Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

Pull btrfs update from Chris Mason:
 "The biggest feature in the pull is the new (and still experimental)
  raid56 code that David Woodhouse started long ago.  I'm still working
  on the parity logging setup that will avoid inconsistent parity after
  a crash, so this is only for testing right now.  But, I'd really like
  to get it out to a broader audience to hammer out any performance
  issues or other problems.

  scrub does not yet correct errors on raid5/6 either.

  Josef has another pass at fsync performance.  The big change here is
  to combine waiting for metadata with waiting for data, which is a big
  latency win.  It is also step one toward using atomics from the
  hardware during a commit.

  Mark Fasheh has a new way to use btrfs send/receive to send only the
  metadata changes.  SUSE is using this to make snapper more efficient
  at finding changes between snapshosts.

  Snapshot-aware defrag is also included.

  Otherwise we have a large number of fixes and cleanups.  Eric Sandeen
  wins the award for removing the most lines, and I'm hoping we steal
  this idea from XFS over and over again."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (118 commits)
  btrfs: fixup/remove module.h usage as required
  Btrfs: delete inline extents when we find them during logging
  btrfs: try harder to allocate raid56 stripe cache
  Btrfs: cleanup to make the function btrfs_delalloc_reserve_metadata more logic
  Btrfs: don't call btrfs_qgroup_free if just btrfs_qgroup_reserve fails
  Btrfs: remove reduplicate check about root in the function btrfs_clean_quota_tree
  Btrfs: return ENOMEM rather than use BUG_ON when btrfs_alloc_path fails
  Btrfs: fix missing deleted items in btrfs_clean_quota_tree
  btrfs: use only inline_pages from extent buffer
  Btrfs: fix wrong reserved space when deleting a snapshot/subvolume
  Btrfs: fix wrong reserved space in qgroup during snap/subv creation
  Btrfs: remove unnecessary dget_parent/dput when creating the pending snapshot
  btrfs: remove a printk from scan_one_device
  Btrfs: fix NULL pointer after aborting a transaction
  Btrfs: fix memory leak of log roots
  Btrfs: copy everything if we've created an inline extent
  btrfs: cleanup for open-coded alignment
  Btrfs: do not change inode flags in rename
  Btrfs: use reserved space for creating a snapshot
  clear chunk_alloc flag on retryable failure
  ...

1 files changed, 890 insertions, 174 deletions

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 55c07b65037..c226daefd65 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -39,12 +39,13 @@
 #include <linux/slab.h>
 #include <linux/ratelimit.h>
 #include <linux/mount.h>
+#include <linux/btrfs.h>
+#include <linux/blkdev.h>
 #include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
 #include "btrfs_inode.h"
-#include "ioctl.h"
 #include "print-tree.h"
 #include "ordered-data.h"
 #include "xattr.h"
@@ -54,6 +55,7 @@
 #include "locking.h"
 #include "free-space-cache.h"
 #include "inode-map.h"
+#include "backref.h"
 
 struct btrfs_iget_args {
 	u64 ino;
@@ -231,8 +233,7 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
 	u64 isize = i_size_read(inode);
 	u64 actual_end = min(end + 1, isize);
 	u64 inline_len = actual_end - start;
-	u64 aligned_end = (end + root->sectorsize - 1) &
-			~((u64)root->sectorsize - 1);
+	u64 aligned_end = ALIGN(end, root->sectorsize);
 	u64 data_len = inline_len;
 	int ret;
 
@@ -265,6 +266,7 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
 		return 1;
 	}
 
+	set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &BTRFS_I(inode)->runtime_flags);
 	btrfs_delalloc_release_metadata(inode, end + 1 - start);
 	btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
 	return 0;
@@ -389,7 +391,7 @@ again:
 	 * a compressed extent to 128k.
 	 */
 	total_compressed = min(total_compressed, max_uncompressed);
-	num_bytes = (end - start + blocksize) & ~(blocksize - 1);
+	num_bytes = ALIGN(end - start + 1, blocksize);
 	num_bytes = max(blocksize,  num_bytes);
 	total_in = 0;
 	ret = 0;
@@ -488,15 +490,13 @@ cont:
 		 * up to a block size boundary so the allocator does sane
 		 * things
 		 */
-		total_compressed = (total_compressed + blocksize - 1) &
-			~(blocksize - 1);
+		total_compressed = ALIGN(total_compressed, blocksize);
 
 		/*
 		 * one last check to make sure the compression is really a
 		 * win, compare the page count read with the blocks on disk
 		 */
-		total_in = (total_in + PAGE_CACHE_SIZE - 1) &
-			~(PAGE_CACHE_SIZE - 1);
+		total_in = ALIGN(total_in, PAGE_CACHE_SIZE);
 		if (total_compressed >= total_in) {
 			will_compress = 0;
 		} else {
@@ -608,7 +608,7 @@ static noinline int submit_compressed_extents(struct inode *inode,
 	if (list_empty(&async_cow->extents))
 		return 0;
 
-
+again:
 	while (!list_empty(&async_cow->extents)) {
 		async_extent = list_entry(async_cow->extents.next,
 					  struct async_extent, list);
@@ -648,6 +648,8 @@ retry:
 						  async_extent->ram_size - 1,
 						  btrfs_get_extent,
 						  WB_SYNC_ALL);
+			else if (ret)
+				unlock_page(async_cow->locked_page);
 			kfree(async_extent);
 			cond_resched();
 			continue;
@@ -672,6 +674,7 @@ retry:
 
 		if (ret) {
 			int i;
+
 			for (i = 0; i < async_extent->nr_pages; i++) {
 				WARN_ON(async_extent->pages[i]->mapping);
 				page_cache_release(async_extent->pages[i]);
@@ -679,12 +682,10 @@ retry:
 			kfree(async_extent->pages);
 			async_extent->nr_pages = 0;
 			async_extent->pages = NULL;
-			unlock_extent(io_tree, async_extent->start,
-				      async_extent->start +
-				      async_extent->ram_size - 1);
+
 			if (ret == -ENOSPC)
 				goto retry;
-			goto out_free; /* JDM: Requeue? */
+			goto out_free;
 		}
 
 		/*
@@ -696,10 +697,13 @@ retry:
 					async_extent->ram_size - 1, 0);
 
 		em = alloc_extent_map();
-		BUG_ON(!em); /* -ENOMEM */
+		if (!em)
+			goto out_free_reserve;
 		em->start = async_extent->start;
 		em->len = async_extent->ram_size;
 		em->orig_start = em->start;
+		em->mod_start = em->start;
+		em->mod_len = em->len;
 
 		em->block_start = ins.objectid;
 		em->block_len = ins.offset;
@@ -726,6 +730,9 @@ retry:
 						async_extent->ram_size - 1, 0);
 		}
 
+		if (ret)
+			goto out_free_reserve;
+
 		ret = btrfs_add_ordered_extent_compress(inode,
 						async_extent->start,
 						ins.objectid,
@@ -733,7 +740,8 @@ retry:
 						ins.offset,
 						BTRFS_ORDERED_COMPRESSED,
 						async_extent->compress_type);
-		BUG_ON(ret); /* -ENOMEM */
+		if (ret)
+			goto out_free_reserve;
 
