Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

Pull ext4 updates from Ted Ts'o:
 "Lots of bugs fixes, including Zheng and Jan's extent status shrinker
  fixes, which should improve CPU utilization and potential soft lockups
  under heavy memory pressure, and Eric Whitney's bigalloc fixes"

* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (26 commits)
  ext4: ext4_da_convert_inline_data_to_extent drop locked page after error
  ext4: fix suboptimal seek_{data,hole} extents traversial
  ext4: ext4_inline_data_fiemap should respect callers argument
  ext4: prevent fsreentrance deadlock for inline_data
  ext4: forbid journal_async_commit in data=ordered mode
  jbd2: remove unnecessary NULL check before iput()
  ext4: Remove an unnecessary check for NULL before iput()
  ext4: remove unneeded code in ext4_unlink
  ext4: don't count external journal blocks as overhead
  ext4: remove never taken branch from ext4_ext_shift_path_extents()
  ext4: create nojournal_checksum mount option
  ext4: update comments regarding ext4_delete_inode()
  ext4: cleanup GFP flags inside resize path
  ext4: introduce aging to extent status tree
  ext4: cleanup flag definitions for extent status tree
  ext4: limit number of scanned extents in status tree shrinker
  ext4: move handling of list of shrinkable inodes into extent status code
  ext4: change LRU to round-robin in extent status tree shrinker
  ext4: cache extent hole in extent status tree for ext4_da_map_blocks()
  ext4: fix block reservation for bigalloc filesystems
  ...

14 files changed, 526 insertions, 518 deletions

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index db3f772e57ae..a75fba67bb1f 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -158,17 +158,8 @@ struct ext4_allocation_request {
 #define EXT4_MAP_MAPPED		(1 << BH_Mapped)
 #define EXT4_MAP_UNWRITTEN	(1 << BH_Unwritten)
 #define EXT4_MAP_BOUNDARY	(1 << BH_Boundary)
-/* Sometimes (in the bigalloc case, from ext4_da_get_block_prep) the caller of
- * ext4_map_blocks wants to know whether or not the underlying cluster has
- * already been accounted for. EXT4_MAP_FROM_CLUSTER conveys to the caller that
- * the requested mapping was from previously mapped (or delayed allocated)
- * cluster. We use BH_AllocFromCluster only for this flag. BH_AllocFromCluster
- * should never appear on buffer_head's state flags.
- */
-#define EXT4_MAP_FROM_CLUSTER	(1 << BH_AllocFromCluster)
 #define EXT4_MAP_FLAGS		(EXT4_MAP_NEW | EXT4_MAP_MAPPED |\
-				 EXT4_MAP_UNWRITTEN | EXT4_MAP_BOUNDARY |\
-				 EXT4_MAP_FROM_CLUSTER)
+				 EXT4_MAP_UNWRITTEN | EXT4_MAP_BOUNDARY)
 
 struct ext4_map_blocks {
 	ext4_fsblk_t m_pblk;
@@ -565,10 +556,8 @@ enum {
 #define EXT4_GET_BLOCKS_KEEP_SIZE		0x0080
 	/* Do not take i_data_sem locking in ext4_map_blocks */
 #define EXT4_GET_BLOCKS_NO_LOCK			0x0100
-	/* Do not put hole in extent cache */
-#define EXT4_GET_BLOCKS_NO_PUT_HOLE		0x0200
 	/* Convert written extents to unwritten */
-#define EXT4_GET_BLOCKS_CONVERT_UNWRITTEN	0x0400
+#define EXT4_GET_BLOCKS_CONVERT_UNWRITTEN	0x0200
 
 /*
  * The bit position of these flags must not overlap with any of the
@@ -889,10 +878,12 @@ struct ext4_inode_info {
 	/* extents status tree */
 	struct ext4_es_tree i_es_tree;
 	rwlock_t i_es_lock;
-	struct list_head i_es_lru;
+	struct list_head i_es_list;
 	unsigned int i_es_all_nr;	/* protected by i_es_lock */
-	unsigned int i_es_lru_nr;	/* protected by i_es_lock */
-	unsigned long i_touch_when;	/* jiffies of last accessing */
+	unsigned int i_es_shk_nr;	/* protected by i_es_lock */
+	ext4_lblk_t i_es_shrink_lblk;	/* Offset where we start searching for
+					   extents to shrink. Protected by
+					   i_es_lock  */
 
 	/* ialloc */
 	ext4_group_t	i_last_alloc_group;
@@ -1337,10 +1328,11 @@ struct ext4_sb_info {
 
 	/* Reclaim extents from extent status tree */
 	struct shrinker s_es_shrinker;
-	struct list_head s_es_lru;
+	struct list_head s_es_list;	/* List of inodes with reclaimable extents */
+	long s_es_nr_inode;
 	struct ext4_es_stats s_es_stats;
 	struct mb_cache *s_mb_cache;
-	spinlock_t s_es_lru_lock ____cacheline_aligned_in_smp;
+	spinlock_t s_es_lock ____cacheline_aligned_in_smp;
 
