Merge: mm: Remember a/d bits for migration entries

MR: https://gitlab.com/redhat/centos-stream/src/kernel/centos-stream-9/-/merge_requests/2294

Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2168392
Testing: KT1-lite + MM regressions as reported on the BZ ticket

This work brings in upstream homonymous series [1] alongside
its missing dependencies and required follow-ups to address
the feature request expressed in Bug 2168392.

This is a micro-optimization for the page migration path,
which would unconditionally mark migrated pages as old &
clean. This changeset teaches the migration code to remember
these PTE bits from the old PTE and properly reset them after
migrating the page over. For the generic use case, the bit
loosing in the new PTE isn't really much of a problem, but
there's measurable performance costs. For specific use cases
(DEVICE MEMORY migrations comes to mind) this might represent
potential problems, though, so we're bringing in the changeset
for correctness and performance.

[1] https://lore.kernel.org/all/20220811161331.37055-6-peterx@redhat.com/T/#me0e2db4c09f9a615cc7f7e21df37ce184f48571d

Omitted-fix: bf2f34a ("LoongArch: Set _PAGE_DIRTY only if _PAGE_WRITE is set in {pmd,pte}_mkdirty()")

Rafael Aquini (13):
  include/linux/swapops.h: remove stub for non_swap_entry()
  mm/swapops: make is_pmd_migration_entry more strict
  mm/x86: use SWP_TYPE_BITS in 3-level swap macros
  mm/swap: comment all the ifdef in swapops.h
  mm/swap: add swp_offset_pfn() to fetch PFN from swap entry
  mm/thp: carry over dirty bit when thp splits on pmd
  mm: remember young/dirty bit for page migrations
  mm/swap: cache maximum swapfile size when init swap
  mm/swap: cache swap migration A/D bits support
  mm/swap: fix SWP_PFN_BITS with CONFIG_PHYS_ADDR_T_64BIT on 32bit
  Partly revert "mm/thp: carry over dirty bit when thp splits on pmd"
  mm/thp: re-apply mkdirty for small pages after split
  mm/migrate: fix wrongly apply write bit after mkdirty on sparc64

Signed-off-by: Rafael Aquini <aquini@redhat.com>

Approved-by: Nico Pache <npache@redhat.com>
Approved-by: Waiman Long <longman@redhat.com>
Approved-by: Donald Dutile <ddutile@redhat.com>
Approved-by: Prarit Bhargava <prarit@redhat.com>

Signed-off-by: Jan Stancek <jstancek@redhat.com>

Author: jstancek

arch/arm64/mm/hugetlbpage.c

@@ -244,7 +244,7 @@ static inline struct folio *hugetlb_swap_entry_to_folio(swp_entry_t entry)
 {
 	VM_BUG_ON(!is_migration_entry(entry) && !is_hwpoison_entry(entry));
 
-	return page_folio(pfn_to_page(swp_offset(entry)));
+	return page_folio(pfn_to_page(swp_offset_pfn(entry)));
 }
 
 void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,

arch/x86/include/asm/pgtable-3level.h

@@ -256,10 +256,10 @@ static inline pud_t native_pudp_get_and_clear(pud_t *pudp)
 /* We always extract/encode the offset by shifting it all the way up, and then down again */
 #define SWP_OFFSET_SHIFT	(SWP_OFFSET_FIRST_BIT + SWP_TYPE_BITS)
 
-#define MAX_SWAPFILES_CHECK() BUILD_BUG_ON(MAX_SWAPFILES_SHIFT > 5)
-#define __swp_type(x)			(((x).val) & 0x1f)
-#define __swp_offset(x)			((x).val >> 5)
-#define __swp_entry(type, offset)	((swp_entry_t){(type) | (offset) << 5})
+#define MAX_SWAPFILES_CHECK() BUILD_BUG_ON(MAX_SWAPFILES_SHIFT > SWP_TYPE_BITS)
+#define __swp_type(x)			(((x).val) & ((1UL << SWP_TYPE_BITS) - 1))
+#define __swp_offset(x)			((x).val >> SWP_TYPE_BITS)
+#define __swp_entry(type, offset)	((swp_entry_t){(type) | (offset) << SWP_TYPE_BITS})
 
 /*
  * Normally, __swp_entry() converts from arch-independent swp_entry_t to

arch/x86/mm/init.c

@@ -1054,7 +1054,7 @@ void update_cache_mode_entry(unsigned entry, enum page_cache_mode cache)
 }
 