 		/*
 		 * clear dirty, set writeback and unlock the pages.
@@ -754,18 +762,30 @@ retry:
 				    ins.objectid,
 				    ins.offset, async_extent->pages,
 				    async_extent->nr_pages);
-
-		BUG_ON(ret); /* -ENOMEM */
 		alloc_hint = ins.objectid + ins.offset;
 		kfree(async_extent);
+		if (ret)
+			goto out;
 		cond_resched();
 	}
 	ret = 0;
 out:
 	return ret;
+out_free_reserve:
+	btrfs_free_reserved_extent(root, ins.objectid, ins.offset);
 out_free:
+	extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
+				     async_extent->start,
+				     async_extent->start +
+				     async_extent->ram_size - 1,
+				     NULL, EXTENT_CLEAR_UNLOCK_PAGE |
+				     EXTENT_CLEAR_UNLOCK |
+				     EXTENT_CLEAR_DELALLOC |
+				     EXTENT_CLEAR_DIRTY |
+				     EXTENT_SET_WRITEBACK |
+				     EXTENT_END_WRITEBACK);
 	kfree(async_extent);
-	goto out;
+	goto again;
 }
 
 static u64 get_extent_allocation_hint(struct inode *inode, u64 start,
@@ -834,7 +854,7 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 
 	BUG_ON(btrfs_is_free_space_inode(inode));
 
-	num_bytes = (end - start + blocksize) & ~(blocksize - 1);
+	num_bytes = ALIGN(end - start + 1, blocksize);
 	num_bytes = max(blocksize,  num_bytes);
 	disk_num_bytes = num_bytes;
 
@@ -892,6 +912,8 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 		em->orig_start = em->start;
 		ram_size = ins.offset;
 		em->len = ins.offset;
+		em->mod_start = em->start;
+		em->mod_len = em->len;
 
 		em->block_start = ins.objectid;
 		em->block_len = ins.offset;
@@ -1338,6 +1360,8 @@ out_check:
 			em->block_start = disk_bytenr;
 			em->orig_block_len = disk_num_bytes;
 			em->bdev = root->fs_info->fs_devices->latest_bdev;
+			em->mod_start = em->start;
+			em->mod_len = em->len;
 			set_bit(EXTENT_FLAG_PINNED, &em->flags);
 			set_bit(EXTENT_FLAG_FILLING, &em->flags);
 			em->generation = -1;
@@ -1508,14 +1532,22 @@ static void btrfs_set_bit_hook(struct inode *inode,
 			spin_unlock(&BTRFS_I(inode)->lock);
 		}
 
-		spin_lock(&root->fs_info->delalloc_lock);
+		__percpu_counter_add(&root->fs_info->delalloc_bytes, len,
+				     root->fs_info->delalloc_batch);
+		spin_lock(&BTRFS_I(inode)->lock);
 		BTRFS_I(inode)->delalloc_bytes += len;
-		root->fs_info->delalloc_bytes += len;
-		if (do_list && list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
-			list_add_tail(&BTRFS_I(inode)->delalloc_inodes,
-				      &root->fs_info->delalloc_inodes);
+		if (do_list && !test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+					 &BTRFS_I(inode)->runtime_flags)) {
+			spin_lock(&root->fs_info->delalloc_lock);
+			if (list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
+				list_add_tail(&BTRFS_I(inode)->delalloc_inodes,
+					      &root->fs_info->delalloc_inodes);
+				set_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+					&BTRFS_I(inode)->runtime_flags);
+			}
+			spin_unlock(&root->fs_info->delalloc_lock);
 		}
-		spin_unlock(&root->fs_info->delalloc_lock);
+		spin_unlock(&BTRFS_I(inode)->lock);
 	}
 }
 
@@ -1550,15 +1582,22 @@ static void btrfs_clear_bit_hook(struct inode *inode,
 		    && do_list)
 			btrfs_free_reserved_data_space(inode, len);
 
-		spin_lock(&root->fs_info->delalloc_lock);
-		root->fs_info->delalloc_bytes -= len;
+		__percpu_counter_add(&root->fs_info->delalloc_bytes, -len,
+				     root->fs_info->delalloc_batch);
+		spin_lock(&BTRFS_I(inode)->lock);
 		BTRFS_I(inode)->delalloc_bytes -= len;
-
 		if (do_list && BTRFS_I(inode)->delalloc_bytes == 0 &&
-		    !list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
-			list_del_init(&BTRFS_I(inode)->delalloc_inodes);
+		    test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+			     &BTRFS_I(inode)->runtime_flags)) {
+			spin_lock(&root->fs_info->delalloc_lock);
+			if (!list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
+				list_del_init(&BTRFS_I(inode)->delalloc_inodes);
+				clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+					  &BTRFS_I(inode)->runtime_flags);
+			}
+			spin_unlock(&root->fs_info->delalloc_lock);
 		}
-		spin_unlock(&root->fs_info->delalloc_lock);
+		spin_unlock(&BTRFS_I(inode)->lock);
 	}
 }
 
@@ -1566,7 +1605,7 @@ static void btrfs_clear_bit_hook(struct inode *inode,
  * extent_io.c merge_bio_hook, this must check the chunk tree to make sure
  * we don't create bios that span stripes or chunks
  */
-int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
+int btrfs_merge_bio_hook(int rw, struct page *page, unsigned long offset,
 			 size_t size, struct bio *bio,
 			 unsigned long bio_flags)
 {
@@ -1581,7 +1620,7 @@ int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
 
 	length = bio->bi_size;
 	map_length = length;
-	ret = btrfs_map_block(root->fs_info, READ, logical,
+	ret = btrfs_map_block(root->fs_info, rw, logical,
 			      &map_length, NULL, 0);
 	/* Will always return 0 with map_multi == NULL */
 	BUG_ON(ret < 0);
@@ -1892,6 +1931,640 @@ out:
 	return ret;
 }
 