 	/* Ratelimit ext4 messages. */
 	struct ratelimit_state s_err_ratelimit_state;
@@ -2196,7 +2188,6 @@ extern int ext4_calculate_overhead(struct super_block *sb);
 extern void ext4_superblock_csum_set(struct super_block *sb);
 extern void *ext4_kvmalloc(size_t size, gfp_t flags);
 extern void *ext4_kvzalloc(size_t size, gfp_t flags);
-extern void ext4_kvfree(void *ptr);
 extern int ext4_alloc_flex_bg_array(struct super_block *sb,
 				    ext4_group_t ngroup);
 extern const char *ext4_decode_error(struct super_block *sb, int errno,
@@ -2647,7 +2638,7 @@ extern struct buffer_head *ext4_get_first_inline_block(struct inode *inode,
 					int *retval);
 extern int ext4_inline_data_fiemap(struct inode *inode,
 				   struct fiemap_extent_info *fieinfo,
-				   int *has_inline);
+				   int *has_inline, __u64 start, __u64 len);
 extern int ext4_try_to_evict_inline_data(handle_t *handle,
 					 struct inode *inode,
 					 int needed);
@@ -2795,16 +2786,6 @@ extern int ext4_bio_write_page(struct ext4_io_submit *io,
 extern int ext4_multi_mount_protect(struct super_block *, ext4_fsblk_t);
 
 /*
- * Note that these flags will never ever appear in a buffer_head's state flag.
- * See EXT4_MAP_... to see where this is used.
- */
-enum ext4_state_bits {
-	BH_AllocFromCluster	/* allocated blocks were part of already
-				 * allocated cluster. */
-	= BH_JBDPrivateStart
-};
-
-/*
  * Add new method to test whether block and inode bitmaps are properly
  * initialized. With uninit_bg reading the block from disk is not enough
  * to mark the bitmap uptodate. We need to also zero-out the bitmap
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 0b16fb4c06d3..e5d3eadf47b1 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2306,16 +2306,16 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
 				ext4_lblk_t block)
 {
 	int depth = ext_depth(inode);
-	unsigned long len = 0;
-	ext4_lblk_t lblock = 0;
+	ext4_lblk_t len;
+	ext4_lblk_t lblock;
 	struct ext4_extent *ex;
+	struct extent_status es;
 
 	ex = path[depth].p_ext;
 	if (ex == NULL) {
-		/*
-		 * there is no extent yet, so gap is [0;-] and we
-		 * don't cache it
-		 */
+		/* there is no extent yet, so gap is [0;-] */
+		lblock = 0;
+		len = EXT_MAX_BLOCKS;
 		ext_debug("cache gap(whole file):");
 	} else if (block < le32_to_cpu(ex->ee_block)) {
 		lblock = block;
@@ -2324,9 +2324,6 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
 				block,
 				le32_to_cpu(ex->ee_block),
 				 ext4_ext_get_actual_len(ex));
-		if (!ext4_find_delalloc_range(inode, lblock, lblock + len - 1))
-			ext4_es_insert_extent(inode, lblock, len, ~0,
-					      EXTENT_STATUS_HOLE);
 	} else if (block >= le32_to_cpu(ex->ee_block)
 			+ ext4_ext_get_actual_len(ex)) {
 		ext4_lblk_t next;
@@ -2340,14 +2337,19 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
 				block);
 		BUG_ON(next == lblock);
 		len = next - lblock;
-		if (!ext4_find_delalloc_range(inode, lblock, lblock + len - 1))
-			ext4_es_insert_extent(inode, lblock, len, ~0,
-					      EXTENT_STATUS_HOLE);
 	} else {
 		BUG();
 	}
 
-	ext_debug(" -> %u:%lu\n", lblock, len);
+	ext4_es_find_delayed_extent_range(inode, lblock, lblock + len - 1, &es);
+	if (es.es_len) {
+		/* There's delayed extent containing lblock? */
+		if (es.es_lblk <= lblock)
+			return;
+		len = min(es.es_lblk - lblock, len);
+	}
+	ext_debug(" -> %u:%u\n", lblock, len);
+	ext4_es_insert_extent(inode, lblock, len, ~0, EXTENT_STATUS_HOLE);
 }
 
 /*
@@ -2481,7 +2483,7 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 			      ext4_lblk_t from, ext4_lblk_t to)
 {
 	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
-	unsigned short ee_len =  ext4_ext_get_actual_len(ex);
+	unsigned short ee_len = ext4_ext_get_actual_len(ex);
 	ext4_fsblk_t pblk;
 	int flags = get_default_free_blocks_flags(inode);
 
@@ -2490,7 +2492,7 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 	 * at the beginning of the extent.  Instead, we make a note
 	 * that we tried freeing the cluster, and check to see if we
 	 * need to free it on a subsequent call to ext4_remove_blocks,
-	 * or at the end of the ext4_truncate() operation.
+	 * or at the end of ext4_ext_rm_leaf or ext4_ext_remove_space.
 	 */
 	flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER;
 
