[PATCH rfc v2 00/10] mm: convert mm counter to take a folio

[PATCH rfc v2 00/10] mm: convert mm counter to take a folio

[Kefeng Wang]

Make all mm_counter() and mm_counter_file() callers to use a folio,
then convert mm counter functions to take a folio, which saves lots
of compound_head() calls.

Kefeng Wang (10):
  mm: swap: introduce pfn_swap_entry_to_folio()
  s390: pgtable: use a folio in ptep_zap_swap_entry()
  mm: huge_memory: use a folio in __split_huge_pmd_locked()
  mm: huge_memory: use a folio in zap_huge_pmd()
  mm: memory: use a folio in copy_nonpresent_pte()
  mm: memory: use a folio in zap_pte_range()
  mm: memory: use a folio in do_set_pmd()
  mm: memory: use a folio in insert_page_into_pte_locked()
  mm: convert mm_counter() to take a folio
  mm: convert mm_counter_file() to take a folio

 arch/s390/mm/pgtable.c  |  4 +-
 include/linux/mm.h      | 12 +++---
 include/linux/swapops.h | 13 +++++++
 kernel/events/uprobes.c |  2 +-
 mm/huge_memory.c        | 34 +++++++++--------
 mm/khugepaged.c         |  4 +-
 mm/memory.c             | 81 +++++++++++++++++++++++------------------
 mm/rmap.c               | 10 ++---
 mm/userfaultfd.c        |  2 +-
 9 files changed, 94 insertions(+), 68 deletions(-)

-- 
2.27.0

[PATCH v2 01/10] mm: swap: introduce pfn_swap_entry_to_folio()

[Kefeng Wang]

Introduce a new pfn_swap_entry_to_folio(), it is similar to
pfn_swap_entry_to_page(), but return a folio, which allow us
to completely replace the struct page variables with struct
folio variables.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 include/linux/swapops.h | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/include/linux/swapops.h b/include/linux/swapops.h
index bff1e8d97de0..85cb84e4be95 100644
--- a/include/linux/swapops.h
+++ b/include/linux/swapops.h
@@ -468,6 +468,19 @@ static inline struct page *pfn_swap_entry_to_page(swp_entry_t entry)
 	return p;
 }
 
+static inline struct folio *pfn_swap_entry_to_folio(swp_entry_t entry)
+{
+	struct folio *folio = pfn_folio(swp_offset_pfn(entry));
+
+	/*
+	 * Any use of migration entries may only occur while the
+	 * corresponding folio is locked
+	 */
+	BUG_ON(is_migration_entry(entry) && !folio_test_locked(folio));
+
+	return folio;
+}
+
 /*
  * A pfn swap entry is a special type of swap entry that always has a pfn stored
  * in the swap offset. They are used to represent unaddressable device memory
-- 
2.27.0

[PATCH v2 02/10] s390: pgtable: use a folio in ptep_zap_swap_entry()

[Kefeng Wang]

Use a folio in ptep_zap_swap_entry(), which is preparetion for
converting mm counter functions to take a folio.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 arch/s390/mm/pgtable.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
index 3bd2ab2a9a34..2f946b493fff 100644
--- a/arch/s390/mm/pgtable.c
+++ b/arch/s390/mm/pgtable.c
@@ -730,9 +730,9 @@ static void ptep_zap_swap_entry(struct mm_struct *mm, swp_entry_t entry)
 	if (!non_swap_entry(entry))
 		dec_mm_counter(mm, MM_SWAPENTS);
 	else if (is_migration_entry(entry)) {
-		struct page *page = pfn_swap_entry_to_page(entry);
+		struct folio *folio = pfn_swap_entry_to_folio(entry);
 
-		dec_mm_counter(mm, mm_counter(page));
+		dec_mm_counter(mm, mm_counter(&folio->page));
 	}
 	free_swap_and_cache(entry);
 }
-- 
2.27.0

[PATCH v2 03/10] mm: huge_memory: use a folio in __split_huge_pmd_locked()

[Kefeng Wang]