+/* snapshot-aware defrag */
+struct sa_defrag_extent_backref {
+	struct rb_node node;
+	struct old_sa_defrag_extent *old;
+	u64 root_id;
+	u64 inum;
+	u64 file_pos;
+	u64 extent_offset;
+	u64 num_bytes;
+	u64 generation;
+};
+
+struct old_sa_defrag_extent {
+	struct list_head list;
+	struct new_sa_defrag_extent *new;
+
+	u64 extent_offset;
+	u64 bytenr;
+	u64 offset;
+	u64 len;
+	int count;
+};
+
+struct new_sa_defrag_extent {
+	struct rb_root root;
+	struct list_head head;
+	struct btrfs_path *path;
+	struct inode *inode;
+	u64 file_pos;
+	u64 len;
+	u64 bytenr;
+	u64 disk_len;
+	u8 compress_type;
+};
+
+static int backref_comp(struct sa_defrag_extent_backref *b1,
+			struct sa_defrag_extent_backref *b2)
+{
+	if (b1->root_id < b2->root_id)
+		return -1;
+	else if (b1->root_id > b2->root_id)
+		return 1;
+
+	if (b1->inum < b2->inum)
+		return -1;
+	else if (b1->inum > b2->inum)
+		return 1;
+
+	if (b1->file_pos < b2->file_pos)
+		return -1;
+	else if (b1->file_pos > b2->file_pos)
+		return 1;
+
+	/*
+	 * [------------------------------] ===> (a range of space)
+	 *     |<--->|   |<---->| =============> (fs/file tree A)
+	 * |<---------------------------->| ===> (fs/file tree B)
+	 *
+	 * A range of space can refer to two file extents in one tree while
+	 * refer to only one file extent in another tree.
+	 *
+	 * So we may process a disk offset more than one time(two extents in A)
+	 * and locate at the same extent(one extent in B), then insert two same
+	 * backrefs(both refer to the extent in B).
+	 */
+	return 0;
+}
+
+static void backref_insert(struct rb_root *root,
+			   struct sa_defrag_extent_backref *backref)
+{
+	struct rb_node **p = &root->rb_node;
+	struct rb_node *parent = NULL;
+	struct sa_defrag_extent_backref *entry;
+	int ret;
+
+	while (*p) {
+		parent = *p;
+		entry = rb_entry(parent, struct sa_defrag_extent_backref, node);
+
+		ret = backref_comp(backref, entry);
+		if (ret < 0)
+			p = &(*p)->rb_left;
+		else
+			p = &(*p)->rb_right;
+	}
+
+	rb_link_node(&backref->node, parent, p);
+	rb_insert_color(&backref->node, root);
+}
+
+/*
+ * Note the backref might has changed, and in this case we just return 0.
+ */
+static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
+				       void *ctx)
+{
+	struct btrfs_file_extent_item *extent;
+	struct btrfs_fs_info *fs_info;
+	struct old_sa_defrag_extent *old = ctx;
+	struct new_sa_defrag_extent *new = old->new;
+	struct btrfs_path *path = new->path;
+	struct btrfs_key key;
+	struct btrfs_root *root;
+	struct sa_defrag_extent_backref *backref;
+	struct extent_buffer *leaf;
+	struct inode *inode = new->inode;
+	int slot;
+	int ret;
+	u64 extent_offset;
+	u64 num_bytes;
+
+	if (BTRFS_I(inode)->root->root_key.objectid == root_id &&
+	    inum == btrfs_ino(inode))
+		return 0;
+
+	key.objectid = root_id;
+	key.type = BTRFS_ROOT_ITEM_KEY;
+	key.offset = (u64)-1;
+
+	fs_info = BTRFS_I(inode)->root->fs_info;
+	root = btrfs_read_fs_root_no_name(fs_info, &key);
+	if (IS_ERR(root)) {
+		if (PTR_ERR(root) == -ENOENT)
+			return 0;
+		WARN_ON(1);
+		pr_debug("inum=%llu, offset=%llu, root_id=%llu\n",
+			 inum, offset, root_id);
+		return PTR_ERR(root);
+	}
+
+	key.objectid = inum;
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	if (offset > (u64)-1 << 32)
+		key.offset = 0;
+	else
+		key.offset = offset;
+
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	if (ret < 0) {
+		WARN_ON(1);
+		return ret;
+	}
+
+	while (1) {
+		cond_resched();
+
+		leaf = path->nodes[0];
+		slot = path->slots[0];
+
+		if (slot >= btrfs_header_nritems(leaf)) {
+			ret = btrfs_next_leaf(root, path);
+			if (ret < 0) {
+				goto out;
+			} else if (ret > 0) {
+				ret = 0;
+				goto out;
+			}
+			continue;
+		}
+
+		path->slots[0]++;
+
+		btrfs_item_key_to_cpu(leaf, &key, slot);
+
+		if (key.objectid > inum)
+			goto out;
+
+		if (key.objectid < inum || key.type != BTRFS_EXTENT_DATA_KEY)
+			continue;
+
+		extent = btrfs_item_ptr(leaf, slot,
+					struct btrfs_file_extent_item);
+
+		if (btrfs_file_extent_disk_bytenr(leaf, extent) != old->bytenr)
+			continue;
+
+		extent_offset = btrfs_file_extent_offset(leaf, extent);
+		if (key.offset - extent_offset != offset)
+			continue;
+
+		num_bytes = btrfs_file_extent_num_bytes(leaf, extent);
+		if (extent_offset >= old->extent_offset + old->offset +
+		    old->len || extent_offset + num_bytes <=
+		    old->extent_offset + old->offset)
+			continue;
+
+		break;
+	}
+
+	backref = kmalloc(sizeof(*backref), GFP_NOFS);
+	if (!backref) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	backref->root_id = root_id;
+	backref->inum = inum;
+	backref->file_pos = offset + extent_offset;
+	backref->num_bytes = num_bytes;
+	backref->extent_offset = extent_offset;
+	backref->generation = btrfs_file_extent_generation(leaf, extent);
+	backref->old = old;
+	backref_insert(&new->root, backref);
+	old->count++;
+out:
+	btrfs_release_path(path);
+	WARN_ON(ret);
+	return ret;
+}
+
+static noinline bool record_extent_backrefs(struct btrfs_path *path,
+				   struct new_sa_defrag_extent *new)
+{
+	struct btrfs_fs_info *fs_info = BTRFS_I(new->inode)->root->fs_info;
+	struct old_sa_defrag_extent *old, *tmp;
+	int ret;
+
+	new->path = path;
+
+	list_for_each_entry_safe(old, tmp, &new->head, list) {
+		ret = iterate_inodes_from_logical(old->bytenr, fs_info,
+						  path, record_one_backref,
+						  old);
+		BUG_ON(ret < 0 && ret != -ENOENT);
+
+		/* no backref to be processed for this extent */
+		if (!old->count) {
+			list_del(&old->list);
+			kfree(old);
+		}
+	}
+
+	if (list_empty(&new->head))
+		return false;
+
+	return true;
+}
+
+static int relink_is_mergable(struct extent_buffer *leaf,
+			      struct btrfs_file_extent_item *fi,
+			      u64 disk_bytenr)
+{
+	if (btrfs_file_extent_disk_bytenr(leaf, fi) != disk_bytenr)
+		return 0;
+
+	if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
+		return 0;
+
+	if (btrfs_file_extent_compression(leaf, fi) ||
+	    btrfs_file_extent_encryption(leaf, fi) ||
+	    btrfs_file_extent_other_encoding(leaf, fi))
+		return 0;
+
+	return 1;
+}
+
+/*
+ * Note the backref might has changed, and in this case we just return 0.
+ */
+static noinline int relink_extent_backref(struct btrfs_path *path,
+				 struct sa_defrag_extent_backref *prev,
+				 struct sa_defrag_extent_backref *backref)
+{
+	struct btrfs_file_extent_item *extent;
+	struct btrfs_file_extent_item *item;
+	struct btrfs_ordered_extent *ordered;
+	struct btrfs_trans_handle *trans;
+	struct btrfs_fs_info *fs_info;
+	struct btrfs_root *root;
+	struct btrfs_key key;
+	struct extent_buffer *leaf;
+	struct old_sa_defrag_extent *old = backref->old;
+	struct new_sa_defrag_extent *new = old->new;
+	struct inode *src_inode = new->inode;
+	struct inode *inode;
+	struct extent_state *cached = NULL;
+	int ret = 0;
+	u64 start;
+	u64 len;
+	u64 lock_start;
+	u64 lock_end;
+	bool merge = false;
+	int index;
+
+	if (prev && prev->root_id == backref->root_id &&
+	    prev->inum == backref->inum &&
+	    prev->file_pos + prev->num_bytes == backref->file_pos)
+		merge = true;
+
+	/* step 1: get root */
+	key.objectid = backref->root_id;
+	key.type = BTRFS_ROOT_ITEM_KEY;
+	key.offset = (u64)-1;
+
+	fs_info = BTRFS_I(src_inode)->root->fs_info;
+	index = srcu_read_lock(&fs_info->subvol_srcu);
+
+	root = btrfs_read_fs_root_no_name(fs_info, &key);
+	if (IS_ERR(root)) {
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
+		if (PTR_ERR(root) == -ENOENT)
+			return 0;
+		return PTR_ERR(root);
+	}
+	if (btrfs_root_refs(&root->root_item) == 0) {
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
+		/* parse ENOENT to 0 */
+		return 0;
+	}
+
+	/* step 2: get inode */
+	key.objectid = backref->inum;
+	key.type = BTRFS_INODE_ITEM_KEY;
+	key.offset = 0;
+
+	inode = btrfs_iget(fs_info->sb, &key, root, NULL);
+	if (IS_ERR(inode)) {
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
+		return 0;
+	}
+
+	srcu_read_unlock(&fs_info->subvol_srcu, index);
+
+	/* step 3: relink backref */
+	lock_start = backref->file_pos;
+	lock_end = backref->file_pos + backref->num_bytes - 1;
+	lock_extent_bits(&BTRFS_I(inode)->io_tree, lock_start, lock_end,
