[PATCH v4 0/3] Optimize folio split in memory failure

[PATCH v4 0/3] Optimize folio split in memory failure

[Zi Yan]

Hi all,

This patchset is a follow-up of "[PATCH v3] mm/huge_memory: do not change
split_huge_page*() target order silently."[1] and
[PATCH v4] mm/huge_memory: preserve PG_has_hwpoisoned if a folio is split
to >0 order[2], since both are separated out as hotfixes. It improves how
memory failure code handles large block size(LBS) folios with
min_order_for_split() > 0. By splitting a large folio containing HW
poisoned pages to min_order_for_split(), the after-split folios without
HW poisoned pages could be freed for reuse. To achieve this, folio split
code needs to set has_hwpoisoned on after-split folios containing HW
poisoned pages and it is done in the hotfix in [2].

This patchset includes:
1. A patch adds split_huge_page_to_order(),
2. Patch 2 and Patch 3 of "[PATCH v2 0/3] Do not change split folio target
   order"[3],

This patchset is based on mm-new.

Changelog
===
From V3[4]:
1. Patch, mm/huge_memory: preserve PG_has_hwpoisoned if a folio is split
   to >0 order, is sent separately as a hotfix[2].
2. made newly added new_order const in memory_failure() and
   soft_offline_in_use_page().
3. explained in a comment why in memory_failure() after-split >0 order
   folios are still treated as if the split failed.


From V2[3]:
1. Patch 1 is sent separately as a hotfix[1].
2. set has_hwpoisoned on after-split folios if any contains HW poisoned
   pages.
3. added split_huge_page_to_order().
4. added a missing newline after variable decalaration.
5. added /* release= */ to try_to_split_thp_page().
6. restructured try_to_split_thp_page() in memory_failure().
7. fixed a typo.
8. reworded the comment in soft_offline_in_use_page() for better
   understanding.


Link: https://lore.kernel.org/all/20251017013630.139907-1-ziy@nvidia.com/ [1]
Link: https://lore.kernel.org/all/20251023030521.473097-1-ziy@nvidia.com/ [2]
Link: https://lore.kernel.org/all/20251016033452.125479-1-ziy@nvidia.com/ [3]
Link: https://lore.kernel.org/all/20251022033531.389351-1-ziy@nvidia.com/ [4]

Zi Yan (3):
  mm/huge_memory: add split_huge_page_to_order()
  mm/memory-failure: improve large block size folio handling.
  mm/huge_memory: fix kernel-doc comments for folio_split() and related.

 include/linux/huge_mm.h | 22 ++++++++++++++++------
 mm/huge_memory.c        | 27 +++++++++++++++------------
 mm/memory-failure.c     | 31 +++++++++++++++++++++++++++----
 3 files changed, 58 insertions(+), 22 deletions(-)

-- 
2.43.0

[PATCH v4 1/3] mm/huge_memory: add split_huge_page_to_order()

[Zi Yan]

When caller does not supply a list to split_huge_page_to_list_to_order(),
use split_huge_page_to_order() instead.

Signed-off-by: Zi Yan <ziy@nvidia.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
---
 include/linux/huge_mm.h | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
index 7698b3542c4f..34f8d8453bf3 100644
--- a/include/linux/huge_mm.h
+++ b/include/linux/huge_mm.h
@@ -381,6 +381,10 @@ static inline int split_huge_page_to_list_to_order(struct page *page, struct lis
 {
 	return __split_huge_page_to_list_to_order(page, list, new_order, false);
 }
+static inline int split_huge_page_to_order(struct page *page, unsigned int new_order)
+{
+	return split_huge_page_to_list_to_order(page, NULL, new_order);
+}
 
 /*
  * try_folio_split_to_order - try to split a @folio at @page to @new_order using
@@ -400,8 +404,7 @@ static inline int try_folio_split_to_order(struct folio *folio,
 		struct page *page, unsigned int new_order)
 {
 	if (!non_uniform_split_supported(folio, new_order, /* warns= */ false))
-		return split_huge_page_to_list_to_order(&folio->page, NULL,
-				new_order);
+		return split_huge_page_to_order(&folio->page, new_order);
 	return folio_split(folio, new_order, page, NULL);
 }
 static inline int split_huge_page(struct page *page)
@@ -590,6 +593,11 @@ split_huge_page_to_list_to_order(struct page *page, struct list_head *list,
 	VM_WARN_ON_ONCE_PAGE(1, page);
 	return -EINVAL;
 }
+static inline int split_huge_page_to_order(struct page *page, unsigned int new_order)
+{
+	VM_WARN_ON_ONCE_PAGE(1, page);
+	return -EINVAL;
+}
 static inline int split_huge_page(struct page *page)
 {
 	VM_WARN_ON_ONCE_PAGE(1, page);
-- 
2.43.0

[PATCH v4 2/3] mm/memory-failure: improve large block size folio handling.

[Zi Yan]

Large block size (LBS) folios cannot be split to order-0 folios but
min_order_for_folio(). Current split fails directly, but that is not
optimal. Split the folio to min_order_for_folio(), so that, after split,
only the folio containing the poisoned page becomes unusable instead.

For soft offline, do not split the large folio if its min_order_for_folio()
is not 0. Since the folio is still accessible from userspace and premature
split might lead to potential performance loss.

