ext4: cache all of an extent tree's leaf block upon reading

When we read in an extent tree leaf block from disk, arrange to have
all of its entries cached.  In nearly all cases the in-memory
representation will be more compact than the on-disk representation in
the buffer cache, and it allows us to get the information without
having to traverse the extent tree for successive extents.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>

fs/ext4/ext4.h

@@ -560,6 +560,17 @@ enum {
 	/* Do not put hole in extent cache */
 #define EXT4_GET_BLOCKS_NO_PUT_HOLE		0x0200
 
+/*
+ * The bit position of this flag must not overlap with any of the
+ * EXT4_GET_BLOCKS_*.  It is used by ext4_ext_find_extent(),
+ * read_extent_tree_block(), ext4_split_extent_at(),
+ * ext4_ext_insert_extent(), and ext4_ext_create_new_leaf() to
+ * indicate that the we shouldn't be caching the extents when reading
+ * from the extent tree while a truncate or punch hole operation
+ * is in progress.
+ */
+#define EXT4_EX_NOCACHE				0x0400
+
 /*
  * Flags used by ext4_free_blocks
  */
@@ -2684,7 +2695,8 @@ extern int ext4_ext_insert_extent(handle_t *, struct inode *,
 				  struct ext4_ext_path *,
 				  struct ext4_extent *, int);
 extern struct ext4_ext_path *ext4_ext_find_extent(struct inode *, ext4_lblk_t,
-						  struct ext4_ext_path *);
+						  struct ext4_ext_path *,
+						  int flags);
 extern void ext4_ext_drop_refs(struct ext4_ext_path *);
 extern int ext4_ext_check_inode(struct inode *inode);
 extern int ext4_find_delalloc_range(struct inode *inode,

fs/ext4/extents.c

@@ -466,7 +466,8 @@ int ext4_ext_check_inode(struct inode *inode)
 
 static struct buffer_head *
 __read_extent_tree_block(const char *function, unsigned int line,
-			 struct inode *inode, ext4_fsblk_t pblk, int depth)
+			 struct inode *inode, ext4_fsblk_t pblk, int depth,
+			 int flags)
 {
 	struct buffer_head		*bh;
 	int				err;
@@ -488,6 +489,32 @@ __read_extent_tree_block(const char *function, unsigned int line,
 	if (err)
 		goto errout;
 	set_buffer_verified(bh);
+	/*
+	 * If this is a leaf block, cache all of its entries
+	 */
+	if (!(flags & EXT4_EX_NOCACHE) && depth == 0) {
+		struct ext4_extent_header *eh = ext_block_hdr(bh);
+		struct ext4_extent *ex = EXT_FIRST_EXTENT(eh);
+		ext4_lblk_t prev = 0;
+		int i;
+
+		for (i = le16_to_cpu(eh->eh_entries); i > 0; i--, ex++) {
+			unsigned int status = EXTENT_STATUS_WRITTEN;
+			ext4_lblk_t lblk = le32_to_cpu(ex->ee_block);
+			int len = ext4_ext_get_actual_len(ex);
+
+			if (prev && (prev != lblk))
+				ext4_es_cache_extent(inode, prev,
+						     lblk - prev, ~0,
+						     EXTENT_STATUS_HOLE);
+
+			if (ext4_ext_is_uninitialized(ex))
+				status = EXTENT_STATUS_UNWRITTEN;
+			ext4_es_cache_extent(inode, lblk, len,
+					     ext4_ext_pblock(ex), status);
+			prev = lblk + len;
+		}
+	}
 	return bh;
 errout:
 	put_bh(bh);
@@ -495,8 +522,9 @@ __read_extent_tree_block(const char *function, unsigned int line,
 
 }
 
-#define read_extent_tree_block(inode, pblk, depth)		\
-	__read_extent_tree_block(__func__, __LINE__, (inode), (pblk), (depth))
+#define read_extent_tree_block(inode, pblk, depth, flags)		\
+	__read_extent_tree_block(__func__, __LINE__, (inode), (pblk),   \
+				 (depth), (flags))
 
 #ifdef EXT_DEBUG
 static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
@@ -730,7 +758,7 @@ int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
 
 struct ext4_ext_path *
 ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
-					struct ext4_ext_path *path)
+		     struct ext4_ext_path *path, int flags)
 {
 	struct ext4_extent_header *eh;
 	struct buffer_head *bh;
@@ -762,7 +790,8 @@ ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
 		path[ppos].p_depth = i;
 		path[ppos].p_ext = NULL;
 
