Re: [PATCH v5 01/14] mm/khugepaged: remove READ_ONLY_THP_FOR_FS check

Re: [PATCH v5 01/14] mm/khugepaged: remove READ_ONLY_THP_FOR_FS check

[Lance Yang]

On Sun, May 03, 2026 at 09:48:40PM -0600, Nico Pache wrote:
>
>
>On 4/29/26 9:29 AM, Zi Yan wrote:
>> collapse_file() requires FSes supporting large folio with at least
>> PMD_ORDER, so replace the READ_ONLY_THP_FOR_FS check with that.
>> MADV_COLLAPSE ignores shmem huge config, so exclude the check for shmem.
>> 
>> While at it, replace VM_BUG_ON with VM_WARN_ON_ONCE.
>> 
>> Add a helper function mapping_pmd_folio_support() for FSes supporting large
>> folio with at least PMD_ORDER.
>> 
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> Reviewed-by: Lance Yang <lance.yang@linux.dev>
>> Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
>> ---
>>   include/linux/pagemap.h | 26 ++++++++++++++++++++++++++
>>   mm/khugepaged.c         | 10 ++++++++--
>>   2 files changed, 34 insertions(+), 2 deletions(-)
>> 
>> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
>> index 1f50991b43e3b..1fed3414fe9b8 100644
>> --- a/include/linux/pagemap.h
>> +++ b/include/linux/pagemap.h
>> @@ -513,6 +513,32 @@ static inline bool mapping_large_folio_support(const struct address_space *mappi
>>   	return mapping_max_folio_order(mapping) > 0;
>>   }
>>   
>> +/**
>> + * mapping_pmd_folio_support() - Check if a mapping support PMD-sized folio
>> + * @mapping: The address_space
>> + *
>> + * Some file supports large folio but does not support as large as PMD order.
>> + * If a PMD-sized pagecache folio is attempted to be created on a filesystem,
>> + * this check needs to be performed first.
>> + *
>> + * Return: true - PMD-sized folio is supported, false - PMD-sized folio is not
>> + * supported.
>> + */
>> +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> +static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
>> +{
>> +	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
>> +	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);
>> +
>> +	return mapping_max_folio_order(mapping) >= PMD_ORDER;
>
>Probably a stupid question, but I dont know FS thats well.
>
>Here we are checking that the max allowed folio order is greater than 
>(or eq) to the PMD_ORDER. Yet the function asks if PMD specifically is 
>supported. In the future could we have some FS that does not support PMD 
>orders, but does support larger orders (eg. PUD)?

Good point.

IIUC, mapping_max_folio_order() means "maximum supported order" not
"the only supported order", so mapping_pmd_folio_support() just means
"PMD order is within the supported range".

Also, mapping_set_large_folios() sets the range to:

mapping_set_folio_order_range(mapping, 0, MAX_PAGECACHE_ORDER);

and __filemap_get_folio_mpol() treats max as a cap, then falls back down
towards min.

That said, if we want the helper name to mean "PMD order specifically is
supported", the more future-proof test would be:

mapping_min_folio_order(mapping) <= PMD_ORDER &&
mapping_max_folio_order(mapping) >= PMD_ORDER

Thoughs?

Re: [PATCH v5 01/14] mm/khugepaged: remove READ_ONLY_THP_FOR_FS check

[Zi Yan]

On 7 May 2026, at 11:29, Lance Yang wrote:

> On Sun, May 03, 2026 at 09:48:40PM -0600, Nico Pache wrote:
>>
>>
>> On 4/29/26 9:29 AM, Zi Yan wrote:
>>> collapse_file() requires FSes supporting large folio with at least
>>> PMD_ORDER, so replace the READ_ONLY_THP_FOR_FS check with that.
>>> MADV_COLLAPSE ignores shmem huge config, so exclude the check for shmem.
>>>
>>> While at it, replace VM_BUG_ON with VM_WARN_ON_ONCE.
>>>
>>> Add a helper function mapping_pmd_folio_support() for FSes supporting large
>>> folio with at least PMD_ORDER.
>>>
>>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>>> Reviewed-by: Lance Yang <lance.yang@linux.dev>
>>> Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
>>> ---
>>>   include/linux/pagemap.h | 26 ++++++++++++++++++++++++++
>>>   mm/khugepaged.c         | 10 ++++++++--
>>>   2 files changed, 34 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
>>> index 1f50991b43e3b..1fed3414fe9b8 100644
>>> --- a/include/linux/pagemap.h
>>> +++ b/include/linux/pagemap.h
>>> @@ -513,6 +513,32 @@ static inline bool mapping_large_folio_support(const struct address_space *mappi
>>>   	return mapping_max_folio_order(mapping) > 0;
>>>   }
>>>
>>> +/**
>>> + * mapping_pmd_folio_support() - Check if a mapping support PMD-sized folio
>>> + * @mapping: The address_space
>>> + *
>>> + * Some file supports large folio but does not support as large as PMD order.
>>> + * If a PMD-sized pagecache folio is attempted to be created on a filesystem,
>>> + * this check needs to be performed first.
>>> + *
>>> + * Return: true - PMD-sized folio is supported, false - PMD-sized folio is not
>>> + * supported.
>>> + */
>>> +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>> +static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
>>> +{
>>> +	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
>>> +	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);
>>> +
>>> +	return mapping_max_folio_order(mapping) >= PMD_ORDER;
>>
>> Probably a stupid question, but I dont know FS thats well.
>>
>> Here we are checking that the max allowed folio order is greater than
>> (or eq) to the PMD_ORDER. Yet the function asks if PMD specifically is
>> supported. In the future could we have some FS that does not support PMD
>> orders, but does support larger orders (eg. PUD)?
>
> Good point.
>
> IIUC, mapping_max_folio_order() means "maximum supported order" not
> "the only supported order", so mapping_pmd_folio_support() just means
> "PMD order is within the supported range".
>
> Also, mapping_set_large_folios() sets the range to:
>
> mapping_set_folio_order_range(mapping, 0, MAX_PAGECACHE_ORDER);
>
> and __filemap_get_folio_mpol() treats max as a cap, then falls back down
> towards min.
>
> That said, if we want the helper name to mean "PMD order specifically is
> supported", the more future-proof test would be:
>
> mapping_min_folio_order(mapping) <= PMD_ORDER &&
> mapping_max_folio_order(mapping) >= PMD_ORDER
>
> Thoughs?

Thank both for raising this up. I agree just <= PMD_ORDER might not cover
the exact requirement here. Let me send a fixup.


Best Regards,
Yan, Zi

Re: [PATCH v5 01/14] mm/khugepaged: remove READ_ONLY_THP_FOR_FS check

[Zi Yan]

On 29 Apr 2026, at 23:29, Zi Yan wrote:

> collapse_file() requires FSes supporting large folio with at least
> PMD_ORDER, so replace the READ_ONLY_THP_FOR_FS check with that.
> MADV_COLLAPSE ignores shmem huge config, so exclude the check for shmem.
>
> While at it, replace VM_BUG_ON with VM_WARN_ON_ONCE.
>
> Add a helper function mapping_pmd_folio_support() for FSes supporting large
> folio with at least PMD_ORDER.
>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lance Yang <lance.yang@linux.dev>
> Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
> ---
>  include/linux/pagemap.h | 26 ++++++++++++++++++++++++++
>  mm/khugepaged.c         | 10 ++++++++--
>  2 files changed, 34 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
> index 1f50991b43e3b..1fed3414fe9b8 100644
> --- a/include/linux/pagemap.h
> +++ b/include/linux/pagemap.h
> @@ -513,6 +513,32 @@ static inline bool mapping_large_folio_support(const struct address_space *mappi
>  	return mapping_max_folio_order(mapping) > 0;
>  }
>
> +/**
> + * mapping_pmd_folio_support() - Check if a mapping support PMD-sized folio
> + * @mapping: The address_space
> + *
> + * Some file supports large folio but does not support as large as PMD order.
> + * If a PMD-sized pagecache folio is attempted to be created on a filesystem,
> + * this check needs to be performed first.
> + *
> + * Return: true - PMD-sized folio is supported, false - PMD-sized folio is not
> + * supported.
> + */
> +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> +static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
> +{
> +	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
> +	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);
> +
> +	return mapping_max_folio_order(mapping) >= PMD_ORDER;

Hi Andrew,

Can you help apply the fixup below? It addresses the concern raised by
Nico and Lance[1]. Thanks.