Use a folio in __split_huge_pmd_locked() which replaces five
compound_head() call with two page_folio() calls.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 mm/huge_memory.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index f31f02472396..34dd01970927 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -2117,6 +2117,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
 	count_vm_event(THP_SPLIT_PMD);
 
 	if (!vma_is_anonymous(vma)) {
+		struct folio *folio;
 		old_pmd = pmdp_huge_clear_flush(vma, haddr, pmd);
 		/*
 		 * We are going to unmap this huge page. So
@@ -2130,17 +2131,17 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
 			swp_entry_t entry;
 
 			entry = pmd_to_swp_entry(old_pmd);
-			page = pfn_swap_entry_to_page(entry);
+			folio = pfn_swap_entry_to_folio(entry);
 		} else {
-			page = pmd_page(old_pmd);
-			if (!PageDirty(page) && pmd_dirty(old_pmd))
-				set_page_dirty(page);
-			if (!PageReferenced(page) && pmd_young(old_pmd))
-				SetPageReferenced(page);
-			page_remove_rmap(page, vma, true);
-			put_page(page);
+			folio = page_folio(pmd_page(old_pmd));
+			if (!folio_test_dirty(folio) && pmd_dirty(old_pmd))
+				folio_set_dirty(folio);
+			if (!folio_test_referenced(folio) && pmd_young(old_pmd))
+				folio_set_referenced(folio);
+			page_remove_rmap(&folio->page, vma, true);
+			folio_put(folio);
 		}
-		add_mm_counter(mm, mm_counter_file(page), -HPAGE_PMD_NR);
+		add_mm_counter(mm, mm_counter_file(&folio->page), -HPAGE_PMD_NR);
 		return;
 	}
 
-- 
2.27.0

[PATCH v2 04/10] mm: huge_memory: use a folio in zap_huge_pmd()

[Kefeng Wang]

Use a folio in zap_huge_pmd(), which is a preparetion for
converting mm counter functions to take a folio.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 mm/huge_memory.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 34dd01970927..78a00fe22c2d 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1716,11 +1716,13 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 		zap_deposited_table(tlb->mm, pmd);
 		spin_unlock(ptl);
 	} else {
-		struct page *page = NULL;
+		struct folio *folio = NULL;
 		int flush_needed = 1;
 
 		if (pmd_present(orig_pmd)) {
-			page = pmd_page(orig_pmd);
+			struct page *page = pmd_page(orig_pmd);
+
+			folio = page_folio(page);
 			page_remove_rmap(page, vma, true);
 			VM_BUG_ON_PAGE(page_mapcount(page) < 0, page);
 			VM_BUG_ON_PAGE(!PageHead(page), page);
@@ -1729,12 +1731,12 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 
 			VM_BUG_ON(!is_pmd_migration_entry(orig_pmd));
 			entry = pmd_to_swp_entry(orig_pmd);
-			page = pfn_swap_entry_to_page(entry);
+			folio = pfn_swap_entry_to_folio(entry);
 			flush_needed = 0;
 		} else
 			WARN_ONCE(1, "Non present huge pmd without pmd migration enabled!");
 
-		if (PageAnon(page)) {
+		if (folio_test_anon(folio)) {
 			zap_deposited_table(tlb->mm, pmd);
 			add_mm_counter(tlb->mm, MM_ANONPAGES, -HPAGE_PMD_NR);
 		} else {
@@ -1745,7 +1747,7 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 
 		spin_unlock(ptl);
 		if (flush_needed)
-			tlb_remove_page_size(tlb, page, HPAGE_PMD_SIZE);
+			tlb_remove_page_size(tlb, &folio->page, HPAGE_PMD_SIZE);
 	}
 	return 1;
 }
-- 
2.27.0

[PATCH v2 05/10] mm: memory: use a folio in copy_nonpresent_pte()

[Kefeng Wang]

Use a folio in copy_nonpresent_pte() to save one compound_head() call.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 mm/memory.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 1f18ed4a5497..d9314dee355e 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -779,7 +779,7 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 	unsigned long vm_flags = dst_vma->vm_flags;
 	pte_t orig_pte = ptep_get(src_pte);
 	pte_t pte = orig_pte;
-	struct page *page;
+	struct folio *folio;
 	swp_entry_t entry = pte_to_swp_entry(orig_pte);
 
 	if (likely(!non_swap_entry(entry))) {
@@ -801,9 +801,9 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 		}
 		rss[MM_SWAPENTS]++;
 	} else if (is_migration_entry(entry)) {
-		page = pfn_swap_entry_to_page(entry);
+		folio = pfn_swap_entry_to_folio(entry);
 
