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

Pull btrfs update frm Chris Mason:
 "This is our usual merge window set of bug fixes, performance
  improvements and cleanups.  Miao Xie has some really nice
  optimizations for writeback.

  Josef also expanded our sanity checks quite a bit; these make up a big
  chunk of the new lines"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (98 commits)
  Btrfs: rename btrfs_start_all_delalloc_inodes
  Btrfs: don't wait for the completion of all the ordered extents
  Btrfs: don't wait for all the async delalloc when shrinking delalloc
  Btrfs: fix the confusion between delalloc bytes and metadata bytes
  Btrfs: pick up the code for the item number calculation in flush_space()
  Btrfs: wait for the ordered extent only when we want
  Btrfs: remove unnecessary initialization and memory barrior in shrink_delalloc()
  Btrfs: avoid unnecessary scrub workers allocation
  Btrfs: check file extent type before anything else
  btrfs: Remove useless variable in write_ctree_super()
  btrfs: Fix checkpatch.pl warning of spacing issues
  btrfs: Replace kmalloc with kmalloc_array
  btrfs: Enclose macros with complex values within parenthesis
  btrfs: Use WARN_ON()'s return value in place of WARN_ON(1)
  btrfs: Remove redundant local zero structure
  btrfs: Pack struct btrfs_device
  btrfs: Replace multiple atomic_inc() with atomic_add()
  btrfs: Add helper function for free_root_pointers()
  Btrfs: fix a crash when running balance and defrag concurrently
  Btrfs: do not run snapshot-aware defragment on error
  ...

fs/btrfs/Makefile

@@ -14,4 +14,6 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \
 btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o
 btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o
 
-btrfs-$(CONFIG_BTRFS_FS_RUN_SANITY_TESTS) += tests/free-space-tests.o
+btrfs-$(CONFIG_BTRFS_FS_RUN_SANITY_TESTS) += tests/free-space-tests.o \
+	tests/extent-buffer-tests.o tests/btrfs-tests.o \
+	tests/extent-io-tests.o tests/inode-tests.o

fs/btrfs/acl.c

@@ -229,7 +229,7 @@ int btrfs_init_acl(struct btrfs_trans_handle *trans,
 		if (ret > 0) {
 			/* we need an acl */
 			ret = btrfs_set_acl(trans, inode, acl, ACL_TYPE_ACCESS);
-		} else {
+		} else if (ret < 0) {
 			cache_no_acl(inode);
 		}
 	} else {

fs/btrfs/async-thread.c

@@ -262,7 +262,7 @@ static struct btrfs_work *get_next_work(struct btrfs_worker_thread *worker,
 	struct btrfs_work *work = NULL;
 	struct list_head *cur = NULL;
 
-	if(!list_empty(prio_head))
+	if (!list_empty(prio_head))
 		cur = prio_head->next;
 
 	smp_mb();

fs/btrfs/backref.c

@@ -185,6 +185,9 @@ static int __add_prelim_ref(struct list_head *head, u64 root_id,
 {
 	struct __prelim_ref *ref;
 
+	if (root_id == BTRFS_DATA_RELOC_TREE_OBJECTID)
+		return 0;
+
 	ref = kmem_cache_alloc(btrfs_prelim_ref_cache, gfp_mask);
 	if (!ref)
 		return -ENOMEM;
@@ -323,8 +326,7 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
 
 	eb = path->nodes[level];
 	while (!eb) {
-		if (!level) {
+		if (WARN_ON(!level)) {
-			WARN_ON(1);
 			ret = 1;
 			goto out;
 		}
@@ -1619,7 +1621,7 @@ static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
 		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
 		btrfs_release_path(path);
 
-		item = btrfs_item_nr(eb, slot);
+		item = btrfs_item_nr(slot);
 		iref = btrfs_item_ptr(eb, slot, struct btrfs_inode_ref);
 
 		for (cur = 0; cur < btrfs_item_size(eb, item); cur += len) {

fs/btrfs/btrfs_inode.h

@@ -19,6 +19,7 @@
 #ifndef __BTRFS_I__
 #define __BTRFS_I__
 
+#include <linux/hash.h>
 #include "extent_map.h"
 #include "extent_io.h"
 #include "ordered-data.h"
@@ -179,6 +180,25 @@ static inline struct btrfs_inode *BTRFS_I(struct inode *inode)
 	return container_of(inode, struct btrfs_inode, vfs_inode);
 }
 
+static inline unsigned long btrfs_inode_hash(u64 objectid,
+					     const struct btrfs_root *root)
+{
+	u64 h = objectid ^ (root->objectid * GOLDEN_RATIO_PRIME);
+
+#if BITS_PER_LONG == 32
+	h = (h >> 32) ^ (h & 0xffffffff);
+#endif
+
+	return (unsigned long)h;
+}
+
+static inline void btrfs_insert_inode_hash(struct inode *inode)
+{
+	unsigned long h = btrfs_inode_hash(inode->i_ino, BTRFS_I(inode)->root);
+
+	__insert_inode_hash(inode, h);
+}
+
 static inline u64 btrfs_ino(struct inode *inode)
 {
 	u64 ino = BTRFS_I(inode)->location.objectid;

fs/btrfs/check-integrity.c

@@ -1038,7 +1038,7 @@ static int btrfsic_process_metablock(
 						     disk_item_offset,
 						     sizeof(struct btrfs_item));
 			item_offset = btrfs_stack_item_offset(&disk_item);
-			item_size = btrfs_stack_item_offset(&disk_item);
+			item_size = btrfs_stack_item_size(&disk_item);
 			disk_key = &disk_item.key;
 			type = btrfs_disk_key_type(disk_key);
 
@@ -1900,7 +1900,9 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
 							       dev_state,
 							       dev_bytenr);
 			}
-			if (block->logical_bytenr != bytenr) {
+			if (block->logical_bytenr != bytenr &&
+			    !(!block->is_metadata &&
+			      block->logical_bytenr == 0))
 				printk(KERN_INFO
 				       "Written block @%llu (%s/%llu/%d)"
 				       " found in hash table, %c,"
@@ -1910,15 +1912,14 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
 				       block->mirror_num,
 				       btrfsic_get_block_type(state, block),
 				       block->logical_bytenr);
-				block->logical_bytenr = bytenr;
+			else if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
-			} else if (state->print_mask &
-				   BTRFSIC_PRINT_MASK_VERBOSE)
 				printk(KERN_INFO
 				       "Written block @%llu (%s/%llu/%d)"
 				       " found in hash table, %c.\n",
 				       bytenr, dev_state->name, dev_bytenr,
 				       block->mirror_num,
 				       btrfsic_get_block_type(state, block));
+			block->logical_bytenr = bytenr;
 		} else {
 			if (num_pages * PAGE_CACHE_SIZE <
 			    state->datablock_size) {
@@ -2463,10 +2464,8 @@ static int btrfsic_process_written_superblock(
 		}
 	}
 
-	if (-1 == btrfsic_check_all_ref_blocks(state, superblock, 0)) {
+	if (WARN_ON(-1 == btrfsic_check_all_ref_blocks(state, superblock, 0)))
-		WARN_ON(1);
 		btrfsic_dump_tree(state);
-	}
 
 	return 0;
 }
@@ -2906,7 +2905,7 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
 		btrfsic_release_block_ctx(&block_ctx);
 	}
 
-	if (!match) {
+	if (WARN_ON(!match)) {
 		printk(KERN_INFO "btrfs: attempt to write M-block which contains logical bytenr that doesn't map to dev+physical bytenr of submit_bio,"
 		       " buffer->log_bytenr=%llu, submit_bio(bdev=%s,"
 		       " phys_bytenr=%llu)!\n",
@@ -2923,7 +2922,6 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
 			       bytenr, block_ctx.dev->name,
 			       block_ctx.dev_bytenr, mirror_num);
 		}
-		WARN_ON(1);
 	}
 }
 

fs/btrfs/compat.h

@@ -1,7 +0,0 @@
-#ifndef _COMPAT_H_
-#define _COMPAT_H_
-
-#define btrfs_drop_nlink(inode) drop_nlink(inode)
-#define btrfs_inc_nlink(inode)	inc_nlink(inode)
-
-#endif /* _COMPAT_H_ */

fs/btrfs/compression.c

@@ -32,7 +32,6 @@
 #include <linux/writeback.h>
 #include <linux/bit_spinlock.h>
 #include <linux/slab.h>
-#include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -360,7 +359,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
 	bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
 
 	bio = compressed_bio_alloc(bdev, first_byte, GFP_NOFS);
-	if(!bio) {
+	if (!bio) {
 		kfree(cb);
 		return -ENOMEM;
 	}

fs/btrfs/ctree.c

@@ -274,7 +274,7 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
 	else
 		btrfs_set_header_owner(cow, new_root_objectid);
 
-	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(cow),
+	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(),
 			    BTRFS_FSID_SIZE);
 
 	WARN_ON(btrfs_header_generation(buf) > trans->transid);
@@ -996,7 +996,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 	else
 		btrfs_set_header_owner(cow, root->root_key.objectid);
 
-	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(cow),
+	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(),
 			    BTRFS_FSID_SIZE);
 
 	ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
@@ -1285,11 +1285,10 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 		free_extent_buffer(eb_root);
 		blocksize = btrfs_level_size(root, old_root->level);
 		old = read_tree_block(root, logical, blocksize, 0);
-		if (!old || !extent_buffer_uptodate(old)) {
+		if (WARN_ON(!old || !extent_buffer_uptodate(old))) {
 			free_extent_buffer(old);
 			pr_warn("btrfs: failed to read tree block %llu from get_old_root\n",
 				logical);
-			WARN_ON(1);
 		} else {
 			eb = btrfs_clone_extent_buffer(old);
 			free_extent_buffer(old);
@@ -2758,7 +2757,7 @@ int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
 	int level;
 	int lowest_unlock = 1;
 	u8 lowest_level = 0;
-	int prev_cmp;
+	int prev_cmp = -1;
 
 	lowest_level = p->lowest_level;
 	WARN_ON(p->nodes[0] != NULL);
@@ -2769,7 +2768,6 @@ int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
 	}
 
 again:
-	prev_cmp = -1;
 	b = get_old_root(root, time_seq);
 	level = btrfs_header_level(b);
 	p->locks[level] = BTRFS_READ_LOCK;
@@ -2787,6 +2785,11 @@ int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
 		 */
 		btrfs_unlock_up_safe(p, level + 1);
 
+		/*
+		 * Since we can unwind eb's we want to do a real search every
+		 * time.
+		 */
+		prev_cmp = -1;
 		ret = key_search(b, key, level, &prev_cmp, &slot);
 
 		if (level != 0) {
@@ -3148,7 +3151,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
 	btrfs_set_header_backref_rev(c, BTRFS_MIXED_BACKREF_REV);
 	btrfs_set_header_owner(c, root->root_key.objectid);
 
-	write_extent_buffer(c, root->fs_info->fsid, btrfs_header_fsid(c),
+	write_extent_buffer(c, root->fs_info->fsid, btrfs_header_fsid(),
 			    BTRFS_FSID_SIZE);
 
 	write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
@@ -3287,7 +3290,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 	btrfs_set_header_backref_rev(split, BTRFS_MIXED_BACKREF_REV);
 	btrfs_set_header_owner(split, root->root_key.objectid);
 	write_extent_buffer(split, root->fs_info->fsid,
-			    btrfs_header_fsid(split), BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(), BTRFS_FSID_SIZE);
 	write_extent_buffer(split, root->fs_info->chunk_tree_uuid,
 			    btrfs_header_chunk_tree_uuid(split),
 			    BTRFS_UUID_SIZE);
@@ -3337,8 +3340,8 @@ static int leaf_space_used(struct extent_buffer *l, int start, int nr)
 	if (!nr)
 		return 0;
 	btrfs_init_map_token(&token);
-	start_item = btrfs_item_nr(l, start);
+	start_item = btrfs_item_nr(start);
-	end_item = btrfs_item_nr(l, end);
+	end_item = btrfs_item_nr(end);
 	data_len = btrfs_token_item_offset(l, start_item, &token) +
 		btrfs_token_item_size(l, start_item, &token);
 	data_len = data_len - btrfs_token_item_offset(l, end_item, &token);
@@ -3406,7 +3409,7 @@ static noinline int __push_leaf_right(struct btrfs_trans_handle *trans,
 	slot = path->slots[1];
 	i = left_nritems - 1;
 	while (i >= nr) {
-		item = btrfs_item_nr(left, i);
+		item = btrfs_item_nr(i);
 
 		if (!empty && push_items > 0) {
 			if (path->slots[0] > i)
@@ -3470,7 +3473,7 @@ static noinline int __push_leaf_right(struct btrfs_trans_handle *trans,
 	btrfs_set_header_nritems(right, right_nritems);
 	push_space = BTRFS_LEAF_DATA_SIZE(root);
 	for (i = 0; i < right_nritems; i++) {
-		item = btrfs_item_nr(right, i);
+		item = btrfs_item_nr(i);
 		push_space -= btrfs_token_item_size(right, item, &token);
 		btrfs_set_token_item_offset(right, item, push_space, &token);
 	}
@@ -3612,7 +3615,7 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 		nr = min(right_nritems - 1, max_slot);
 
 	for (i = 0; i < nr; i++) {
-		item = btrfs_item_nr(right, i);
+		item = btrfs_item_nr(i);
 
 		if (!empty && push_items > 0) {
 			if (path->slots[0] < i)
@@ -3639,8 +3642,7 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 		ret = 1;
 		goto out;
 	}
-	if (!empty && push_items == btrfs_header_nritems(right))
+	WARN_ON(!empty && push_items == btrfs_header_nritems(right));
-		WARN_ON(1);
 