@@ -2501,8 +2503,8 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 	 * partial cluster here.
 	 */
 	pblk = ext4_ext_pblock(ex) + ee_len - 1;
-	if ((*partial_cluster > 0) &&
-	    (EXT4_B2C(sbi, pblk) != *partial_cluster)) {
+	if (*partial_cluster > 0 &&
+	    *partial_cluster != (long long) EXT4_B2C(sbi, pblk)) {
 		ext4_free_blocks(handle, inode, NULL,
 				 EXT4_C2B(sbi, *partial_cluster),
 				 sbi->s_cluster_ratio, flags);
@@ -2528,7 +2530,7 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 	    && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
 		/* tail removal */
 		ext4_lblk_t num;
-		unsigned int unaligned;
+		long long first_cluster;
 
 		num = le32_to_cpu(ex->ee_block) + ee_len - from;
 		pblk = ext4_ext_pblock(ex) + ee_len - num;
@@ -2538,7 +2540,7 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 		 * used by any other extent (partial_cluster is negative).
 		 */
 		if (*partial_cluster < 0 &&
-		    -(*partial_cluster) == EXT4_B2C(sbi, pblk + num - 1))
+		    *partial_cluster == -(long long) EXT4_B2C(sbi, pblk+num-1))
 			flags |= EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER;
 
 		ext_debug("free last %u blocks starting %llu partial %lld\n",
@@ -2549,21 +2551,24 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 		 * beginning of a cluster, and we removed the entire
 		 * extent and the cluster is not used by any other extent,
 		 * save the partial cluster here, since we might need to
-		 * delete if we determine that the truncate operation has
-		 * removed all of the blocks in the cluster.
+		 * delete if we determine that the truncate or punch hole
+		 * operation has removed all of the blocks in the cluster.
+		 * If that cluster is used by another extent, preserve its
+		 * negative value so it isn't freed later on.
 		 *
-		 * On the other hand, if we did not manage to free the whole
-		 * extent, we have to mark the cluster as used (store negative
-		 * cluster number in partial_cluster).
+		 * If the whole extent wasn't freed, we've reached the
+		 * start of the truncated/punched region and have finished
+		 * removing blocks.  If there's a partial cluster here it's
+		 * shared with the remainder of the extent and is no longer
+		 * a candidate for removal.
 		 */
-		unaligned = EXT4_PBLK_COFF(sbi, pblk);
-		if (unaligned && (ee_len == num) &&
-		    (*partial_cluster != -((long long)EXT4_B2C(sbi, pblk))))
-			*partial_cluster = EXT4_B2C(sbi, pblk);
-		else if (unaligned)
-			*partial_cluster = -((long long)EXT4_B2C(sbi, pblk));
-		else if (*partial_cluster > 0)
+		if (EXT4_PBLK_COFF(sbi, pblk) && ee_len == num) {
+			first_cluster = (long long) EXT4_B2C(sbi, pblk);
+			if (first_cluster != -*partial_cluster)
+				*partial_cluster = first_cluster;
+		} else {
 			*partial_cluster = 0;
+		}
 	} else
 		ext4_error(sbi->s_sb, "strange request: removal(2) "
 			   "%u-%u from %u:%u\n",
@@ -2574,15 +2579,16 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
 
 /*
  * ext4_ext_rm_leaf() Removes the extents associated with the
- * blocks appearing between "start" and "end", and splits the extents
- * if "start" and "end" appear in the same extent
+ * blocks appearing between "start" and "end".  Both "start"
+ * and "end" must appear in the same extent or EIO is returned.
  *
  * @handle: The journal handle
  * @inode:  The files inode
  * @path:   The path to the leaf
  * @partial_cluster: The cluster which we'll have to free if all extents
- *                   has been released from it. It gets negative in case
- *                   that the cluster is still used.
+ *                   has been released from it.  However, if this value is
+ *                   negative, it's a cluster just to the right of the
+ *                   punched region and it must not be freed.
  * @start:  The first block to remove
  * @end:   The last block to remove
  */
@@ -2621,27 +2627,6 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 	ex_ee_block = le32_to_cpu(ex->ee_block);
 	ex_ee_len = ext4_ext_get_actual_len(ex);
 
