Skip to contentSkip to conversation

mm: avoid poison consumption when splitting THP

PATCH In review Andrew Zaborowski 4 participants 4 messages Sep 11, 2025

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mm: avoid poison consumption when splitting THP

Handling a memory failure pointing inside a huge page requires splitting the page. The splitting logic uses a mechanism, implemented in migrate.c:try_to_map_unused_to_zeropage(), that inspects contents of individual pages to find zero-filled pages. The read access to the contents may cause a new, synchronous exception like an x86 Machine Check, delivered before the initial memory_failure() finishes, ending in a crash.

Luckily memory_failure() already sets the has_hwpoisoned flag on the folio right before try_to_split_thp_page(). Don't enable the shared zeropage mechanism (RMP_USE_SHARED_ZEROPAGE flag) down in __split_unmapped_folio() when the original folio has has_hwpoisoned.

Note: we're disabling a potentially useful feature, some of the individual pages that aren't poisoned might be zero-filled. One argument for not trying to add a mechanism to maybe re-scan them later, apart from code cost, is that the owning process is likely being killed and the memory released.

---
 mm/huge_memory.c    | 3 ++-
 mm/memory-failure.c | 6 ++++--
 2 files changed, 6 insertions(+), 3 deletions(-)
mm/huge_memory.c +2 -1
index 9c38a95e9f0..1568f0308b9 100644
@@ -3588,6 +3588,7 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
 		struct list_head *list, bool uniform_split)
 {
 	struct deferred_split *ds_queue = get_deferred_split_queue(folio);
+	bool has_hwpoisoned = folio_test_has_hwpoisoned(folio);
 	XA_STATE(xas, &folio->mapping->i_pages, folio->index);
 	struct folio *end_folio = folio_next(folio);
 	bool is_anon = folio_test_anon(folio);
@@ -3858,7 +3859,7 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
 	if (nr_shmem_dropped)
 		shmem_uncharge(mapping->host, nr_shmem_dropped);
 
-	if (!ret && is_anon)
+	if (!ret && is_anon && !has_hwpoisoned)
 		remap_flags = RMP_USE_SHARED_ZEROPAGE;
 	remap_page(folio, 1 << order, remap_flags);
 
mm/memory-failure.c +4 -2
index fc30ca4804b..2d755493de9 100644
@@ -2352,8 +2352,10 @@ int memory_failure(unsigned long pfn, int flags)
 		 * otherwise it may race with THP split.
 		 * And the flag can't be set in get_hwpoison_page() since
 		 * it is called by soft offline too and it is just called
-		 * for !MF_COUNT_INCREASED.  So here seems to be the best
-		 * place.
+		 * for !MF_COUNT_INCREASED.
+		 * It also tells __split_unmapped_folio() to not bother
+		 * using the shared zeropage -- the all-zeros check would
+		 * consume the poison.  So here seems to be the best place.
 		 *
 		 * Don't need care about the above error handling paths for
 		 * get_hwpoison_page() since they handle either free page
Signature · 1 line
2.45.2
reviewer
#1 lore raw
Quoted 13 lines · Andrew Zaborowski

Handling a memory failure pointing inside a huge page requires splitting the page. The splitting logic uses a mechanism, implemented in migrate.c:try_to_map_unused_to_zeropage(), that inspects contents of individual pages to find zero-filled pages. The read access to the contents may cause a new, synchronous exception like an x86 Machine Check, delivered before the initial memory_failure() finishes, ending in a crash.

Luckily memory_failure() already sets the has_hwpoisoned flag on the folio right before try_to_split_thp_page(). Don't enable the shared zeropage mechanism (RMP_USE_SHARED_ZEROPAGE flag) down in __split_unmapped_folio() when the original folio has has_hwpoisoned.

Quoted 5 lines

Note: we're disabling a potentially useful feature, some of the individual pages that aren't poisoned might be zero-filled. One argument for not trying to add a mechanism to maybe re-scan them later, apart from code cost, is that the owning process is likely being killed and the memory released.

Sounds reasonable to me.