+			 0, &cached);
+
+	ordered = btrfs_lookup_first_ordered_extent(inode, lock_end);
+	if (ordered) {
+		btrfs_put_ordered_extent(ordered);
+		goto out_unlock;
+	}
+
+	trans = btrfs_join_transaction(root);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		goto out_unlock;
+	}
+
+	key.objectid = backref->inum;
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = backref->file_pos;
+
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	if (ret < 0) {
+		goto out_free_path;
+	} else if (ret > 0) {
+		ret = 0;
+		goto out_free_path;
+	}
+
+	extent = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				struct btrfs_file_extent_item);
+
+	if (btrfs_file_extent_generation(path->nodes[0], extent) !=
+	    backref->generation)
+		goto out_free_path;
+
+	btrfs_release_path(path);
+
+	start = backref->file_pos;
+	if (backref->extent_offset < old->extent_offset + old->offset)
+		start += old->extent_offset + old->offset -
+			 backref->extent_offset;
+
+	len = min(backref->extent_offset + backref->num_bytes,
+		  old->extent_offset + old->offset + old->len);
+	len -= max(backref->extent_offset, old->extent_offset + old->offset);
+
+	ret = btrfs_drop_extents(trans, root, inode, start,
+				 start + len, 1);
+	if (ret)
+		goto out_free_path;
+again:
+	key.objectid = btrfs_ino(inode);
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = start;
+
+	if (merge) {
+		struct btrfs_file_extent_item *fi;
+		u64 extent_len;
+		struct btrfs_key found_key;
+
+		ret = btrfs_search_slot(trans, root, &key, path, 1, 1);
+		if (ret < 0)
+			goto out_free_path;
+
+		path->slots[0]--;
+		leaf = path->nodes[0];
+		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
+
+		fi = btrfs_item_ptr(leaf, path->slots[0],
+				    struct btrfs_file_extent_item);
+		extent_len = btrfs_file_extent_num_bytes(leaf, fi);
+
+		if (relink_is_mergable(leaf, fi, new->bytenr) &&
+		    extent_len + found_key.offset == start) {
+			btrfs_set_file_extent_num_bytes(leaf, fi,
+							extent_len + len);
+			btrfs_mark_buffer_dirty(leaf);
+			inode_add_bytes(inode, len);
+
+			ret = 1;
+			goto out_free_path;
+		} else {
+			merge = false;
+			btrfs_release_path(path);
+			goto again;
+		}
+	}
+
+	ret = btrfs_insert_empty_item(trans, root, path, &key,
+					sizeof(*extent));
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto out_free_path;
+	}
+
+	leaf = path->nodes[0];
+	item = btrfs_item_ptr(leaf, path->slots[0],
+				struct btrfs_file_extent_item);
+	btrfs_set_file_extent_disk_bytenr(leaf, item, new->bytenr);
+	btrfs_set_file_extent_disk_num_bytes(leaf, item, new->disk_len);
+	btrfs_set_file_extent_offset(leaf, item, start - new->file_pos);
+	btrfs_set_file_extent_num_bytes(leaf, item, len);
+	btrfs_set_file_extent_ram_bytes(leaf, item, new->len);
+	btrfs_set_file_extent_generation(leaf, item, trans->transid);
+	btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG);
+	btrfs_set_file_extent_compression(leaf, item, new->compress_type);
+	btrfs_set_file_extent_encryption(leaf, item, 0);
+	btrfs_set_file_extent_other_encoding(leaf, item, 0);
+
+	btrfs_mark_buffer_dirty(leaf);
+	inode_add_bytes(inode, len);
+
+	ret = btrfs_inc_extent_ref(trans, root, new->bytenr,
+			new->disk_len, 0,
+			backref->root_id, backref->inum,
+			new->file_pos, 0);	/* start - extent_offset */
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto out_free_path;
+	}
+
+	ret = 1;
+out_free_path:
+	btrfs_release_path(path);
+	btrfs_end_transaction(trans, root);
+out_unlock:
+	unlock_extent_cached(&BTRFS_I(inode)->io_tree, lock_start, lock_end,
+			     &cached, GFP_NOFS);
+	iput(inode);
+	return ret;
+}
+
+static void relink_file_extents(struct new_sa_defrag_extent *new)
+{
+	struct btrfs_path *path;
+	struct old_sa_defrag_extent *old, *tmp;
+	struct sa_defrag_extent_backref *backref;
+	struct sa_defrag_extent_backref *prev = NULL;
+	struct inode *inode;
+	struct btrfs_root *root;
+	struct rb_node *node;
+	int ret;
+
+	inode = new->inode;
+	root = BTRFS_I(inode)->root;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return;
+
+	if (!record_extent_backrefs(path, new)) {
+		btrfs_free_path(path);
+		goto out;
+	}
+	btrfs_release_path(path);
+
+	while (1) {
+		node = rb_first(&new->root);
+		if (!node)
+			break;
+		rb_erase(node, &new->root);
+
+		backref = rb_entry(node, struct sa_defrag_extent_backref, node);
+
+		ret = relink_extent_backref(path, prev, backref);
+		WARN_ON(ret < 0);
+
+		kfree(prev);
+
+		if (ret == 1)
+			prev = backref;
+		else
+			prev = NULL;
+		cond_resched();
+	}
+	kfree(prev);
+
+	btrfs_free_path(path);
+
+	list_for_each_entry_safe(old, tmp, &new->head, list) {
+		list_del(&old->list);
+		kfree(old);
+	}
+out:
+	atomic_dec(&root->fs_info->defrag_running);
+	wake_up(&root->fs_info->transaction_wait);
+
+	kfree(new);
+}
+
+static struct new_sa_defrag_extent *
+record_old_file_extents(struct inode *inode,
+			struct btrfs_ordered_extent *ordered)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct btrfs_path *path;
+	struct btrfs_key key;
+	struct old_sa_defrag_extent *old, *tmp;
+	struct new_sa_defrag_extent *new;
+	int ret;
+
+	new = kmalloc(sizeof(*new), GFP_NOFS);
+	if (!new)
+		return NULL;
+
+	new->inode = inode;
+	new->file_pos = ordered->file_offset;
+	new->len = ordered->len;
+	new->bytenr = ordered->start;
+	new->disk_len = ordered->disk_len;
+	new->compress_type = ordered->compress_type;
+	new->root = RB_ROOT;
+	INIT_LIST_HEAD(&new->head);
+
+	path = btrfs_alloc_path();
+	if (!path)
+		goto out_kfree;
+
+	key.objectid = btrfs_ino(inode);
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = new->file_pos;
+
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	if (ret < 0)
+		goto out_free_path;
+	if (ret > 0 && path->slots[0] > 0)
+		path->slots[0]--;
+
+	/* find out all the old extents for the file range */
+	while (1) {
+		struct btrfs_file_extent_item *extent;
+		struct extent_buffer *l;
+		int slot;
+		u64 num_bytes;
+		u64 offset;
+		u64 end;
+		u64 disk_bytenr;
+		u64 extent_offset;
+
+		l = path->nodes[0];
+		slot = path->slots[0];
+
+		if (slot >= btrfs_header_nritems(l)) {
+			ret = btrfs_next_leaf(root, path);
+			if (ret < 0)
+				goto out_free_list;
+			else if (ret > 0)
+				break;
+			continue;
+		}
+
+		btrfs_item_key_to_cpu(l, &key, slot);
+
+		if (key.objectid != btrfs_ino(inode))
+			break;
+		if (key.type != BTRFS_EXTENT_DATA_KEY)
+			break;
+		if (key.offset >= new->file_pos + new->len)
+			break;
+
+		extent = btrfs_item_ptr(l, slot, struct btrfs_file_extent_item);
+
+		num_bytes = btrfs_file_extent_num_bytes(l, extent);
+		if (key.offset + num_bytes < new->file_pos)
+			goto next;
+
+		disk_bytenr = btrfs_file_extent_disk_bytenr(l, extent);
+		if (!disk_bytenr)
+			goto next;
+
+		extent_offset = btrfs_file_extent_offset(l, extent);
+
+		old = kmalloc(sizeof(*old), GFP_NOFS);
+		if (!old)
+			goto out_free_list;
+
+		offset = max(new->file_pos, key.offset);
+		end = min(new->file_pos + new->len, key.offset + num_bytes);
+
+		old->bytenr = disk_bytenr;
+		old->extent_offset = extent_offset;
+		old->offset = offset - key.offset;
+		old->len = end - offset;
+		old->new = new;
+		old->count = 0;
+		list_add_tail(&old->list, &new->head);
+next:
+		path->slots[0]++;
+		cond_resched();
+	}
+
+	btrfs_free_path(path);
+	atomic_inc(&root->fs_info->defrag_running);
+
+	return new;
+
+out_free_list:
+	list_for_each_entry_safe(old, tmp, &new->head, list) {
+		list_del(&old->list);
+		kfree(old);
+	}
+out_free_path:
+	btrfs_free_path(path);
+out_kfree:
+	kfree(new);
+	return NULL;
+}
+
 /*
  * helper function for btrfs_finish_ordered_io, this
  * just reads in some of the csum leaves to prime them into ram
@@ -1909,6 +2582,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	struct btrfs_trans_handle *trans = NULL;
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	struct extent_state *cached_state = NULL;
+	struct new_sa_defrag_extent *new = NULL;
 	int compress_type = 0;
 	int ret;
 	bool nolock;
@@ -1943,6 +2617,20 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 			 ordered_extent->file_offset + ordered_extent->len - 1,
 			 0, &cached_state);
 