 #ifdef CONFIG_SWAP
-unsigned long max_swapfile_size(void)
+unsigned long arch_max_swapfile_size(void)
 {
 	unsigned long pages;
 

fs/proc/task_mmu.c

@@ -1410,9 +1410,19 @@ static pagemap_entry_t pte_to_pagemap_entry(struct pagemapread *pm,
 		if (pte_swp_uffd_wp(pte))
 			flags |= PM_UFFD_WP;
 		entry = pte_to_swp_entry(pte);
-		if (pm->show_pfn)
+		if (pm->show_pfn) {
+			pgoff_t offset;
+			/*
+			 * For PFN swap offsets, keeping the offset field
+			 * to be PFN only to be compatible with old smaps.
+			 */
+			if (is_pfn_swap_entry(entry))
+				offset = swp_offset_pfn(entry);
+			else
+				offset = swp_offset(entry);
 			frame = swp_type(entry) |
-				(swp_offset(entry) << MAX_SWAPFILES_SHIFT);
+			    (offset << MAX_SWAPFILES_SHIFT);
+		}
 		flags |= PM_SWAP;
 		migration = is_migration_entry(entry);
 		if (is_pfn_swap_entry(entry))
@@ -1467,7 +1477,11 @@ static int pagemap_pmd_range(pmd_t *pmdp, unsigned long addr, unsigned long end,
 			unsigned long offset;
 
 			if (pm->show_pfn) {
-				offset = swp_offset(entry) +
+				if (is_pfn_swap_entry(entry))
+					offset = swp_offset_pfn(entry);
+				else
+					offset = swp_offset(entry);
+				offset = offset +
 					((addr & ~PMD_MASK) >> PAGE_SHIFT);
 				frame = swp_type(entry) |
 					(offset << MAX_SWAPFILES_SHIFT);

include/linux/swapfile.h

@@ -8,6 +8,11 @@
  */
 extern struct swap_info_struct *swap_info[];
 extern unsigned long generic_max_swapfile_size(void);
-extern unsigned long max_swapfile_size(void);
+unsigned long arch_max_swapfile_size(void);
+
+/* Maximum swapfile size supported for the arch (not inclusive). */
+extern unsigned long swapfile_maximum_size;
+/* Whether swap migration entry supports storing A/D bits for the arch */
+extern bool swap_migration_ad_supported;
 
 #endif /* _LINUX_SWAPFILE_H */

include/linux/swapops.h

@@ -8,6 +8,10 @@
 
 #ifdef CONFIG_MMU
 
+#ifdef CONFIG_SWAP
+#include <linux/swapfile.h>
+#endif	/* CONFIG_SWAP */
+
 /*
  * swapcache pages are stored in the swapper_space radix tree.  We want to
  * get good packing density in that tree, so the index should be dense in
@@ -23,6 +27,47 @@
 #define SWP_TYPE_SHIFT	(BITS_PER_XA_VALUE - MAX_SWAPFILES_SHIFT)
 #define SWP_OFFSET_MASK	((1UL << SWP_TYPE_SHIFT) - 1)
 
+/*
+ * Definitions only for PFN swap entries (see is_pfn_swap_entry()).  To
+ * store PFN, we only need SWP_PFN_BITS bits.  Each of the pfn swap entries
+ * can use the extra bits to store other information besides PFN.
+ */
+#ifdef MAX_PHYSMEM_BITS
+#define SWP_PFN_BITS		(MAX_PHYSMEM_BITS - PAGE_SHIFT)
+#else  /* MAX_PHYSMEM_BITS */
+#define SWP_PFN_BITS		min_t(int, \
+				      sizeof(phys_addr_t) * 8 - PAGE_SHIFT, \
+				      SWP_TYPE_SHIFT)
+#endif	/* MAX_PHYSMEM_BITS */
+#define SWP_PFN_MASK		(BIT(SWP_PFN_BITS) - 1)
+
+/**
+ * Migration swap entry specific bitfield definitions.  Layout:
+ *
+ *   |----------+--------------------|
+ *   | swp_type | swp_offset         |
+ *   |----------+--------+-+-+-------|
+ *   |          | resv   |D|A|  PFN  |
+ *   |----------+--------+-+-+-------|
+ *
+ * @SWP_MIG_YOUNG_BIT: Whether the page used to have young bit set (bit A)
+ * @SWP_MIG_DIRTY_BIT: Whether the page used to have dirty bit set (bit D)
+ *
+ * Note: A/D bits will be stored in migration entries iff there're enough
+ * free bits in arch specific swp offset.  By default we'll ignore A/D bits
+ * when migrating a page.  Please refer to migration_entry_supports_ad()
+ * for more information.  If there're more bits besides PFN and A/D bits,
+ * they should be reserved and always be zeros.
+ */
+#define SWP_MIG_YOUNG_BIT		(SWP_PFN_BITS)
+#define SWP_MIG_DIRTY_BIT		(SWP_PFN_BITS + 1)
+#define SWP_MIG_TOTAL_BITS		(SWP_PFN_BITS + 2)
+
+#define SWP_MIG_YOUNG			BIT(SWP_MIG_YOUNG_BIT)
+#define SWP_MIG_DIRTY			BIT(SWP_MIG_DIRTY_BIT)
+
+static inline bool is_pfn_swap_entry(swp_entry_t entry);
+
 /* Clear all flags but only keep swp_entry_t related information */
 static inline pte_t pte_swp_clear_flags(pte_t pte)
 {
@@ -64,6 +109,17 @@ static inline pgoff_t swp_offset(swp_entry_t entry)
 	return entry.val & SWP_OFFSET_MASK;
 }
 