-		bh = read_extent_tree_block(inode, path[ppos].p_block, --i);
+		bh = read_extent_tree_block(inode, path[ppos].p_block, --i,
+					    flags);
 		if (IS_ERR(bh)) {
 			ret = PTR_ERR(bh);
 			goto err;
@@ -1199,7 +1228,8 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
  * if no free index is found, then it requests in-depth growing.
  */
 static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
-				    unsigned int flags,
+				    unsigned int mb_flags,
+				    unsigned int gb_flags,
 				    struct ext4_ext_path *path,
 				    struct ext4_extent *newext)
 {
@@ -1221,7 +1251,7 @@ static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
 	if (EXT_HAS_FREE_INDEX(curp)) {
 		/* if we found index with free entry, then use that
 		 * entry: create all needed subtree and add new leaf */
-		err = ext4_ext_split(handle, inode, flags, path, newext, i);
+		err = ext4_ext_split(handle, inode, mb_flags, path, newext, i);
 		if (err)
 			goto out;
 
@@ -1229,12 +1259,12 @@ static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
 		ext4_ext_drop_refs(path);
 		path = ext4_ext_find_extent(inode,
 				    (ext4_lblk_t)le32_to_cpu(newext->ee_block),
-				    path);
+				    path, gb_flags);
 		if (IS_ERR(path))
 			err = PTR_ERR(path);
 	} else {
 		/* tree is full, time to grow in depth */
-		err = ext4_ext_grow_indepth(handle, inode, flags, newext);
+		err = ext4_ext_grow_indepth(handle, inode, mb_flags, newext);
 		if (err)
 			goto out;
 
@@ -1242,7 +1272,7 @@ static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
 		ext4_ext_drop_refs(path);
 		path = ext4_ext_find_extent(inode,
 				   (ext4_lblk_t)le32_to_cpu(newext->ee_block),
-				    path);
+				    path, gb_flags);
 		if (IS_ERR(path)) {
 			err = PTR_ERR(path);
 			goto out;
@@ -1415,7 +1445,7 @@ static int ext4_ext_search_right(struct inode *inode,
 	while (++depth < path->p_depth) {
 		/* subtract from p_depth to get proper eh_depth */
 		bh = read_extent_tree_block(inode, block,
-					    path->p_depth - depth);
+					    path->p_depth - depth, 0);
 		if (IS_ERR(bh))
 			return PTR_ERR(bh);
 		eh = ext_block_hdr(bh);
@@ -1424,7 +1454,7 @@ static int ext4_ext_search_right(struct inode *inode,
 		put_bh(bh);
 	}
 
-	bh = read_extent_tree_block(inode, block, path->p_depth - depth);
+	bh = read_extent_tree_block(inode, block, path->p_depth - depth, 0);
 	if (IS_ERR(bh))
 		return PTR_ERR(bh);
 	eh = ext_block_hdr(bh);
@@ -1786,7 +1816,7 @@ static unsigned int ext4_ext_check_overlap(struct ext4_sb_info *sbi,
  */
 int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 				struct ext4_ext_path *path,
-				struct ext4_extent *newext, int flag)
+				struct ext4_extent *newext, int gb_flags)
 {
 	struct ext4_extent_header *eh;
 	struct ext4_extent *ex, *fex;
@@ -1795,7 +1825,7 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 	int depth, len, err;
 	ext4_lblk_t next;
 	unsigned uninitialized = 0;
-	int flags = 0;
+	int mb_flags = 0;
 
 	if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
 		EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
@@ -1810,7 +1840,7 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 	}
 
 	/* try to insert block into found extent and return */
-	if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)) {
+	if (ex && !(gb_flags & EXT4_GET_BLOCKS_PRE_IO)) {
 
