The folio_referenced() is used to test whether a folio was referenced during
reclaim. Moreover, ZONE_DEVICE folios are controlled by their device driver,
have a lifetime tied to that driver, and are never placed on the LRU list.
That means we should never try to reclaim ZONE_DEVICE folios, so add a warning
to catch this unexpected behavior in folio_referenced() to avoid confusion,
as discussed in the previous thread[1].

[1] https://lore.kernel.org/all/16fb7985-ec0f-4b56-91e7-404c5114f899@kernel.org/
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 mm/rmap.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mm/rmap.c b/mm/rmap.c
index bff8f222004e..be785dfc9336 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1065,6 +1065,7 @@ int folio_referenced(struct folio *folio, int is_locked,
 		.invalid_vma = invalid_folio_referenced_vma,
 	};
 
+	VM_WARN_ON_FOLIO(folio_is_zone_device(folio), folio);
 	*vm_flags = 0;
 	if (!pra.mapcount)
 		return 0;
-- 
2.47.3

On 2026-02-24 at 12:56 +1100, Baolin Wang <baolin.wang@linux.alibaba.com> wrote...
> The folio_referenced() is used to test whether a folio was referenced during
> reclaim. Moreover, ZONE_DEVICE folios are controlled by their device driver,
> have a lifetime tied to that driver, and are never placed on the LRU list.
> That means we should never try to reclaim ZONE_DEVICE folios, so add a warning
> to catch this unexpected behavior in folio_referenced() to avoid confusion,
> as discussed in the previous thread[1].

Agree with this so one you've handled Rik's comment feel free to add:

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

> [1] https://lore.kernel.org/all/16fb7985-ec0f-4b56-91e7-404c5114f899@kernel.org/
> Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
> ---
>  mm/rmap.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/mm/rmap.c b/mm/rmap.c
> index bff8f222004e..be785dfc9336 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -1065,6 +1065,7 @@ int folio_referenced(struct folio *folio, int is_locked,
>  		.invalid_vma = invalid_folio_referenced_vma,
>  	};
>  
> +	VM_WARN_ON_FOLIO(folio_is_zone_device(folio), folio);
>  	*vm_flags = 0;
>  	if (!pra.mapcount)
>  		return 0;
> -- 
> 2.47.3
> 
>

On Tue, 2026-02-24 at 09:56 +0800, Baolin Wang wrote:
> 
> diff --git a/mm/rmap.c b/mm/rmap.c
> index bff8f222004e..be785dfc9336 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -1065,6 +1065,7 @@ int folio_referenced(struct folio *folio, int
> is_locked,
>  		.invalid_vma = invalid_folio_referenced_vma,
>  	};
>  
> +	VM_WARN_ON_FOLIO(folio_is_zone_device(folio), folio);
>  	*vm_flags = 0;
>  	if (!pra.mapcount)
>  		return 0;

Should be a VM_WARN_ON_ONCE_FOLIO so we do not cause
a softlockup if we try to print information about a
million ZONE_DEVICE pages?

-- 
All Rights Reversed.

On 2/24/26 10:38 AM, Rik van Riel wrote:
> On Tue, 2026-02-24 at 09:56 +0800, Baolin Wang wrote:
>>
>> diff --git a/mm/rmap.c b/mm/rmap.c
>> index bff8f222004e..be785dfc9336 100644
>> --- a/mm/rmap.c
>> +++ b/mm/rmap.c
>> @@ -1065,6 +1065,7 @@ int folio_referenced(struct folio *folio, int
>> is_locked,
>>   		.invalid_vma = invalid_folio_referenced_vma,
>>   	};
>>   
>> +	VM_WARN_ON_FOLIO(folio_is_zone_device(folio), folio);
>>   	*vm_flags = 0;
>>   	if (!pra.mapcount)
>>   		return 0;
> 
> Should be a VM_WARN_ON_ONCE_FOLIO so we do not cause
> a softlockup if we try to print information about a
> million ZONE_DEVICE pages?

Yes, sounds reasonable. Will do in next version. Thanks for reviewing.

On 2/24/26 06:49, Baolin Wang wrote:
> 
> 
> On 2/24/26 10:38 AM, Rik van Riel wrote:
>> On Tue, 2026-02-24 at 09:56 +0800, Baolin Wang wrote:
>>>
>>> diff --git a/mm/rmap.c b/mm/rmap.c
>>> index bff8f222004e..be785dfc9336 100644
>>> --- a/mm/rmap.c
>>> +++ b/mm/rmap.c
>>> @@ -1065,6 +1065,7 @@ int folio_referenced(struct folio *folio, int
>>> is_locked,
>>>           .invalid_vma = invalid_folio_referenced_vma,
>>>       };
>>>   +    VM_WARN_ON_FOLIO(folio_is_zone_device(folio), folio);
>>>       *vm_flags = 0;
>>>       if (!pra.mapcount)
>>>           return 0;
>>
>> Should be a VM_WARN_ON_ONCE_FOLIO so we do not cause
>> a softlockup if we try to print information about a
>> million ZONE_DEVICE pages?
> 
> Yes, sounds reasonable. Will do in next version. Thanks for reviewing.

Feel free to add my

Acked-by: David Hildenbrand (Arm) <david@kernel.org>

to that.

-- 
Cheers,

David