Quoted 12 lines
---
 mm/huge_memory.c    | 3 ++-
 mm/memory-failure.c | 6 ++++--
 2 files changed, 6 insertions(+), 3 deletions(-)
mm/huge_memory.c +0 -0
index 9c38a95e9f0..1568f0308b9 100644

mm/huge_memory.c static int __folio_split(struct folio *folio, unsigned int new_order, · line 3591

Quoted 4 lines
mm/huge_memory.c +1 -0
@@ -3588,6 +3588,7 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
 		struct list_head *list, bool uniform_split)
 {
 	struct deferred_split *ds_queue = get_deferred_split_queue(folio);
+	bool has_hwpoisoned = folio_test_has_hwpoisoned(folio);

The state needs to be stored here because __split_unmapped_folio() clears the flag. Maybe add a comment here to prevent people from “optimizing” it by calling folio_test_has_hwpoisoned(folio) in the code below.

(I wanted to until I checked the definition of folio_test_has_hwpoisoned())

Quoted 17 lines
mm/huge_memory.c +1 -1
 	XA_STATE(xas, &folio->mapping->i_pages, folio->index);
 	struct folio *end_folio = folio_next(folio);
 	bool is_anon = folio_test_anon(folio);
@@ -3858,7 +3859,7 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
 	if (nr_shmem_dropped)
 		shmem_uncharge(mapping->host, nr_shmem_dropped);
 
-	if (!ret && is_anon)
+	if (!ret && is_anon && !has_hwpoisoned)
 		remap_flags = RMP_USE_SHARED_ZEROPAGE;
 	remap_page(folio, 1 << order, remap_flags);
 
mm/memory-failure.c +0 -0
index fc30ca4804b..2d755493de9 100644

mm/memory-failure.c int memory_failure(unsigned long pfn, int flags) · line 2356

Quoted 7 lines
mm/memory-failure.c +2 -2
@@ -2352,8 +2352,10 @@ int memory_failure(unsigned long pfn, int flags)
 		 * otherwise it may race with THP split.
 		 * And the flag can't be set in get_hwpoison_page() since
 		 * it is called by soft offline too and it is just called
-		 * for !MF_COUNT_INCREASED.  So here seems to be the best
-		 * place.
+		 * for !MF_COUNT_INCREASED.
+		 * It also tells __split_unmapped_folio() to not bother

s/__split_unmapped_folio/__folio_split/, since remap_page() is called in __folio_split().

Quoted 7 lines
mm/memory-failure.c +2 -0
+		 * using the shared zeropage -- the all-zeros check would
+		 * consume the poison.  So here seems to be the best place.
 		 *
 		 * Don't need care about the above error handling paths for
 		 * get_hwpoison_page() since they handle either free page
Signature · 1 line
2.45.2

Otherwise, Acked-by: Zi Yan <[email protected]>

Best Regards,
Yan, Zi

reviewer
#2 lore raw
Quoted 11 lines · Zi Yan
Quoted 8 lines · Andrew Zaborowski

Handling a memory failure pointing inside a huge page requires splitting the page. The splitting logic uses a mechanism, implemented in migrate.c:try_to_map_unused_to_zeropage(), that inspects contents of individual pages to find zero-filled pages. The read access to the contents may cause a new, synchronous exception like an x86 Machine Check, delivered before the initial memory_failure() finishes, ending in a crash.

Quoted 4 lines
Quoted 4 lines

Luckily memory_failure() already sets the has_hwpoisoned flag on the folio right before try_to_split_thp_page(). Don't enable the shared zeropage mechanism (RMP_USE_SHARED_ZEROPAGE flag) down in __split_unmapped_folio() when the original folio has has_hwpoisoned.

Nit: s/__split_unmapped_folio/__folio_split/

As Zi mentioned, remap_page() is called in __folio_split() ;)

Quoted 8 lines
Quoted 6 lines

Note: we're disabling a potentially useful feature, some of the individual pages that aren't poisoned might be zero-filled. One argument for not trying to add a mechanism to maybe re-scan them later, apart from code cost, is that the owning process is likely being killed and the memory released.

Sounds reasonable to me.

Makes sense to me as well!