 	/* push data from right to left */
 	copy_extent_buffer(left, right,
@@ -3663,7 +3665,7 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 	for (i = old_left_nritems; i < old_left_nritems + push_items; i++) {
 		u32 ioff;
 
-		item = btrfs_item_nr(left, i);
+		item = btrfs_item_nr(i);
 
 		ioff = btrfs_token_item_offset(left, item, &token);
 		btrfs_set_token_item_offset(left, item,
@@ -3694,7 +3696,7 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 	btrfs_set_header_nritems(right, right_nritems);
 	push_space = BTRFS_LEAF_DATA_SIZE(root);
 	for (i = 0; i < right_nritems; i++) {
-		item = btrfs_item_nr(right, i);
+		item = btrfs_item_nr(i);
 
 		push_space = push_space - btrfs_token_item_size(right,
 								item, &token);
@@ -3835,7 +3837,7 @@ static noinline void copy_for_split(struct btrfs_trans_handle *trans,
 		      btrfs_item_end_nr(l, mid);
 
 	for (i = 0; i < nritems; i++) {
-		struct btrfs_item *item = btrfs_item_nr(right, i);
+		struct btrfs_item *item = btrfs_item_nr(i);
 		u32 ioff;
 
 		ioff = btrfs_token_item_offset(right, item, &token);
@@ -4016,7 +4018,7 @@ static noinline int split_leaf(struct btrfs_trans_handle *trans,
 				    data_size > BTRFS_LEAF_DATA_SIZE(root)) {
 					if (data_size && !tried_avoid_double)
 						goto push_for_double;
-					split = 2 ;
+					split = 2;
 				}
 			}
 		}
@@ -4042,7 +4044,7 @@ static noinline int split_leaf(struct btrfs_trans_handle *trans,
 	btrfs_set_header_owner(right, root->root_key.objectid);
 	btrfs_set_header_level(right, 0);
 	write_extent_buffer(right, root->fs_info->fsid,
-			    btrfs_header_fsid(right), BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(), BTRFS_FSID_SIZE);
 
 	write_extent_buffer(right, root->fs_info->chunk_tree_uuid,
 			    btrfs_header_chunk_tree_uuid(right),
@@ -4177,7 +4179,7 @@ static noinline int split_item(struct btrfs_trans_handle *trans,
 
 	btrfs_set_path_blocking(path);
 
-	item = btrfs_item_nr(leaf, path->slots[0]);
+	item = btrfs_item_nr(path->slots[0]);
 	orig_offset = btrfs_item_offset(leaf, item);
 	item_size = btrfs_item_size(leaf, item);
 
@@ -4200,7 +4202,7 @@ static noinline int split_item(struct btrfs_trans_handle *trans,
 	btrfs_cpu_key_to_disk(&disk_key, new_key);
 	btrfs_set_item_key(leaf, &disk_key, slot);
 
-	new_item = btrfs_item_nr(leaf, slot);
+	new_item = btrfs_item_nr(slot);
 
 	btrfs_set_item_offset(leaf, new_item, orig_offset);
 	btrfs_set_item_size(leaf, new_item, item_size - split_offset);
@@ -4339,7 +4341,7 @@ void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
 	/* first correct the data pointers */
 	for (i = slot; i < nritems; i++) {
 		u32 ioff;
-		item = btrfs_item_nr(leaf, i);
+		item = btrfs_item_nr(i);
 
 		ioff = btrfs_token_item_offset(leaf, item, &token);
 		btrfs_set_token_item_offset(leaf, item,
@@ -4387,7 +4389,7 @@ void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
 			fixup_low_keys(root, path, &disk_key, 1);
 	}
 
-	item = btrfs_item_nr(leaf, slot);
+	item = btrfs_item_nr(slot);
 	btrfs_set_item_size(leaf, item, new_size);
 	btrfs_mark_buffer_dirty(leaf);
 
@@ -4441,7 +4443,7 @@ void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
 	/* first correct the data pointers */
 	for (i = slot; i < nritems; i++) {
 		u32 ioff;
-		item = btrfs_item_nr(leaf, i);
+		item = btrfs_item_nr(i);
 
 		ioff = btrfs_token_item_offset(leaf, item, &token);
 		btrfs_set_token_item_offset(leaf, item,
@@ -4455,7 +4457,7 @@ void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
 
 	data_end = old_data;
 	old_size = btrfs_item_size_nr(leaf, slot);
-	item = btrfs_item_nr(leaf, slot);
+	item = btrfs_item_nr(slot);
 	btrfs_set_item_size(leaf, item, old_size + data_size);
 	btrfs_mark_buffer_dirty(leaf);
 
@@ -4514,7 +4516,7 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 		for (i = slot; i < nritems; i++) {
 			u32 ioff;
 
-			item = btrfs_item_nr(leaf, i);
+			item = btrfs_item_nr( i);
 			ioff = btrfs_token_item_offset(leaf, item, &token);
 			btrfs_set_token_item_offset(leaf, item,
 						    ioff - total_data, &token);
@@ -4535,7 +4537,7 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 	for (i = 0; i < nr; i++) {
 		btrfs_cpu_key_to_disk(&disk_key, cpu_key + i);
 		btrfs_set_item_key(leaf, &disk_key, slot + i);
-		item = btrfs_item_nr(leaf, slot + i);
+		item = btrfs_item_nr(slot + i);
 		btrfs_set_token_item_offset(leaf, item,
 					    data_end - data_size[i], &token);
 		data_end -= data_size[i];
@@ -4730,7 +4732,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		for (i = slot + nr; i < nritems; i++) {
 			u32 ioff;
 
-			item = btrfs_item_nr(leaf, i);
+			item = btrfs_item_nr(i);
 			ioff = btrfs_token_item_offset(leaf, item, &token);
 			btrfs_set_token_item_offset(leaf, item,
 						    ioff + dsize, &token);
@@ -4823,14 +4825,18 @@ static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
 
 	btrfs_item_key_to_cpu(path->nodes[0], &key, 0);
 
-	if (key.offset > 0)
+	if (key.offset > 0) {
 		key.offset--;
-	else if (key.type > 0)
+	} else if (key.type > 0) {
 		key.type--;
-	else if (key.objectid > 0)
+		key.offset = (u64)-1;
+	} else if (key.objectid > 0) {
 		key.objectid--;
-	else
+		key.type = (u8)-1;
+		key.offset = (u64)-1;
+	} else {
 		return 1;
+	}
 
 	btrfs_release_path(path);
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
@@ -4866,7 +4872,6 @@ static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
  * was nothing in the tree that matched the search criteria.
  */
 int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
-			 struct btrfs_key *max_key,
 			 struct btrfs_path *path,
 			 u64 min_trans)
 {
@@ -4911,10 +4916,8 @@ int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
 		 * If it is too old, old, skip to the next one.
 		 */
 		while (slot < nritems) {
-			u64 blockptr;
 			u64 gen;
 
-			blockptr = btrfs_node_blockptr(cur, slot);
 			gen = btrfs_node_ptr_generation(cur, slot);
 			if (gen < min_trans) {
 				slot++;

fs/btrfs/ctree.h

@@ -47,6 +47,12 @@ extern struct kmem_cache *btrfs_path_cachep;
 extern struct kmem_cache *btrfs_free_space_cachep;
 struct btrfs_ordered_sum;
 
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+#define STATIC noinline
+#else
+#define STATIC static noinline
+#endif
+
 #define BTRFS_MAGIC 0x4D5F53665248425FULL /* ascii _BHRfS_M, no null */
 
 #define BTRFS_MAX_MIRRORS 3
@@ -1580,7 +1586,6 @@ struct btrfs_fs_info {
 	atomic_t scrubs_paused;
 	atomic_t scrub_cancel_req;
 	wait_queue_head_t scrub_pause_wait;
-	struct rw_semaphore scrub_super_lock;
 	int scrub_workers_refcnt;
 	struct btrfs_workers scrub_workers;
 	struct btrfs_workers scrub_wr_completion_workers;
@@ -1724,7 +1729,9 @@ struct btrfs_root {
 	int ref_cows;
 	int track_dirty;
 	int in_radix;
-
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+	int dummy_root;
+#endif
 	u64 defrag_trans_start;
 	struct btrfs_key defrag_progress;
 	struct btrfs_key defrag_max;
@@ -2461,8 +2468,7 @@ static inline unsigned long btrfs_item_nr_offset(int nr)
 		sizeof(struct btrfs_item) * nr;
 }
 
-static inline struct btrfs_item *btrfs_item_nr(struct extent_buffer *eb,
+static inline struct btrfs_item *btrfs_item_nr(int nr)
-					       int nr)
 {
 	return (struct btrfs_item *)btrfs_item_nr_offset(nr);
 }
@@ -2475,30 +2481,30 @@ static inline u32 btrfs_item_end(struct extent_buffer *eb,
 
 static inline u32 btrfs_item_end_nr(struct extent_buffer *eb, int nr)
 {
-	return btrfs_item_end(eb, btrfs_item_nr(eb, nr));
+	return btrfs_item_end(eb, btrfs_item_nr(nr));
 }
 
 static inline u32 btrfs_item_offset_nr(struct extent_buffer *eb, int nr)
 {
-	return btrfs_item_offset(eb, btrfs_item_nr(eb, nr));
+	return btrfs_item_offset(eb, btrfs_item_nr(nr));
 }
 
 static inline u32 btrfs_item_size_nr(struct extent_buffer *eb, int nr)
 {
-	return btrfs_item_size(eb, btrfs_item_nr(eb, nr));
+	return btrfs_item_size(eb, btrfs_item_nr(nr));
 }
 
 static inline void btrfs_item_key(struct extent_buffer *eb,
 			   struct btrfs_disk_key *disk_key, int nr)
 {
-	struct btrfs_item *item = btrfs_item_nr(eb, nr);
+	struct btrfs_item *item = btrfs_item_nr(nr);
 	read_eb_member(eb, item, struct btrfs_item, key, disk_key);
 }
 
 static inline void btrfs_set_item_key(struct extent_buffer *eb,
 			       struct btrfs_disk_key *disk_key, int nr)
 {
-	struct btrfs_item *item = btrfs_item_nr(eb, nr);
+	struct btrfs_item *item = btrfs_item_nr(nr);
 	write_eb_member(eb, item, struct btrfs_item, key, disk_key);
 }
 
@@ -2666,7 +2672,7 @@ static inline void btrfs_set_header_backref_rev(struct extent_buffer *eb,
 	btrfs_set_header_flags(eb, flags);
 }
 
-static inline unsigned long btrfs_header_fsid(struct extent_buffer *eb)
+static inline unsigned long btrfs_header_fsid(void)
 {
 	return offsetof(struct btrfs_header, fsid);
 }
@@ -3308,7 +3314,6 @@ int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path,
 			struct btrfs_key *key, int lowest_level,
 			u64 min_trans);
 int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
-			 struct btrfs_key *max_key,
 			 struct btrfs_path *path,
 			 u64 min_trans);
 enum btrfs_compare_tree_result {
@@ -3675,8 +3680,7 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,
 			       u32 min_type);
 
 int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput);
-int btrfs_start_all_delalloc_inodes(struct btrfs_fs_info *fs_info,
+int btrfs_start_delalloc_roots(struct btrfs_fs_info *fs_info, int delay_iput);
-				    int delay_iput);
 int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
 			      struct extent_state **cached_state);
 int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
@@ -3944,9 +3948,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		    u64 end, struct btrfs_scrub_progress *progress,
 		    int readonly, int is_dev_replace);
 void btrfs_scrub_pause(struct btrfs_root *root);
-void btrfs_scrub_pause_super(struct btrfs_root *root);
 void btrfs_scrub_continue(struct btrfs_root *root);
-void btrfs_scrub_continue_super(struct btrfs_root *root);
 int btrfs_scrub_cancel(struct btrfs_fs_info *info);
 int btrfs_scrub_cancel_dev(struct btrfs_fs_info *info,
 			   struct btrfs_device *dev);
@@ -4028,5 +4030,9 @@ static inline int btrfs_defrag_cancelled(struct btrfs_fs_info *fs_info)
 	return signal_pending(current);
 }
 
+/* Sanity test specific functions */
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+void btrfs_test_destroy_inode(struct inode *inode);
+#endif
 