 		/*
 		 * Try to see whether we should rather test the extent on
@@ -1913,7 +1943,7 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 	if (next != EXT_MAX_BLOCKS) {
 		ext_debug("next leaf block - %u\n", next);
 		BUG_ON(npath != NULL);
-		npath = ext4_ext_find_extent(inode, next, NULL);
+		npath = ext4_ext_find_extent(inode, next, NULL, 0);
 		if (IS_ERR(npath))
 			return PTR_ERR(npath);
 		BUG_ON(npath->p_depth != path->p_depth);
@@ -1932,9 +1962,10 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 	 * There is no free space in the found leaf.
 	 * We're gonna add a new leaf in the tree.
 	 */
-	if (flag & EXT4_GET_BLOCKS_METADATA_NOFAIL)
-		flags = EXT4_MB_USE_RESERVED;
-	err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext);
+	if (gb_flags & EXT4_GET_BLOCKS_METADATA_NOFAIL)
+		mb_flags = EXT4_MB_USE_RESERVED;
+	err = ext4_ext_create_new_leaf(handle, inode, mb_flags, gb_flags,
+				       path, newext);
 	if (err)
 		goto cleanup;
 	depth = ext_depth(inode);
@@ -2000,7 +2031,7 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 
 merge:
 	/* try to merge extents */
-	if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
+	if (!(gb_flags & EXT4_GET_BLOCKS_PRE_IO))
 		ext4_ext_try_to_merge(handle, inode, path, nearex);
 
 
@@ -2043,7 +2074,7 @@ static int ext4_fill_fiemap_extents(struct inode *inode,
 			path = NULL;
 		}
 
-		path = ext4_ext_find_extent(inode, block, path);
+		path = ext4_ext_find_extent(inode, block, path, 0);
 		if (IS_ERR(path)) {
 			up_read(&EXT4_I(inode)->i_data_sem);
 			err = PTR_ERR(path);
@@ -2705,7 +2736,7 @@ int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
 		ext4_lblk_t ee_block;
 
 		/* find extent for this block */
-		path = ext4_ext_find_extent(inode, end, NULL);
+		path = ext4_ext_find_extent(inode, end, NULL, EXT4_EX_NOCACHE);
 		if (IS_ERR(path)) {
 			ext4_journal_stop(handle);
 			return PTR_ERR(path);
@@ -2747,6 +2778,7 @@ int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
 			 */
 			err = ext4_split_extent_at(handle, inode, path,
 					end + 1, split_flag,
+					EXT4_EX_NOCACHE |
 					EXT4_GET_BLOCKS_PRE_IO |
 					EXT4_GET_BLOCKS_METADATA_NOFAIL);
 
@@ -2823,7 +2855,8 @@ int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
 				  i + 1, ext4_idx_pblock(path[i].p_idx));
 			memset(path + i + 1, 0, sizeof(*path));
 			bh = read_extent_tree_block(inode,
-				ext4_idx_pblock(path[i].p_idx), depth - i - 1);
+				ext4_idx_pblock(path[i].p_idx), depth - i - 1,
+				EXT4_EX_NOCACHE);
 			if (IS_ERR(bh)) {
 				/* should we reset i_size? */
 				err = PTR_ERR(bh);
@@ -3170,7 +3203,7 @@ static int ext4_split_extent(handle_t *handle,
 	 * result in split of original leaf or extent zeroout.
 	 */
 	ext4_ext_drop_refs(path);
-	path = ext4_ext_find_extent(inode, map->m_lblk, path);
+	path = ext4_ext_find_extent(inode, map->m_lblk, path, 0);
 	if (IS_ERR(path))
 		return PTR_ERR(path);
 	depth = ext_depth(inode);
@@ -3554,7 +3587,7 @@ static int ext4_convert_unwritten_extents_endio(handle_t *handle,
 		if (err < 0)
 			goto out;
 		ext4_ext_drop_refs(path);
-		path = ext4_ext_find_extent(inode, map->m_lblk, path);
+		path = ext4_ext_find_extent(inode, map->m_lblk, path, 0);
 		if (IS_ERR(path)) {
 			err = PTR_ERR(path);
 			goto out;
@@ -4041,7 +4074,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
 	trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags);
 
 	/* find extent for this block */
-	path = ext4_ext_find_extent(inode, map->m_lblk, NULL);
+	path = ext4_ext_find_extent(inode, map->m_lblk, NULL, 0);
 	if (IS_ERR(path)) {
 		err = PTR_ERR(path);
 		path = NULL;
@@ -4760,6 +4793,6 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		error = ext4_fill_fiemap_extents(inode, start_blk,
 						 len_blks, fieinfo);
 	}
-
+	ext4_es_lru_add(inode);
 	return error;
 }

fs/ext4/extents_status.c

@@ -419,7 +419,7 @@ static void ext4_es_insert_extent_ext_check(struct inode *inode,
 	unsigned short ee_len;
 	int depth, ee_status, es_status;
 