+	ret = test_range_bit(io_tree, ordered_extent->file_offset,
+			ordered_extent->file_offset + ordered_extent->len - 1,
+			EXTENT_DEFRAG, 1, cached_state);
+	if (ret) {
+		u64 last_snapshot = btrfs_root_last_snapshot(&root->root_item);
+		if (last_snapshot >= BTRFS_I(inode)->generation)
+			/* the inode is shared */
+			new = record_old_file_extents(inode, ordered_extent);
+
+		clear_extent_bit(io_tree, ordered_extent->file_offset,
+			ordered_extent->file_offset + ordered_extent->len - 1,
+			EXTENT_DEFRAG, 0, 0, &cached_state, GFP_NOFS);
+	}
+
 	if (nolock)
 		trans = btrfs_join_transaction_nolock(root);
 	else
@@ -2001,17 +2689,33 @@ out:
 	if (trans)
 		btrfs_end_transaction(trans, root);
 
-	if (ret)
+	if (ret) {
 		clear_extent_uptodate(io_tree, ordered_extent->file_offset,
 				      ordered_extent->file_offset +
 				      ordered_extent->len - 1, NULL, GFP_NOFS);
 
+		/*
+		 * If the ordered extent had an IOERR or something else went
+		 * wrong we need to return the space for this ordered extent
+		 * back to the allocator.
+		 */
+		if (!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) &&
+		    !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags))
+			btrfs_free_reserved_extent(root, ordered_extent->start,
+						   ordered_extent->disk_len);
+	}
+
+
 	/*
 	 * This needs to be done to make sure anybody waiting knows we are done
 	 * updating everything for this ordered extent.
 	 */
 	btrfs_remove_ordered_extent(inode, ordered_extent);
 
+	/* for snapshot-aware defrag */
+	if (new)
+		relink_file_extents(new);
+
 	/* once for us */
 	btrfs_put_ordered_extent(ordered_extent);
 	/* once for the tree */
@@ -2062,7 +2766,7 @@ static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
 static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 			       struct extent_state *state, int mirror)
 {
-	size_t offset = start - ((u64)page->index << PAGE_CACHE_SHIFT);
+	size_t offset = start - page_offset(page);
 	struct inode *inode = page->mapping->host;
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	char *kaddr;
@@ -2167,11 +2871,6 @@ void btrfs_run_delayed_iputs(struct btrfs_root *root)
 	}
 }
 
-enum btrfs_orphan_cleanup_state {
-	ORPHAN_CLEANUP_STARTED	= 1,
-	ORPHAN_CLEANUP_DONE	= 2,
-};
-
 /*
  * This is called in transaction commit time. If there are no orphan
  * files in the subvolume, it removes orphan item and frees block_rsv
@@ -2469,6 +3168,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 		 */
 		set_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
 			&BTRFS_I(inode)->runtime_flags);
+		atomic_inc(&root->orphan_inodes);
 