[1] https://lore.kernel.org/all/aa778cfc-b7f8-4100-89bb-d2b2ef8e1138@redhat.com/

From 9c1749df6516c06b1628ac7804bc7a6ab5709f37 Mon Sep 17 00:00:00 2001
From: Zi Yan <ziy@nvidia.com>
Date: Thu, 7 May 2026 02:02:28 -0400
Subject: [PATCH] fix mapping_pmd_folio_support() to represent its exact
 meaning

Signed-off-by: Zi Yan <ziy@nvidia.com>
---
 include/linux/pagemap.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index c6a4ecd3d6ed1..52895896f3357 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -530,7 +530,8 @@ static inline bool mapping_pmd_folio_support(const struct address_space *mapping
 	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
 	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);

-	return mapping_max_folio_order(mapping) >= PMD_ORDER;
+	return mapping_min_folio_order(mapping) <= PMD_ORDER &&
+	       mapping_max_folio_order(mapping) >= PMD_ORDER
 }
 #else
 static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
-- 
2.53.0


Best Regards,
Yan, Zi

Re: [PATCH v5 01/14] mm/khugepaged: remove READ_ONLY_THP_FOR_FS check

[Zi Yan]

On 7 May 2026, at 14:08, Zi Yan wrote:

> On 29 Apr 2026, at 23:29, Zi Yan wrote:
>
>> collapse_file() requires FSes supporting large folio with at least
>> PMD_ORDER, so replace the READ_ONLY_THP_FOR_FS check with that.
>> MADV_COLLAPSE ignores shmem huge config, so exclude the check for shmem.
>>
>> While at it, replace VM_BUG_ON with VM_WARN_ON_ONCE.
>>
>> Add a helper function mapping_pmd_folio_support() for FSes supporting large
>> folio with at least PMD_ORDER.
>>
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> Reviewed-by: Lance Yang <lance.yang@linux.dev>
>> Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
>> ---
>>  include/linux/pagemap.h | 26 ++++++++++++++++++++++++++
>>  mm/khugepaged.c         | 10 ++++++++--
>>  2 files changed, 34 insertions(+), 2 deletions(-)
>>
>> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
>> index 1f50991b43e3b..1fed3414fe9b8 100644
>> --- a/include/linux/pagemap.h
>> +++ b/include/linux/pagemap.h
>> @@ -513,6 +513,32 @@ static inline bool mapping_large_folio_support(const struct address_space *mappi
>>  	return mapping_max_folio_order(mapping) > 0;
>>  }
>>
>> +/**
>> + * mapping_pmd_folio_support() - Check if a mapping support PMD-sized folio
>> + * @mapping: The address_space
>> + *
>> + * Some file supports large folio but does not support as large as PMD order.
>> + * If a PMD-sized pagecache folio is attempted to be created on a filesystem,
>> + * this check needs to be performed first.
>> + *
>> + * Return: true - PMD-sized folio is supported, false - PMD-sized folio is not
>> + * supported.
>> + */
>> +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> +static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
>> +{
>> +	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
>> +	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);
>> +
>> +	return mapping_max_folio_order(mapping) >= PMD_ORDER;
>
> Hi Andrew,
>
> Can you help apply the fixup below? It addresses the concern raised by
> Nico and Lance[1]. Thanks.
>
> [1] https://lore.kernel.org/all/aa778cfc-b7f8-4100-89bb-d2b2ef8e1138@redhat.com/
>

I was missing a semicolon. Here is the right fixup. Sorry for the noise.

From fbd183f7528a3d0bdb421018af4aef45f6366682 Mon Sep 17 00:00:00 2001
From: Zi Yan <ziy@nvidia.com>
Date: Thu, 7 May 2026 02:02:28 -0400
Subject: [PATCH] fix mapping_pmd_folio_support() to represent its exact
 meaning.