Suggested-by: Jane Chu <jane.chu@oracle.com>
Signed-off-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
---
 mm/memory-failure.c | 31 +++++++++++++++++++++++++++----
 1 file changed, 27 insertions(+), 4 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index f698df156bf8..acc35c881547 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1656,12 +1656,13 @@ static int identify_page_state(unsigned long pfn, struct page *p,
  * there is still more to do, hence the page refcount we took earlier
  * is still needed.
  */
-static int try_to_split_thp_page(struct page *page, bool release)
+static int try_to_split_thp_page(struct page *page, unsigned int new_order,
+		bool release)
 {
 	int ret;
 
 	lock_page(page);
-	ret = split_huge_page(page);
+	ret = split_huge_page_to_order(page, new_order);
 	unlock_page(page);
 
 	if (ret && release)
@@ -2280,6 +2281,9 @@ int memory_failure(unsigned long pfn, int flags)
 	folio_unlock(folio);
 
 	if (folio_test_large(folio)) {
+		const int new_order = min_order_for_split(folio);
+		int err;
+
 		/*
 		 * The flag must be set after the refcount is bumped
 		 * otherwise it may race with THP split.
@@ -2294,7 +2298,16 @@ int memory_failure(unsigned long pfn, int flags)
 		 * page is a valid handlable page.
 		 */
 		folio_set_has_hwpoisoned(folio);
-		if (try_to_split_thp_page(p, false) < 0) {
+		err = try_to_split_thp_page(p, new_order, /* release= */ false);
+		/*
+		 * If splitting a folio to order-0 fails, kill the process.
+		 * Split the folio regardless to minimize unusable pages.
+		 * Because the memory failure code cannot handle large
+		 * folios, this split is always treated as if it failed.
+		 */
+		if (err || new_order) {
+			/* get folio again in case the original one is split */
+			folio = page_folio(p);
 			res = -EHWPOISON;
 			kill_procs_now(p, pfn, flags, folio);
 			put_page(p);
@@ -2621,7 +2634,17 @@ static int soft_offline_in_use_page(struct page *page)
 	};
 
 	if (!huge && folio_test_large(folio)) {
-		if (try_to_split_thp_page(page, true)) {
+		const int new_order = min_order_for_split(folio);
+
+		/*
+		 * If new_order (target split order) is not 0, do not split the
+		 * folio at all to retain the still accessible large folio.
+		 * NOTE: if minimizing the number of soft offline pages is
+		 * preferred, split it to non-zero new_order like it is done in
+		 * memory_failure().
+		 */
+		if (new_order || try_to_split_thp_page(page, /* new_order= */ 0,
+						       /* release= */ true)) {
 			pr_info("%#lx: thp split failed\n", pfn);
 			return -EBUSY;
 		}
-- 
2.43.0

[PATCH v4 3/3] mm/huge_memory: fix kernel-doc comments for folio_split() and related.

[Zi Yan]

try_folio_split_to_order(), folio_split, __folio_split(), and
__split_unmapped_folio() do not have correct kernel-doc comment format.
Fix them.

Signed-off-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Acked-by: David Hildenbrand <david@redhat.com>
---
 include/linux/huge_mm.h | 10 ++++++----
 mm/huge_memory.c        | 27 +++++++++++++++------------
 2 files changed, 21 insertions(+), 16 deletions(-)

diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
index 34f8d8453bf3..cbb2243f8e56 100644
--- a/include/linux/huge_mm.h
+++ b/include/linux/huge_mm.h
@@ -386,9 +386,9 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
 	return split_huge_page_to_list_to_order(page, NULL, new_order);
 }
 
-/*
- * try_folio_split_to_order - try to split a @folio at @page to @new_order using
- * non uniform split.
+/**
+ * try_folio_split_to_order() - try to split a @folio at @page to @new_order
+ * using non uniform split.
  * @folio: folio to be split
  * @page: split to @new_order at the given page
  * @new_order: the target split order
@@ -398,7 +398,7 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
  * folios are put back to LRU list. Use min_order_for_split() to get the lower
  * bound of @new_order.
  *
- * Return: 0: split is successful, otherwise split failed.
+ * Return: 0 - split is successful, otherwise split failed.
  */
 static inline int try_folio_split_to_order(struct folio *folio,
 		struct page *page, unsigned int new_order)
@@ -486,6 +486,8 @@ static inline spinlock_t *pud_trans_huge_lock(pud_t *pud,
 /**
  * folio_test_pmd_mappable - Can we map this folio with a PMD?
  * @folio: The folio to test
+ *
+ * Return: true - @folio can be mapped, false - @folio cannot be mapped.
  */
 static inline bool folio_test_pmd_mappable(struct folio *folio)
 {
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 0e24bb7e90d0..381a49c5ac3f 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -3567,8 +3567,9 @@ static void __split_folio_to_order(struct folio *folio, int old_order,
 		ClearPageCompound(&folio->page);
 }
 