 		/* if we have links, this was a truncate, lets do that */
 		if (inode->i_nlink) {
@@ -2491,6 +3191,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 				goto out;
 
 			ret = btrfs_truncate(inode);
+			if (ret)
+				btrfs_orphan_del(NULL, inode);
 		} else {
 			nr_unlink++;
 		}
@@ -2709,34 +3411,41 @@ static void fill_inode_item(struct btrfs_trans_handle *trans,
 			    struct btrfs_inode_item *item,
 			    struct inode *inode)
 {
-	btrfs_set_inode_uid(leaf, item, i_uid_read(inode));
-	btrfs_set_inode_gid(leaf, item, i_gid_read(inode));
-	btrfs_set_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size);
-	btrfs_set_inode_mode(leaf, item, inode->i_mode);
-	btrfs_set_inode_nlink(leaf, item, inode->i_nlink);
+	struct btrfs_map_token token;
+
+	btrfs_init_map_token(&token);
+
+	btrfs_set_token_inode_uid(leaf, item, i_uid_read(inode), &token);
+	btrfs_set_token_inode_gid(leaf, item, i_gid_read(inode), &token);
+	btrfs_set_token_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size,
+				   &token);
+	btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token);
+	btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token);
 
-	btrfs_set_timespec_sec(leaf, btrfs_inode_atime(item),
-			       inode->i_atime.tv_sec);
-	btrfs_set_timespec_nsec(leaf, btrfs_inode_atime(item),
-				inode->i_atime.tv_nsec);
+	btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item),
+				     inode->i_atime.tv_sec, &token);
+	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item),
+				      inode->i_atime.tv_nsec, &token);
 
-	btrfs_set_timespec_sec(leaf, btrfs_inode_mtime(item),
-			       inode->i_mtime.tv_sec);
-	btrfs_set_timespec_nsec(leaf, btrfs_inode_mtime(item),
-				inode->i_mtime.tv_nsec);
+	btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item),
+				     inode->i_mtime.tv_sec, &token);
+	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item),
+				      inode->i_mtime.tv_nsec, &token);
 
-	btrfs_set_timespec_sec(leaf, btrfs_inode_ctime(item),
-			       inode->i_ctime.tv_sec);
-	btrfs_set_timespec_nsec(leaf, btrfs_inode_ctime(item),
-				inode->i_ctime.tv_nsec);
+	btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item),
+				     inode->i_ctime.tv_sec, &token);
+	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item),
+				      inode->i_ctime.tv_nsec, &token);
 
-	btrfs_set_inode_nbytes(leaf, item, inode_get_bytes(inode));
-	btrfs_set_inode_generation(leaf, item, BTRFS_I(inode)->generation);
-	btrfs_set_inode_sequence(leaf, item, inode->i_version);
-	btrfs_set_inode_transid(leaf, item, trans->transid);
-	btrfs_set_inode_rdev(leaf, item, inode->i_rdev);
-	btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags);
-	btrfs_set_inode_block_group(leaf, item, 0);
+	btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode),
+				     &token);
+	btrfs_set_token_inode_generation(leaf, item, BTRFS_I(inode)->generation,
+					 &token);
+	btrfs_set_token_inode_sequence(leaf, item, inode->i_version, &token);
+	btrfs_set_token_inode_transid(leaf, item, trans->transid, &token);
+	btrfs_set_token_inode_rdev(leaf, item, inode->i_rdev, &token);
+	btrfs_set_token_inode_flags(leaf, item, BTRFS_I(inode)->flags, &token);
+	btrfs_set_token_inode_block_group(leaf, item, 0, &token);
 }
 
 /*
@@ -3304,7 +4013,6 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,
 	u64 extent_num_bytes = 0;
 	u64 extent_offset = 0;
 	u64 item_end = 0;
-	u64 mask = root->sectorsize - 1;
 	u32 found_type = (u8)-1;
 	int found_extent;
 	int del_item;
@@ -3328,7 +4036,8 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,
 	 * extent just the way it is.
 	 */
 	if (root->ref_cows || root == root->fs_info->tree_root)
-		btrfs_drop_extent_cache(inode, (new_size + mask) & (~mask), (u64)-1, 0);
+		btrfs_drop_extent_cache(inode, ALIGN(new_size,
+					root->sectorsize), (u64)-1, 0);
 
 	/*
 	 * This function is also used to drop the items in the log tree before
@@ -3407,10 +4116,9 @@ search_again:
 			if (!del_item) {
 				u64 orig_num_bytes =
 					btrfs_file_extent_num_bytes(leaf, fi);
-				extent_num_bytes = new_size -
-					found_key.offset + root->sectorsize - 1;
-				extent_num_bytes = extent_num_bytes &
-					~((u64)root->sectorsize - 1);
+				extent_num_bytes = ALIGN(new_size -
+						found_key.offset,
+						root->sectorsize);
 				btrfs_set_file_extent_num_bytes(leaf, fi,
 							 extent_num_bytes);
 				num_dec = (orig_num_bytes -
@@ -3646,9 +4354,8 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 	struct extent_map *em = NULL;
 	struct extent_state *cached_state = NULL;
 	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
-	u64 mask = root->sectorsize - 1;
-	u64 hole_start = (oldsize + mask) & ~mask;
-	u64 block_end = (size + mask) & ~mask;
+	u64 hole_start = ALIGN(oldsize, root->sectorsize);
+	u64 block_end = ALIGN(size, root->sectorsize);
 	u64 last_byte;
 	u64 cur_offset;
 	u64 hole_size;
@@ -3681,7 +4388,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			break;
 		}
 		last_byte = min(extent_map_end(em), block_end);
-		last_byte = (last_byte + mask) & ~mask;
+		last_byte = ALIGN(last_byte , root->sectorsize);
 		if (!test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
 			struct extent_map *hole_em;
 			hole_size = last_byte - cur_offset;
@@ -3832,6 +4539,12 @@ static int btrfs_setsize(struct inode *inode, struct iattr *attr)
 
 		/* we don't support swapfiles, so vmtruncate shouldn't fail */
 		truncate_setsize(inode, newsize);
+
+		/* Disable nonlocked read DIO to avoid the end less truncate */
+		btrfs_inode_block_unlocked_dio(inode);
+		inode_dio_wait(inode);
+		btrfs_inode_resume_unlocked_dio(inode);
+
 		ret = btrfs_truncate(inode);
 		if (ret && inode->i_nlink)
 			btrfs_orphan_del(NULL, inode);
@@ -3904,6 +4617,12 @@ void btrfs_evict_inode(struct inode *inode)
 		goto no_delete;
 	}
 
+	ret = btrfs_commit_inode_delayed_inode(inode);
+	if (ret) {
+		btrfs_orphan_del(NULL, inode);
+		goto no_delete;
+	}
+
 	rsv = btrfs_alloc_block_rsv(root, BTRFS_BLOCK_RSV_TEMP);
 	if (!rsv) {
 		btrfs_orphan_del(NULL, inode);
@@ -3941,7 +4660,7 @@ void btrfs_evict_inode(struct inode *inode)
 			goto no_delete;
 		}
 
