mm: page cache: store only head pages in i_pages

Transparent Huge Pages are currently stored in i_pages as pointers to
consecutive subpages.  This patch changes that to storing consecutive
pointers to the head page in preparation for storing huge pages more
efficiently in i_pages.

Large parts of this are "inspired" by Kirill's patch
https://lore.kernel.org/lkml/20170126115819.58875-2-kirill.shutemov@linux.intel.com/

Kirill and Huang Ying contributed several fixes.

[willy@infradead.org: use compound_nr, squish uninit-var warning]
Link: http://lkml.kernel.org/r/20190731210400.7419-1-willy@infradead.org
Signed-off-by: Matthew Wilcox <willy@infradead.org>
Acked-by: Jan Kara <jack@suse.cz>
Reviewed-by: Kirill Shutemov <kirill@shutemov.name>
Reviewed-by: Song Liu <songliubraving@fb.com>
Tested-by: Song Liu <songliubraving@fb.com>
Tested-by: William Kucharski <william.kucharski@oracle.com>
Reviewed-by: William Kucharski <william.kucharski@oracle.com>
Tested-by: Qian Cai <cai@lca.pw>
Tested-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Song Liu <songliubraving@fb.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/pagemap.h

@@ -333,6 +333,16 @@ static inline struct page *grab_cache_page_nowait(struct address_space *mapping,
 			mapping_gfp_mask(mapping));
 }
 
+static inline struct page *find_subpage(struct page *page, pgoff_t offset)
+{
+	if (PageHuge(page))
+		return page;
+
+	VM_BUG_ON_PAGE(PageTail(page), page);
+
+	return page + (offset & (compound_nr(page) - 1));
+}
+
 struct page *find_get_entry(struct address_space *mapping, pgoff_t offset);
 struct page *find_lock_entry(struct address_space *mapping, pgoff_t offset);
 unsigned find_get_entries(struct address_space *mapping, pgoff_t start,

mm/filemap.c

@@ -281,11 +281,11 @@ EXPORT_SYMBOL(delete_from_page_cache);
  * @pvec: pagevec with pages to delete
  *
  * The function walks over mapping->i_pages and removes pages passed in @pvec
- * from the mapping. The function expects @pvec to be sorted by page index.
+ * from the mapping. The function expects @pvec to be sorted by page index
+ * and is optimised for it to be dense.
  * It tolerates holes in @pvec (mapping entries at those indices are not
  * modified). The function expects only THP head pages to be present in the
- * @pvec and takes care to delete all corresponding tail pages from the
+ * @pvec.
- * mapping as well.
  *
  * The function expects the i_pages lock to be held.
  */
@@ -294,40 +294,43 @@ static void page_cache_delete_batch(struct address_space *mapping,
 {
 	XA_STATE(xas, &mapping->i_pages, pvec->pages[0]->index);
 	int total_pages = 0;
-	int i = 0, tail_pages = 0;
+	int i = 0;
 	struct page *page;
 
 	mapping_set_update(&xas, mapping);
 	xas_for_each(&xas, page, ULONG_MAX) {
-		if (i >= pagevec_count(pvec) && !tail_pages)
+		if (i >= pagevec_count(pvec))
 			break;
+
+		/* A swap/dax/shadow entry got inserted? Skip it. */
 		if (xa_is_value(page))
 			continue;
-		if (!tail_pages) {
 		/*
-			 * Some page got inserted in our range? Skip it. We
+		 * A page got inserted in our range? Skip it. We have our
-			 * have our pages locked so they are protected from
+		 * pages locked so they are protected from being removed.
-			 * being removed.
+		 * If we see a page whose index is higher than ours, it
+		 * means our page has been removed, which shouldn't be
+		 * possible because we're holding the PageLock.
 		 */
 		if (page != pvec->pages[i]) {
-				VM_BUG_ON_PAGE(page->index >
+			VM_BUG_ON_PAGE(page->index > pvec->pages[i]->index,
-						pvec->pages[i]->index, page);
+					page);
 			continue;
 		}
+
 		WARN_ON_ONCE(!PageLocked(page));
-			if (PageTransHuge(page) && !PageHuge(page))
+
-				tail_pages = HPAGE_PMD_NR - 1;
+		if (page->index == xas.xa_index)
 			page->mapping = NULL;
+		/* Leave page->index set: truncation lookup relies on it */
+
 		/*
-			 * Leave page->index set: truncation lookup relies
+		 * Move to the next page in the vector if this is a regular
-			 * upon it
+		 * page or the index is of the last sub-page of this compound
+		 * page.
 		 */
+		if (page->index + compound_nr(page) - 1 == xas.xa_index)
 			i++;
-		} else {
-			VM_BUG_ON_PAGE(page->index + HPAGE_PMD_NR - tail_pages
-					!= pvec->pages[i]->index, page);
-			tail_pages--;
-		}
 		xas_store(&xas, NULL);
 		total_pages++;
 	}
@@ -1520,7 +1523,7 @@ EXPORT_SYMBOL(page_cache_prev_miss);
 struct page *find_get_entry(struct address_space *mapping, pgoff_t offset)
 {
 	XA_STATE(xas, &mapping->i_pages, offset);
-	struct page *head, *page;
+	struct page *page;
 