-		rss[mm_counter(page)]++;
+		rss[mm_counter(&folio->page)]++;
 
 		if (!is_readable_migration_entry(entry) &&
 				is_cow_mapping(vm_flags)) {
@@ -822,7 +822,7 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 			set_pte_at(src_mm, addr, src_pte, pte);
 		}
 	} else if (is_device_private_entry(entry)) {
-		page = pfn_swap_entry_to_page(entry);
+		folio = pfn_swap_entry_to_folio(entry);
 
 		/*
 		 * Update rss count even for unaddressable pages, as
@@ -833,10 +833,10 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 		 * for unaddressable pages, at some point. But for now
 		 * keep things as they are.
 		 */
-		get_page(page);
-		rss[mm_counter(page)]++;
+		folio_get(folio);
+		rss[mm_counter(&folio->page)]++;
 		/* Cannot fail as these pages cannot get pinned. */
-		BUG_ON(page_try_dup_anon_rmap(page, false, src_vma));
+		BUG_ON(page_try_dup_anon_rmap(&folio->page, false, src_vma));
 
 		/*
 		 * We do not preserve soft-dirty information, because so
-- 
2.27.0

[PATCH v2 06/10] mm: memory: use a folio in zap_pte_range()

[Kefeng Wang]

Make should_zap_page() to take a folio and use a folio in
zap_pte_range(), which save several compound_head() calls.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 mm/memory.c | 45 +++++++++++++++++++++++++--------------------
 1 file changed, 25 insertions(+), 20 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index d9314dee355e..806568f9605b 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -1358,19 +1358,19 @@ static inline bool should_zap_cows(struct zap_details *details)
 	return details->even_cows;
 }
 
-/* Decides whether we should zap this page with the page pointer specified */
-static inline bool should_zap_page(struct zap_details *details, struct page *page)
+/* Decides whether we should zap this folio with the folio pointer specified */
+static inline bool should_zap_page(struct zap_details *details, struct folio *folio)
 {
-	/* If we can make a decision without *page.. */
+	/* If we can make a decision without *folio.. */
 	if (should_zap_cows(details))
 		return true;
 
-	/* E.g. the caller passes NULL for the case of a zero page */
-	if (!page)
+	/* E.g. the caller passes NULL for the case of a zero folio */
+	if (!folio)
 		return true;
 
-	/* Otherwise we should only zap non-anon pages */
-	return !PageAnon(page);
+	/* Otherwise we should only zap non-anon folios */
+	return !folio_test_anon(folio);
 }
 
 static inline bool zap_drop_file_uffd_wp(struct zap_details *details)
@@ -1423,7 +1423,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	arch_enter_lazy_mmu_mode();
 	do {
 		pte_t ptent = ptep_get(pte);
-		struct page *page;
+		struct folio *folio = NULL;
 
 		if (pte_none(ptent))
 			continue;
@@ -1433,9 +1433,13 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 
 		if (pte_present(ptent)) {
 			unsigned int delay_rmap;
+			struct page *page;
 
 			page = vm_normal_page(vma, addr, ptent);
-			if (unlikely(!should_zap_page(details, page)))
+			if (page)
+				folio = page_folio(page);
+
+			if (unlikely(!should_zap_page(details, folio)))
 				continue;
 			ptent = ptep_get_and_clear_full(mm, addr, pte,
 							tlb->fullmm);
@@ -1449,16 +1453,16 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			}
 