-	/*
-	 * If we're starting with an extent other than the last one in the
-	 * node, we need to see if it shares a cluster with the extent to
-	 * the right (towards the end of the file). If its leftmost cluster
-	 * is this extent's rightmost cluster and it is not cluster aligned,
-	 * we'll mark it as a partial that is not to be deallocated.
-	 */
-
-	if (ex != EXT_LAST_EXTENT(eh)) {
-		ext4_fsblk_t current_pblk, right_pblk;
-		long long current_cluster, right_cluster;
-
-		current_pblk = ext4_ext_pblock(ex) + ex_ee_len - 1;
-		current_cluster = (long long)EXT4_B2C(sbi, current_pblk);
-		right_pblk = ext4_ext_pblock(ex + 1);
-		right_cluster = (long long)EXT4_B2C(sbi, right_pblk);
-		if (current_cluster == right_cluster &&
-			EXT4_PBLK_COFF(sbi, right_pblk))
-			*partial_cluster = -right_cluster;
-	}
-
 	trace_ext4_ext_rm_leaf(inode, start, ex, *partial_cluster);
 
 	while (ex >= EXT_FIRST_EXTENT(eh) &&
@@ -2666,14 +2651,16 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 		if (end < ex_ee_block) {
 			/*
 			 * We're going to skip this extent and move to another,
-			 * so if this extent is not cluster aligned we have
-			 * to mark the current cluster as used to avoid
-			 * accidentally freeing it later on
+			 * so note that its first cluster is in use to avoid
+			 * freeing it when removing blocks.  Eventually, the
+			 * right edge of the truncated/punched region will
+			 * be just to the left.
 			 */
-			pblk = ext4_ext_pblock(ex);
-			if (EXT4_PBLK_COFF(sbi, pblk))
+			if (sbi->s_cluster_ratio > 1) {
+				pblk = ext4_ext_pblock(ex);
 				*partial_cluster =
-					-((long long)EXT4_B2C(sbi, pblk));
+					-(long long) EXT4_B2C(sbi, pblk);
+			}
 			ex--;
 			ex_ee_block = le32_to_cpu(ex->ee_block);
 			ex_ee_len = ext4_ext_get_actual_len(ex);
@@ -2749,8 +2736,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 					sizeof(struct ext4_extent));
 			}
 			le16_add_cpu(&eh->eh_entries, -1);
-		} else if (*partial_cluster > 0)
-			*partial_cluster = 0;
+		}
 
 		err = ext4_ext_dirty(handle, inode, path + depth);
 		if (err)
@@ -2769,20 +2755,18 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 	/*
 	 * If there's a partial cluster and at least one extent remains in
 	 * the leaf, free the partial cluster if it isn't shared with the
-	 * current extent.  If there's a partial cluster and no extents
-	 * remain in the leaf, it can't be freed here.  It can only be
-	 * freed when it's possible to determine if it's not shared with
-	 * any other extent - when the next leaf is processed or when space
-	 * removal is complete.
+	 * current extent.  If it is shared with the current extent
+	 * we zero partial_cluster because we've reached the start of the
+	 * truncated/punched region and we're done removing blocks.
 	 */
-	if (*partial_cluster > 0 && eh->eh_entries &&
-	    (EXT4_B2C(sbi, ext4_ext_pblock(ex) + ex_ee_len - 1) !=
-	     *partial_cluster)) {
-		int flags = get_default_free_blocks_flags(inode);
-
-		ext4_free_blocks(handle, inode, NULL,
-				 EXT4_C2B(sbi, *partial_cluster),
-				 sbi->s_cluster_ratio, flags);
+	if (*partial_cluster > 0 && ex >= EXT_FIRST_EXTENT(eh)) {
+		pblk = ext4_ext_pblock(ex) + ex_ee_len - 1;
+		if (*partial_cluster != (long long) EXT4_B2C(sbi, pblk)) {
+			ext4_free_blocks(handle, inode, NULL,
+					 EXT4_C2B(sbi, *partial_cluster),
+					 sbi->s_cluster_ratio,
+					 get_default_free_blocks_flags(inode));
+		}
 		*partial_cluster = 0;
 	}
 
@@ -2819,7 +2803,7 @@ ext4_ext_more_to_rm(struct ext4_ext_path *path)
 int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
 			  ext4_lblk_t end)
 {
-	struct super_block *sb = inode->i_sb;
+	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 	int depth = ext_depth(inode);
 	struct ext4_ext_path *path = NULL;
 	long long partial_cluster = 0;
@@ -2845,9 +2829,10 @@ again:
 	 */
 	if (end < EXT_MAX_BLOCKS - 1) {
 		struct ext4_extent *ex;
-		ext4_lblk_t ee_block;
+		ext4_lblk_t ee_block, ex_end, lblk;
+		ext4_fsblk_t pblk;
 
-		/* find extent for this block */
+		/* find extent for or closest extent to this block */
 		path = ext4_find_extent(inode, end, NULL, EXT4_EX_NOCACHE);
 		if (IS_ERR(path)) {
 			ext4_journal_stop(handle);
@@ -2867,6 +2852,7 @@ again:
 		}
 
 		ee_block = le32_to_cpu(ex->ee_block);
+		ex_end = ee_block + ext4_ext_get_actual_len(ex) - 1;
 