 #endif

fs/btrfs/delayed-inode.c

@@ -108,8 +108,8 @@ static struct btrfs_delayed_node *btrfs_get_delayed_node(struct inode *inode)
 			return node;
 		}
 		btrfs_inode->delayed_node = node;
-		atomic_inc(&node->refs);	/* can be accessed */
+		/* can be accessed and cached in the inode */
-		atomic_inc(&node->refs);	/* cached in the inode */
+		atomic_add(2, &node->refs);
 		spin_unlock(&root->inode_lock);
 		return node;
 	}
@@ -138,8 +138,8 @@ static struct btrfs_delayed_node *btrfs_get_or_create_delayed_node(
 		return ERR_PTR(-ENOMEM);
 	btrfs_init_delayed_node(node, root, ino);
 
-	atomic_inc(&node->refs);	/* cached in the btrfs inode */
+	/* cached in the btrfs inode and can be accessed */
-	atomic_inc(&node->refs);	/* can be accessed */
+	atomic_add(2, &node->refs);
 
 	ret = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM);
 	if (ret) {
@@ -649,14 +649,13 @@ static int btrfs_delayed_inode_reserve_metadata(
 			goto out;
 
 		ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes);
-		if (!ret)
+		if (!WARN_ON(ret))
 			goto out;
 
 		/*
 		 * Ok this is a problem, let's just steal from the global rsv
 		 * since this really shouldn't happen that often.
 		 */
-		WARN_ON(1);
 		ret = btrfs_block_rsv_migrate(&root->fs_info->global_block_rsv,
 					      dst_rsv, num_bytes);
 		goto out;
@@ -771,13 +770,13 @@ static int btrfs_batch_insert_items(struct btrfs_root *root,
 	 */
 	btrfs_set_path_blocking(path);
 
-	keys = kmalloc(sizeof(struct btrfs_key) * nitems, GFP_NOFS);
+	keys = kmalloc_array(nitems, sizeof(struct btrfs_key), GFP_NOFS);
 	if (!keys) {
 		ret = -ENOMEM;
 		goto out;
 	}
 
-	data_size = kmalloc(sizeof(u32) * nitems, GFP_NOFS);
+	data_size = kmalloc_array(nitems, sizeof(u32), GFP_NOFS);
 	if (!data_size) {
 		ret = -ENOMEM;
 		goto error;
@@ -1174,8 +1173,10 @@ int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 	mutex_unlock(&delayed_node->mutex);
 
 	path = btrfs_alloc_path();
-	if (!path)
+	if (!path) {
+		btrfs_release_delayed_node(delayed_node);
 		return -ENOMEM;
+	}
 	path->leave_spinning = 1;
 
 	block_rsv = trans->block_rsv;

fs/btrfs/dev-replace.c

@@ -26,7 +26,6 @@
 #include <linux/kthread.h>
 #include <linux/math64.h>
 #include <asm/div64.h>
-#include "compat.h"
 #include "ctree.h"
 #include "extent_map.h"
 #include "disk-io.h"
@@ -38,7 +37,6 @@
 #include "rcu-string.h"
 #include "dev-replace.h"
 
-static u64 btrfs_get_seconds_since_1970(void);
 static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 				       int scrub_ret);
 static void btrfs_dev_replace_update_device_in_mapping_tree(
@@ -296,13 +294,6 @@ void btrfs_after_dev_replace_commit(struct btrfs_fs_info *fs_info)
 		dev_replace->cursor_left_last_write_of_item;
 }
 
-static u64 btrfs_get_seconds_since_1970(void)
-{
-	struct timespec t = CURRENT_TIME_SEC;
-
-	return t.tv_sec;
-}
-
 int btrfs_dev_replace_start(struct btrfs_root *root,
 			    struct btrfs_ioctl_dev_replace_args *args)
 {
@@ -390,7 +381,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	 * go to the tgtdev as well (refer to btrfs_map_block()).
 	 */
 	dev_replace->replace_state = BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED;
-	dev_replace->time_started = btrfs_get_seconds_since_1970();
+	dev_replace->time_started = get_seconds();
 	dev_replace->cursor_left = 0;
 	dev_replace->committed_cursor_left = 0;
 	dev_replace->cursor_left_last_write_of_item = 0;
@@ -400,7 +391,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	args->result = BTRFS_IOCTL_DEV_REPLACE_RESULT_NO_ERROR;
 	btrfs_dev_replace_unlock(dev_replace);
 
-	btrfs_wait_all_ordered_extents(root->fs_info);
+	btrfs_wait_ordered_roots(root->fs_info, -1);
 
 	/* force writing the updated state information to disk */
 	trans = btrfs_start_transaction(root, 0);
@@ -470,12 +461,12 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	 * flush all outstanding I/O and inode extent mappings before the
 	 * copy operation is declared as being finished
 	 */
-	ret = btrfs_start_all_delalloc_inodes(root->fs_info, 0);
+	ret = btrfs_start_delalloc_roots(root->fs_info, 0);
 	if (ret) {
 		mutex_unlock(&dev_replace->lock_finishing_cancel_unmount);
 		return ret;
 	}
-	btrfs_wait_all_ordered_extents(root->fs_info);
+	btrfs_wait_ordered_roots(root->fs_info, -1);
 
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
@@ -493,7 +484,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 			  : BTRFS_IOCTL_DEV_REPLACE_STATE_FINISHED;
 	dev_replace->tgtdev = NULL;
 	dev_replace->srcdev = NULL;
-	dev_replace->time_stopped = btrfs_get_seconds_since_1970();
+	dev_replace->time_stopped = get_seconds();
 	dev_replace->item_needs_writeback = 1;
 
 	if (scrub_ret) {
@@ -650,6 +641,9 @@ static u64 __btrfs_dev_replace_cancel(struct btrfs_fs_info *fs_info)
 	u64 result;
 	int ret;
 
+	if (fs_info->sb->s_flags & MS_RDONLY)
+		return -EROFS;
+
 	mutex_lock(&dev_replace->lock_finishing_cancel_unmount);
 	btrfs_dev_replace_lock(dev_replace);
 	switch (dev_replace->replace_state) {
@@ -668,7 +662,7 @@ static u64 __btrfs_dev_replace_cancel(struct btrfs_fs_info *fs_info)
 		break;
 	}
 	dev_replace->replace_state = BTRFS_IOCTL_DEV_REPLACE_STATE_CANCELED;
-	dev_replace->time_stopped = btrfs_get_seconds_since_1970();
+	dev_replace->time_stopped = get_seconds();
 	dev_replace->item_needs_writeback = 1;
 	btrfs_dev_replace_unlock(dev_replace);
 	btrfs_scrub_cancel(fs_info);
@@ -703,7 +697,7 @@ void btrfs_dev_replace_suspend_for_unmount(struct btrfs_fs_info *fs_info)
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED:
 		dev_replace->replace_state =
 			BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED;
-		dev_replace->time_stopped = btrfs_get_seconds_since_1970();
+		dev_replace->time_stopped = get_seconds();
 		dev_replace->item_needs_writeback = 1;
 		pr_info("btrfs: suspending dev_replace for unmount\n");
 		break;

fs/btrfs/dir-item.c

@@ -58,7 +58,7 @@ static struct btrfs_dir_item *insert_with_overflow(struct btrfs_trans_handle
 		return ERR_PTR(ret);
 	WARN_ON(ret > 0);
 	leaf = path->nodes[0];
-	item = btrfs_item_nr(leaf, path->slots[0]);
+	item = btrfs_item_nr(path->slots[0]);
 	ptr = btrfs_item_ptr(leaf, path->slots[0], char);
 	BUG_ON(data_size > btrfs_item_size(leaf, item));
 	ptr += btrfs_item_size(leaf, item) - data_size;
@@ -474,8 +474,10 @@ int verify_dir_item(struct btrfs_root *root,
 	}
 
 	/* BTRFS_MAX_XATTR_SIZE is the same for all dir items */
-	if (btrfs_dir_data_len(leaf, dir_item) > BTRFS_MAX_XATTR_SIZE(root)) {
+	if ((btrfs_dir_data_len(leaf, dir_item) +
-		printk(KERN_CRIT "btrfs: invalid dir item data len: %u\n",
+	     btrfs_dir_name_len(leaf, dir_item)) > BTRFS_MAX_XATTR_SIZE(root)) {
+		printk(KERN_CRIT "btrfs: invalid dir item name + data len: %u + %u\n",
+		       (unsigned)btrfs_dir_name_len(leaf, dir_item),
 		       (unsigned)btrfs_dir_data_len(leaf, dir_item));
 		return 1;
 	}

fs/btrfs/disk-io.c

@@ -33,7 +33,6 @@
 #include <linux/uuid.h>
 #include <linux/semaphore.h>
 #include <asm/unaligned.h>
-#include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -64,7 +63,6 @@ static void btrfs_destroy_ordered_operations(struct btrfs_transaction *t,
 static void btrfs_destroy_ordered_extents(struct btrfs_root *root);
 static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 				      struct btrfs_root *root);
-static void btrfs_evict_pending_snapshots(struct btrfs_transaction *t);
 static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root);
 static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 					struct extent_io_tree *dirty_pages,
@@ -477,14 +475,8 @@ static int csum_dirty_buffer(struct btrfs_root *root, struct page *page)
 	if (page != eb->pages[0])
 		return 0;
 	found_start = btrfs_header_bytenr(eb);
-	if (found_start != start) {
+	if (WARN_ON(found_start != start || !PageUptodate(page)))
-		WARN_ON(1);
 		return 0;
-	}
-	if (!PageUptodate(page)) {
-		WARN_ON(1);
-		return 0;
-	}
 	csum_tree_block(root, eb, 0);
 	return 0;
 }
@@ -496,7 +488,7 @@ static int check_tree_block_fsid(struct btrfs_root *root,
 	u8 fsid[BTRFS_UUID_SIZE];
 	int ret = 1;
 
-	read_extent_buffer(eb, fsid, btrfs_header_fsid(eb), BTRFS_FSID_SIZE);
+	read_extent_buffer(eb, fsid, btrfs_header_fsid(), BTRFS_FSID_SIZE);
 	while (fs_devices) {
 		if (!memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE)) {
 			ret = 0;
@@ -1105,8 +1097,7 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
 {
 	struct inode *btree_inode = root->fs_info->btree_inode;
 	struct extent_buffer *eb;
-	eb = find_extent_buffer(&BTRFS_I(btree_inode)->io_tree,
+	eb = find_extent_buffer(&BTRFS_I(btree_inode)->io_tree, bytenr);
-				bytenr, blocksize);
 	return eb;
 }
 
@@ -1229,14 +1220,18 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	atomic_set(&root->refs, 1);
 	root->log_transid = 0;
 	root->last_log_commit = 0;
-	extent_io_tree_init(&root->dirty_log_pages,
+	if (fs_info)
-			     fs_info->btree_inode->i_mapping);
+		extent_io_tree_init(&root->dirty_log_pages,
+				     fs_info->btree_inode->i_mapping);
 
 	memset(&root->root_key, 0, sizeof(root->root_key));
 	memset(&root->root_item, 0, sizeof(root->root_item));
 	memset(&root->defrag_progress, 0, sizeof(root->defrag_progress));
 	memset(&root->root_kobj, 0, sizeof(root->root_kobj));
-	root->defrag_trans_start = fs_info->generation;
+	if (fs_info)
+		root->defrag_trans_start = fs_info->generation;
+	else
+		root->defrag_trans_start = 0;
 	init_completion(&root->kobj_unregister);
 	root->defrag_running = 0;
 	root->root_key.objectid = objectid;
@@ -1253,6 +1248,22 @@ static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info)
 	return root;
 }
 
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+/* Should only be used by the testing infrastructure */
+struct btrfs_root *btrfs_alloc_dummy_root(void)
+{
+	struct btrfs_root *root;
+
+	root = btrfs_alloc_root(NULL);
+	if (!root)
+		return ERR_PTR(-ENOMEM);
+	__setup_root(4096, 4096, 4096, 4096, root, NULL, 1);
+	root->dummy_root = 1;
+
+	return root;
+}
+#endif
+
 struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 				     struct btrfs_fs_info *fs_info,
 				     u64 objectid)
@@ -1292,7 +1303,7 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	btrfs_set_header_owner(leaf, objectid);
 	root->node = leaf;
 
-	write_extent_buffer(leaf, fs_info->fsid, btrfs_header_fsid(leaf),
+	write_extent_buffer(leaf, fs_info->fsid, btrfs_header_fsid(),
 			    BTRFS_FSID_SIZE);
 	write_extent_buffer(leaf, fs_info->chunk_tree_uuid,
 			    btrfs_header_chunk_tree_uuid(leaf),
@@ -1379,7 +1390,7 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 	root->node = leaf;
 
 	write_extent_buffer(root->node, root->fs_info->fsid,
-			    btrfs_header_fsid(root->node), BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(), BTRFS_FSID_SIZE);
 	btrfs_mark_buffer_dirty(root->node);
 	btrfs_tree_unlock(root->node);
 	return root;
@@ -1780,6 +1791,9 @@ static int transaction_kthread(void *arg)
 		wake_up_process(root->fs_info->cleaner_kthread);
 		mutex_unlock(&root->fs_info->transaction_kthread_mutex);
 