 			delay_rmap = 0;
-			if (!PageAnon(page)) {
+			if (!folio_test_anon(folio)) {
 				if (pte_dirty(ptent)) {
-					set_page_dirty(page);
+					folio_set_dirty(folio);
 					if (tlb_delay_rmap(tlb)) {
 						delay_rmap = 1;
 						force_flush = 1;
 					}
 				}
 				if (pte_young(ptent) && likely(vma_has_recency(vma)))
-					mark_page_accessed(page);
+					folio_mark_accessed(folio);
 			}
 			rss[mm_counter(page)]--;
 			if (!delay_rmap) {
@@ -1477,9 +1481,10 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 		entry = pte_to_swp_entry(ptent);
 		if (is_device_private_entry(entry) ||
 		    is_device_exclusive_entry(entry)) {
-			page = pfn_swap_entry_to_page(entry);
-			if (unlikely(!should_zap_page(details, page)))
+			folio = pfn_swap_entry_to_folio(entry);
+			if (unlikely(!should_zap_page(details, folio)))
 				continue;
+
 			/*
 			 * Both device private/exclusive mappings should only
 			 * work with anonymous page so far, so we don't need to
@@ -1487,10 +1492,10 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			 * see zap_install_uffd_wp_if_needed().
 			 */
 			WARN_ON_ONCE(!vma_is_anonymous(vma));
-			rss[mm_counter(page)]--;
+			rss[mm_counter(&folio->page)]--;
 			if (is_device_private_entry(entry))
-				page_remove_rmap(page, vma, false);
-			put_page(page);
+				page_remove_rmap(&folio->page, vma, false);
+			folio_put(folio);
 		} else if (!non_swap_entry(entry)) {
 			/* Genuine swap entry, hence a private anon page */
 			if (!should_zap_cows(details))
@@ -1499,10 +1504,10 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			if (unlikely(!free_swap_and_cache(entry)))
 				print_bad_pte(vma, addr, ptent, NULL);
 		} else if (is_migration_entry(entry)) {
-			page = pfn_swap_entry_to_page(entry);
-			if (!should_zap_page(details, page))
+			folio = pfn_swap_entry_to_folio(entry);
+			if (!should_zap_page(details, folio))
 				continue;
-			rss[mm_counter(page)]--;
+			rss[mm_counter(&folio->page)]--;
 		} else if (pte_marker_entry_uffd_wp(entry)) {
 			/*
 			 * For anon: always drop the marker; for file: only
-- 
2.27.0

[PATCH v2 07/10] mm: memory: use a folio in do_set_pmd()

[Kefeng Wang]

Use a folio in do_set_pmd(), which is a preparetion for
converting mm counter functions to take a folio.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 mm/memory.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 806568f9605b..ac247850919a 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -4318,12 +4318,13 @@ vm_fault_t do_set_pmd(struct vm_fault *vmf, struct page *page)
 	unsigned long haddr = vmf->address & HPAGE_PMD_MASK;
 	pmd_t entry;
 	vm_fault_t ret = VM_FAULT_FALLBACK;
+	struct folio *folio;
 
 	if (!transhuge_vma_suitable(vma, haddr))
 		return ret;
 
-	page = compound_head(page);
-	if (compound_order(page) != HPAGE_PMD_ORDER)
+	folio = page_folio(page);
+	if (folio_order(folio) != HPAGE_PMD_ORDER)
 		return ret;
 
 	/*
-- 
2.27.0

[PATCH v2 08/10] mm: memory: use a folio in insert_page_into_pte_locked()

[Kefeng Wang]

Use a folio in insert_page_into_pte_locked(), which is preparetion
for converting mm counter functions to take a folio.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 mm/memory.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/mm/memory.c b/mm/memory.c
index ac247850919a..a2cf240b1975 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -1850,10 +1850,13 @@ static int validate_page_before_insert(struct page *page)
 static int insert_page_into_pte_locked(struct vm_area_struct *vma, pte_t *pte,
 			unsigned long addr, struct page *page, pgprot_t prot)
 {
+	struct folio *folio;
+
 	if (!pte_none(ptep_get(pte)))
 		return -EBUSY;
+	folio = page_folio(page);
 	/* Ok, finally just insert the thing.. */
-	get_page(page);
+	folio_get(folio);
 	inc_mm_counter(vma->vm_mm, mm_counter_file(page));
 	page_add_file_rmap(page, vma, false);
 	set_pte_at(vma->vm_mm, addr, pte, mk_pte(page, prot));
-- 
2.27.0

[PATCH v2 09/10] mm: convert mm_counter() to take a folio

[Kefeng Wang]

Since all mm_counter() callers with a folio, let's convert
mm_counter() to take a folio.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 arch/s390/mm/pgtable.c |  2 +-
 include/linux/mm.h     |  6 +++---
 mm/memory.c            | 10 +++++-----
 mm/rmap.c              |  8 ++++----
 mm/userfaultfd.c       |  2 +-
 5 files changed, 14 insertions(+), 14 deletions(-)

diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
index 2f946b493fff..54b184648db6 100644
--- a/arch/s390/mm/pgtable.c
+++ b/arch/s390/mm/pgtable.c
@@ -732,7 +732,7 @@ static void ptep_zap_swap_entry(struct mm_struct *mm, swp_entry_t entry)
 	else if (is_migration_entry(entry)) {
 		struct folio *folio = pfn_swap_entry_to_folio(entry);
 