Signed-off-by: Zi Yan <ziy@nvidia.com>
---
 include/linux/pagemap.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index c6a4ecd3d6ed1..41dbb55a47d8e 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -530,7 +530,8 @@ static inline bool mapping_pmd_folio_support(const struct address_space *mapping
 	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
 	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);

-	return mapping_max_folio_order(mapping) >= PMD_ORDER;
+	return mapping_min_folio_order(mapping) <= PMD_ORDER &&
+	       mapping_max_folio_order(mapping) >= PMD_ORDER;
 }
 #else
 static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
-- 
2.53.0




Best Regards,
Yan, Zi

Re: [PATCH v5 01/14] mm/khugepaged: remove READ_ONLY_THP_FOR_FS check

[David Hildenbrand (Arm)]

On 4/29/26 17:29, Zi Yan wrote:
> collapse_file() requires FSes supporting large folio with at least
> PMD_ORDER, so replace the READ_ONLY_THP_FOR_FS check with that.
> MADV_COLLAPSE ignores shmem huge config, so exclude the check for shmem.
> 
> While at it, replace VM_BUG_ON with VM_WARN_ON_ONCE.
> 
> Add a helper function mapping_pmd_folio_support() for FSes supporting large
> folio with at least PMD_ORDER.
> 
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lance Yang <lance.yang@linux.dev>
> Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
> ---
>  include/linux/pagemap.h | 26 ++++++++++++++++++++++++++
>  mm/khugepaged.c         | 10 ++++++++--
>  2 files changed, 34 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
> index 1f50991b43e3b..1fed3414fe9b8 100644
> --- a/include/linux/pagemap.h
> +++ b/include/linux/pagemap.h
> @@ -513,6 +513,32 @@ static inline bool mapping_large_folio_support(const struct address_space *mappi
>  	return mapping_max_folio_order(mapping) > 0;
>  }

Some smaller doc improvements:

>  
> +/**
> + * mapping_pmd_folio_support() - Check if a mapping support PMD-sized folio

"supports"

> + * @mapping: The address_space
> + *
> + * Some file supports large folio but does not support as large as PMD order.

"While some mappings support large folios, they might not support PMD-sized folios."

> + * If a PMD-sized pagecache folio is attempted to be created on a filesystem,
> + * this check needs to be performed first.

I'd rather just add (relate to the previous sentence):

"This function checks whether a mapping supports PMD-sized folios. For example,
khugepaged needs this information before attempting to collapsing THPs".

> + *
> + * Return: true - PMD-sized folio is supported, false - PMD-sized folio is not
> + * supported.

I'd say "folios are supported" and end up with:

"True if PMD-sized folios are supported, otherwise false."

> + */
> +#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> +static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
> +{
> +	/* AS_FOLIO_ORDER is only reasonable for pagecache folios */
> +	VM_WARN_ON_ONCE((unsigned long)mapping & FOLIO_MAPPING_ANON);
> +
> +	return mapping_max_folio_order(mapping) >= PMD_ORDER;

Agreed that this should check for ==, although in practice it would currently
not make a difference.

> +}
> +#else
> +static inline bool mapping_pmd_folio_support(const struct address_space *mapping)
> +{
> +	return false;
> +}
> +#endif
> +
>  /* Return the maximum folio size for this pagecache mapping, in bytes. */
>  static inline size_t mapping_max_folio_size(const struct address_space *mapping)
>  {
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index e112525c4aa9c..6808f2b48d864 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -2235,8 +2235,14 @@ static enum scan_result collapse_file(struct mm_struct *mm, unsigned long addr,
>  	int nr_none = 0;
>  	bool is_shmem = shmem_file(file);
>  
> -	VM_BUG_ON(!IS_ENABLED(CONFIG_READ_ONLY_THP_FOR_FS) && !is_shmem);
> -	VM_BUG_ON(start & (HPAGE_PMD_NR - 1));
> +	/*
> +	 * MADV_COLLAPSE ignores shmem huge config, so do not check shmem
> +	 *
> +	 * TODO: once shmem always calls mapping_set_large_folios() on its
> +	 * mapping, the shmem check can be removed.
> +	 */

TODO LGTM.

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

-- 
Cheers,

David