+/*
+ * This should only be called upon a pfn swap entry to get the PFN stored
+ * in the swap entry.  Please refers to is_pfn_swap_entry() for definition
+ * of pfn swap entry.
+ */
+static inline unsigned long swp_offset_pfn(swp_entry_t entry)
+{
+	VM_BUG_ON(!is_pfn_swap_entry(entry));
+	return swp_offset(entry) & SWP_PFN_MASK;
+}
+
 /* check whether a pte points to a swap entry */
 static inline int is_swap_pte(pte_t pte)
 {
@@ -240,15 +296,61 @@ static inline swp_entry_t make_writable_migration_entry(pgoff_t offset)
 	return swp_entry(SWP_MIGRATION_WRITE, offset);
 }
 
+/*
+ * Returns whether the host has large enough swap offset field to support
+ * carrying over pgtable A/D bits for page migrations.  The result is
+ * pretty much arch specific.
+ */
+static inline bool migration_entry_supports_ad(void)
+{
+#ifdef CONFIG_SWAP
+	return swap_migration_ad_supported;
+#else  /* CONFIG_SWAP */
+	return false;
+#endif	/* CONFIG_SWAP */
+}
+
+static inline swp_entry_t make_migration_entry_young(swp_entry_t entry)
+{
+	if (migration_entry_supports_ad())
+		return swp_entry(swp_type(entry),
+				 swp_offset(entry) | SWP_MIG_YOUNG);
+	return entry;
+}
+
+static inline bool is_migration_entry_young(swp_entry_t entry)
+{
+	if (migration_entry_supports_ad())
+		return swp_offset(entry) & SWP_MIG_YOUNG;
+	/* Keep the old behavior of aging page after migration */
+	return false;
+}
+
+static inline swp_entry_t make_migration_entry_dirty(swp_entry_t entry)
+{
+	if (migration_entry_supports_ad())
+		return swp_entry(swp_type(entry),
+				 swp_offset(entry) | SWP_MIG_DIRTY);
+	return entry;
+}
+
+static inline bool is_migration_entry_dirty(swp_entry_t entry)
+{
+	if (migration_entry_supports_ad())
+		return swp_offset(entry) & SWP_MIG_DIRTY;
+	/* Keep the old behavior of clean page after migration */
+	return false;
+}
+
 extern void __migration_entry_wait(struct mm_struct *mm, pte_t *ptep,
 					spinlock_t *ptl);
 extern void migration_entry_wait(struct mm_struct *mm, pmd_t *pmd,
 					unsigned long address);
 #ifdef CONFIG_HUGETLB_PAGE
 extern void __migration_entry_wait_huge(pte_t *ptep, spinlock_t *ptl);
 extern void migration_entry_wait_huge(struct vm_area_struct *vma, pte_t *pte);
-#endif
-#else
+#endif	/* CONFIG_HUGETLB_PAGE */
+#else  /* CONFIG_MIGRATION */
 static inline swp_entry_t make_readable_migration_entry(pgoff_t offset)
 {
 	return swp_entry(0, 0);
@@ -276,7 +378,7 @@ static inline void migration_entry_wait(struct mm_struct *mm, pmd_t *pmd,
 #ifdef CONFIG_HUGETLB_PAGE
 static inline void __migration_entry_wait_huge(pte_t *ptep, spinlock_t *ptl) { }
 static inline void migration_entry_wait_huge(struct vm_area_struct *vma, pte_t *pte) { }
-#endif
+#endif	/* CONFIG_HUGETLB_PAGE */
 static inline int is_writable_migration_entry(swp_entry_t entry)
 {
 	return 0;
@@ -286,7 +388,26 @@ static inline int is_readable_migration_entry(swp_entry_t entry)
 	return 0;
 }
 