-		dec_mm_counter(mm, mm_counter(&folio->page));
+		dec_mm_counter(mm, mm_counter(folio));
 	}
 	free_swap_and_cache(entry);
 }
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 418d26608ece..fea78900bf84 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2591,11 +2591,11 @@ static inline int mm_counter_file(struct page *page)
 	return MM_FILEPAGES;
 }
 
-static inline int mm_counter(struct page *page)
+static inline int mm_counter(struct folio *folio)
 {
-	if (PageAnon(page))
+	if (folio_test_anon(folio))
 		return MM_ANONPAGES;
-	return mm_counter_file(page);
+	return mm_counter_file(&folio->page);
 }
 
 static inline unsigned long get_mm_rss(struct mm_struct *mm)
diff --git a/mm/memory.c b/mm/memory.c
index a2cf240b1975..ad30d4ad2223 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -803,7 +803,7 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 	} else if (is_migration_entry(entry)) {
 		folio = pfn_swap_entry_to_folio(entry);
 
-		rss[mm_counter(&folio->page)]++;
+		rss[mm_counter(folio)]++;
 
 		if (!is_readable_migration_entry(entry) &&
 				is_cow_mapping(vm_flags)) {
@@ -834,7 +834,7 @@ copy_nonpresent_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
 		 * keep things as they are.
 		 */
 		folio_get(folio);
-		rss[mm_counter(&folio->page)]++;
+		rss[mm_counter(folio)]++;
 		/* Cannot fail as these pages cannot get pinned. */
 		BUG_ON(page_try_dup_anon_rmap(&folio->page, false, src_vma));
 
@@ -1464,7 +1464,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				if (pte_young(ptent) && likely(vma_has_recency(vma)))
 					folio_mark_accessed(folio);
 			}
-			rss[mm_counter(page)]--;
+			rss[mm_counter(folio)]--;
 			if (!delay_rmap) {
 				page_remove_rmap(page, vma, false);
 				if (unlikely(page_mapcount(page) < 0))
@@ -1492,7 +1492,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			 * see zap_install_uffd_wp_if_needed().
 			 */
 			WARN_ON_ONCE(!vma_is_anonymous(vma));
-			rss[mm_counter(&folio->page)]--;
+			rss[mm_counter(folio)]--;
 			if (is_device_private_entry(entry))
 				page_remove_rmap(&folio->page, vma, false);
 			folio_put(folio);
@@ -1507,7 +1507,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			folio = pfn_swap_entry_to_folio(entry);
 			if (!should_zap_page(details, folio))
 				continue;
-			rss[mm_counter(&folio->page)]--;
+			rss[mm_counter(folio)]--;
 		} else if (pte_marker_entry_uffd_wp(entry)) {
 			/*
 			 * For anon: always drop the marker; for file: only
diff --git a/mm/rmap.c b/mm/rmap.c
index 7a27a2b41802..7a563490ce08 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1678,7 +1678,7 @@ static bool try_to_unmap_one(struct folio *folio, struct vm_area_struct *vma,
 				set_huge_pte_at(mm, address, pvmw.pte, pteval,
 						hsz);
 			} else {
-				dec_mm_counter(mm, mm_counter(&folio->page));
+				dec_mm_counter(mm, mm_counter(folio));
 				set_pte_at(mm, address, pvmw.pte, pteval);
 			}
 
@@ -1693,7 +1693,7 @@ static bool try_to_unmap_one(struct folio *folio, struct vm_area_struct *vma,
 			 * migration) will not expect userfaults on already
 			 * copied pages.
 			 */
-			dec_mm_counter(mm, mm_counter(&folio->page));
+			dec_mm_counter(mm, mm_counter(folio));
 		} else if (folio_test_anon(folio)) {
 			swp_entry_t entry = page_swap_entry(subpage);
 			pte_t swp_pte;
@@ -2075,7 +2075,7 @@ static bool try_to_migrate_one(struct folio *folio, struct vm_area_struct *vma,
 				set_huge_pte_at(mm, address, pvmw.pte, pteval,
 						hsz);
 			} else {
-				dec_mm_counter(mm, mm_counter(&folio->page));
+				dec_mm_counter(mm, mm_counter(folio));
 				set_pte_at(mm, address, pvmw.pte, pteval);
 			}
 