+		if (unlikely(test_bit(BTRFS_FS_STATE_ERROR,
+				      &root->fs_info->fs_state)))
+			btrfs_cleanup_transaction(root);
 		if (!try_to_freeze()) {
 			set_current_state(TASK_INTERRUPTIBLE);
 			if (!kthread_should_stop() &&
@@ -2013,50 +2027,28 @@ static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info)
 	btrfs_stop_workers(&fs_info->qgroup_rescan_workers);
 }
 
+static void free_root_extent_buffers(struct btrfs_root *root)
+{
+	if (root) {
+		free_extent_buffer(root->node);
+		free_extent_buffer(root->commit_root);
+		root->node = NULL;
+		root->commit_root = NULL;
+	}
+}
+
 /* helper to cleanup tree roots */
 static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
 {
-	free_extent_buffer(info->tree_root->node);
+	free_root_extent_buffers(info->tree_root);
-	free_extent_buffer(info->tree_root->commit_root);
-	info->tree_root->node = NULL;
-	info->tree_root->commit_root = NULL;
 
-	if (info->dev_root) {
+	free_root_extent_buffers(info->dev_root);
-		free_extent_buffer(info->dev_root->node);
+	free_root_extent_buffers(info->extent_root);
-		free_extent_buffer(info->dev_root->commit_root);
+	free_root_extent_buffers(info->csum_root);
-		info->dev_root->node = NULL;
+	free_root_extent_buffers(info->quota_root);
-		info->dev_root->commit_root = NULL;
+	free_root_extent_buffers(info->uuid_root);
-	}
+	if (chunk_root)
-	if (info->extent_root) {
+		free_root_extent_buffers(info->chunk_root);
-		free_extent_buffer(info->extent_root->node);
-		free_extent_buffer(info->extent_root->commit_root);
-		info->extent_root->node = NULL;
-		info->extent_root->commit_root = NULL;
-	}
-	if (info->csum_root) {
-		free_extent_buffer(info->csum_root->node);
-		free_extent_buffer(info->csum_root->commit_root);
-		info->csum_root->node = NULL;
-		info->csum_root->commit_root = NULL;
-	}
-	if (info->quota_root) {
-		free_extent_buffer(info->quota_root->node);
-		free_extent_buffer(info->quota_root->commit_root);
-		info->quota_root->node = NULL;
-		info->quota_root->commit_root = NULL;
-	}
-	if (info->uuid_root) {
-		free_extent_buffer(info->uuid_root->node);
-		free_extent_buffer(info->uuid_root->commit_root);
-		info->uuid_root->node = NULL;
-		info->uuid_root->commit_root = NULL;
-	}
-	if (chunk_root) {
-		free_extent_buffer(info->chunk_root->node);
-		free_extent_buffer(info->chunk_root->commit_root);
-		info->chunk_root->node = NULL;
-		info->chunk_root->commit_root = NULL;
-	}
 }
 
 static void del_fs_roots(struct btrfs_fs_info *fs_info)
@@ -2230,7 +2222,6 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->scrubs_paused, 0);
 	atomic_set(&fs_info->scrub_cancel_req, 0);
 	init_waitqueue_head(&fs_info->scrub_pause_wait);
-	init_rwsem(&fs_info->scrub_super_lock);
 	fs_info->scrub_workers_refcnt = 0;
 #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
 	fs_info->check_integrity_print_mask = 0;
@@ -2272,7 +2263,7 @@ int open_ctree(struct super_block *sb,
 	       sizeof(struct btrfs_key));
 	set_bit(BTRFS_INODE_DUMMY,
 		&BTRFS_I(fs_info->btree_inode)->runtime_flags);
-	insert_inode_hash(fs_info->btree_inode);
+	btrfs_insert_inode_hash(fs_info->btree_inode);
 
 	spin_lock_init(&fs_info->block_group_cache_lock);
 	fs_info->block_group_cache_tree = RB_ROOT;
@@ -2670,6 +2661,7 @@ int open_ctree(struct super_block *sb,
 
 	btrfs_set_root_node(&tree_root->root_item, tree_root->node);
 	tree_root->commit_root = btrfs_root_node(tree_root);
+	btrfs_set_root_refs(&tree_root->root_item, 1);
 
 	location.objectid = BTRFS_EXTENT_TREE_OBJECTID;
 	location.type = BTRFS_ROOT_ITEM_KEY;
@@ -3448,10 +3440,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 int write_ctree_super(struct btrfs_trans_handle *trans,
 		      struct btrfs_root *root, int max_mirrors)
 {
-	int ret;
+	return write_all_supers(root, max_mirrors);
-
-	ret = write_all_supers(root, max_mirrors);
-	return ret;
 }
 
 /* Drop a fs root from the radix tree and free it. */
@@ -3614,12 +3603,12 @@ int close_ctree(struct btrfs_root *root)
 		       percpu_counter_sum(&fs_info->delalloc_bytes));
 	}
 
+	del_fs_roots(fs_info);
+
 	btrfs_free_block_groups(fs_info);
 
 	btrfs_stop_all_workers(fs_info);
 
-	del_fs_roots(fs_info);
-
 	free_root_pointers(fs_info, 1);
 
 	iput(fs_info->btree_inode);
@@ -3669,10 +3658,20 @@ int btrfs_set_buffer_uptodate(struct extent_buffer *buf)
 
 void btrfs_mark_buffer_dirty(struct extent_buffer *buf)
 {
-	struct btrfs_root *root = BTRFS_I(buf->pages[0]->mapping->host)->root;
+	struct btrfs_root *root;
 	u64 transid = btrfs_header_generation(buf);
 	int was_dirty;
 
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+	/*
+	 * This is a fast path so only do this check if we have sanity tests
+	 * enabled.  Normal people shouldn't be marking dummy buffers as dirty
+	 * outside of the sanity tests.
+	 */
+	if (unlikely(test_bit(EXTENT_BUFFER_DUMMY, &buf->bflags)))
+		return;
+#endif
+	root = BTRFS_I(buf->pages[0]->mapping->host)->root;
 	btrfs_assert_tree_locked(buf);
 	if (transid != root->fs_info->generation)
 		WARN(1, KERN_CRIT "btrfs transid mismatch buffer %llu, "
@@ -3802,7 +3801,8 @@ static void btrfs_destroy_all_ordered_extents(struct btrfs_fs_info *fs_info)
 	while (!list_empty(&splice)) {
 		root = list_first_entry(&splice, struct btrfs_root,
 					ordered_root);
-		list_del_init(&root->ordered_root);
+		list_move_tail(&root->ordered_root,
+			       &fs_info->ordered_roots);
 
 		btrfs_destroy_ordered_extents(root);
 
@@ -3880,24 +3880,6 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 	return ret;
 }
 
-static void btrfs_evict_pending_snapshots(struct btrfs_transaction *t)
-{
-	struct btrfs_pending_snapshot *snapshot;
-	struct list_head splice;
-
-	INIT_LIST_HEAD(&splice);
-
-	list_splice_init(&t->pending_snapshots, &splice);
-
-	while (!list_empty(&splice)) {
-		snapshot = list_entry(splice.next,
-				      struct btrfs_pending_snapshot,
-				      list);
-		snapshot->error = -ECANCELED;
-		list_del_init(&snapshot->list);
-	}
-}
-
 static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root)
 {
 	struct btrfs_inode *btrfs_inode;
@@ -4027,15 +4009,13 @@ static int btrfs_destroy_pinned_extent(struct btrfs_root *root,
 void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 				   struct btrfs_root *root)
 {
+	btrfs_destroy_ordered_operations(cur_trans, root);
+
 	btrfs_destroy_delayed_refs(cur_trans, root);
-	btrfs_block_rsv_release(root, &root->fs_info->trans_block_rsv,
-				cur_trans->dirty_pages.dirty_bytes);
 
 	cur_trans->state = TRANS_STATE_COMMIT_START;
 	wake_up(&root->fs_info->transaction_blocked_wait);
 
-	btrfs_evict_pending_snapshots(cur_trans);
-
 	cur_trans->state = TRANS_STATE_UNBLOCKED;
 	wake_up(&root->fs_info->transaction_wait);
 
@@ -4059,63 +4039,51 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 static int btrfs_cleanup_transaction(struct btrfs_root *root)
 {
 	struct btrfs_transaction *t;
-	LIST_HEAD(list);
 
 	mutex_lock(&root->fs_info->transaction_kthread_mutex);
 
 	spin_lock(&root->fs_info->trans_lock);
-	list_splice_init(&root->fs_info->trans_list, &list);
+	while (!list_empty(&root->fs_info->trans_list)) {
-	root->fs_info->running_transaction = NULL;
+		t = list_first_entry(&root->fs_info->trans_list,
-	spin_unlock(&root->fs_info->trans_lock);
+				     struct btrfs_transaction, list);
+		if (t->state >= TRANS_STATE_COMMIT_START) {
+			atomic_inc(&t->use_count);
+			spin_unlock(&root->fs_info->trans_lock);
+			btrfs_wait_for_commit(root, t->transid);
+			btrfs_put_transaction(t);
+			spin_lock(&root->fs_info->trans_lock);
+			continue;
+		}
+		if (t == root->fs_info->running_transaction) {
+			t->state = TRANS_STATE_COMMIT_DOING;
+			spin_unlock(&root->fs_info->trans_lock);
+			/*
+			 * We wait for 0 num_writers since we don't hold a trans
+			 * handle open currently for this transaction.
+			 */
+			wait_event(t->writer_wait,
+				   atomic_read(&t->num_writers) == 0);
+		} else {
+			spin_unlock(&root->fs_info->trans_lock);
+		}
+		btrfs_cleanup_one_transaction(t, root);
 
-	while (!list_empty(&list)) {
+		spin_lock(&root->fs_info->trans_lock);
-		t = list_entry(list.next, struct btrfs_transaction, list);
+		if (t == root->fs_info->running_transaction)
-
+			root->fs_info->running_transaction = NULL;
-		btrfs_destroy_ordered_operations(t, root);
-
-		btrfs_destroy_all_ordered_extents(root->fs_info);
-
-		btrfs_destroy_delayed_refs(t, root);
-
-		/*
-		 *  FIXME: cleanup wait for commit
-		 *  We needn't acquire the lock here, because we are during
-		 *  the umount, there is no other task which will change it.
-		 */
-		t->state = TRANS_STATE_COMMIT_START;
-		smp_mb();
-		if (waitqueue_active(&root->fs_info->transaction_blocked_wait))
-			wake_up(&root->fs_info->transaction_blocked_wait);
-
-		btrfs_evict_pending_snapshots(t);
-
-		t->state = TRANS_STATE_UNBLOCKED;
-		smp_mb();
-		if (waitqueue_active(&root->fs_info->transaction_wait))
-			wake_up(&root->fs_info->transaction_wait);
-
-		btrfs_destroy_delayed_inodes(root);
-		btrfs_assert_delayed_root_empty(root);
-
-		btrfs_destroy_all_delalloc_inodes(root->fs_info);
-
-		btrfs_destroy_marked_extents(root, &t->dirty_pages,
-					     EXTENT_DIRTY);
-
-		btrfs_destroy_pinned_extent(root,
-					    root->fs_info->pinned_extents);
-
-		t->state = TRANS_STATE_COMPLETED;
-		smp_mb();
-		if (waitqueue_active(&t->commit_wait))
-			wake_up(&t->commit_wait);
-
-		atomic_set(&t->use_count, 0);
 		list_del_init(&t->list);
-		memset(t, 0, sizeof(*t));
+		spin_unlock(&root->fs_info->trans_lock);
-		kmem_cache_free(btrfs_transaction_cachep, t);
-	}
 
+		btrfs_put_transaction(t);
+		trace_btrfs_transaction_commit(root);
+		spin_lock(&root->fs_info->trans_lock);
+	}
+	spin_unlock(&root->fs_info->trans_lock);
+	btrfs_destroy_all_ordered_extents(root->fs_info);
+	btrfs_destroy_delayed_inodes(root);
+	btrfs_assert_delayed_root_empty(root);
+	btrfs_destroy_pinned_extent(root, root->fs_info->pinned_extents);
+	btrfs_destroy_all_delalloc_inodes(root->fs_info);
 	mutex_unlock(&root->fs_info->transaction_kthread_mutex);
 
 	return 0;

fs/btrfs/disk-io.h

@@ -86,6 +86,10 @@ void btrfs_drop_and_free_fs_root(struct btrfs_fs_info *fs_info,
 				 struct btrfs_root *root);
 void btrfs_free_fs_root(struct btrfs_root *root);
 
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+struct btrfs_root *btrfs_alloc_dummy_root(void);
+#endif
+
 /*
  * This function is used to grab the root, and avoid it is freed when we
  * access it. But it doesn't ensure that the tree is not dropped.

fs/btrfs/export.c

@@ -5,7 +5,6 @@
 #include "btrfs_inode.h"
 #include "print-tree.h"
 #include "export.h"
-#include "compat.h"
 