Quoted 23 lines
Quoted 13 lines
---
 mm/huge_memory.c    | 3 ++-
 mm/memory-failure.c | 6 ++++--
 2 files changed, 6 insertions(+), 3 deletions(-)
mm/huge_memory.c +0 -0
index 9c38a95e9f0..1568f0308b9 100644

mm/huge_memory.c static int __folio_split(struct folio *folio, unsigned int new_order, · line 3591

Quoted 4 lines
mm/huge_memory.c +1 -0
@@ -3588,6 +3588,7 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
              struct list_head *list, bool uniform_split)
 {
      struct deferred_split *ds_queue = get_deferred_split_queue(folio);
+     bool has_hwpoisoned = folio_test_has_hwpoisoned(folio);

The state needs to be stored here because __split_unmapped_folio() clears the flag. Maybe add a comment here to prevent people from “optimizing” it by calling folio_test_has_hwpoisoned(folio) in the code below.

Quoted 1 line

(I wanted to until I checked the definition of folio_test_has_hwpoisoned())

folio_test_has_hwpoisoned() requires a large folio. That is safe in this context, since this path is only ever called for large folios.

Cheers,
Lance

Quoted 42 lines
Quoted 18 lines
mm/huge_memory.c +1 -1
      XA_STATE(xas, &folio->mapping->i_pages, folio->index);
      struct folio *end_folio = folio_next(folio);
      bool is_anon = folio_test_anon(folio);
@@ -3858,7 +3859,7 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
      if (nr_shmem_dropped)
              shmem_uncharge(mapping->host, nr_shmem_dropped);
 
-     if (!ret && is_anon)
+     if (!ret && is_anon && !has_hwpoisoned)
              remap_flags = RMP_USE_SHARED_ZEROPAGE;
      remap_page(folio, 1 << order, remap_flags);
 
mm/memory-failure.c +0 -0
index fc30ca4804b..2d755493de9 100644

mm/memory-failure.c int memory_failure(unsigned long pfn, int flags) · line 2356

Quoted 7 lines
mm/memory-failure.c +2 -2
@@ -2352,8 +2352,10 @@ int memory_failure(unsigned long pfn, int flags)
               * otherwise it may race with THP split.
               * And the flag can't be set in get_hwpoison_page() since
               * it is called by soft offline too and it is just called
-              * for !MF_COUNT_INCREASED.  So here seems to be the best
-              * place.
+              * for !MF_COUNT_INCREASED.
+              * It also tells __split_unmapped_folio() to not bother

s/__split_unmapped_folio/__folio_split/, since remap_page() is called in __folio_split().

Quoted 7 lines
mm/memory-failure.c +2 -1
+              * using the shared zeropage -- the all-zeros check would
+              * consume the poison.  So here seems to be the best place.
               *
               * Don't need care about the above error handling paths for
               * get_hwpoison_page() since they handle either free page
--

2.45.2

Otherwise, Acked-by: Zi Yan <[email protected]>

Best Regards,
Yan, Zi

#3 lore raw
In reply to Andrew Zaborowski
Quoted 13 lines · Andrew Zaborowski

Handling a memory failure pointing inside a huge page requires splitting the page. The splitting logic uses a mechanism, implemented in migrate.c:try_to_map_unused_to_zeropage(), that inspects contents of individual pages to find zero-filled pages. The read access to the contents may cause a new, synchronous exception like an x86 Machine Check, delivered before the initial memory_failure() finishes, ending in a crash.

Luckily memory_failure() already sets the has_hwpoisoned flag on the folio right before try_to_split_thp_page(). Don't enable the shared zeropage mechanism (RMP_USE_SHARED_ZEROPAGE flag) down in __split_unmapped_folio() when the original folio has has_hwpoisoned.

Quoted 8 lines

Note: we're disabling a potentially useful feature, some of the individual pages that aren't poisoned might be zero-filled. One argument for not trying to add a mechanism to maybe re-scan them later, apart from code cost, is that the owning process is likely being killed and the memory released.

---

I would suggest just checking whether the page (PageHWPoison()) is poisoned before doing the check for zero. If set, just treat it as non-zero.

No need to stop the split.

You'll have to do that in two locations.

No need to mess with RMP_USE_SHARED_ZEROPAGE

Signature · 3 lines
Cheers

David / dhildenb