 	rcu_read_lock();
 repeat:
@@ -1535,25 +1538,19 @@ struct page *find_get_entry(struct address_space *mapping, pgoff_t offset)
 	if (!page || xa_is_value(page))
 		goto out;
 
-	head = compound_head(page);
+	if (!page_cache_get_speculative(page))
-	if (!page_cache_get_speculative(head))
 		goto repeat;
 
-	/* The page was split under us? */
-	if (compound_head(page) != head) {
-		put_page(head);
-		goto repeat;
-	}
-
 	/*
-	 * Has the page moved?
+	 * Has the page moved or been split?
 	 * This is part of the lockless pagecache protocol. See
 	 * include/linux/pagemap.h for details.
 	 */
 	if (unlikely(page != xas_reload(&xas))) {
-		put_page(head);
+		put_page(page);
 		goto repeat;
 	}
+	page = find_subpage(page, offset);
 out:
 	rcu_read_unlock();
 
@@ -1735,7 +1732,6 @@ unsigned find_get_entries(struct address_space *mapping,
 
 	rcu_read_lock();
 	xas_for_each(&xas, page, ULONG_MAX) {
-		struct page *head;
 		if (xas_retry(&xas, page))
 			continue;
 		/*
@@ -1746,17 +1742,13 @@ unsigned find_get_entries(struct address_space *mapping,
 		if (xa_is_value(page))
 			goto export;
 
-		head = compound_head(page);
+		if (!page_cache_get_speculative(page))
-		if (!page_cache_get_speculative(head))
 			goto retry;
 
-		/* The page was split under us? */
+		/* Has the page moved or been split? */
-		if (compound_head(page) != head)
-			goto put_page;
-
-		/* Has the page moved? */
 		if (unlikely(page != xas_reload(&xas)))
 			goto put_page;
+		page = find_subpage(page, xas.xa_index);
 
 export:
 		indices[ret] = xas.xa_index;
@@ -1765,7 +1757,7 @@ unsigned find_get_entries(struct address_space *mapping,
 			break;
 		continue;
 put_page:
-		put_page(head);
+		put_page(page);
 retry:
 		xas_reset(&xas);
 	}
@@ -1807,33 +1799,27 @@ unsigned find_get_pages_range(struct address_space *mapping, pgoff_t *start,
 
 	rcu_read_lock();
 	xas_for_each(&xas, page, end) {
-		struct page *head;
 		if (xas_retry(&xas, page))
 			continue;
 		/* Skip over shadow, swap and DAX entries */
 		if (xa_is_value(page))
 			continue;
 
-		head = compound_head(page);
+		if (!page_cache_get_speculative(page))
-		if (!page_cache_get_speculative(head))
 			goto retry;
 
-		/* The page was split under us? */
+		/* Has the page moved or been split? */
-		if (compound_head(page) != head)
-			goto put_page;
-
-		/* Has the page moved? */
 		if (unlikely(page != xas_reload(&xas)))
 			goto put_page;
 
-		pages[ret] = page;
+		pages[ret] = find_subpage(page, xas.xa_index);
 		if (++ret == nr_pages) {
 			*start = xas.xa_index + 1;
 			goto out;
 		}
 		continue;
 put_page:
-		put_page(head);
+		put_page(page);
 retry:
 		xas_reset(&xas);
 	}
@@ -1878,7 +1864,6 @@ unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t index,
 
 	rcu_read_lock();
 	for (page = xas_load(&xas); page; page = xas_next(&xas)) {
-		struct page *head;
 		if (xas_retry(&xas, page))
 			continue;
 		/*
@@ -1888,24 +1873,19 @@ unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t index,
 		if (xa_is_value(page))
 			break;
 