 		/*
 		 * See if the last block is inside the extent, if so split
@@ -2874,8 +2860,19 @@ again:
 		 * tail of the first part of the split extent in
 		 * ext4_ext_rm_leaf().
 		 */
-		if (end >= ee_block &&
-		    end < ee_block + ext4_ext_get_actual_len(ex) - 1) {
+		if (end >= ee_block && end < ex_end) {
+
+			/*
+			 * If we're going to split the extent, note that
+			 * the cluster containing the block after 'end' is
+			 * in use to avoid freeing it when removing blocks.
+			 */
+			if (sbi->s_cluster_ratio > 1) {
+				pblk = ext4_ext_pblock(ex) + end - ee_block + 2;
+				partial_cluster =
+					-(long long) EXT4_B2C(sbi, pblk);
+			}
+
 			/*
 			 * Split the extent in two so that 'end' is the last
 			 * block in the first new extent. Also we should not
@@ -2886,6 +2883,24 @@ again:
 							 end + 1, 1);
 			if (err < 0)
 				goto out;
+
+		} else if (sbi->s_cluster_ratio > 1 && end >= ex_end) {
+			/*
+			 * If there's an extent to the right its first cluster
+			 * contains the immediate right boundary of the
+			 * truncated/punched region.  Set partial_cluster to
+			 * its negative value so it won't be freed if shared
+			 * with the current extent.  The end < ee_block case
+			 * is handled in ext4_ext_rm_leaf().
+			 */
+			lblk = ex_end + 1;
+			err = ext4_ext_search_right(inode, path, &lblk, &pblk,
+						    &ex);
+			if (err)
+				goto out;
+			if (pblk)
+				partial_cluster =
+					-(long long) EXT4_B2C(sbi, pblk);
 		}
 	}
 	/*
@@ -2996,16 +3011,18 @@ again:
 	trace_ext4_ext_remove_space_done(inode, start, end, depth,
 			partial_cluster, path->p_hdr->eh_entries);
 
-	/* If we still have something in the partial cluster and we have removed
+	/*
+	 * If we still have something in the partial cluster and we have removed
 	 * even the first extent, then we should free the blocks in the partial
-	 * cluster as well. */
-	if (partial_cluster > 0 && path->p_hdr->eh_entries == 0) {
-		int flags = get_default_free_blocks_flags(inode);
-
+	 * cluster as well.  (This code will only run when there are no leaves
+	 * to the immediate left of the truncated/punched region.)
+	 */
+	if (partial_cluster > 0 && err == 0) {
+		/* don't zero partial_cluster since it's not used afterwards */
 		ext4_free_blocks(handle, inode, NULL,
-				 EXT4_C2B(EXT4_SB(sb), partial_cluster),
-				 EXT4_SB(sb)->s_cluster_ratio, flags);
-		partial_cluster = 0;
+				 EXT4_C2B(sbi, partial_cluster),
+				 sbi->s_cluster_ratio,
+				 get_default_free_blocks_flags(inode));
 	}
 
 	/* TODO: flexible tree reduction should be here */
@@ -4267,6 +4284,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
 	ext4_io_end_t *io = ext4_inode_aio(inode);
 	ext4_lblk_t cluster_offset;
 	int set_unwritten = 0;
+	bool map_from_cluster = false;
 
 	ext_debug("blocks %u/%u requested for inode %lu\n",
 		  map->m_lblk, map->m_len, inode->i_ino);
@@ -4343,10 +4361,6 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
 		}
 	}
 
-	if ((sbi->s_cluster_ratio > 1) &&
-	    ext4_find_delalloc_cluster(inode, map->m_lblk))
-		map->m_flags |= EXT4_MAP_FROM_CLUSTER;
-
 	/*
 	 * requested block isn't allocated yet;
 	 * we couldn't try to create block if create flag is zero
@@ -4356,15 +4370,13 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
 		 * put just found gap into cache to speed up
 		 * subsequent requests
 		 */
-		if ((flags & EXT4_GET_BLOCKS_NO_PUT_HOLE) == 0)
-			ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
+		ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
 		goto out2;
 	}
 
 	/*
 	 * Okay, we need to do block allocation.
 	 */
-	map->m_flags &= ~EXT4_MAP_FROM_CLUSTER;
 	newex.ee_block = cpu_to_le32(map->m_lblk);
 	cluster_offset = EXT4_LBLK_COFF(sbi, map->m_lblk);
 
@@ -4376,7 +4388,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
 	    get_implied_cluster_alloc(inode->i_sb, map, ex, path)) {
 		ar.len = allocated = map->m_len;
 		newblock = map->m_pblk;
-		map->m_flags |= EXT4_MAP_FROM_CLUSTER;
+		map_from_cluster = true;
 		goto got_allocated_blocks;
 	}
 
@@ -4397,7 +4409,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
 	    get_implied_cluster_alloc(inode->i_sb, map, ex2, path)) {
 		ar.len = allocated = map->m_len;
 		newblock = map->m_pblk;
-		map->m_flags |= EXT4_MAP_FROM_CLUSTER;
+		map_from_cluster = true;
 		goto got_allocated_blocks;
 	}
 