-		trans = btrfs_start_transaction_lflush(root, 1);
+		trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans)) {
 			btrfs_orphan_del(NULL, inode);
 			btrfs_free_block_rsv(root, rsv);
@@ -3955,9 +4674,6 @@ void btrfs_evict_inode(struct inode *inode)
 			break;
 
 		trans->block_rsv = &root->fs_info->trans_block_rsv;
-		ret = btrfs_update_inode(trans, root, inode);
-		BUG_ON(ret);
-
 		btrfs_end_transaction(trans, root);
 		trans = NULL;
 		btrfs_btree_balance_dirty(root);
@@ -4854,7 +5570,8 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 		if (btrfs_test_opt(root, NODATASUM))
 			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
 		if (btrfs_test_opt(root, NODATACOW))
-			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
+			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW |
+				BTRFS_INODE_NODATASUM;
 	}
 
 	insert_inode_hash(inode);
@@ -5006,12 +5723,6 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
-	err = btrfs_update_inode(trans, root, inode);
-	if (err) {
-		drop_inode = 1;
-		goto out_unlock;
-	}
-
 	/*
 	* If the active LSM wants to access the inode during
 	* d_instantiate it needs these. Smack checks to see
@@ -5396,8 +6107,7 @@ again:
 	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
 		size_t size;
 		size = btrfs_file_extent_inline_len(leaf, item);
-		extent_end = (extent_start + size + root->sectorsize - 1) &
-			~((u64)root->sectorsize - 1);
+		extent_end = ALIGN(extent_start + size, root->sectorsize);
 	}
 
 	if (start >= extent_end) {
@@ -5469,8 +6179,7 @@ again:
 		copy_size = min_t(u64, PAGE_CACHE_SIZE - pg_offset,
 				size - extent_offset);
 		em->start = extent_start + extent_offset;
-		em->len = (copy_size + root->sectorsize - 1) &
-			~((u64)root->sectorsize - 1);
+		em->len = ALIGN(copy_size, root->sectorsize);
 		em->orig_block_len = em->len;
 		em->orig_start = em->start;
 		if (compress_type) {
@@ -5949,6 +6658,8 @@ static struct extent_map *create_pinned_em(struct inode *inode, u64 start,
 
 	em->start = start;
 	em->orig_start = orig_start;
+	em->mod_start = start;
+	em->mod_len = len;
 	em->len = len;
 	em->block_len = block_len;
 	em->block_start = block_start;
@@ -5990,16 +6701,12 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 	u64 len = bh_result->b_size;
 	struct btrfs_trans_handle *trans;
 	int unlock_bits = EXTENT_LOCKED;
-	int ret;
+	int ret = 0;
 
-	if (create) {
-		ret = btrfs_delalloc_reserve_space(inode, len);
-		if (ret)
-			return ret;
+	if (create)
 		unlock_bits |= EXTENT_DELALLOC | EXTENT_DIRTY;
-	} else {
+	else
 		len = min_t(u64, len, root->sectorsize);
-	}
 
 	lockstart = start;
 	lockend = start + len - 1;
@@ -6011,14 +6718,6 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 	if (lock_extent_direct(inode, lockstart, lockend, &cached_state, create))
 		return -ENOTBLK;
 
-	if (create) {
-		ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
-				     lockend, EXTENT_DELALLOC, NULL,
-				     &cached_state, GFP_NOFS);
-		if (ret)
-			goto unlock_err;
-	}
-
 	em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
 	if (IS_ERR(em)) {
 		ret = PTR_ERR(em);
@@ -6050,7 +6749,6 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 	if (!create && (em->block_start == EXTENT_MAP_HOLE ||
 			test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) {
 		free_extent_map(em);
-		ret = 0;
 		goto unlock_err;
 	}
 
@@ -6148,6 +6846,15 @@ unlock:
 		 */
 		if (start + len > i_size_read(inode))
 			i_size_write(inode, start + len);
+
+		spin_lock(&BTRFS_I(inode)->lock);
+		BTRFS_I(inode)->outstanding_extents++;
+		spin_unlock(&BTRFS_I(inode)->lock);
+
+		ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
+				     lockstart + len - 1, EXTENT_DELALLOC, NULL,
+				     &cached_state, GFP_NOFS);
+		BUG_ON(ret);
 	}
 
 	/*
@@ -6156,24 +6863,9 @@ unlock:
 	 * aren't using if there is any left over space.
 	 */
 	if (lockstart < lockend) {
-		if (create && len < lockend - lockstart) {
-			clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
-					 lockstart + len - 1,
-					 unlock_bits | EXTENT_DEFRAG, 1, 0,
-					 &cached_state, GFP_NOFS);
-			/*
-			 * Beside unlock, we also need to cleanup reserved space
-			 * for the left range by attaching EXTENT_DO_ACCOUNTING.
-			 */
-			clear_extent_bit(&BTRFS_I(inode)->io_tree,
-					 lockstart + len, lockend,
-					 unlock_bits | EXTENT_DO_ACCOUNTING |
-					 EXTENT_DEFRAG, 1, 0, NULL, GFP_NOFS);
-		} else {
-			clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
-					 lockend, unlock_bits, 1, 0,
-					 &cached_state, GFP_NOFS);
-		}
+		clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
+				 lockend, unlock_bits, 1, 0,
+				 &cached_state, GFP_NOFS);
 	} else {
 		free_extent_state(cached_state);
 	}
@@ -6183,9 +6875,6 @@ unlock:
 	return 0;
 
 unlock_err:
-	if (create)
-		unlock_bits |= EXTENT_DO_ACCOUNTING;
-
 	clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend,
 			 unlock_bits, 1, 0, &cached_state, GFP_NOFS);
 	return ret;
@@ -6426,19 +7115,24 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 	int async_submit = 0;
 
 	map_length = orig_bio->bi_size;
-	ret = btrfs_map_block(root->fs_info, READ, start_sector << 9,
+	ret = btrfs_map_block(root->fs_info, rw, start_sector << 9,
 			      &map_length, NULL, 0);
 	if (ret) {
 		bio_put(orig_bio);
 		return -EIO;
 	}
-
 	if (map_length >= orig_bio->bi_size) {
 		bio = orig_bio;
 		goto submit;
 	}
 
-	async_submit = 1;
+	/* async crcs make it difficult to collect full stripe writes. */
+	if (btrfs_get_alloc_profile(root, 1) &
+	    (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6))
+		async_submit = 0;
+	else
+		async_submit = 1;
+
 	bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS);
 	if (!bio)
 		return -ENOMEM;
@@ -6480,7 +7174,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 			bio->bi_end_io = btrfs_end_dio_bio;
 
 			map_length = orig_bio->bi_size;
-			ret = btrfs_map_block(root->fs_info, READ,
+			ret = btrfs_map_block(root->fs_info, rw,
 					      start_sector << 9,
 					      &map_length, NULL, 0);
 			if (ret) {
@@ -6623,15 +7317,60 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 {
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file->f_mapping->host;
+	size_t count = 0;
+	int flags = 0;
+	bool wakeup = true;
+	bool relock = false;
+	ssize_t ret;
 
 	if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iov,
 			    offset, nr_segs))
 		return 0;
 