-		head = compound_head(page);
+		if (!page_cache_get_speculative(page))
-		if (!page_cache_get_speculative(head))
 			goto retry;
 
-		/* The page was split under us? */
+		/* Has the page moved or been split? */
-		if (compound_head(page) != head)
-			goto put_page;
-
-		/* Has the page moved? */
 		if (unlikely(page != xas_reload(&xas)))
 			goto put_page;
 
-		pages[ret] = page;
+		pages[ret] = find_subpage(page, xas.xa_index);
 		if (++ret == nr_pages)
 			break;
 		continue;
 put_page:
-		put_page(head);
+		put_page(page);
 retry:
 		xas_reset(&xas);
 	}
@@ -1941,7 +1921,6 @@ unsigned find_get_pages_range_tag(struct address_space *mapping, pgoff_t *index,
 
 	rcu_read_lock();
 	xas_for_each_marked(&xas, page, end, tag) {
-		struct page *head;
 		if (xas_retry(&xas, page))
 			continue;
 		/*
@@ -1952,26 +1931,21 @@ unsigned find_get_pages_range_tag(struct address_space *mapping, pgoff_t *index,
 		if (xa_is_value(page))
 			continue;
 
-		head = compound_head(page);
+		if (!page_cache_get_speculative(page))
-		if (!page_cache_get_speculative(head))
 			goto retry;
 
-		/* The page was split under us? */
+		/* Has the page moved or been split? */
-		if (compound_head(page) != head)
-			goto put_page;
-
-		/* Has the page moved? */
 		if (unlikely(page != xas_reload(&xas)))
 			goto put_page;
 
-		pages[ret] = page;
+		pages[ret] = find_subpage(page, xas.xa_index);
 		if (++ret == nr_pages) {
 			*index = xas.xa_index + 1;
 			goto out;
 		}
 		continue;
 put_page:
-		put_page(head);
+		put_page(page);
 retry:
 		xas_reset(&xas);
 	}
@@ -2652,7 +2626,7 @@ void filemap_map_pages(struct vm_fault *vmf,
 	pgoff_t last_pgoff = start_pgoff;
 	unsigned long max_idx;
 	XA_STATE(xas, &mapping->i_pages, start_pgoff);
-	struct page *head, *page;
+	struct page *page;
 
 	rcu_read_lock();
 	xas_for_each(&xas, page, end_pgoff) {
@@ -2661,24 +2635,19 @@ void filemap_map_pages(struct vm_fault *vmf,
 		if (xa_is_value(page))
 			goto next;
 
-		head = compound_head(page);
-
 		/*
 		 * Check for a locked page first, as a speculative
 		 * reference may adversely influence page migration.
 		 */
-		if (PageLocked(head))
+		if (PageLocked(page))
 			goto next;
-		if (!page_cache_get_speculative(head))
+		if (!page_cache_get_speculative(page))
 			goto next;
 
-		/* The page was split under us? */
+		/* Has the page moved or been split? */
-		if (compound_head(page) != head)
-			goto skip;
-
-		/* Has the page moved? */
 		if (unlikely(page != xas_reload(&xas)))
 			goto skip;
+		page = find_subpage(page, xas.xa_index);
 
 		if (!PageUptodate(page) ||
 				PageReadahead(page) ||

mm/huge_memory.c

@@ -2497,6 +2497,8 @@ static void __split_huge_page(struct page *page, struct list_head *list,
 	struct page *head = compound_head(page);
 	pg_data_t *pgdat = page_pgdat(head);
 	struct lruvec *lruvec;
+	struct address_space *swap_cache = NULL;
+	unsigned long offset = 0;
 	int i;
 
 	lruvec = mem_cgroup_page_lruvec(head, pgdat);
@@ -2504,6 +2506,14 @@ static void __split_huge_page(struct page *page, struct list_head *list,
 	/* complete memcg works before add pages to LRU */
 	mem_cgroup_split_huge_fixup(head);
 