 #define BTRFS_FID_SIZE_NON_CONNECTABLE (offsetof(struct btrfs_fid, \
 						 parent_objectid) / 4)

fs/btrfs/extent-tree.c

@@ -25,7 +25,6 @@
 #include <linux/slab.h>
 #include <linux/ratelimit.h>
 #include <linux/percpu_counter.h>
-#include "compat.h"
 #include "hash.h"
 #include "ctree.h"
 #include "disk-io.h"
@@ -1551,9 +1550,8 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans,
 	if (ret && !insert) {
 		err = -ENOENT;
 		goto out;
-	} else if (ret) {
+	} else if (WARN_ON(ret)) {
 		err = -EIO;
-		WARN_ON(1);
 		goto out;
 	}
 
@@ -1979,7 +1977,6 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 	struct btrfs_extent_item *item;
 	u64 refs;
 	int ret;
-	int err = 0;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -1992,14 +1989,9 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 					   path, bytenr, num_bytes, parent,
 					   root_objectid, owner, offset,
 					   refs_to_add, extent_op);
-	if (ret == 0)
+	if (ret != -EAGAIN)
 		goto out;
 
-	if (ret != -EAGAIN) {
-		err = ret;
-		goto out;
-	}
-
 	leaf = path->nodes[0];
 	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
 	refs = btrfs_extent_refs(leaf, item);
@@ -2021,7 +2013,7 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 		btrfs_abort_transaction(trans, root, ret);
 out:
 	btrfs_free_path(path);
-	return err;
+	return ret;
 }
 
 static int run_delayed_data_ref(struct btrfs_trans_handle *trans,
@@ -2137,15 +2129,28 @@ static int run_delayed_extent_op(struct btrfs_trans_handle *trans,
 	}
 	if (ret > 0) {
 		if (metadata) {
-			btrfs_release_path(path);
+			if (path->slots[0] > 0) {
-			metadata = 0;
+				path->slots[0]--;
+				btrfs_item_key_to_cpu(path->nodes[0], &key,
+						      path->slots[0]);
+				if (key.objectid == node->bytenr &&
+				    key.type == BTRFS_EXTENT_ITEM_KEY &&
+				    key.offset == node->num_bytes)
+					ret = 0;
+			}
+			if (ret > 0) {
+				btrfs_release_path(path);
+				metadata = 0;
 
-			key.offset = node->num_bytes;
+				key.objectid = node->bytenr;
-			key.type = BTRFS_EXTENT_ITEM_KEY;
+				key.offset = node->num_bytes;
-			goto again;
+				key.type = BTRFS_EXTENT_ITEM_KEY;
+				goto again;
+			}
+		} else {
+			err = -EIO;
+			goto out;
 		}
-		err = -EIO;
-		goto out;
 	}
 
 	leaf = path->nodes[0];
@@ -2234,8 +2239,12 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 {
 	int ret = 0;
 
-	if (trans->aborted)
+	if (trans->aborted) {
+		if (insert_reserved)
+			btrfs_pin_extent(root, node->bytenr,
+					 node->num_bytes, 1);
 		return 0;
+	}
 
 	if (btrfs_delayed_ref_is_head(node)) {
 		struct btrfs_delayed_ref_head *head;
@@ -2411,6 +2420,14 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 				btrfs_free_delayed_extent_op(extent_op);
 
 				if (ret) {
+					/*
+					 * Need to reset must_insert_reserved if
+					 * there was an error so the abort stuff
+					 * can cleanup the reserved space
+					 * properly.
+					 */
+					if (must_insert_reserved)
+						locked_ref->must_insert_reserved = 1;
 					btrfs_debug(fs_info, "run_delayed_extent_op returned %d", ret);
 					spin_lock(&delayed_refs->lock);
 					btrfs_delayed_ref_unlock(locked_ref);
@@ -3197,8 +3214,7 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group,
 		if (ret)
 			goto out_put;
 
-		ret = btrfs_truncate_free_space_cache(root, trans, path,
+		ret = btrfs_truncate_free_space_cache(root, trans, inode);
-						      inode);
 		if (ret)
 			goto out_put;
 	}
@@ -3318,10 +3334,9 @@ int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
 		last = cache->key.objectid + cache->key.offset;
 
 		err = write_one_cache_group(trans, root, path, cache);
+		btrfs_put_block_group(cache);
 		if (err) /* File system offline */
 			goto out;
-
-		btrfs_put_block_group(cache);
 	}
 
 	while (1) {
@@ -3605,10 +3620,9 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes)
 	/* make sure bytes are sectorsize aligned */
 	bytes = ALIGN(bytes, root->sectorsize);
 
-	if (root == root->fs_info->tree_root ||
+	if (btrfs_is_free_space_inode(inode)) {
-	    BTRFS_I(inode)->location.objectid == BTRFS_FREE_INO_OBJECTID) {
-		alloc_chunk = 0;
 		committed = 1;
+		ASSERT(current->journal_info);
 	}
 
 	data_sinfo = fs_info->data_sinfo;
@@ -3636,6 +3650,16 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes)
 			spin_unlock(&data_sinfo->lock);
 alloc:
 			alloc_target = btrfs_get_alloc_profile(root, 1);
+			/*
+			 * It is ugly that we don't call nolock join
+			 * transaction for the free space inode case here.
+			 * But it is safe because we only do the data space
+			 * reservation for the free space cache in the
+			 * transaction context, the common join transaction
+			 * just increase the counter of the current transaction
+			 * handler, doesn't try to acquire the trans_lock of
+			 * the fs.
+			 */
 			trans = btrfs_join_transaction(root);
 			if (IS_ERR(trans))
 				return PTR_ERR(trans);
@@ -3681,6 +3705,9 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes)
 			goto again;
 		}
 
+		trace_btrfs_space_reservation(root->fs_info,
+					      "space_info:enospc",
+					      data_sinfo->flags, bytes, 1);
 		return -ENOSPC;
 	}
 	data_sinfo->bytes_may_use += bytes;
@@ -3989,12 +4016,26 @@ static void btrfs_writeback_inodes_sb_nr(struct btrfs_root *root,
 		 * the filesystem is readonly(all dirty pages are written to
 		 * the disk).
 		 */
-		btrfs_start_all_delalloc_inodes(root->fs_info, 0);
+		btrfs_start_delalloc_roots(root->fs_info, 0);
 		if (!current->journal_info)
-			btrfs_wait_all_ordered_extents(root->fs_info);
+			btrfs_wait_ordered_roots(root->fs_info, -1);
 	}
 }
 
+static inline int calc_reclaim_items_nr(struct btrfs_root *root, u64 to_reclaim)
+{
+	u64 bytes;
+	int nr;
+
+	bytes = btrfs_calc_trans_metadata_size(root, 1);
+	nr = (int)div64_u64(to_reclaim, bytes);
+	if (!nr)
+		nr = 1;
+	return nr;
+}
+
+#define EXTENT_SIZE_PER_ITEM	(256 * 1024)
+
 /*
  * shrink metadata reservation for delalloc
  */
@@ -4007,24 +4048,30 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig,
 	u64 delalloc_bytes;
 	u64 max_reclaim;
 	long time_left;
-	unsigned long nr_pages = (2 * 1024 * 1024) >> PAGE_CACHE_SHIFT;
+	unsigned long nr_pages;
-	int loops = 0;
+	int loops;
+	int items;
 	enum btrfs_reserve_flush_enum flush;
 
+	/* Calc the number of the pages we need flush for space reservation */
+	items = calc_reclaim_items_nr(root, to_reclaim);
+	to_reclaim = items * EXTENT_SIZE_PER_ITEM;
+
 	trans = (struct btrfs_trans_handle *)current->journal_info;
 	block_rsv = &root->fs_info->delalloc_block_rsv;
 	space_info = block_rsv->space_info;
 
-	smp_mb();
 	delalloc_bytes = percpu_counter_sum_positive(
 						&root->fs_info->delalloc_bytes);
 	if (delalloc_bytes == 0) {
 		if (trans)
 			return;
-		btrfs_wait_all_ordered_extents(root->fs_info);
+		if (wait_ordered)
+			btrfs_wait_ordered_roots(root->fs_info, items);
 		return;
 	}
 
+	loops = 0;
 	while (delalloc_bytes && loops < 3) {
 		max_reclaim = min(delalloc_bytes, to_reclaim);
 		nr_pages = max_reclaim >> PAGE_CACHE_SHIFT;
@@ -4033,9 +4080,19 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig,
 		 * We need to wait for the async pages to actually start before
 		 * we do anything.
 		 */
-		wait_event(root->fs_info->async_submit_wait,
+		max_reclaim = atomic_read(&root->fs_info->async_delalloc_pages);
-			   !atomic_read(&root->fs_info->async_delalloc_pages));
+		if (!max_reclaim)
+			goto skip_async;
 
+		if (max_reclaim <= nr_pages)
+			max_reclaim = 0;
+		else
+			max_reclaim -= nr_pages;
+
+		wait_event(root->fs_info->async_submit_wait,
+			   atomic_read(&root->fs_info->async_delalloc_pages) <=
+			   (int)max_reclaim);
+skip_async:
 		if (!trans)
 			flush = BTRFS_RESERVE_FLUSH_ALL;
 		else
@@ -4049,13 +4106,12 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig,
 
 		loops++;
 		if (wait_ordered && !trans) {
-			btrfs_wait_all_ordered_extents(root->fs_info);
+			btrfs_wait_ordered_roots(root->fs_info, items);
 		} else {
 			time_left = schedule_timeout_killable(1);
 			if (time_left)
 				break;
 		}
-		smp_mb();
 		delalloc_bytes = percpu_counter_sum_positive(
 						&root->fs_info->delalloc_bytes);
 	}
@@ -4140,16 +4196,11 @@ static int flush_space(struct btrfs_root *root,
 	switch (state) {
 	case FLUSH_DELAYED_ITEMS_NR:
 	case FLUSH_DELAYED_ITEMS:
-		if (state == FLUSH_DELAYED_ITEMS_NR) {
+		if (state == FLUSH_DELAYED_ITEMS_NR)
-			u64 bytes = btrfs_calc_trans_metadata_size(root, 1);
+			nr = calc_reclaim_items_nr(root, num_bytes) * 2;
-
+		else
-			nr = (int)div64_u64(num_bytes, bytes);
-			if (!nr)
-				nr = 1;
-			nr *= 2;
-		} else {
 			nr = -1;
-		}
+
 		trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans)) {
 			ret = PTR_ERR(trans);
@@ -4332,6 +4383,10 @@ static int reserve_metadata_bytes(struct btrfs_root *root,
 		    !block_rsv_use_bytes(global_rsv, orig_bytes))
 			ret = 0;
 	}
+	if (ret == -ENOSPC)
+		trace_btrfs_space_reservation(root->fs_info,
+					      "space_info:enospc",
+					      space_info->flags, orig_bytes, 1);
 	if (flushing) {
 		spin_lock(&space_info->lock);
 		space_info->flush = 0;
@@ -4986,7 +5041,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 		mutex_unlock(&BTRFS_I(inode)->delalloc_mutex);
 
 	if (to_reserve)
-		trace_btrfs_space_reservation(root->fs_info,"delalloc",
+		trace_btrfs_space_reservation(root->fs_info, "delalloc",
 					      btrfs_ino(inode), to_reserve, 1);
 	block_rsv_add_bytes(block_rsv, to_reserve, 1);
 
@@ -5264,6 +5319,8 @@ static int pin_down_extent(struct btrfs_root *root,
 
 	set_extent_dirty(root->fs_info->pinned_extents, bytenr,
 			 bytenr + num_bytes - 1, GFP_NOFS | __GFP_NOFAIL);
+	if (reserved)
+		trace_btrfs_reserved_extent_free(root, bytenr, num_bytes);
 	return 0;
 }
 
@@ -5718,9 +5775,8 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 			}
 			extent_slot = path->slots[0];
 		}
-	} else if (ret == -ENOENT) {
+	} else if (WARN_ON(ret == -ENOENT)) {
 		btrfs_print_leaf(extent_root, path->nodes[0]);
-		WARN_ON(1);
 		btrfs_err(info,
 			"unable to find ref byte nr %llu parent %llu root %llu  owner %llu offset %llu",
 			bytenr, parent, root_objectid, owner_objectid,
@@ -5967,6 +6023,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 
 		btrfs_add_free_space(cache, buf->start, buf->len);
 		btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE);
+		trace_btrfs_reserved_extent_free(root, buf->start, buf->len);
 		pin = 0;
 	}
 out:
@@ -6594,8 +6651,6 @@ int btrfs_reserve_extent(struct btrfs_root *root,
 		}
 	}
 
-	trace_btrfs_reserved_extent_alloc(root, ins->objectid, ins->offset);
-
 	return ret;
 }
 
@@ -6707,6 +6762,7 @@ static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 			ins->objectid, ins->offset);
 		BUG();
 	}
+	trace_btrfs_reserved_extent_alloc(root, ins->objectid, ins->offset);
 	return ret;
 }
 