-	return __blockdev_direct_IO(rw, iocb, inode,
-		   BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev,
-		   iov, offset, nr_segs, btrfs_get_blocks_direct, NULL,
-		   btrfs_submit_direct, 0);
+	atomic_inc(&inode->i_dio_count);
+	smp_mb__after_atomic_inc();
+
+	if (rw & WRITE) {
+		count = iov_length(iov, nr_segs);
+		/*
+		 * If the write DIO is beyond the EOF, we need update
+		 * the isize, but it is protected by i_mutex. So we can
+		 * not unlock the i_mutex at this case.
+		 */
+		if (offset + count <= inode->i_size) {
+			mutex_unlock(&inode->i_mutex);
+			relock = true;
+		}
+		ret = btrfs_delalloc_reserve_space(inode, count);
+		if (ret)
+			goto out;
+	} else if (unlikely(test_bit(BTRFS_INODE_READDIO_NEED_LOCK,
+				     &BTRFS_I(inode)->runtime_flags))) {
+		inode_dio_done(inode);
+		flags = DIO_LOCKING | DIO_SKIP_HOLES;
+		wakeup = false;
+	}
+
+	ret = __blockdev_direct_IO(rw, iocb, inode,
+			BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev,
+			iov, offset, nr_segs, btrfs_get_blocks_direct, NULL,
+			btrfs_submit_direct, flags);
+	if (rw & WRITE) {
+		if (ret < 0 && ret != -EIOCBQUEUED)
+			btrfs_delalloc_release_space(inode, count);
+		else if (ret >= 0 && (size_t)ret < count)
+			btrfs_delalloc_release_space(inode,
+						     count - (size_t)ret);
+		else
+			btrfs_delalloc_release_metadata(inode, 0);
+	}
+out:
+	if (wakeup)
+		inode_dio_done(inode);
+	if (relock)
+		mutex_lock(&inode->i_mutex);
+
+	return ret;
 }
 
 #define BTRFS_FIEMAP_FLAGS	(FIEMAP_FLAG_SYNC)
@@ -6735,8 +7474,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
 		return;
 	}
 	lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
-	ordered = btrfs_lookup_ordered_extent(inode,
-					   page_offset(page));
+	ordered = btrfs_lookup_ordered_extent(inode, page_offset(page));
 	if (ordered) {
 		/*
 		 * IO on this page will never be started, so we need
@@ -7216,8 +7954,9 @@ int btrfs_drop_inode(struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 
+	/* the snap/subvol tree is on deleting */
 	if (btrfs_root_refs(&root->root_item) == 0 &&
-	    !btrfs_is_free_space_inode(inode))
+	    root != root->fs_info->tree_root)
 		return 1;
 	else
 		return generic_drop_inode(inode);
@@ -7299,40 +8038,22 @@ fail:
 static int btrfs_getattr(struct vfsmount *mnt,
 			 struct dentry *dentry, struct kstat *stat)
 {
+	u64 delalloc_bytes;
 	struct inode *inode = dentry->d_inode;
 	u32 blocksize = inode->i_sb->s_blocksize;
 
 	generic_fillattr(inode, stat);
 	stat->dev = BTRFS_I(inode)->root->anon_dev;
 	stat->blksize = PAGE_CACHE_SIZE;
+
+	spin_lock(&BTRFS_I(inode)->lock);
+	delalloc_bytes = BTRFS_I(inode)->delalloc_bytes;
+	spin_unlock(&BTRFS_I(inode)->lock);
 	stat->blocks = (ALIGN(inode_get_bytes(inode), blocksize) +
-		ALIGN(BTRFS_I(inode)->delalloc_bytes, blocksize)) >> 9;
+			ALIGN(delalloc_bytes, blocksize)) >> 9;
 	return 0;
 }
 
-/*
- * If a file is moved, it will inherit the cow and compression flags of the new
- * directory.
- */
-static void fixup_inode_flags(struct inode *dir, struct inode *inode)
-{
-	struct btrfs_inode *b_dir = BTRFS_I(dir);
-	struct btrfs_inode *b_inode = BTRFS_I(inode);
-
-	if (b_dir->flags & BTRFS_INODE_NODATACOW)
-		b_inode->flags |= BTRFS_INODE_NODATACOW;
-	else
-		b_inode->flags &= ~BTRFS_INODE_NODATACOW;
-
-	if (b_dir->flags & BTRFS_INODE_COMPRESS) {
-		b_inode->flags |= BTRFS_INODE_COMPRESS;
-		b_inode->flags &= ~BTRFS_INODE_NOCOMPRESS;
-	} else {
-		b_inode->flags &= ~(BTRFS_INODE_COMPRESS |
-				    BTRFS_INODE_NOCOMPRESS);
-	}
-}
-
 static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 			   struct inode *new_dir, struct dentry *new_dentry)
 {
@@ -7498,8 +8219,6 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		}
 	}
 
-	fixup_inode_flags(new_dir, old_inode);
-
 	ret = btrfs_add_link(trans, new_dir, old_inode,
 			     new_dentry->d_name.name,
 			     new_dentry->d_name.len, 0, index);
@@ -7583,7 +8302,7 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput)
 
 	INIT_LIST_HEAD(&works);
 	INIT_LIST_HEAD(&splice);
-again:
+
 	spin_lock(&root->fs_info->delalloc_lock);
 	list_splice_init(&root->fs_info->delalloc_inodes, &splice);
 	while (!list_empty(&splice)) {
@@ -7593,8 +8312,11 @@ again:
 		list_del_init(&binode->delalloc_inodes);
 
 		inode = igrab(&binode->vfs_inode);
-		if (!inode)
+		if (!inode) {
+			clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+				  &binode->runtime_flags);
 			continue;
+		}
 
 		list_add_tail(&binode->delalloc_inodes,
 			      &root->fs_info->delalloc_inodes);
@@ -7619,13 +8341,6 @@ again:
 		btrfs_wait_and_free_delalloc_work(work);
 	}
 
-	spin_lock(&root->fs_info->delalloc_lock);
-	if (!list_empty(&root->fs_info->delalloc_inodes)) {
-		spin_unlock(&root->fs_info->delalloc_lock);
-		goto again;
-	}
-	spin_unlock(&root->fs_info->delalloc_lock);
-
 	/* the filemap_flush will queue IO into the worker threads, but
 	 * we have to make sure the IO is actually started and that
 	 * ordered extents get created before we return
@@ -7801,8 +8516,9 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 			}
 		}
 
-		ret = btrfs_reserve_extent(trans, root, num_bytes, min_size,
-					   0, *alloc_hint, &ins, 1);
+		ret = btrfs_reserve_extent(trans, root,
+					   min(num_bytes, 256ULL * 1024 * 1024),
+					   min_size, 0, *alloc_hint, &ins, 1);
 		if (ret) {
 			if (own_trans)
 				btrfs_end_transaction(trans, root);