@@ -4523,7 +4535,7 @@ got_allocated_blocks:
 		 */
 		reserved_clusters = get_reserved_cluster_alloc(inode,
 						map->m_lblk, allocated);
-		if (map->m_flags & EXT4_MAP_FROM_CLUSTER) {
+		if (map_from_cluster) {
 			if (reserved_clusters) {
 				/*
 				 * We have clusters reserved for this range.
@@ -4620,7 +4632,6 @@ out2:
 
 	trace_ext4_ext_map_blocks_exit(inode, flags, map,
 				       err ? err : allocated);
-	ext4_es_lru_add(inode);
 	return err ? err : allocated;
 }
 
@@ -5140,7 +5151,8 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	if (ext4_has_inline_data(inode)) {
 		int has_inline = 1;
 
-		error = ext4_inline_data_fiemap(inode, fieinfo, &has_inline);
+		error = ext4_inline_data_fiemap(inode, fieinfo, &has_inline,
+						start, len);
 
 		if (has_inline)
 			return error;
@@ -5154,8 +5166,8 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 
 	/* fallback to generic here if not in extents fmt */
 	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
-		return generic_block_fiemap(inode, fieinfo, start, len,
-			ext4_get_block);
+		return __generic_block_fiemap(inode, fieinfo, start, len,
+					      ext4_get_block);
 
 	if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
 		return -EBADR;
@@ -5179,7 +5191,6 @@ 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;
 }
 
@@ -5239,8 +5250,6 @@ ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift,
 				return -EIO;
 
 			ex_last = EXT_LAST_EXTENT(path[depth].p_hdr);
-			if (!ex_last)
-				return -EIO;
 
 			err = ext4_access_path(handle, inode, path + depth);
 			if (err)
diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c
index 94e7855ae71b..e04d45733976 100644
--- a/fs/ext4/extents_status.c
+++ b/fs/ext4/extents_status.c
@@ -147,10 +147,9 @@ static struct kmem_cache *ext4_es_cachep;
 static int __es_insert_extent(struct inode *inode, struct extent_status *newes);
 static int __es_remove_extent(struct inode *inode, ext4_lblk_t lblk,
 			      ext4_lblk_t end);
-static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei,
-				       int nr_to_scan);
-static int __ext4_es_shrink(struct ext4_sb_info *sbi, int nr_to_scan,
-			    struct ext4_inode_info *locked_ei);
+static int es_reclaim_extents(struct ext4_inode_info *ei, int *nr_to_scan);
+static int __es_shrink(struct ext4_sb_info *sbi, int nr_to_scan,
+		       struct ext4_inode_info *locked_ei);
 
 int __init ext4_init_es(void)
 {
@@ -298,6 +297,36 @@ out:
 	trace_ext4_es_find_delayed_extent_range_exit(inode, es);
 }
 
+static void ext4_es_list_add(struct inode *inode)
+{
+	struct ext4_inode_info *ei = EXT4_I(inode);
+	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+
+	if (!list_empty(&ei->i_es_list))
+		return;
+
+	spin_lock(&sbi->s_es_lock);
+	if (list_empty(&ei->i_es_list)) {
+		list_add_tail(&ei->i_es_list, &sbi->s_es_list);
+		sbi->s_es_nr_inode++;
+	}
+	spin_unlock(&sbi->s_es_lock);
+}
+
+static void ext4_es_list_del(struct inode *inode)
+{
+	struct ext4_inode_info *ei = EXT4_I(inode);
+	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
+
+	spin_lock(&sbi->s_es_lock);
+	if (!list_empty(&ei->i_es_list)) {
+		list_del_init(&ei->i_es_list);
+		sbi->s_es_nr_inode--;
+		WARN_ON_ONCE(sbi->s_es_nr_inode < 0);
+	}
+	spin_unlock(&sbi->s_es_lock);
+}
+
 static struct extent_status *
 ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len,
 		     ext4_fsblk_t pblk)
@@ -314,9 +343,10 @@ ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len,
 	 * We don't count delayed extent because we never try to reclaim them
 	 */
 	if (!ext4_es_is_delayed(es)) {
-		EXT4_I(inode)->i_es_lru_nr++;
+		if (!EXT4_I(inode)->i_es_shk_nr++)
+			ext4_es_list_add(inode);
 		percpu_counter_inc(&EXT4_SB(inode->i_sb)->
-					s_es_stats.es_stats_lru_cnt);
+					s_es_stats.es_stats_shk_cnt);
 	}
 
 	EXT4_I(inode)->i_es_all_nr++;
@@ -330,12 +360,13 @@ static void ext4_es_free_extent(struct inode *inode, struct extent_status *es)
 	EXT4_I(inode)->i_es_all_nr--;
 	percpu_counter_dec(&EXT4_SB(inode->i_sb)->s_es_stats.es_stats_all_cnt);
 