@@ -6731,13 +6787,18 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 		size += sizeof(*block_info);
 
 	path = btrfs_alloc_path();
-	if (!path)
+	if (!path) {
+		btrfs_free_and_pin_reserved_extent(root, ins->objectid,
+						   root->leafsize);
 		return -ENOMEM;
+	}
 
 	path->leave_spinning = 1;
 	ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
 				      ins, size);
 	if (ret) {
+		btrfs_free_and_pin_reserved_extent(root, ins->objectid,
+						   root->leafsize);
 		btrfs_free_path(path);
 		return ret;
 	}
@@ -6779,6 +6840,8 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 			ins->objectid, ins->offset);
 		BUG();
 	}
+
+	trace_btrfs_reserved_extent_alloc(root, ins->objectid, root->leafsize);
 	return ret;
 }
 
@@ -7983,7 +8046,7 @@ u64 btrfs_account_ro_block_groups_free_space(struct btrfs_space_info *sinfo)
 
 	spin_lock(&sinfo->lock);
 
-	for(i = 0; i < BTRFS_NR_RAID_TYPES; i++)
+	for (i = 0; i < BTRFS_NR_RAID_TYPES; i++)
 		if (!list_empty(&sinfo->block_groups[i]))
 			free_bytes += __btrfs_get_ro_block_group_free_space(
 						&sinfo->block_groups[i]);
@@ -8271,15 +8334,14 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 
 	release_global_block_rsv(info);
 
-	while(!list_empty(&info->space_info)) {
+	while (!list_empty(&info->space_info)) {
 		space_info = list_entry(info->space_info.next,
 					struct btrfs_space_info,
 					list);
 		if (btrfs_test_opt(info->tree_root, ENOSPC_DEBUG)) {
-			if (space_info->bytes_pinned > 0 ||
+			if (WARN_ON(space_info->bytes_pinned > 0 ||
 			    space_info->bytes_reserved > 0 ||
-			    space_info->bytes_may_use > 0) {
+			    space_info->bytes_may_use > 0)) {
-				WARN_ON(1);
 				dump_space_info(space_info, 0, 0);
 			}
 		}

fs/btrfs/extent_io.c

@@ -13,13 +13,13 @@
 #include <linux/cleancache.h>
 #include "extent_io.h"
 #include "extent_map.h"
-#include "compat.h"
 #include "ctree.h"
 #include "btrfs_inode.h"
 #include "volumes.h"
 #include "check-integrity.h"
 #include "locking.h"
 #include "rcu-string.h"
+#include "backref.h"
 
 static struct kmem_cache *extent_state_cache;
 static struct kmem_cache *extent_buffer_cache;
@@ -1597,11 +1597,10 @@ static noinline int lock_delalloc_pages(struct inode *inode,
  *
  * 1 is returned if we find something, 0 if nothing was in the tree
  */
-static noinline u64 find_lock_delalloc_range(struct inode *inode,
+STATIC u64 find_lock_delalloc_range(struct inode *inode,
-					     struct extent_io_tree *tree,
+				    struct extent_io_tree *tree,
-					     struct page *locked_page,
+				    struct page *locked_page, u64 *start,
-					     u64 *start, u64 *end,
+				    u64 *end, u64 max_bytes)
-					     u64 max_bytes)
 {
 	u64 delalloc_start;
 	u64 delalloc_end;
@@ -1740,10 +1739,8 @@ u64 count_range_bits(struct extent_io_tree *tree,
 	u64 last = 0;
 	int found = 0;
 
-	if (search_end <= cur_start) {
+	if (WARN_ON(search_end <= cur_start))
-		WARN_ON(1);
 		return 0;
-	}
 
 	spin_lock(&tree->lock);
 	if (cur_start == 0 && bits == EXTENT_DIRTY) {
@@ -3569,9 +3566,8 @@ int btree_write_cache_pages(struct address_space *mapping,
 			 * but no sense in crashing the users box for something
 			 * we can survive anyway.
 			 */
-			if (!eb) {
+			if (WARN_ON(!eb)) {
 				spin_unlock(&mapping->private_lock);
-				WARN_ON(1);
 				continue;
 			}
 
@@ -4038,7 +4034,7 @@ static struct extent_map *get_extent_skip_holes(struct inode *inode,
 	if (offset >= last)
 		return NULL;
 
-	while(1) {
+	while (1) {
 		len = last - offset;
 		if (len == 0)
 			break;
@@ -4062,6 +4058,19 @@ static struct extent_map *get_extent_skip_holes(struct inode *inode,
 	return NULL;
 }
 
+static noinline int count_ext_ref(u64 inum, u64 offset, u64 root_id, void *ctx)
+{
+	unsigned long cnt = *((unsigned long *)ctx);
+
+	cnt++;
+	*((unsigned long *)ctx) = cnt;
+
+	/* Now we're sure that the extent is shared. */
+	if (cnt > 1)
+		return 1;
+	return 0;
+}
+
 int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len, get_extent_t *get_extent)
 {
@@ -4128,7 +4137,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		last = found_key.offset;
 		last_for_get_extent = last + 1;
 	}
-	btrfs_free_path(path);
+	btrfs_release_path(path);
 
 	/*
 	 * we might have some extents allocated but more delalloc past those
@@ -4198,7 +4207,24 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 			flags |= (FIEMAP_EXTENT_DELALLOC |
 				  FIEMAP_EXTENT_UNKNOWN);
 		} else {
+			unsigned long ref_cnt = 0;
+
 			disko = em->block_start + offset_in_extent;
+
+			/*
+			 * As btrfs supports shared space, this information
+			 * can be exported to userspace tools via
+			 * flag FIEMAP_EXTENT_SHARED.
+			 */
+			ret = iterate_inodes_from_logical(
+					em->block_start,
+					BTRFS_I(inode)->root->fs_info,
+					path, count_ext_ref, &ref_cnt);
+			if (ret < 0 && ret != -ENOENT)
+				goto out_free;
+
+			if (ref_cnt > 1)
+				flags |= FIEMAP_EXTENT_SHARED;
 		}
 		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags))
 			flags |= FIEMAP_EXTENT_ENCODED;
@@ -4230,6 +4256,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 out_free:
 	free_extent_map(em);
 out:
+	btrfs_free_path(path);
 	unlock_extent_cached(&BTRFS_I(inode)->io_tree, start, start + len - 1,
 			     &cached_state, GFP_NOFS);
 	return ret;
@@ -4455,6 +4482,23 @@ static void mark_extent_buffer_accessed(struct extent_buffer *eb)
 	}
 }
 
+struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
+					 		u64 start)
+{
+	struct extent_buffer *eb;
+
+	rcu_read_lock();
+	eb = radix_tree_lookup(&tree->buffer, start >> PAGE_CACHE_SHIFT);
+	if (eb && atomic_inc_not_zero(&eb->refs)) {
+		rcu_read_unlock();
+		mark_extent_buffer_accessed(eb);
+		return eb;
+	}
+	rcu_read_unlock();
+
+	return NULL;
+}
+
 struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
 					  u64 start, unsigned long len)
 {
@@ -4468,14 +4512,10 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
 	int uptodate = 1;
 	int ret;
 
-	rcu_read_lock();
+
-	eb = radix_tree_lookup(&tree->buffer, start >> PAGE_CACHE_SHIFT);
+	eb = find_extent_buffer(tree, start);
-	if (eb && atomic_inc_not_zero(&eb->refs)) {
+	if (eb)
-		rcu_read_unlock();
-		mark_extent_buffer_accessed(eb);
 		return eb;
-	}
-	rcu_read_unlock();
 
 	eb = __alloc_extent_buffer(tree, start, len, GFP_NOFS);
 	if (!eb)
@@ -4534,24 +4574,17 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
 
 	spin_lock(&tree->buffer_lock);
 	ret = radix_tree_insert(&tree->buffer, start >> PAGE_CACHE_SHIFT, eb);
+	spin_unlock(&tree->buffer_lock);
+	radix_tree_preload_end();
 	if (ret == -EEXIST) {
-		exists = radix_tree_lookup(&tree->buffer,
+		exists = find_extent_buffer(tree, start);
-						start >> PAGE_CACHE_SHIFT);
+		if (exists)
-		if (!atomic_inc_not_zero(&exists->refs)) {
+			goto free_eb;
-			spin_unlock(&tree->buffer_lock);
+		else
-			radix_tree_preload_end();
-			exists = NULL;
 			goto again;
-		}
-		spin_unlock(&tree->buffer_lock);
-		radix_tree_preload_end();
-		mark_extent_buffer_accessed(exists);
-		goto free_eb;
 	}
 	/* add one reference for the tree */
 	check_buffer_tree_ref(eb);
-	spin_unlock(&tree->buffer_lock);
-	radix_tree_preload_end();
 
 	/*
 	 * there is a race where release page may have
@@ -4582,23 +4615,6 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
 	return exists;
 }
 
-struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
-					 u64 start, unsigned long len)
-{
-	struct extent_buffer *eb;
-
-	rcu_read_lock();
-	eb = radix_tree_lookup(&tree->buffer, start >> PAGE_CACHE_SHIFT);
-	if (eb && atomic_inc_not_zero(&eb->refs)) {
-		rcu_read_unlock();
-		mark_extent_buffer_accessed(eb);
-		return eb;
-	}
-	rcu_read_unlock();
-
-	return NULL;
-}
-
 static inline void btrfs_release_extent_buffer_rcu(struct rcu_head *head)
 {
 	struct extent_buffer *eb =
@@ -5062,23 +5078,6 @@ void copy_extent_buffer(struct extent_buffer *dst, struct extent_buffer *src,
 	}
 }
 
-static void move_pages(struct page *dst_page, struct page *src_page,
-		       unsigned long dst_off, unsigned long src_off,
-		       unsigned long len)
-{
-	char *dst_kaddr = page_address(dst_page);
-	if (dst_page == src_page) {
-		memmove(dst_kaddr + dst_off, dst_kaddr + src_off, len);
-	} else {
-		char *src_kaddr = page_address(src_page);
-		char *p = dst_kaddr + dst_off + len;
-		char *s = src_kaddr + src_off + len;
-
-		while (len--)
-			*--p = *--s;
-	}
-}
-
 static inline bool areas_overlap(unsigned long src, unsigned long dst, unsigned long len)
 {
 	unsigned long distance = (src > dst) ? src - dst : dst - src;
@@ -5189,7 +5188,7 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 
 		cur = min_t(unsigned long, len, src_off_in_page + 1);
 		cur = min(cur, dst_off_in_page + 1);
-		move_pages(extent_buffer_page(dst, dst_i),
+		copy_pages(extent_buffer_page(dst, dst_i),
 			   extent_buffer_page(dst, src_i),
 			   dst_off_in_page - cur + 1,
 			   src_off_in_page - cur + 1, cur);

fs/btrfs/extent_io.h

@@ -271,7 +271,7 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
 struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len);
 struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src);
 struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
-					 u64 start, unsigned long len);
+					 u64 start);
 void free_extent_buffer(struct extent_buffer *eb);
 void free_extent_buffer_stale(struct extent_buffer *eb);
 #define WAIT_NONE	0
@@ -345,4 +345,10 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
 int end_extent_writepage(struct page *page, int err, u64 start, u64 end);
 int repair_eb_io_failure(struct btrfs_root *root, struct extent_buffer *eb,
 			 int mirror_num);
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+noinline u64 find_lock_delalloc_range(struct inode *inode,
+				      struct extent_io_tree *tree,
+				      struct page *locked_page, u64 *start,
+				      u64 *end, u64 max_bytes);
+#endif
 #endif

fs/btrfs/extent_map.h

@@ -3,10 +3,10 @@
 
 #include <linux/rbtree.h>
 
-#define EXTENT_MAP_LAST_BYTE (u64)-4
+#define EXTENT_MAP_LAST_BYTE ((u64)-4)
-#define EXTENT_MAP_HOLE (u64)-3
+#define EXTENT_MAP_HOLE ((u64)-3)
-#define EXTENT_MAP_INLINE (u64)-2
+#define EXTENT_MAP_INLINE ((u64)-2)
-#define EXTENT_MAP_DELALLOC (u64)-1
+#define EXTENT_MAP_DELALLOC ((u64)-1)
 
 /* bits for the flags field */
 #define EXTENT_FLAG_PINNED 0 /* this entry not yet on disk, don't free it */

fs/btrfs/file-item.c

@@ -329,6 +329,9 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 	u64 csum_end;
 	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
 
