mm: save soft-dirty bits on swapped pages

Andy Lutomirski reported that if a page with _PAGE_SOFT_DIRTY bit set
get swapped out, the bit is getting lost and no longer available when
pte read back.

To resolve this we introduce _PTE_SWP_SOFT_DIRTY bit which is saved in
pte entry for the page being swapped out.  When such page is to be read
back from a swap cache we check for bit presence and if it's there we
clear it and restore the former _PAGE_SOFT_DIRTY bit back.

One of the problem was to find a place in pte entry where we can save
the _PTE_SWP_SOFT_DIRTY bit while page is in swap.  The _PAGE_PSE was
chosen for that, it doesn't intersect with swap entry format stored in
pte.

Reported-by: Andy Lutomirski <luto@amacapital.net>
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Acked-by: Pavel Emelyanov <xemul@parallels.com>
Cc: Matt Mackall <mpm@selenic.com>
Cc: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@gmail.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Reviewed-by: Minchan Kim <minchan@kernel.org>
Reviewed-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 files changed, 84 insertions, 9 deletions

diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h
index 7dc305a..bd0518a 100644
--- a/arch/x86/include/asm/pgtable.h
+++ b/arch/x86/include/asm/pgtable.h
@@ -314,6 +314,21 @@ static inline pmd_t pmd_mksoft_dirty(pmd_t pmd)
 	return pmd_set_flags(pmd, _PAGE_SOFT_DIRTY);
 }
 
+static inline pte_t pte_swp_mksoft_dirty(pte_t pte)
+{
+	return pte_set_flags(pte, _PAGE_SWP_SOFT_DIRTY);
+}
+
+static inline int pte_swp_soft_dirty(pte_t pte)
+{
+	return pte_flags(pte) & _PAGE_SWP_SOFT_DIRTY;
+}
+
+static inline pte_t pte_swp_clear_soft_dirty(pte_t pte)
+{
+	return pte_clear_flags(pte, _PAGE_SWP_SOFT_DIRTY);
+}
+
 /*
  * Mask out unsupported bits in a present pgprot.  Non-present pgprots
  * can use those bits for other purposes, so leave them be.
diff --git a/arch/x86/include/asm/pgtable_types.h b/arch/x86/include/asm/pgtable_types.h
index c98ac63..5e8442f 100644
--- a/arch/x86/include/asm/pgtable_types.h
+++ b/arch/x86/include/asm/pgtable_types.h
@@ -67,6 +67,19 @@
 #define _PAGE_SOFT_DIRTY	(_AT(pteval_t, 0))
 #endif
 
+/*
+ * Tracking soft dirty bit when a page goes to a swap is tricky.
+ * We need a bit which can be stored in pte _and_ not conflict
+ * with swap entry format. On x86 bits 6 and 7 are *not* involved
+ * into swap entry computation, but bit 6 is used for nonlinear
+ * file mapping, so we borrow bit 7 for soft dirty tracking.
+ */
+#ifdef CONFIG_MEM_SOFT_DIRTY
+#define _PAGE_SWP_SOFT_DIRTY	_PAGE_PSE
+#else
+#define _PAGE_SWP_SOFT_DIRTY	(_AT(pteval_t, 0))
+#endif
+
 #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
 #define _PAGE_NX	(_AT(pteval_t, 1) << _PAGE_BIT_NX)
 #else
diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index dbf61f6..e2d9bdc 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -730,8 +730,14 @@ static inline void clear_soft_dirty(struct vm_area_struct *vma,
 	 * of how soft-dirty works.
 	 */
 	pte_t ptent = *pte;
-	ptent = pte_wrprotect(ptent);
-	ptent = pte_clear_flags(ptent, _PAGE_SOFT_DIRTY);
+
+	if (pte_present(ptent)) {
+		ptent = pte_wrprotect(ptent);
+		ptent = pte_clear_flags(ptent, _PAGE_SOFT_DIRTY);
+	} else if (is_swap_pte(ptent)) {
+		ptent = pte_swp_clear_soft_dirty(ptent);
+	}
+
 	set_pte_at(vma->vm_mm, addr, pte, ptent);
 #endif
 }
@@ -752,14 +758,15 @@ static int clear_refs_pte_range(pmd_t *pmd, unsigned long addr,
 	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
 	for (; addr != end; pte++, addr += PAGE_SIZE) {
 		ptent = *pte;
-		if (!pte_present(ptent))
-			continue;
 
 		if (cp->type == CLEAR_REFS_SOFT_DIRTY) {
 			clear_soft_dirty(vma, addr, pte);
 			continue;
 		}
 