-	/* Decrease the lru counter when this es is not delayed */
+	/* Decrease the shrink counter when this es is not delayed */
 	if (!ext4_es_is_delayed(es)) {
-		BUG_ON(EXT4_I(inode)->i_es_lru_nr == 0);
-		EXT4_I(inode)->i_es_lru_nr--;
+		BUG_ON(EXT4_I(inode)->i_es_shk_nr == 0);
+		if (!--EXT4_I(inode)->i_es_shk_nr)
+			ext4_es_list_del(inode);
 		percpu_counter_dec(&EXT4_SB(inode->i_sb)->
-					s_es_stats.es_stats_lru_cnt);
+					s_es_stats.es_stats_shk_cnt);
 	}
 
 	kmem_cache_free(ext4_es_cachep, es);
@@ -351,7 +382,7 @@ static void ext4_es_free_extent(struct inode *inode, struct extent_status *es)
 static int ext4_es_can_be_merged(struct extent_status *es1,
 				 struct extent_status *es2)
 {
-	if (ext4_es_status(es1) != ext4_es_status(es2))
+	if (ext4_es_type(es1) != ext4_es_type(es2))
 		return 0;
 
 	if (((__u64) es1->es_len) + es2->es_len > EXT_MAX_BLOCKS) {
@@ -394,6 +425,8 @@ ext4_es_try_to_merge_left(struct inode *inode, struct extent_status *es)
 	es1 = rb_entry(node, struct extent_status, rb_node);
 	if (ext4_es_can_be_merged(es1, es)) {
 		es1->es_len += es->es_len;
+		if (ext4_es_is_referenced(es))
+			ext4_es_set_referenced(es1);
 		rb_erase(&es->rb_node, &tree->root);
 		ext4_es_free_extent(inode, es);
 		es = es1;
@@ -416,6 +449,8 @@ ext4_es_try_to_merge_right(struct inode *inode, struct extent_status *es)
 	es1 = rb_entry(node, struct extent_status, rb_node);
 	if (ext4_es_can_be_merged(es, es1)) {
 		es->es_len += es1->es_len;
+		if (ext4_es_is_referenced(es1))
+			ext4_es_set_referenced(es);
 		rb_erase(node, &tree->root);
 		ext4_es_free_extent(inode, es1);
 	}
@@ -683,8 +718,8 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk,
 		goto error;
 retry:
 	err = __es_insert_extent(inode, &newes);
-	if (err == -ENOMEM && __ext4_es_shrink(EXT4_SB(inode->i_sb), 1,
-					       EXT4_I(inode)))
+	if (err == -ENOMEM && __es_shrink(EXT4_SB(inode->i_sb),
+					  128, EXT4_I(inode)))
 		goto retry;
 	if (err == -ENOMEM && !ext4_es_is_delayed(&newes))
 		err = 0;
@@ -782,6 +817,8 @@ out:
 		es->es_lblk = es1->es_lblk;
 		es->es_len = es1->es_len;
 		es->es_pblk = es1->es_pblk;
+		if (!ext4_es_is_referenced(es))
+			ext4_es_set_referenced(es);
 		stats->es_stats_cache_hits++;
 	} else {
 		stats->es_stats_cache_misses++;
@@ -841,8 +878,8 @@ retry:
 				es->es_lblk = orig_es.es_lblk;
 				es->es_len = orig_es.es_len;
 				if ((err == -ENOMEM) &&
-				    __ext4_es_shrink(EXT4_SB(inode->i_sb), 1,
-						     EXT4_I(inode)))
+				    __es_shrink(EXT4_SB(inode->i_sb),
+							128, EXT4_I(inode)))
 					goto retry;
 				goto out;
 			}
@@ -914,6 +951,11 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk,
 	end = lblk + len - 1;
 	BUG_ON(end < lblk);
 
+	/*
+	 * ext4_clear_inode() depends on us taking i_es_lock unconditionally
+	 * so that we are sure __es_shrink() is done with the inode before it
+	 * is reclaimed.
+	 */
 	write_lock(&EXT4_I(inode)->i_es_lock);
 	err = __es_remove_extent(inode, lblk, end);
 	write_unlock(&EXT4_I(inode)->i_es_lock);
@@ -921,114 +963,75 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk,
 	return err;
 }
 
-static int ext4_inode_touch_time_cmp(void *priv, struct list_head *a,
-				     struct list_head *b)
-{
-	struct ext4_inode_info *eia, *eib;
-	eia = list_entry(a, struct ext4_inode_info, i_es_lru);
-	eib = list_entry(b, struct ext4_inode_info, i_es_lru);
-
-	if (ext4_test_inode_state(&eia->vfs_inode, EXT4_STATE_EXT_PRECACHED) &&
-	    !ext4_test_inode_state(&eib->vfs_inode, EXT4_STATE_EXT_PRECACHED))
-		return 1;
-	if (!ext4_test_inode_state(&eia->vfs_inode, EXT4_STATE_EXT_PRECACHED) &&
-	    ext4_test_inode_state(&eib->vfs_inode, EXT4_STATE_EXT_PRECACHED))
-		return -1;
-	if (eia->i_touch_when == eib->i_touch_when)
-		return 0;
-	if (time_after(eia->i_touch_when, eib->i_touch_when))
-		return 1;
-	else
-		return -1;
-}
-
-static int __ext4_es_shrink(struct ext4_sb_info *sbi, int nr_to_scan,
-			    struct ext4_inode_info *locked_ei)
+static int __es_shrink(struct ext4_sb_info *sbi, int nr_to_scan,
+		       struct ext4_inode_info *locked_ei)
 {
 	struct ext4_inode_info *ei;
 	struct ext4_es_stats *es_stats;
-	struct list_head *cur, *tmp;
-	LIST_HEAD(skipped);
 	ktime_t start_time;
 	u64 scan_time;
+	int nr_to_walk;
 	int nr_shrunk = 0;
-	int retried = 0, skip_precached = 1, nr_skipped = 0;
+	int retried = 0, nr_skipped = 0;
 