+	ASSERT(start == ALIGN(start, root->sectorsize) &&
+	       (end + 1) == ALIGN(end + 1, root->sectorsize));
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -846,10 +849,8 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
 	path->leave_spinning = 0;
 	if (ret < 0)
 		goto fail_unlock;
-	if (ret != 0) {
+	if (WARN_ON(ret != 0))
-		WARN_ON(1);
 		goto fail_unlock;
-	}
 	leaf = path->nodes[0];
 csum:
 	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item);

fs/btrfs/file.c

@@ -39,7 +39,6 @@
 #include "print-tree.h"
 #include "tree-log.h"
 #include "locking.h"
-#include "compat.h"
 #include "volumes.h"
 
 static struct kmem_cache *btrfs_inode_defrag_cachep;
@@ -370,7 +369,7 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
 	u64 root_objectid = 0;
 
 	atomic_inc(&fs_info->defrag_running);
-	while(1) {
+	while (1) {
 		/* Pause the auto defragger. */
 		if (test_bit(BTRFS_FS_STATE_REMOUNTING,
 			     &fs_info->fs_state))
@@ -1281,6 +1280,7 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file,
 		}
 		wait_on_page_writeback(pages[i]);
 	}
+	faili = num_pages - 1;
 	err = 0;
 	if (start_pos < inode->i_size) {
 		struct btrfs_ordered_extent *ordered;
@@ -1299,8 +1299,10 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file,
 				unlock_page(pages[i]);
 				page_cache_release(pages[i]);
 			}
-			btrfs_wait_ordered_range(inode, start_pos,
+			err = btrfs_wait_ordered_range(inode, start_pos,
-						 last_pos - start_pos);
+						       last_pos - start_pos);
+			if (err)
+				goto fail;
 			goto again;
 		}
 		if (ordered)
@@ -1809,8 +1811,13 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	atomic_inc(&root->log_batch);
 	full_sync = test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
 			     &BTRFS_I(inode)->runtime_flags);
-	if (full_sync)
+	if (full_sync) {
-		btrfs_wait_ordered_range(inode, start, end - start + 1);
+		ret = btrfs_wait_ordered_range(inode, start, end - start + 1);
+		if (ret) {
+			mutex_unlock(&inode->i_mutex);
+			goto out;
+		}
+	}
 	atomic_inc(&root->log_batch);
 
 	/*
@@ -1876,27 +1883,20 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	mutex_unlock(&inode->i_mutex);
 
 	if (ret != BTRFS_NO_LOG_SYNC) {
-		if (ret > 0) {
+		if (!ret) {
-			/*
-			 * If we didn't already wait for ordered extents we need
-			 * to do that now.
-			 */
-			if (!full_sync)
-				btrfs_wait_ordered_range(inode, start,
-							 end - start + 1);
-			ret = btrfs_commit_transaction(trans, root);
-		} else {
 			ret = btrfs_sync_log(trans, root);
-			if (ret == 0) {
+			if (!ret) {
 				ret = btrfs_end_transaction(trans, root);
-			} else {
+				goto out;
-				if (!full_sync)
-					btrfs_wait_ordered_range(inode, start,
-								 end -
-								 start + 1);
-				ret = btrfs_commit_transaction(trans, root);
 			}
 		}
+		if (!full_sync) {
+			ret = btrfs_wait_ordered_range(inode, start,
+						       end - start + 1);
+			if (ret)
+				goto out;
+		}
+		ret = btrfs_commit_transaction(trans, root);
 	} else {
 		ret = btrfs_end_transaction(trans, root);
 	}
@@ -2067,7 +2067,9 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	bool same_page = ((offset >> PAGE_CACHE_SHIFT) ==
 			  ((offset + len - 1) >> PAGE_CACHE_SHIFT));
 
-	btrfs_wait_ordered_range(inode, offset, len);
+	ret = btrfs_wait_ordered_range(inode, offset, len);
+	if (ret)
+		return ret;
 
 	mutex_lock(&inode->i_mutex);
 	/*
@@ -2136,8 +2138,12 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 			btrfs_put_ordered_extent(ordered);
 		unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart,
 				     lockend, &cached_state, GFP_NOFS);
-		btrfs_wait_ordered_range(inode, lockstart,
+		ret = btrfs_wait_ordered_range(inode, lockstart,
-					 lockend - lockstart + 1);
+					       lockend - lockstart + 1);
+		if (ret) {
+			mutex_unlock(&inode->i_mutex);
+			return ret;
+		}
 	}
 
 	path = btrfs_alloc_path();
@@ -2308,7 +2314,10 @@ static long btrfs_fallocate(struct file *file, int mode,
 	 * wait for ordered IO before we have any locks.  We'll loop again
 	 * below with the locks held.
 	 */
-	btrfs_wait_ordered_range(inode, alloc_start, alloc_end - alloc_start);
+	ret = btrfs_wait_ordered_range(inode, alloc_start,
+				       alloc_end - alloc_start);
+	if (ret)
+		goto out;
 
 	locked_end = alloc_end - 1;
 	while (1) {
@@ -2332,8 +2341,10 @@ static long btrfs_fallocate(struct file *file, int mode,
 			 * we can't wait on the range with the transaction
 			 * running or with the extent lock held
 			 */
-			btrfs_wait_ordered_range(inode, alloc_start,
+			ret = btrfs_wait_ordered_range(inode, alloc_start,
-						 alloc_end - alloc_start);
+						       alloc_end - alloc_start);
+			if (ret)
+				goto out;
 		} else {
 			if (ordered)
 				btrfs_put_ordered_extent(ordered);
@@ -2405,14 +2416,12 @@ static long btrfs_fallocate(struct file *file, int mode,
 static int find_desired_extent(struct inode *inode, loff_t *offset, int whence)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct extent_map *em;
+	struct extent_map *em = NULL;
 	struct extent_state *cached_state = NULL;
 	u64 lockstart = *offset;
 	u64 lockend = i_size_read(inode);
 	u64 start = *offset;
-	u64 orig_start = *offset;
 	u64 len = i_size_read(inode);
-	u64 last_end = 0;
 	int ret = 0;
 
 	lockend = max_t(u64, root->sectorsize, lockend);
@@ -2429,89 +2438,35 @@ static int find_desired_extent(struct inode *inode, loff_t *offset, int whence)
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend, 0,
 			 &cached_state);
 
-	/*
+	while (start < inode->i_size) {
-	 * Delalloc is such a pain.  If we have a hole and we have pending
-	 * delalloc for a portion of the hole we will get back a hole that
-	 * exists for the entire range since it hasn't been actually written
-	 * yet.  So to take care of this case we need to look for an extent just
-	 * before the position we want in case there is outstanding delalloc
-	 * going on here.
-	 */
-	if (whence == SEEK_HOLE && start != 0) {
-		if (start <= root->sectorsize)
-			em = btrfs_get_extent_fiemap(inode, NULL, 0, 0,
-						     root->sectorsize, 0);
-		else
-			em = btrfs_get_extent_fiemap(inode, NULL, 0,
-						     start - root->sectorsize,
-						     root->sectorsize, 0);
-		if (IS_ERR(em)) {
-			ret = PTR_ERR(em);
-			goto out;
-		}
-		last_end = em->start + em->len;
-		if (em->block_start == EXTENT_MAP_DELALLOC)
-			last_end = min_t(u64, last_end, inode->i_size);
-		free_extent_map(em);
-	}
-
-	while (1) {
 		em = btrfs_get_extent_fiemap(inode, NULL, 0, start, len, 0);
 		if (IS_ERR(em)) {
 			ret = PTR_ERR(em);
+			em = NULL;
 			break;
 		}
 
-		if (em->block_start == EXTENT_MAP_HOLE) {
+		if (whence == SEEK_HOLE &&
-			if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
+		    (em->block_start == EXTENT_MAP_HOLE ||
-				if (last_end <= orig_start) {
+		     test_bit(EXTENT_FLAG_PREALLOC, &em->flags)))
-					free_extent_map(em);
+			break;
-					ret = -ENXIO;
+		else if (whence == SEEK_DATA &&
-					break;
+			   (em->block_start != EXTENT_MAP_HOLE &&
-				}
+			    !test_bit(EXTENT_FLAG_PREALLOC, &em->flags)))
-			}
+			break;
-
-			if (whence == SEEK_HOLE) {
-				*offset = start;
-				free_extent_map(em);
-				break;
-			}
-		} else {
-			if (whence == SEEK_DATA) {
-				if (em->block_start == EXTENT_MAP_DELALLOC) {
-					if (start >= inode->i_size) {
-						free_extent_map(em);
-						ret = -ENXIO;
-						break;
-					}
-				}
-
-				if (!test_bit(EXTENT_FLAG_PREALLOC,
-					      &em->flags)) {
-					*offset = start;
-					free_extent_map(em);
-					break;
-				}
-			}
-		}
 
 		start = em->start + em->len;
-		last_end = em->start + em->len;
-
-		if (em->block_start == EXTENT_MAP_DELALLOC)
-			last_end = min_t(u64, last_end, inode->i_size);
-
-		if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
-			free_extent_map(em);
-			ret = -ENXIO;
-			break;
-		}
 		free_extent_map(em);
+		em = NULL;
 		cond_resched();
 	}
-	if (!ret)
+	free_extent_map(em);
-		*offset = min(*offset, inode->i_size);
+	if (!ret) {
-out:
+		if (whence == SEEK_DATA && start >= inode->i_size)
+			ret = -ENXIO;
+		else
+			*offset = min_t(loff_t, start, inode->i_size);
+	}
 	unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend,
 			     &cached_state, GFP_NOFS);
 	return ret;

fs/btrfs/free-space-cache.c

@@ -218,7 +218,6 @@ int btrfs_check_trunc_cache_free_space(struct btrfs_root *root,
 
 int btrfs_truncate_free_space_cache(struct btrfs_root *root,
 				    struct btrfs_trans_handle *trans,
-				    struct btrfs_path *path,
 				    struct inode *inode)
 {
 	int ret = 0;
@@ -1009,8 +1008,13 @@ static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 	if (ret)
 		goto out;
 
-
+	ret = btrfs_wait_ordered_range(inode, 0, (u64)-1);
-	btrfs_wait_ordered_range(inode, 0, (u64)-1);
+	if (ret) {
+		clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, inode->i_size - 1,
+				 EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, NULL,
+				 GFP_NOFS);
+		goto out;
+	}
 
 	key.objectid = BTRFS_FREE_SPACE_OBJECTID;
 	key.offset = offset;
@@ -2276,7 +2280,7 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group,
 		goto out;
 
 	entry = rb_entry(node, struct btrfs_free_space, offset_index);
-	while(1) {
+	while (1) {
 		if (entry->bytes < bytes && entry->bytes > *max_extent_size)
 			*max_extent_size = entry->bytes;
 
@@ -2967,19 +2971,15 @@ int load_free_ino_cache(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
 
 int btrfs_write_out_ino_cache(struct btrfs_root *root,
 			      struct btrfs_trans_handle *trans,
-			      struct btrfs_path *path)
+			      struct btrfs_path *path,
+			      struct inode *inode)
 {
 	struct btrfs_free_space_ctl *ctl = root->free_ino_ctl;
-	struct inode *inode;
 	int ret;
 
 	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
 		return 0;
 
-	inode = lookup_free_ino_inode(root, path);
-	if (IS_ERR(inode))
-		return 0;
-
 	ret = __btrfs_write_out_cache(root, inode, ctl, NULL, trans, path, 0);
 	if (ret) {
 		btrfs_delalloc_release_metadata(inode, inode->i_size);
@@ -2990,7 +2990,6 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 #endif
 	}
 
-	iput(inode);
 	return ret;
 }
 

fs/btrfs/free-space-cache.h

@@ -58,7 +58,6 @@ int btrfs_check_trunc_cache_free_space(struct btrfs_root *root,
 				       struct btrfs_block_rsv *rsv);
 int btrfs_truncate_free_space_cache(struct btrfs_root *root,
 				    struct btrfs_trans_handle *trans,
-				    struct btrfs_path *path,
 				    struct inode *inode);
 int load_free_space_cache(struct btrfs_fs_info *fs_info,
 			  struct btrfs_block_group_cache *block_group);
@@ -76,7 +75,8 @@ int load_free_ino_cache(struct btrfs_fs_info *fs_info,
 			struct btrfs_root *root);
 int btrfs_write_out_ino_cache(struct btrfs_root *root,
 			      struct btrfs_trans_handle *trans,
-			      struct btrfs_path *path);
+			      struct btrfs_path *path,
+			      struct inode *inode);
 
 void btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group);
 int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl,

fs/btrfs/inode-item.c

@@ -369,7 +369,7 @@ static int btrfs_insert_inode_extref(struct btrfs_trans_handle *trans,
 		goto out;
 
 	leaf = path->nodes[0];
-	item = btrfs_item_nr(leaf, path->slots[0]);
+	item = btrfs_item_nr(path->slots[0]);
 	ptr = (unsigned long)btrfs_item_ptr(leaf, path->slots[0], char);
 	ptr += btrfs_item_size(leaf, item) - ins_len;
 	extref = (struct btrfs_inode_extref *)ptr;

fs/btrfs/inode-map.c

@@ -78,10 +78,8 @@ static int caching_kthread(void *data)
 			    btrfs_transaction_in_commit(fs_info)) {
 				leaf = path->nodes[0];
 
-				if (btrfs_header_nritems(leaf) == 0) {
+				if (WARN_ON(btrfs_header_nritems(leaf) == 0))
-					WARN_ON(1);
 					break;
-				}
 