-/*
- * It splits an unmapped @folio to lower order smaller folios in two ways.
+/**
+ * __split_unmapped_folio() - splits an unmapped @folio to lower order folios in
+ * two ways: uniform split or non-uniform split.
  * @folio: the to-be-split folio
  * @new_order: the smallest order of the after split folios (since buddy
  *             allocator like split generates folios with orders from @folio's
@@ -3603,8 +3604,8 @@ static void __split_folio_to_order(struct folio *folio, int old_order,
  * folio containing @page. The caller needs to unlock and/or free after-split
  * folios if necessary.
  *
- * For !uniform_split, when -ENOMEM is returned, the original folio might be
- * split. The caller needs to check the input folio.
+ * Return: 0 - successful, <0 - failed (if -ENOMEM is returned, @folio might be
+ * split but not to @new_order, the caller needs to check)
  */
 static int __split_unmapped_folio(struct folio *folio, int new_order,
 		struct page *split_at, struct xa_state *xas,
@@ -3722,8 +3723,8 @@ bool uniform_split_supported(struct folio *folio, unsigned int new_order,
 	return true;
 }
 
-/*
- * __folio_split: split a folio at @split_at to a @new_order folio
+/**
+ * __folio_split() - split a folio at @split_at to a @new_order folio
  * @folio: folio to split
  * @new_order: the order of the new folio
  * @split_at: a page within the new folio
@@ -3741,7 +3742,7 @@ bool uniform_split_supported(struct folio *folio, unsigned int new_order,
  * 1. for uniform split, @lock_at points to one of @folio's subpages;
  * 2. for buddy allocator like (non-uniform) split, @lock_at points to @folio.
  *
- * return: 0: successful, <0 failed (if -ENOMEM is returned, @folio might be
+ * Return: 0 - successful, <0 - failed (if -ENOMEM is returned, @folio might be
  * split but not to @new_order, the caller needs to check)
  */
 static int __folio_split(struct folio *folio, unsigned int new_order,
@@ -4130,14 +4131,13 @@ int __split_huge_page_to_list_to_order(struct page *page, struct list_head *list
 				unmapped);
 }
 
-/*
- * folio_split: split a folio at @split_at to a @new_order folio
+/**
+ * folio_split() - split a folio at @split_at to a @new_order folio
  * @folio: folio to split
  * @new_order: the order of the new folio
  * @split_at: a page within the new folio
- *
- * return: 0: successful, <0 failed (if -ENOMEM is returned, @folio might be
- * split but not to @new_order, the caller needs to check)
+ * @list: after-split folios are added to @list if not null, otherwise to LRU
+ *        list
  *
  * It has the same prerequisites and returns as
  * split_huge_page_to_list_to_order().
@@ -4151,6 +4151,9 @@ int __split_huge_page_to_list_to_order(struct page *page, struct list_head *list
  * [order-4, {order-3}, order-3, order-5, order-6, order-7, order-8].
  *
  * After split, folio is left locked for caller.
+ *
+ * Return: 0 - successful, <0 - failed (if -ENOMEM is returned, @folio might be
+ * split but not to @new_order, the caller needs to check)
  */
 int folio_split(struct folio *folio, unsigned int new_order,
 		struct page *split_at, struct list_head *list)
-- 
2.43.0

Re: [PATCH v4 1/3] mm/huge_memory: add split_huge_page_to_order()

[Lance Yang]

On 2025/10/30 09:40, Zi Yan wrote:
> When caller does not supply a list to split_huge_page_to_list_to_order(),
> use split_huge_page_to_order() instead.
> 
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> ---

Reviewed-by: Lance Yang <lance.yang@linux.dev>

Re: [PATCH v4 2/3] mm/memory-failure: improve large block size folio handling.

[Lance Yang]

On 2025/10/30 09:40, Zi Yan wrote:
> Large block size (LBS) folios cannot be split to order-0 folios but
> min_order_for_folio(). Current split fails directly, but that is not
> optimal. Split the folio to min_order_for_folio(), so that, after split,
> only the folio containing the poisoned page becomes unusable instead.
> 
> For soft offline, do not split the large folio if its min_order_for_folio()
> is not 0. Since the folio is still accessible from userspace and premature
> split might lead to potential performance loss.
> 
> Suggested-by: Jane Chu <jane.chu@oracle.com>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> ---

LGTM! Feel free to add:

Reviewed-by: Lance Yang <lance.yang@linux.dev>

>   mm/memory-failure.c | 31 +++++++++++++++++++++++++++----
>   1 file changed, 27 insertions(+), 4 deletions(-)
> 
> diff --git a/mm/memory-failure.c b/mm/memory-failure.c
> index f698df156bf8..acc35c881547 100644
> --- a/mm/memory-failure.c
> +++ b/mm/memory-failure.c
> @@ -1656,12 +1656,13 @@ static int identify_page_state(unsigned long pfn, struct page *p,
>    * there is still more to do, hence the page refcount we took earlier
>    * is still needed.
>    */
> -static int try_to_split_thp_page(struct page *page, bool release)
> +static int try_to_split_thp_page(struct page *page, unsigned int new_order,
> +		bool release)
>   {
>   	int ret;
>   
>   	lock_page(page);
> -	ret = split_huge_page(page);
> +	ret = split_huge_page_to_order(page, new_order);
>   	unlock_page(page);
>   
>   	if (ret && release)
> @@ -2280,6 +2281,9 @@ int memory_failure(unsigned long pfn, int flags)
>   	folio_unlock(folio);
>   
>   	if (folio_test_large(folio)) {
> +		const int new_order = min_order_for_split(folio);
> +		int err;
> +
>   		/*
>   		 * The flag must be set after the refcount is bumped
>   		 * otherwise it may race with THP split.
> @@ -2294,7 +2298,16 @@ int memory_failure(unsigned long pfn, int flags)
>   		 * page is a valid handlable page.
>   		 */
>   		folio_set_has_hwpoisoned(folio);
> -		if (try_to_split_thp_page(p, false) < 0) {
> +		err = try_to_split_thp_page(p, new_order, /* release= */ false);
> +		/*
> +		 * If splitting a folio to order-0 fails, kill the process.
> +		 * Split the folio regardless to minimize unusable pages.
> +		 * Because the memory failure code cannot handle large
> +		 * folios, this split is always treated as if it failed.
> +		 */
> +		if (err || new_order) {
> +			/* get folio again in case the original one is split */
> +			folio = page_folio(p);
>   			res = -EHWPOISON;
>   			kill_procs_now(p, pfn, flags, folio);
>   			put_page(p);
> @@ -2621,7 +2634,17 @@ static int soft_offline_in_use_page(struct page *page)
>   	};
>   
>   	if (!huge && folio_test_large(folio)) {
> -		if (try_to_split_thp_page(page, true)) {
> +		const int new_order = min_order_for_split(folio);
> +
> +		/*
> +		 * If new_order (target split order) is not 0, do not split the
> +		 * folio at all to retain the still accessible large folio.
> +		 * NOTE: if minimizing the number of soft offline pages is
> +		 * preferred, split it to non-zero new_order like it is done in
> +		 * memory_failure().
> +		 */
> +		if (new_order || try_to_split_thp_page(page, /* new_order= */ 0,
> +						       /* release= */ true)) {
>   			pr_info("%#lx: thp split failed\n", pfn);
>   			return -EBUSY;
>   		}

Re: [PATCH v4 3/3] mm/huge_memory: fix kernel-doc comments for folio_split() and related.

[Lance Yang]

On 2025/10/30 09:40, Zi Yan wrote:
> try_folio_split_to_order(), folio_split, __folio_split(), and
> __split_unmapped_folio() do not have correct kernel-doc comment format.
> Fix them.
> 
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> ---

LGTM.

Reviewed-by: Lance Yang <lance.yang@linux.dev>

>   include/linux/huge_mm.h | 10 ++++++----
>   mm/huge_memory.c        | 27 +++++++++++++++------------
>   2 files changed, 21 insertions(+), 16 deletions(-)
> 
> diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
> index 34f8d8453bf3..cbb2243f8e56 100644
> --- a/include/linux/huge_mm.h
> +++ b/include/linux/huge_mm.h
> @@ -386,9 +386,9 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
>   	return split_huge_page_to_list_to_order(page, NULL, new_order);
>   }
>   
> -/*
> - * try_folio_split_to_order - try to split a @folio at @page to @new_order using
> - * non uniform split.
> +/**
> + * try_folio_split_to_order() - try to split a @folio at @page to @new_order
> + * using non uniform split.
>    * @folio: folio to be split
>    * @page: split to @new_order at the given page
>    * @new_order: the target split order
> @@ -398,7 +398,7 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
>    * folios are put back to LRU list. Use min_order_for_split() to get the lower
>    * bound of @new_order.
>    *
> - * Return: 0: split is successful, otherwise split failed.
> + * Return: 0 - split is successful, otherwise split failed.
>    */
>   static inline int try_folio_split_to_order(struct folio *folio,
>   		struct page *page, unsigned int new_order)
> @@ -486,6 +486,8 @@ static inline spinlock_t *pud_trans_huge_lock(pud_t *pud,
>   /**
>    * folio_test_pmd_mappable - Can we map this folio with a PMD?
>    * @folio: The folio to test
> + *
> + * Return: true - @folio can be mapped, false - @folio cannot be mapped.
>    */
>   static inline bool folio_test_pmd_mappable(struct folio *folio)
>   {
> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> index 0e24bb7e90d0..381a49c5ac3f 100644
> --- a/mm/huge_memory.c
> +++ b/mm/huge_memory.c
> @@ -3567,8 +3567,9 @@ static void __split_folio_to_order(struct folio *folio, int old_order,
>   		ClearPageCompound(&folio->page);
>   }
>   
> -/*
> - * It splits an unmapped @folio to lower order smaller folios in two ways.
> +/**
> + * __split_unmapped_folio() - splits an unmapped @folio to lower order folios in
> + * two ways: uniform split or non-uniform split.
>    * @folio: the to-be-split folio
>    * @new_order: the smallest order of the after split folios (since buddy
>    *             allocator like split generates folios with orders from @folio's
> @@ -3603,8 +3604,8 @@ static void __split_folio_to_order(struct folio *folio, int old_order,
>    * folio containing @page. The caller needs to unlock and/or free after-split
>    * folios if necessary.
>    *
> - * For !uniform_split, when -ENOMEM is returned, the original folio might be
> - * split. The caller needs to check the input folio.
> + * Return: 0 - successful, <0 - failed (if -ENOMEM is returned, @folio might be
> + * split but not to @new_order, the caller needs to check)
>    */
>   static int __split_unmapped_folio(struct folio *folio, int new_order,
>   		struct page *split_at, struct xa_state *xas,
> @@ -3722,8 +3723,8 @@ bool uniform_split_supported(struct folio *folio, unsigned int new_order,
>   	return true;
>   }
>   
> -/*
> - * __folio_split: split a folio at @split_at to a @new_order folio
> +/**
> + * __folio_split() - split a folio at @split_at to a @new_order folio
>    * @folio: folio to split
>    * @new_order: the order of the new folio
>    * @split_at: a page within the new folio
> @@ -3741,7 +3742,7 @@ bool uniform_split_supported(struct folio *folio, unsigned int new_order,
>    * 1. for uniform split, @lock_at points to one of @folio's subpages;
>    * 2. for buddy allocator like (non-uniform) split, @lock_at points to @folio.
>    *
> - * return: 0: successful, <0 failed (if -ENOMEM is returned, @folio might be
> + * Return: 0 - successful, <0 - failed (if -ENOMEM is returned, @folio might be
>    * split but not to @new_order, the caller needs to check)
>    */
>   static int __folio_split(struct folio *folio, unsigned int new_order,
> @@ -4130,14 +4131,13 @@ int __split_huge_page_to_list_to_order(struct page *page, struct list_head *list
>   				unmapped);
>   }
>   
> -/*
> - * folio_split: split a folio at @split_at to a @new_order folio
> +/**
> + * folio_split() - split a folio at @split_at to a @new_order folio
>    * @folio: folio to split
>    * @new_order: the order of the new folio
>    * @split_at: a page within the new folio
> - *
> - * return: 0: successful, <0 failed (if -ENOMEM is returned, @folio might be
> - * split but not to @new_order, the caller needs to check)
> + * @list: after-split folios are added to @list if not null, otherwise to LRU
> + *        list
>    *
>    * It has the same prerequisites and returns as
>    * split_huge_page_to_list_to_order().
> @@ -4151,6 +4151,9 @@ int __split_huge_page_to_list_to_order(struct page *page, struct list_head *list
>    * [order-4, {order-3}, order-3, order-5, order-6, order-7, order-8].
>    *
>    * After split, folio is left locked for caller.
> + *
> + * Return: 0 - successful, <0 - failed (if -ENOMEM is returned, @folio might be
> + * split but not to @new_order, the caller needs to check)
>    */
>   int folio_split(struct folio *folio, unsigned int new_order,
>   		struct page *split_at, struct list_head *list)