@@ -2090,7 +2090,7 @@ static bool try_to_migrate_one(struct folio *folio, struct vm_area_struct *vma,
 			 * migration) will not expect userfaults on already
 			 * copied pages.
 			 */
-			dec_mm_counter(mm, mm_counter(&folio->page));
+			dec_mm_counter(mm, mm_counter(folio));
 		} else {
 			swp_entry_t entry;
 			pte_t swp_pte;
diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c
index 96d9eae5c7cc..9a6759fa9b06 100644
--- a/mm/userfaultfd.c
+++ b/mm/userfaultfd.c
@@ -124,7 +124,7 @@ int mfill_atomic_install_pte(pmd_t *dst_pmd,
 	 * Must happen after rmap, as mm_counter() checks mapping (via
 	 * PageAnon()), which is set by __page_set_anon_rmap().
 	 */
-	inc_mm_counter(dst_mm, mm_counter(page));
+	inc_mm_counter(dst_mm, mm_counter(folio));
 
 	set_pte_at(dst_mm, dst_addr, dst_pte, _dst_pte);
 
-- 
2.27.0

[PATCH v2 10/10] mm: convert mm_counter_file() to take a folio

[Kefeng Wang]

Since all mm_counter_file() callers with a folio, let's convert
mm_counter_file() to take a folio.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
 include/linux/mm.h      |  8 ++++----
 kernel/events/uprobes.c |  2 +-
 mm/huge_memory.c        |  5 +++--
 mm/khugepaged.c         |  4 ++--
 mm/memory.c             | 10 +++++-----
 mm/rmap.c               |  2 +-
 6 files changed, 16 insertions(+), 15 deletions(-)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index fea78900bf84..95573065a46b 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2583,10 +2583,10 @@ static inline void dec_mm_counter(struct mm_struct *mm, int member)
 	mm_trace_rss_stat(mm, member);
 }
 
-/* Optimized variant when page is already known not to be PageAnon */
-static inline int mm_counter_file(struct page *page)
+/* Optimized variant when folio is already known not to be anon */
+static inline int mm_counter_file(struct folio *folio)
 {
-	if (PageSwapBacked(page))
+	if (folio_test_swapbacked(folio))
 		return MM_SHMEMPAGES;
 	return MM_FILEPAGES;
 }
@@ -2595,7 +2595,7 @@ static inline int mm_counter(struct folio *folio)
 {
 	if (folio_test_anon(folio))
 		return MM_ANONPAGES;
-	return mm_counter_file(&folio->page);
+	return mm_counter_file(folio);
 }
 
 static inline unsigned long get_mm_rss(struct mm_struct *mm)
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 435aac1d8c27..ce251e3a4ae6 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -188,7 +188,7 @@ static int __replace_page(struct vm_area_struct *vma, unsigned long addr,
 		dec_mm_counter(mm, MM_ANONPAGES);
 
 	if (!folio_test_anon(old_folio)) {
-		dec_mm_counter(mm, mm_counter_file(old_page));
+		dec_mm_counter(mm, mm_counter_file(old_folio));
 		inc_mm_counter(mm, MM_ANONPAGES);
 	}
 
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 78a00fe22c2d..88420d067477 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1742,7 +1742,8 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 		} else {
 			if (arch_needs_pgtable_deposit())
 				zap_deposited_table(tlb->mm, pmd);
-			add_mm_counter(tlb->mm, mm_counter_file(page), -HPAGE_PMD_NR);
+			add_mm_counter(tlb->mm, mm_counter_file(folio),
+				       -HPAGE_PMD_NR);
 		}
 
 		spin_unlock(ptl);
@@ -2143,7 +2144,7 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd,
 			page_remove_rmap(&folio->page, vma, true);
 			folio_put(folio);
 		}
-		add_mm_counter(mm, mm_counter_file(&folio->page), -HPAGE_PMD_NR);
+		add_mm_counter(mm, mm_counter_file(folio), -HPAGE_PMD_NR);
 		return;
 	}
 
diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index 064654717843..39393f4262b2 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -1630,7 +1630,7 @@ int collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr,
 	/* step 3: set proper refcount and mm_counters. */
 	if (nr_ptes) {
 		folio_ref_sub(folio, nr_ptes);
-		add_mm_counter(mm, mm_counter_file(&folio->page), -nr_ptes);
+		add_mm_counter(mm, mm_counter_file(folio), -nr_ptes);
 	}
 