 	es_stats = &sbi->s_es_stats;
 	start_time = ktime_get();
-	spin_lock(&sbi->s_es_lru_lock);
 
 retry:
-	list_for_each_safe(cur, tmp, &sbi->s_es_lru) {
-		int shrunk;
-
-		/*
-		 * If we have already reclaimed all extents from extent
-		 * status tree, just stop the loop immediately.
-		 */
-		if (percpu_counter_read_positive(
-				&es_stats->es_stats_lru_cnt) == 0)
-			break;
-
-		ei = list_entry(cur, struct ext4_inode_info, i_es_lru);
+	spin_lock(&sbi->s_es_lock);
+	nr_to_walk = sbi->s_es_nr_inode;
+	while (nr_to_walk-- > 0) {
+		if (list_empty(&sbi->s_es_list)) {
+			spin_unlock(&sbi->s_es_lock);
+			goto out;
+		}
+		ei = list_first_entry(&sbi->s_es_list, struct ext4_inode_info,
+				      i_es_list);
+		/* Move the inode to the tail */
+		list_move_tail(&ei->i_es_list, &sbi->s_es_list);
 
 		/*
-		 * Skip the inode that is newer than the last_sorted
-		 * time.  Normally we try hard to avoid shrinking
-		 * precached inodes, but we will as a last resort.
+		 * Normally we try hard to avoid shrinking precached inodes,
+		 * but we will as a last resort.
 		 */
-		if ((es_stats->es_stats_last_sorted < ei->i_touch_when) ||
-		    (skip_precached && ext4_test_inode_state(&ei->vfs_inode,
-						EXT4_STATE_EXT_PRECACHED))) {
+		if (!retried && ext4_test_inode_state(&ei->vfs_inode,
+						EXT4_STATE_EXT_PRECACHED)) {
 			nr_skipped++;
-			list_move_tail(cur, &skipped);
 			continue;
 		}
 
-		if (ei->i_es_lru_nr == 0 || ei == locked_ei ||
-		    !write_trylock(&ei->i_es_lock))
+		if (ei == locked_ei || !write_trylock(&ei->i_es_lock)) {
+			nr_skipped++;
 			continue;
+		}
+		/*
+		 * Now we hold i_es_lock which protects us from inode reclaim
+		 * freeing inode under us
+		 */
+		spin_unlock(&sbi->s_es_lock);
 
-		shrunk = __es_try_to_reclaim_extents(ei, nr_to_scan);
-		if (ei->i_es_lru_nr == 0)
-			list_del_init(&ei->i_es_lru);
+		nr_shrunk += es_reclaim_extents(ei, &nr_to_scan);
 		write_unlock(&ei->i_es_lock);
 
-		nr_shrunk += shrunk;
-		nr_to_scan -= shrunk;
-		if (nr_to_scan == 0)
-			break;
+		if (nr_to_scan <= 0)
+			goto out;
+		spin_lock(&sbi->s_es_lock);
 	}
-
-	/* Move the newer inodes into the tail of the LRU list. */
-	list_splice_tail(&skipped, &sbi->s_es_lru);
-	INIT_LIST_HEAD(&skipped);
+	spin_unlock(&sbi->s_es_lock);
 
 	/*
 	 * If we skipped any inodes, and we weren't able to make any
-	 * forward progress, sort the list and try again.
+	 * forward progress, try again to scan precached inodes.
 	 */
 	if ((nr_shrunk == 0) && nr_skipped && !retried) {
 		retried++;
-		list_sort(NULL, &sbi->s_es_lru, ext4_inode_touch_time_cmp);
-		es_stats->es_stats_last_sorted = jiffies;
-		ei = list_first_entry(&sbi->s_es_lru, struct ext4_inode_info,
-				      i_es_lru);
-		/*
-		 * If there are no non-precached inodes left on the
-		 * list, start releasing precached extents.
-		 */
-		if (ext4_test_inode_state(&ei->vfs_inode,
-					  EXT4_STATE_EXT_PRECACHED))
-			skip_precached = 0;
 		goto retry;
 	}
 
-	spin_unlock(&sbi->s_es_lru_lock);
-
 	if (locked_ei && nr_shrunk == 0)
-		nr_shrunk = __es_try_to_reclaim_extents(locked_ei, nr_to_scan);
+		nr_shrunk = es_reclaim_extents(locked_ei, &nr_to_scan);
 
+out