+		if (!pte_present(ptent))
+			continue;
+
 		page = vm_normal_page(vma, addr, ptent);
 		if (!page)
 			continue;
@@ -930,8 +937,10 @@ static void pte_to_pagemap_entry(pagemap_entry_t *pme, struct pagemapread *pm,
 		flags = PM_PRESENT;
 		page = vm_normal_page(vma, addr, pte);
 	} else if (is_swap_pte(pte)) {
-		swp_entry_t entry = pte_to_swp_entry(pte);
-
+		swp_entry_t entry;
+		if (pte_swp_soft_dirty(pte))
+			flags2 |= __PM_SOFT_DIRTY;
+		entry = pte_to_swp_entry(pte);
 		frame = swp_type(entry) |
 			(swp_offset(entry) << MAX_SWAPFILES_SHIFT);
 		flags = PM_SWAP;
diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
index 2f47ade..2a7e0d1 100644
--- a/include/asm-generic/pgtable.h
+++ b/include/asm-generic/pgtable.h
@@ -417,6 +417,21 @@ static inline pmd_t pmd_mksoft_dirty(pmd_t pmd)
 {
 	return pmd;
 }
+
+static inline pte_t pte_swp_mksoft_dirty(pte_t pte)
+{
+	return pte;
+}
+
+static inline int pte_swp_soft_dirty(pte_t pte)
+{
+	return 0;
+}
+
+static inline pte_t pte_swp_clear_soft_dirty(pte_t pte)
+{
+	return pte;
+}
 #endif
 
 #ifndef __HAVE_PFNMAP_TRACKING
diff --git a/include/linux/swapops.h b/include/linux/swapops.h
index c5fd30d..8d4fa82 100644
--- a/include/linux/swapops.h
+++ b/include/linux/swapops.h
@@ -67,6 +67,8 @@ static inline swp_entry_t pte_to_swp_entry(pte_t pte)
 	swp_entry_t arch_entry;
 
 	BUG_ON(pte_file(pte));
+	if (pte_swp_soft_dirty(pte))
+		pte = pte_swp_clear_soft_dirty(pte);
 	arch_entry = __pte_to_swp_entry(pte);
 	return swp_entry(__swp_type(arch_entry), __swp_offset(arch_entry));
 }
diff --git a/mm/memory.c b/mm/memory.c
index 1ce2e2a..e98ecad2 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3115,6 +3115,8 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
 		exclusive = 1;
 	}
 	flush_icache_page(vma, page);
+	if (pte_swp_soft_dirty(orig_pte))
+		pte = pte_mksoft_dirty(pte);
 	set_pte_at(mm, address, page_table, pte);
 	if (page == swapcache)
 		do_page_add_anon_rmap(page, vma, address, exclusive);
diff --git a/mm/rmap.c b/mm/rmap.c
index cd356df..83325b8 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1236,6 +1236,7 @@ int try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
 			   swp_entry_to_pte(make_hwpoison_entry(page)));
 	} else if (PageAnon(page)) {
 		swp_entry_t entry = { .val = page_private(page) };
+		pte_t swp_pte;
 
 		if (PageSwapCache(page)) {
 			/*
@@ -1264,7 +1265,10 @@ int try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
 			BUG_ON(TTU_ACTION(flags) != TTU_MIGRATION);
 			entry = make_migration_entry(page, pte_write(pteval));
 		}
-		set_pte_at(mm, address, pte, swp_entry_to_pte(entry));
+		swp_pte = swp_entry_to_pte(entry);
+		if (pte_soft_dirty(pteval))
+			swp_pte = pte_swp_mksoft_dirty(swp_pte);
+		set_pte_at(mm, address, pte, swp_pte);
 		BUG_ON(pte_file(*pte));
 	} else if (IS_ENABLED(CONFIG_MIGRATION) &&
 		   (TTU_ACTION(flags) == TTU_MIGRATION)) {
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 36af6ee..6cf2e60 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -866,6 +866,21 @@ unsigned int count_swap_pages(int type, int free)
 }
 #endif /* CONFIG_HIBERNATION */
 
+static inline int maybe_same_pte(pte_t pte, pte_t swp_pte)
+{
+#ifdef CONFIG_MEM_SOFT_DIRTY
+	/*
+	 * When pte keeps soft dirty bit the pte generated
+	 * from swap entry does not has it, still it's same
+	 * pte from logical point of view.
+	 */
+	pte_t swp_pte_dirty = pte_swp_mksoft_dirty(swp_pte);
+	return pte_same(pte, swp_pte) || pte_same(pte, swp_pte_dirty);
+#else
+	return pte_same(pte, swp_pte);
+#endif
+}
+
 /*
  * No need to decide whether this PTE shares the swap entry with others,
  * just let do_wp_page work it out if a write is requested later - to
@@ -892,7 +907,7 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 	}
 
 	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
-	if (unlikely(!pte_same(*pte, swp_entry_to_pte(entry)))) {
+	if (unlikely(!maybe_same_pte(*pte, swp_entry_to_pte(entry)))) {
 		mem_cgroup_cancel_charge_swapin(memcg);
 		ret = 0;
 		goto out;
@@ -947,7 +962,7 @@ static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
 		 * swapoff spends a _lot_ of time in this loop!
 		 * Test inline before going to call unuse_pte.
 		 */
-		if (unlikely(pte_same(*pte, swp_pte))) {
+		if (unlikely(maybe_same_pte(*pte, swp_pte))) {
 			pte_unmap(pte);
 			ret = unuse_pte(vma, pmd, addr, entry, page);
 			if (ret)