 	/* step 4: remove empty page table */
@@ -1661,7 +1661,7 @@ int collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr,
 	if (nr_ptes) {
 		flush_tlb_mm(mm);
 		folio_ref_sub(folio, nr_ptes);
-		add_mm_counter(mm, mm_counter_file(&folio->page), -nr_ptes);
+		add_mm_counter(mm, mm_counter_file(folio), -nr_ptes);
 	}
 	if (start_pte)
 		pte_unmap_unlock(start_pte, ptl);
diff --git a/mm/memory.c b/mm/memory.c
index ad30d4ad2223..3418ace5e0ad 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -960,7 +960,7 @@ copy_present_pte(struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma,
 	} else if (page) {
 		folio_get(folio);
 		page_dup_file_rmap(page, false);
-		rss[mm_counter_file(page)]++;
+		rss[mm_counter_file(folio)]++;
 	}
 
 	/*
@@ -1857,7 +1857,7 @@ static int insert_page_into_pte_locked(struct vm_area_struct *vma, pte_t *pte,
 	folio = page_folio(page);
 	/* Ok, finally just insert the thing.. */
 	folio_get(folio);
-	inc_mm_counter(vma->vm_mm, mm_counter_file(page));
+	inc_mm_counter(vma->vm_mm, mm_counter_file(folio));
 	page_add_file_rmap(page, vma, false);
 	set_pte_at(vma->vm_mm, addr, pte, mk_pte(page, prot));
 	return 0;
@@ -3166,7 +3166,7 @@ static vm_fault_t wp_page_copy(struct vm_fault *vmf)
 	if (likely(vmf->pte && pte_same(ptep_get(vmf->pte), vmf->orig_pte))) {
 		if (old_folio) {
 			if (!folio_test_anon(old_folio)) {
-				dec_mm_counter(mm, mm_counter_file(&old_folio->page));
+				dec_mm_counter(mm, mm_counter_file(old_folio));
 				inc_mm_counter(mm, MM_ANONPAGES);
 			}
 		} else {
@@ -4359,7 +4359,7 @@ vm_fault_t do_set_pmd(struct vm_fault *vmf, struct page *page)
 	if (write)
 		entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma);
 
-	add_mm_counter(vma->vm_mm, mm_counter_file(page), HPAGE_PMD_NR);
+	add_mm_counter(vma->vm_mm, mm_counter_file(folio), HPAGE_PMD_NR);
 	page_add_file_rmap(page, vma, true);
 
 	/*
@@ -4422,7 +4422,7 @@ void set_pte_range(struct vm_fault *vmf, struct folio *folio,
 		folio_add_new_anon_rmap(folio, vma, addr);
 		folio_add_lru_vma(folio, vma);
 	} else {
-		add_mm_counter(vma->vm_mm, mm_counter_file(page), nr);
+		add_mm_counter(vma->vm_mm, mm_counter_file(folio), nr);
 		folio_add_file_rmap_range(folio, page, nr, vma, false);
 	}
 	set_ptes(vma->vm_mm, addr, vmf->pte, entry, nr);
diff --git a/mm/rmap.c b/mm/rmap.c
index 7a563490ce08..9e3d0eff8b05 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1801,7 +1801,7 @@ static bool try_to_unmap_one(struct folio *folio, struct vm_area_struct *vma,
 			 *
 			 * See Documentation/mm/mmu_notifier.rst
 			 */
-			dec_mm_counter(mm, mm_counter_file(&folio->page));
+			dec_mm_counter(mm, mm_counter_file(folio));
 		}
 discard:
 		page_remove_rmap(subpage, vma, folio_test_hugetlb(folio));
-- 
2.27.0

Re: [PATCH v2 06/10] mm: memory: use a folio in zap_pte_range()

[Matthew Wilcox]

On Sat, Nov 04, 2023 at 11:55:18AM +0800, Kefeng Wang wrote:
> -/* Decides whether we should zap this page with the page pointer specified */
> -static inline bool should_zap_page(struct zap_details *details, struct page *page)
> +/* Decides whether we should zap this folio with the folio pointer specified */
> +static inline bool should_zap_page(struct zap_details *details, struct folio *folio)

Surely we should rename this to should_zap_folio()?