-#endif
+static inline swp_entry_t make_migration_entry_young(swp_entry_t entry)
+{
+	return entry;
+}
+
+static inline bool is_migration_entry_young(swp_entry_t entry)
+{
+	return false;
+}
+
+static inline swp_entry_t make_migration_entry_dirty(swp_entry_t entry)
+{
+	return entry;
+}
+
+static inline bool is_migration_entry_dirty(swp_entry_t entry)
+{
+	return false;
+}
+#endif	/* CONFIG_MIGRATION */
 
 typedef unsigned long pte_marker;
 
@@ -369,7 +490,7 @@ static inline int pte_none_mostly(pte_t pte)
 
 static inline struct page *pfn_swap_entry_to_page(swp_entry_t entry)
 {
-	struct page *p = pfn_to_page(swp_offset(entry));
+	struct page *p = pfn_to_page(swp_offset_pfn(entry));
 
 	/*
 	 * Any use of migration entries may only occur while the
@@ -387,6 +508,9 @@ static inline struct page *pfn_swap_entry_to_page(swp_entry_t entry)
  */
 static inline bool is_pfn_swap_entry(swp_entry_t entry)
 {
+	/* Make sure the swp offset can always store the needed fields */
+	BUILD_BUG_ON(SWP_TYPE_SHIFT < SWP_PFN_BITS);
+
 	return is_migration_entry(entry) || is_device_private_entry(entry) ||
 	       is_device_exclusive_entry(entry);
 }
@@ -424,9 +548,9 @@ static inline pmd_t swp_entry_to_pmd(swp_entry_t entry)
 
 static inline int is_pmd_migration_entry(pmd_t pmd)
 {
-	return !pmd_present(pmd) && is_migration_entry(pmd_to_swp_entry(pmd));
+	return is_swap_pmd(pmd) && is_migration_entry(pmd_to_swp_entry(pmd));
 }
-#else
+#else  /* CONFIG_ARCH_ENABLE_THP_MIGRATION */
 static inline int set_pmd_migration_entry(struct page_vma_mapped_walk *pvmw,
 		struct page *page)
 {
@@ -455,7 +579,7 @@ static inline int is_pmd_migration_entry(pmd_t pmd)
 {
 	return 0;
 }
-#endif
+#endif  /* CONFIG_ARCH_ENABLE_THP_MIGRATION */
 
 #ifdef CONFIG_MEMORY_FAILURE
 
@@ -475,11 +599,6 @@ static inline int is_hwpoison_entry(swp_entry_t entry)
 	return swp_type(entry) == SWP_HWPOISON;
 }
 
-static inline unsigned long hwpoison_entry_to_pfn(swp_entry_t entry)
-{
-	return swp_offset(entry);
-}
-
 static inline void num_poisoned_pages_inc(void)
 {
 	atomic_long_inc(&num_poisoned_pages);
@@ -495,7 +614,7 @@ static inline void num_poisoned_pages_sub(long i)
 	atomic_long_sub(i, &num_poisoned_pages);
 }
 
-#else
+#else  /* CONFIG_MEMORY_FAILURE */
 
 static inline swp_entry_t make_hwpoison_entry(struct page *page)
 {
@@ -514,20 +633,12 @@ static inline void num_poisoned_pages_inc(void)
 static inline void num_poisoned_pages_sub(long i)
 {
 }
-#endif
+#endif  /* CONFIG_MEMORY_FAILURE */
 
-#if defined(CONFIG_MEMORY_FAILURE) || defined(CONFIG_MIGRATION) || \
-    defined(CONFIG_DEVICE_PRIVATE)
 static inline int non_swap_entry(swp_entry_t entry)
 {
 	return swp_type(entry) >= MAX_SWAPFILES;
 }
-#else
-static inline int non_swap_entry(swp_entry_t entry)
-{
-	return 0;
-}
-#endif
 
 #endif /* CONFIG_MMU */
 #endif /* _LINUX_SWAPOPS_H */

mm/hmm.c

@@ -253,7 +253,7 @@ static int hmm_vma_handle_pte(struct mm_walk *walk, unsigned long addr,
 			cpu_flags = HMM_PFN_VALID;
 			if (is_writable_device_private_entry(entry))
 				cpu_flags |= HMM_PFN_WRITE;
-			*hmm_pfn = swp_offset(entry) | cpu_flags;
+			*hmm_pfn = swp_offset_pfn(entry) | cpu_flags;
 			return 0;
 		}