+	if (PageAnon(head) && PageSwapCache(head)) {
+		swp_entry_t entry = { .val = page_private(head) };
+
+		offset = swp_offset(entry);
+		swap_cache = swap_address_space(entry);
+		xa_lock(&swap_cache->i_pages);
+	}
+
 	for (i = HPAGE_PMD_NR - 1; i >= 1; i--) {
 		__split_huge_page_tail(head, i, lruvec, list);
 		/* Some pages can be beyond i_size: drop them from page cache */
@@ -2513,6 +2523,12 @@ static void __split_huge_page(struct page *page, struct list_head *list,
 			if (IS_ENABLED(CONFIG_SHMEM) && PageSwapBacked(head))
 				shmem_uncharge(head->mapping->host, 1);
 			put_page(head + i);
+		} else if (!PageAnon(page)) {
+			__xa_store(&head->mapping->i_pages, head[i].index,
+					head + i, 0);
+		} else if (swap_cache) {
+			__xa_store(&swap_cache->i_pages, offset + i,
+					head + i, 0);
 		}
 	}
 
@@ -2523,10 +2539,12 @@ static void __split_huge_page(struct page *page, struct list_head *list,
 	/* See comment in __split_huge_page_tail() */
 	if (PageAnon(head)) {
 		/* Additional pin to swap cache */
-		if (PageSwapCache(head))
+		if (PageSwapCache(head)) {
 			page_ref_add(head, 2);
-		else
+			xa_unlock(&swap_cache->i_pages);
+		} else {
 			page_ref_inc(head);
+		}
 	} else {
 		/* Additional pin to page cache */
 		page_ref_add(head, 2);

mm/khugepaged.c

@@ -1378,7 +1378,7 @@ static void collapse_shmem(struct mm_struct *mm,
 				result = SCAN_FAIL;
 				goto xa_locked;
 			}
-			xas_store(&xas, new_page + (index % HPAGE_PMD_NR));
+			xas_store(&xas, new_page);
 			nr_none++;
 			continue;
 		}
@@ -1454,7 +1454,7 @@ static void collapse_shmem(struct mm_struct *mm,
 		list_add_tail(&page->lru, &pagelist);
 
 		/* Finally, replace with the new page. */
-		xas_store(&xas, new_page + (index % HPAGE_PMD_NR));
+		xas_store(&xas, new_page);
 		continue;
 out_unlock:
 		unlock_page(page);

mm/memfd.c

@@ -39,6 +39,7 @@ static void memfd_tag_pins(struct xa_state *xas)
 	xas_for_each(xas, page, ULONG_MAX) {
 		if (xa_is_value(page))
 			continue;
+		page = find_subpage(page, xas->xa_index);
 		if (page_count(page) - page_mapcount(page) > 1)
 			xas_set_mark(xas, MEMFD_TAG_PINNED);
 
@@ -88,6 +89,7 @@ static int memfd_wait_for_pins(struct address_space *mapping)
 			bool clear = true;
 			if (xa_is_value(page))
 				continue;
+			page = find_subpage(page, xas.xa_index);
 			if (page_count(page) - page_mapcount(page) != 1) {
 				/*
 				 * On the last scan, we clean up all those tags

mm/migrate.c

@@ -460,7 +460,7 @@ int migrate_page_move_mapping(struct address_space *mapping,
 
 		for (i = 1; i < HPAGE_PMD_NR; i++) {
 			xas_next(&xas);
-			xas_store(&xas, newpage + i);
+			xas_store(&xas, newpage);
 		}
 	}
 

mm/shmem.c

@@ -631,7 +631,7 @@ static int shmem_add_to_page_cache(struct page *page,
 		if (xas_error(&xas))
 			goto unlock;
 next:
-		xas_store(&xas, page + i);
+		xas_store(&xas, page);
 		if (++i < nr) {
 			xas_next(&xas);
 			goto next;

mm/swap_state.c

@@ -133,7 +133,7 @@ int add_to_swap_cache(struct page *page, swp_entry_t entry, gfp_t gfp)
 		for (i = 0; i < nr; i++) {
 			VM_BUG_ON_PAGE(xas.xa_index != idx + i, page);
 			set_page_private(page + i, entry.val + i);
-			xas_store(&xas, page + i);
+			xas_store(&xas, page);
 			xas_next(&xas);
 		}
 		address_space->nrpages += nr;
@@ -168,7 +168,7 @@ void __delete_from_swap_cache(struct page *page, swp_entry_t entry)
 
 	for (i = 0; i < nr; i++) {
 		void *entry = xas_store(&xas, NULL);
-		VM_BUG_ON_PAGE(entry != page + i, entry);
+		VM_BUG_ON_PAGE(entry != page, entry);
 		set_page_private(page + i, 0);
 		xas_next(&xas);
 	}