> @@ -1487,10 +1492,10 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
>  			 * see zap_install_uffd_wp_if_needed().
>  			 */
>  			WARN_ON_ONCE(!vma_is_anonymous(vma));
> -			rss[mm_counter(page)]--;
> +			rss[mm_counter(&folio->page)]--;
>  			if (is_device_private_entry(entry))
> -				page_remove_rmap(page, vma, false);
> -			put_page(page);
> +				page_remove_rmap(&folio->page, vma, false);
> +			folio_put(folio);

This is wrong.  If we have a PTE-mapped THP, you'll remove the head page
N times instead of removing each of N pages.

I suspect you're going to collide with Ryan's work by doing this ...

Re: [PATCH v2 06/10] mm: memory: use a folio in zap_pte_range()

[Kefeng Wang]

On 2023/11/5 1:20, Matthew Wilcox wrote:
> On Sat, Nov 04, 2023 at 11:55:18AM +0800, Kefeng Wang wrote:
>> -/* Decides whether we should zap this page with the page pointer specified */
>> -static inline bool should_zap_page(struct zap_details *details, struct page *page)
>> +/* Decides whether we should zap this folio with the folio pointer specified */
>> +static inline bool should_zap_page(struct zap_details *details, struct folio *folio)
> 
> Surely we should rename this to should_zap_folio()?
Will update.
> 
>> @@ -1487,10 +1492,10 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
>>   			 * see zap_install_uffd_wp_if_needed().
>>   			 */
>>   			WARN_ON_ONCE(!vma_is_anonymous(vma));
>> -			rss[mm_counter(page)]--;
>> +			rss[mm_counter(&folio->page)]--;
>>   			if (is_device_private_entry(entry))
>> -				page_remove_rmap(page, vma, false);
>> -			put_page(page);
>> +				page_remove_rmap(&folio->page, vma, false);
>> +			folio_put(folio);
> 
> This is wrong.  If we have a PTE-mapped THP, you'll remove the head page
> N times instead of removing each of N pages.

This is device private entry, I suppose that it won't be a THP and large 
folio when check migrate_vma_check_page() and migrate_vma_insert_page(),
right?

> 
> I suspect you're going to collide with Ryan's work by doing this ...
> 
Maybe not if the above is true, at least for now.

Thanks.

Re: [PATCH v2 06/10] mm: memory: use a folio in zap_pte_range()

[Matthew Wilcox]

On Mon, Nov 06, 2023 at 10:30:59AM +0800, Kefeng Wang wrote:
> On 2023/11/5 1:20, Matthew Wilcox wrote:
> > > -				page_remove_rmap(page, vma, false);
> > > -			put_page(page);
> > > +				page_remove_rmap(&folio->page, vma, false);
> > > +			folio_put(folio);
> > 
> > This is wrong.  If we have a PTE-mapped THP, you'll remove the head page
> > N times instead of removing each of N pages.
> 
> This is device private entry, I suppose that it won't be a THP and large
> folio when check migrate_vma_check_page() and migrate_vma_insert_page(),
> right?

I don't want to leave that kind of booby-trap in the code.  Both places
which currently call page_remove_rmap() should be left as referring to
the page, not the folio.

Re: [PATCH v2 06/10] mm: memory: use a folio in zap_pte_range()

[Kefeng Wang]

On 2023/11/6 22:20, Matthew Wilcox wrote:
> On Mon, Nov 06, 2023 at 10:30:59AM +0800, Kefeng Wang wrote:
>> On 2023/11/5 1:20, Matthew Wilcox wrote:
>>>> -				page_remove_rmap(page, vma, false);
>>>> -			put_page(page);
>>>> +				page_remove_rmap(&folio->page, vma, false);
>>>> +			folio_put(folio);
>>>
>>> This is wrong.  If we have a PTE-mapped THP, you'll remove the head page
>>> N times instead of removing each of N pages.
>>
>> This is device private entry, I suppose that it won't be a THP and large
>> folio when check migrate_vma_check_page() and migrate_vma_insert_page(),
>> right?
> 
> I don't want to leave that kind of booby-trap in the code.  Both places
> which currently call page_remove_rmap() should be left as referring to
> the page, not the folio.

Sure, I will fix this, also page_try_dup_anon_rmap() for device private
entry in copy_nonpresent_pte of patch5.