Re: [PATCH v4 1/3] mm/huge_memory: add split_huge_page_to_order()

[Barry Song]

On Thu, Oct 30, 2025 at 9:40 AM Zi Yan <ziy@nvidia.com> wrote:
>
> When caller does not supply a list to split_huge_page_to_list_to_order(),
> use split_huge_page_to_order() instead.
>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

LGTM,

Reviewed-by: Barry Song <baohua@kernel.org>

> ---
>  include/linux/huge_mm.h | 12 ++++++++++--
>  1 file changed, 10 insertions(+), 2 deletions(-)
>

Re: [PATCH v4 2/3] mm/memory-failure: improve large block size folio handling.

[Barry Song]

On Thu, Oct 30, 2025 at 9:40 AM Zi Yan <ziy@nvidia.com> wrote:
>
> Large block size (LBS) folios cannot be split to order-0 folios but
> min_order_for_folio(). Current split fails directly, but that is not
> optimal. Split the folio to min_order_for_folio(), so that, after split,
> only the folio containing the poisoned page becomes unusable instead.
>
> For soft offline, do not split the large folio if its min_order_for_folio()
> is not 0. Since the folio is still accessible from userspace and premature
> split might lead to potential performance loss.
>
> Suggested-by: Jane Chu <jane.chu@oracle.com>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Reviewed-by: Barry Song <baohua@kernel.org>

> ---
>  mm/memory-failure.c | 31 +++++++++++++++++++++++++++----
>  1 file changed, 27 insertions(+), 4 deletions(-)
>
> diff --git a/mm/memory-failure.c b/mm/memory-failure.c
> index f698df156bf8..acc35c881547 100644
> --- a/mm/memory-failure.c
> +++ b/mm/memory-failure.c
...
> @@ -2294,7 +2298,16 @@ int memory_failure(unsigned long pfn, int flags)
>                  * page is a valid handlable page.
>                  */
>                 folio_set_has_hwpoisoned(folio);
> -               if (try_to_split_thp_page(p, false) < 0) {
> +               err = try_to_split_thp_page(p, new_order, /* release= */ false);
> +               /*
> +                * If splitting a folio to order-0 fails, kill the process.
> +                * Split the folio regardless to minimize unusable pages.
> +                * Because the memory failure code cannot handle large
> +                * folios, this split is always treated as if it failed.
> +                */
> +               if (err || new_order) {
> +                       /* get folio again in case the original one is split */
> +                       folio = page_folio(p);

It’s a bit hard to follow that we implicitly use p to get its original
folio for splitting in try_to_split_thp_page(), and then again use p to
get its new folio for kill_procs_now(). It might be more readable to move
try_to_split_thp_page() into a helper like try_to_split_folio(folio, …),
so it’s explicit that we’re splitting a folio rather than a page?

Thanks
Barry

Re: [PATCH v4 3/3] mm/huge_memory: fix kernel-doc comments for folio_split() and related.

