan]

On 28 Feb 2025, at 12:02, Shivank Garg wrote:

> On 2/28/2025 9:12 PM, Zi Yan wrote:
>> Pagecache uses multi-index entries for large folio, so does shmem. Only
>> swap cache still stores multiple entries for a single large folio.
>> Commit fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
>> fixed swap cache but got shmem wrong by storing multiple entries for
>> a large shmem folio. Fix it by storing a single entry for a shmem
>> folio.
>>
>> Fixes: fc346d0a70a1 ("mm: migrate high-order folios in swap cache correctly")
>> Reported-by: Liu Shixin <liushixin2@huawei.com>
>> Closes: https://lore.kernel.org/all/28546fb4-5210-bf75-16d6-43e1f8646080@huawei.com/
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> ---
>>  mm/migrate.c | 7 ++++++-
>>  1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/mm/migrate.c b/mm/migrate.c
>> index 365c6daa8d1b..9db26f5527a8 100644
>> --- a/mm/migrate.c
>> +++ b/mm/migrate.c
>> @@ -44,6 +44,7 @@
>>  #include <linux/sched/sysctl.h>
>>  #include <linux/memory-tiers.h>
>>  #include <linux/pagewalk.h>
>> +#include <linux/shmem_fs.h>
>>
>>  #include <asm/tlbflush.h>
>>
>> @@ -524,7 +525,11 @@ static int __folio_migrate_mapping(struct address_space *mapping,
>>  			folio_set_swapcache(newfolio);
>>  			newfolio->private = folio_get_private(folio);
>>  		}
>> -		entries = nr;
>> +		/* shmem uses high-order entry */
>> +		if (shmem_mapping(mapping))
>> +			entries = 1;
>> +		else
>> +			entries = nr;
>
> LGTM functionally.
> As a minor style suggestion, can we consider using a ternary operator:
>
> 	entries = shmem_mapping(mapping) ? 1 : nr;  /* shmem uses high-order entry */
>
> This looks cleaner to me.
>
> Reviewed-by: Shivank Garg <shivankg@amd.com>

Thanks.

Best Regards,
Yan, Zi