-	path = ext4_ext_find_extent(inode, es->es_lblk, NULL);
+	path = ext4_ext_find_extent(inode, es->es_lblk, NULL, EXT4_EX_NOCACHE);
 	if (IS_ERR(path))
 		return;
 
@@ -683,6 +683,41 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk,
 	return err;
 }
 
+/*
+ * ext4_es_cache_extent() inserts information into the extent status
+ * tree if and only if there isn't information about the range in
+ * question already.
+ */
+void ext4_es_cache_extent(struct inode *inode, ext4_lblk_t lblk,
+			  ext4_lblk_t len, ext4_fsblk_t pblk,
+			  unsigned int status)
+{
+	struct extent_status *es;
+	struct extent_status newes;
+	ext4_lblk_t end = lblk + len - 1;
+
+	newes.es_lblk = lblk;
+	newes.es_len = len;
+	ext4_es_store_pblock(&newes, pblk);
+	ext4_es_store_status(&newes, status);
+	trace_ext4_es_cache_extent(inode, &newes);
+
+	if (!len)
+		return;
+
+	BUG_ON(end < lblk);
+
+	write_lock(&EXT4_I(inode)->i_es_lock);
+
+	es = __es_tree_search(&EXT4_I(inode)->i_es_tree.root, lblk);
+	if (es && ((es->es_lblk <= lblk) || (es->es_lblk <= end)))
+		goto out;
+
+	__es_insert_extent(inode, &newes);
+out:
+	write_unlock(&EXT4_I(inode)->i_es_lock);
+}
+
 /*
  * ext4_es_lookup_extent() looks up an extent in extent status tree.
  *

fs/ext4/extents_status.h

@@ -71,6 +71,9 @@ extern void ext4_es_init_tree(struct ext4_es_tree *tree);
 extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk,
 				 ext4_lblk_t len, ext4_fsblk_t pblk,
 				 unsigned int status);
+extern void ext4_es_cache_extent(struct inode *inode, ext4_lblk_t lblk,
+				 ext4_lblk_t len, ext4_fsblk_t pblk,
+				 unsigned int status);
 extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk,
 				 ext4_lblk_t len);
 extern void ext4_es_find_delayed_extent_range(struct inode *inode,

fs/ext4/migrate.c

@@ -39,7 +39,7 @@ static int finish_range(handle_t *handle, struct inode *inode,
 	newext.ee_block = cpu_to_le32(lb->first_block);
 	newext.ee_len   = cpu_to_le16(lb->last_block - lb->first_block + 1);
 	ext4_ext_store_pblock(&newext, lb->first_pblock);
-	path = ext4_ext_find_extent(inode, lb->first_block, NULL);
+	path = ext4_ext_find_extent(inode, lb->first_block, NULL, 0);
 
 	if (IS_ERR(path)) {
 		retval = PTR_ERR(path);

fs/ext4/move_extent.c

@@ -37,7 +37,7 @@ get_ext_path(struct inode *inode, ext4_lblk_t lblock,
 	int ret = 0;
 	struct ext4_ext_path *path;
 
-	path = ext4_ext_find_extent(inode, lblock, *orig_path);
+	path = ext4_ext_find_extent(inode, lblock, *orig_path, EXT4_EX_NOCACHE);
 	if (IS_ERR(path))
 		ret = PTR_ERR(path);
 	else if (path[ext_depth(inode)].p_ext == NULL)

include/trace/events/ext4.h

@@ -2192,7 +2192,7 @@ TRACE_EVENT(ext4_ext_remove_space_done,
 		  (unsigned short) __entry->eh_entries)
 );
 
-TRACE_EVENT(ext4_es_insert_extent,
+DECLARE_EVENT_CLASS(ext4__es_extent,
 	TP_PROTO(struct inode *inode, struct extent_status *es),
 
 	TP_ARGS(inode, es),
@@ -2222,6 +2222,18 @@ TRACE_EVENT(ext4_es_insert_extent,
 		  __entry->pblk, show_extent_status(__entry->status))
 );
 
+DEFINE_EVENT(ext4__es_extent, ext4_es_insert_extent,
+	TP_PROTO(struct inode *inode, struct extent_status *es),
+
+	TP_ARGS(inode, es)
+);
+
+DEFINE_EVENT(ext4__es_extent, ext4_es_cache_extent,
+	TP_PROTO(struct inode *inode, struct extent_status *es),
+
+	TP_ARGS(inode, es)
+);
+
 TRACE_EVENT(ext4_es_remove_extent,
 	TP_PROTO(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len),