[Barry Song]

On Thu, Oct 30, 2025 at 9:40 AM Zi Yan <ziy@nvidia.com> wrote:
>
> try_folio_split_to_order(), folio_split, __folio_split(), and
> __split_unmapped_folio() do not have correct kernel-doc comment format.
> Fix them.
>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> ---

LGTM,
Reviewed-by: Barry Song <baohua@kernel.org>

Re: [PATCH v4 1/3] mm/huge_memory: add split_huge_page_to_order()

[Miaohe Lin]

On 2025/10/30 9:40, Zi Yan wrote:
> When caller does not supply a list to split_huge_page_to_list_to_order(),
> use split_huge_page_to_order() instead.
> 
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>

Thanks.
.

Re: [PATCH v4 2/3] mm/memory-failure: improve large block size folio handling.

[Miaohe Lin]

On 2025/10/30 9:40, Zi Yan wrote:
> Large block size (LBS) folios cannot be split to order-0 folios but
> min_order_for_folio(). Current split fails directly, but that is not
> optimal. Split the folio to min_order_for_folio(), so that, after split,
> only the folio containing the poisoned page becomes unusable instead.
> 
> For soft offline, do not split the large folio if its min_order_for_folio()
> is not 0. Since the folio is still accessible from userspace and premature
> split might lead to potential performance loss.
> 
> Suggested-by: Jane Chu <jane.chu@oracle.com>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>

Thanks.
.

Re: [PATCH v4 3/3] mm/huge_memory: fix kernel-doc comments for folio_split() and related.

[Miaohe Lin]

On 2025/10/30 9:40, Zi Yan wrote:
> try_folio_split_to_order(), folio_split, __folio_split(), and
> __split_unmapped_folio() do not have correct kernel-doc comment format.
> Fix them.
> 
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> Acked-by: David Hildenbrand <david@redhat.com>

Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>

Thanks.
.

Re: [PATCH v4 1/3] mm/huge_memory: add split_huge_page_to_order()

[Wei Yang]

On Wed, Oct 29, 2025 at 09:40:18PM -0400, Zi Yan wrote:
>When caller does not supply a list to split_huge_page_to_list_to_order(),
>use split_huge_page_to_order() instead.
>
>Signed-off-by: Zi Yan <ziy@nvidia.com>
>Acked-by: David Hildenbrand <david@redhat.com>
>Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Reviewed-by: Wei Yang <richard.weiyang@gmail.com>

-- 
Wei Yang
Help you, Help me

Re: [PATCH v4 2/3] mm/memory-failure: improve large block size folio handling.

[Wei Yang]

On Wed, Oct 29, 2025 at 09:40:19PM -0400, Zi Yan wrote:
>Large block size (LBS) folios cannot be split to order-0 folios but
>min_order_for_folio(). Current split fails directly, but that is not
>optimal. Split the folio to min_order_for_folio(), so that, after split,
>only the folio containing the poisoned page becomes unusable instead.
>
>For soft offline, do not split the large folio if its min_order_for_folio()
>is not 0. Since the folio is still accessible from userspace and premature
>split might lead to potential performance loss.
>
>Suggested-by: Jane Chu <jane.chu@oracle.com>
>Signed-off-by: Zi Yan <ziy@nvidia.com>
>Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
>Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

Looks reasonable.

Reviewed-by: Wei Yang <richard.weiyang@gmail.com>

-- 
Wei Yang
Help you, Help me

Re: [PATCH v4 3/3] mm/huge_memory: fix kernel-doc comments for folio_split() and related.

[Wei Yang]

On Wed, Oct 29, 2025 at 09:40:20PM -0400, Zi Yan wrote:
>try_folio_split_to_order(), folio_split, __folio_split(), and
>__split_unmapped_folio() do not have correct kernel-doc comment format.
>Fix them.
>
>Signed-off-by: Zi Yan <ziy@nvidia.com>
>Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
>Acked-by: David Hildenbrand <david@redhat.com>

Generally looks good, while some nit below.

>---
> include/linux/huge_mm.h | 10 ++++++----
> mm/huge_memory.c        | 27 +++++++++++++++------------
> 2 files changed, 21 insertions(+), 16 deletions(-)
>
>diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
>index 34f8d8453bf3..cbb2243f8e56 100644
>--- a/include/linux/huge_mm.h
>+++ b/include/linux/huge_mm.h
>@@ -386,9 +386,9 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
> 	return split_huge_page_to_list_to_order(page, NULL, new_order);
> }
> 
>-/*
>- * try_folio_split_to_order - try to split a @folio at @page to @new_order using
>- * non uniform split.
>+/**
>+ * try_folio_split_to_order() - try to split a @folio at @page to @new_order
>+ * using non uniform split.

This looks try_folio_split_to_order() only perform non uniform split, while the
following comment mentions it will try uniform split if non uniform split is
not supported. 

Do you think this is a little confusing?

>  * @folio: folio to be split
>  * @page: split to @new_order at the given page
>  * @new_order: the target split order
>@@ -398,7 +398,7 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
>  * folios are put back to LRU list. Use min_order_for_split() to get the lower
>  * bound of @new_order.
>  *
>- * Return: 0: split is successful, otherwise split failed.
>+ * Return: 0 - split is successful, otherwise split failed.
>  */
> static inline int try_folio_split_to_order(struct folio *folio,
> 		struct page *page, unsigned int new_order)
>@@ -486,6 +486,8 @@ static inline spinlock_t *pud_trans_huge_lock(pud_t *pud,
> /**
>  * folio_test_pmd_mappable - Can we map this folio with a PMD?
>  * @folio: The folio to test
>+ *
>+ * Return: true - @folio can be mapped, false - @folio cannot be mapped.
>  */
> static inline bool folio_test_pmd_mappable(struct folio *folio)
> {
>diff --git a/mm/huge_memory.c b/mm/huge_memory.c
>index 0e24bb7e90d0..381a49c5ac3f 100644
>--- a/mm/huge_memory.c
>+++ b/mm/huge_memory.c
>@@ -3567,8 +3567,9 @@ static void __split_folio_to_order(struct folio *folio, int old_order,
> 		ClearPageCompound(&folio->page);
> }
> 
>-/*
>- * It splits an unmapped @folio to lower order smaller folios in two ways.
>+/**
>+ * __split_unmapped_folio() - splits an unmapped @folio to lower order folios in
>+ * two ways: uniform split or non-uniform split.
>  * @folio: the to-be-split folio
>  * @new_order: the smallest order of the after split folios (since buddy
>  *             allocator like split generates folios with orders from @folio's

In the comment of __split_unmapped_folio(), we have some description about the
split behavior, e.g. update stat, unfreeze.

Is this out-dated?

-- 
Wei Yang
Help you, Help me

Re: [PATCH v4 0/3] Optimize folio split in memory failure

[Andrew Morton]

On Wed, 29 Oct 2025 21:40:17 -0400 Zi Yan <ziy@nvidia.com> wrote:

> This patchset is a follow-up of "[PATCH v3] mm/huge_memory: do not change
> split_huge_page*() target order silently."[1] and
> [PATCH v4] mm/huge_memory: preserve PG_has_hwpoisoned if a folio is split
> to >0 order[2], since both are separated out as hotfixes. It improves how
> memory failure code handles large block size(LBS) folios with
> min_order_for_split() > 0. By splitting a large folio containing HW
> poisoned pages to min_order_for_split(), the after-split folios without
> HW poisoned pages could be freed for reuse. To achieve this, folio split
> code needs to set has_hwpoisoned on after-split folios containing HW
> poisoned pages and it is done in the hotfix in [2].
> 
> This patchset includes:
> 1. A patch adds split_huge_page_to_order(),
> 2. Patch 2 and Patch 3 of "[PATCH v2 0/3] Do not change split folio target
>    order"[3],

Sorry, but best I can tell, none of this tells anyone anything about
this patchset!

Could we please have a [0/N] which provides the usual overview of these
three patches?

Please put yourself in the position of someone reading Linus's tree in
2028 wondering "hm, what does this series do".  All this short-term
transient patch-timing development-time stuff is of no interest to
them and is best placed below the ^---$ separator.

Thanks.

Re: [PATCH v4 1/3] mm/huge_memory: add split_huge_page_to_order()

[David Hildenbrand]

On 30.10.25 02:40, Zi Yan wrote:
> When caller does not supply a list to split_huge_page_to_list_to_order(),
> use split_huge_page_to_order() instead.
> 
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> ---
>   include/linux/huge_mm.h | 12 ++++++++++--
>   1 file changed, 10 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
> index 7698b3542c4f..34f8d8453bf3 100644
> --- a/include/linux/huge_mm.h
> +++ b/include/linux/huge_mm.h
> @@ -381,6 +381,10 @@ static inline int split_huge_page_to_list_to_order(struct page *page, struct lis
>   {
>   	return __split_huge_page_to_list_to_order(page, list, new_order, false);
>   }
> +static inline int split_huge_page_to_order(struct page *page, unsigned int new_order)
> +{
> +	return split_huge_page_to_list_to_order(page, NULL, new_order);
> +}
>   

Scanning this once again, I guess in the future all these interfaces 
should rather be folio-based, and if we want to split at a specific page 
where we want the reference to be held later, pass in a page:

int folio_split_to_order(struct folio *folio, struct page *page,
			 unsigned int new_order);

With the hope that we could end up with all folio split functions to
look similar in that regard ... and remove all the "huge_page" terminology.

Of course, that can be done as cleanups on top, because there seems to 
be quite some inconsistency already.


-- 
Cheers

David / dhildenb

Re: [PATCH v4 2/3] mm/memory-failure: improve large block size folio handling.

[David Hildenbrand]

On 30.10.25 02:40, Zi Yan wrote:
> Large block size (LBS) folios cannot be split to order-0 folios but
> min_order_for_folio(). Current split fails directly, but that is not
> optimal. Split the folio to min_order_for_folio(), so that, after split,
> only the folio containing the poisoned page becomes unusable instead.
> 
> For soft offline, do not split the large folio if its min_order_for_folio()
> is not 0. Since the folio is still accessible from userspace and premature
> split might lead to potential performance loss.
> 
> Suggested-by: Jane Chu <jane.chu@oracle.com>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> ---

Acked-by: David Hildenbrand <david@redhat.com>

-- 
Cheers

David / dhildenb

Re: [PATCH v4 0/3] Optimize folio split in memory failure

[Zi Yan]

On 30 Oct 2025, at 23:42, Andrew Morton wrote:

> On Wed, 29 Oct 2025 21:40:17 -0400 Zi Yan <ziy@nvidia.com> wrote:
>
>> This patchset is a follow-up of "[PATCH v3] mm/huge_memory: do not change
>> split_huge_page*() target order silently."[1] and
>> [PATCH v4] mm/huge_memory: preserve PG_has_hwpoisoned if a folio is split
>> to >0 order[2], since both are separated out as hotfixes. It improves how
>> memory failure code handles large block size(LBS) folios with
>> min_order_for_split() > 0. By splitting a large folio containing HW
>> poisoned pages to min_order_for_split(), the after-split folios without
>> HW poisoned pages could be freed for reuse. To achieve this, folio split
>> code needs to set has_hwpoisoned on after-split folios containing HW
>> poisoned pages and it is done in the hotfix in [2].
>>
>> This patchset includes:
>> 1. A patch adds split_huge_page_to_order(),
>> 2. Patch 2 and Patch 3 of "[PATCH v2 0/3] Do not change split folio target
>>    order"[3],
>
> Sorry, but best I can tell, none of this tells anyone anything about
> this patchset!
>
> Could we please have a [0/N] which provides the usual overview of these
> three patches?
>
> Please put yourself in the position of someone reading Linus's tree in
> 2028 wondering "hm, what does this series do".  All this short-term
> transient patch-timing development-time stuff is of no interest to
> them and is best placed below the ^---$ separator.
>

How about?

The patchset optimizes folio split operations in memory failure code by:
always splitting a folio to min_order_for_split() to minimize unusable
pages, even if min_order_for_split() is non zero and memory failure code
would take the failed path eventually. This means instead of making
the entire original folio unusable memory failure code would only make
its after-split folio, with min_order_for_split() and containing the
page marked as HWPoisoned, unusable. For soft offline case, since the
original folio is still accessible, do not split it. In addition,
add split_huge_page_to_order() to improve code readability and fix
kernel-doc comment format for folio_split() and other related functions.


--
Best Regards,
Yan, Zi

Re: [PATCH v4 3/3] mm/huge_memory: fix kernel-doc comments for folio_split() and related.

[Zi Yan]

On 30 Oct 2025, at 22:55, Wei Yang wrote:

> On Wed, Oct 29, 2025 at 09:40:20PM -0400, Zi Yan wrote:
>> try_folio_split_to_order(), folio_split, __folio_split(), and
>> __split_unmapped_folio() do not have correct kernel-doc comment format.
>> Fix them.
>>
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
>> Acked-by: David Hildenbrand <david@redhat.com>
>
> Generally looks good, while some nit below.
>
>> ---
>> include/linux/huge_mm.h | 10 ++++++----
>> mm/huge_memory.c        | 27 +++++++++++++++------------
>> 2 files changed, 21 insertions(+), 16 deletions(-)
>>
>> diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
>> index 34f8d8453bf3..cbb2243f8e56 100644
>> --- a/include/linux/huge_mm.h
>> +++ b/include/linux/huge_mm.h
>> @@ -386,9 +386,9 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
>> 	return split_huge_page_to_list_to_order(page, NULL, new_order);
>> }
>>
>> -/*
>> - * try_folio_split_to_order - try to split a @folio at @page to @new_order using
>> - * non uniform split.
>> +/**
>> + * try_folio_split_to_order() - try to split a @folio at @page to @new_order
>> + * using non uniform split.
>
> This looks try_folio_split_to_order() only perform non uniform split, while the
> following comment mentions it will try uniform split if non uniform split is
> not supported.
>
> Do you think this is a little confusing?

It says "try to", so it is possible that an alternative can be used.

>
>>  * @folio: folio to be split
>>  * @page: split to @new_order at the given page
>>  * @new_order: the target split order
>> @@ -398,7 +398,7 @@ static inline int split_huge_page_to_order(struct page *page, unsigned int new_o
>>  * folios are put back to LRU list. Use min_order_for_split() to get the lower
>>  * bound of @new_order.
>>  *
>> - * Return: 0: split is successful, otherwise split failed.
>> + * Return: 0 - split is successful, otherwise split failed.
>>  */
>> static inline int try_folio_split_to_order(struct folio *folio,
>> 		struct page *page, unsigned int new_order)
>> @@ -486,6 +486,8 @@ static inline spinlock_t *pud_trans_huge_lock(pud_t *pud,
>> /**
>>  * folio_test_pmd_mappable - Can we map this folio with a PMD?
>>  * @folio: The folio to test
>> + *
>> + * Return: true - @folio can be mapped, false - @folio cannot be mapped.
>>  */
>> static inline bool folio_test_pmd_mappable(struct folio *folio)
>> {
>> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
>> index 0e24bb7e90d0..381a49c5ac3f 100644
>> --- a/mm/huge_memory.c
>> +++ b/mm/huge_memory.c
>> @@ -3567,8 +3567,9 @@ static void __split_folio_to_order(struct folio *folio, int old_order,
>> 		ClearPageCompound(&folio->page);
>> }
>>
>> -/*
>> - * It splits an unmapped @folio to lower order smaller folios in two ways.
>> +/**
>> + * __split_unmapped_folio() - splits an unmapped @folio to lower order folios in
>> + * two ways: uniform split or non-uniform split.
>>  * @folio: the to-be-split folio
>>  * @new_order: the smallest order of the after split folios (since buddy
>>  *             allocator like split generates folios with orders from @folio's
>
> In the comment of __split_unmapped_folio(), we have some description about the
> split behavior, e.g. update stat, unfreeze.
>
> Is this out-dated?

OK, I will update it.

--
Best Regards,
Yan, Zi