 				/*
 				 * Save the key so we can advances forward
@@ -237,7 +235,7 @@ void btrfs_return_ino(struct btrfs_root *root, u64 objectid)
 		start_caching(root);
 
 		if (objectid <= root->cache_progress ||
-		    objectid > root->highest_objectid)
+		    objectid >= root->highest_objectid)
 			__btrfs_add_free_space(ctl, objectid, 1);
 		else
 			__btrfs_add_free_space(pinned, objectid, 1);
@@ -412,8 +410,7 @@ int btrfs_save_ino_cache(struct btrfs_root *root,
 		return 0;
 
 	/* Don't save inode cache if we are deleting this root */
-	if (btrfs_root_refs(&root->root_item) == 0 &&
+	if (btrfs_root_refs(&root->root_item) == 0)
-	    root != root->fs_info->tree_root)
 		return 0;
 
 	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
@@ -467,7 +464,7 @@ int btrfs_save_ino_cache(struct btrfs_root *root,
 	}
 
 	if (i_size_read(inode) > 0) {
-		ret = btrfs_truncate_free_space_cache(root, trans, path, inode);
+		ret = btrfs_truncate_free_space_cache(root, trans, inode);
 		if (ret) {
 			if (ret != -ENOSPC)
 				btrfs_abort_transaction(trans, root, ret);
@@ -504,7 +501,7 @@ int btrfs_save_ino_cache(struct btrfs_root *root,
 	}
 	btrfs_free_reserved_data_space(inode, prealloc);
 
-	ret = btrfs_write_out_ino_cache(root, trans, path);
+	ret = btrfs_write_out_ino_cache(root, trans, path, inode);
 out_put:
 	iput(inode);
 out_release:

fs/btrfs/inode.c

@@ -43,7 +43,6 @@
 #include <linux/btrfs.h>
 #include <linux/blkdev.h>
 #include <linux/posix_acl_xattr.h>
-#include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -844,7 +843,10 @@ static noinline int cow_file_range(struct inode *inode,
 	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
 	int ret = 0;
 
-	BUG_ON(btrfs_is_free_space_inode(inode));
+	if (btrfs_is_free_space_inode(inode)) {
+		WARN_ON_ONCE(1);
+		return -EINVAL;
+	}
 
 	num_bytes = ALIGN(end - start + 1, blocksize);
 	num_bytes = max(blocksize,  num_bytes);
@@ -1178,10 +1180,8 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 	while (1) {
 		ret = btrfs_lookup_file_extent(trans, root, path, ino,
 					       cur_offset, 0);
-		if (ret < 0) {
+		if (ret < 0)
-			btrfs_abort_transaction(trans, root, ret);
 			goto error;
-		}
 		if (ret > 0 && path->slots[0] > 0 && check_prev) {
 			leaf = path->nodes[0];
 			btrfs_item_key_to_cpu(leaf, &found_key,
@@ -1195,10 +1195,8 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 		leaf = path->nodes[0];
 		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
 			ret = btrfs_next_leaf(root, path);
-			if (ret < 0) {
+			if (ret < 0)
-				btrfs_abort_transaction(trans, root, ret);
 				goto error;
-			}
 			if (ret > 0)
 				break;
 			leaf = path->nodes[0];
@@ -1289,10 +1287,8 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 			ret = cow_file_range(inode, locked_page,
 					     cow_start, found_key.offset - 1,
 					     page_started, nr_written, 1);
-			if (ret) {
+			if (ret)
-				btrfs_abort_transaction(trans, root, ret);
 				goto error;
-			}
 			cow_start = (u64)-1;
 		}
 
@@ -1339,10 +1335,8 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 		    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 			ret = btrfs_reloc_clone_csums(inode, cur_offset,
 						      num_bytes);
-			if (ret) {
+			if (ret)
-				btrfs_abort_transaction(trans, root, ret);
 				goto error;
-			}
 		}
 
 		extent_clear_unlock_delalloc(inode, cur_offset,
@@ -1364,10 +1358,8 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 	if (cow_start != (u64)-1) {
 		ret = cow_file_range(inode, locked_page, cow_start, end,
 				     page_started, nr_written, 1);
-		if (ret) {
+		if (ret)
-			btrfs_abort_transaction(trans, root, ret);
 			goto error;
-		}
 	}
 
 error:
@@ -1551,7 +1543,13 @@ static void btrfs_clear_bit_hook(struct inode *inode,
 			spin_unlock(&BTRFS_I(inode)->lock);
 		}
 
-		if (*bits & EXTENT_DO_ACCOUNTING)
+		/*
+		 * We don't reserve metadata space for space cache inodes so we
+		 * don't need to call dellalloc_release_metadata if there is an
+		 * error.
+		 */
+		if (*bits & EXTENT_DO_ACCOUNTING &&
+		    root != root->fs_info->tree_root)
 			btrfs_delalloc_release_metadata(inode, len);
 
 		if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID
@@ -2041,10 +2039,8 @@ static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
 		key.offset = offset;
 
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
-	if (ret < 0) {
+	if (WARN_ON(ret < 0))
-		WARN_ON(1);
 		return ret;
-	}
 	ret = 0;
 
 	while (1) {
@@ -2367,10 +2363,23 @@ static noinline int relink_extent_backref(struct btrfs_path *path,
 	return ret;
 }
 
+static void free_sa_defrag_extent(struct new_sa_defrag_extent *new)
+{
+	struct old_sa_defrag_extent *old, *tmp;
+
+	if (!new)
+		return;
+
+	list_for_each_entry_safe(old, tmp, &new->head, list) {
+		list_del(&old->list);
+		kfree(old);
+	}
+	kfree(new);
+}
+
 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;
@@ -2413,16 +2422,11 @@ static void relink_file_extents(struct new_sa_defrag_extent *new)
 	kfree(prev);
 
 	btrfs_free_path(path);
-
-	list_for_each_entry_safe(old, tmp, &new->head, list) {
-		list_del(&old->list);
-		kfree(old);
-	}
 out:
+	free_sa_defrag_extent(new);
+
 	atomic_dec(&root->fs_info->defrag_running);
 	wake_up(&root->fs_info->transaction_wait);
-
-	kfree(new);
 }
 
 static struct new_sa_defrag_extent *
@@ -2432,7 +2436,7 @@ record_old_file_extents(struct inode *inode,
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_path *path;
 	struct btrfs_key key;
-	struct old_sa_defrag_extent *old, *tmp;
+	struct old_sa_defrag_extent *old;
 	struct new_sa_defrag_extent *new;
 	int ret;
 
@@ -2480,7 +2484,7 @@ record_old_file_extents(struct inode *inode,
 		if (slot >= btrfs_header_nritems(l)) {
 			ret = btrfs_next_leaf(root, path);
 			if (ret < 0)
-				goto out_free_list;
+				goto out_free_path;
 			else if (ret > 0)
 				break;
 			continue;
@@ -2509,7 +2513,7 @@ record_old_file_extents(struct inode *inode,
 
 		old = kmalloc(sizeof(*old), GFP_NOFS);
 		if (!old)
-			goto out_free_list;
+			goto out_free_path;
 
 		offset = max(new->file_pos, key.offset);
 		end = min(new->file_pos + new->len, key.offset + num_bytes);
@@ -2531,15 +2535,10 @@ record_old_file_extents(struct inode *inode,
 
 	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);
+	free_sa_defrag_extent(new);
 	return NULL;
 }
 
@@ -2710,8 +2709,14 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	btrfs_remove_ordered_extent(inode, ordered_extent);
 
 	/* for snapshot-aware defrag */
-	if (new)
+	if (new) {
-		relink_file_extents(new);
+		if (ret) {
+			free_sa_defrag_extent(new);
+			atomic_dec(&root->fs_info->defrag_running);
+		} else {
+			relink_file_extents(new);
+		}
+	}
 
 	/* once for us */
 	btrfs_put_ordered_extent(ordered_extent);
@@ -2969,6 +2974,7 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
 	if (insert >= 1) {
 		ret = btrfs_insert_orphan_item(trans, root, btrfs_ino(inode));
 		if (ret) {
+			atomic_dec(&root->orphan_inodes);
 			if (reserve) {
 				clear_bit(BTRFS_INODE_ORPHAN_META_RESERVED,
 					  &BTRFS_I(inode)->runtime_flags);
@@ -3018,14 +3024,16 @@ static int btrfs_orphan_del(struct btrfs_trans_handle *trans,
 		release_rsv = 1;
 	spin_unlock(&root->orphan_lock);
 
-	if (trans && delete_item)
+	if (delete_item) {
-		ret = btrfs_del_orphan_item(trans, root, btrfs_ino(inode));
-
-	if (release_rsv) {
-		btrfs_orphan_release_metadata(inode);
 		atomic_dec(&root->orphan_inodes);
+		if (trans)
+			ret = btrfs_del_orphan_item(trans, root,
+						    btrfs_ino(inode));
 	}
 
+	if (release_rsv)
+		btrfs_orphan_release_metadata(inode);
+
 	return ret;
 }
 
@@ -3172,8 +3180,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 
 		/* if we have links, this was a truncate, lets do that */
 		if (inode->i_nlink) {
-			if (!S_ISREG(inode->i_mode)) {
+			if (WARN_ON(!S_ISREG(inode->i_mode))) {
-				WARN_ON(1);
 				iput(inode);
 				continue;
 			}
@@ -3636,7 +3643,7 @@ int btrfs_unlink_inode(struct btrfs_trans_handle *trans,
 	int ret;
 	ret = __btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
 	if (!ret) {
-		btrfs_drop_nlink(inode);
+		drop_nlink(inode);
 		ret = btrfs_update_inode(trans, root, inode);
 	}
 	return ret;
@@ -4230,15 +4237,16 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 
 	while (1) {
 		struct btrfs_ordered_extent *ordered;
-		btrfs_wait_ordered_range(inode, hole_start,
+
-					 block_end - hole_start);
 		lock_extent_bits(io_tree, hole_start, block_end - 1, 0,
 				 &cached_state);
-		ordered = btrfs_lookup_ordered_extent(inode, hole_start);
+		ordered = btrfs_lookup_ordered_range(inode, hole_start,
+						     block_end - hole_start);
 		if (!ordered)
 			break;
 		unlock_extent_cached(io_tree, hole_start, block_end - 1,
 				     &cached_state, GFP_NOFS);
+		btrfs_start_ordered_extent(inode, ordered, 1);
 		btrfs_put_ordered_extent(ordered);
 	}
 
@@ -4472,8 +4480,10 @@ void btrfs_evict_inode(struct inode *inode)
 	trace_btrfs_inode_evict(inode);
 
 	truncate_inode_pages(&inode->i_data, 0);
-	if (inode->i_nlink && (btrfs_root_refs(&root->root_item) != 0 ||
+	if (inode->i_nlink &&
-			       btrfs_is_free_space_inode(inode)))
+	    ((btrfs_root_refs(&root->root_item) != 0 &&
+	      root->root_key.objectid != BTRFS_ROOT_TREE_OBJECTID) ||
+	     btrfs_is_free_space_inode(inode)))
 		goto no_delete;
 
 	if (is_bad_inode(inode)) {
@@ -4490,7 +4500,8 @@ void btrfs_evict_inode(struct inode *inode)
 	}
 
 	if (inode->i_nlink > 0) {
-		BUG_ON(btrfs_root_refs(&root->root_item) != 0);
+		BUG_ON(btrfs_root_refs(&root->root_item) != 0 &&
+		       root->root_key.objectid != BTRFS_ROOT_TREE_OBJECTID);
 		goto no_delete;
 	}
 
@@ -4731,14 +4742,7 @@ static void inode_tree_del(struct inode *inode)
 	}
 	spin_unlock(&root->inode_lock);
 
-	/*
+	if (empty && btrfs_root_refs(&root->root_item) == 0) {
-	 * Free space cache has inodes in the tree root, but the tree root has a
-	 * root_refs of 0, so this could end up dropping the tree root as a
-	 * snapshot, so we need the extra !root->fs_info->tree_root check to
-	 * make sure we don't drop it.
-	 */
-	if (empty && btrfs_root_refs(&root->root_item) == 0 &&
-	    root != root->fs_info->tree_root) {
 		synchronize_srcu(&root->fs_info->subvol_srcu);
 		spin_lock(&root->inode_lock);
 		empty = RB_EMPTY_ROOT(&root->inode_tree);
@@ -4831,10 +4835,12 @@ static struct inode *btrfs_iget_locked(struct super_block *s,
 {
 	struct inode *inode;
 	struct btrfs_iget_args args;
+	unsigned long hashval = btrfs_inode_hash(objectid, root);
+
 	args.ino = objectid;
 	args.root = root;
 
-	inode = iget5_locked(s, objectid, btrfs_find_actor,
+	inode = iget5_locked(s, hashval, btrfs_find_actor,
 			     btrfs_init_locked_inode,
 			     (void *)&args);
 	return inode;
@@ -5048,7 +5054,7 @@ static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
 			continue;
 		}
 
-		item = btrfs_item_nr(leaf, slot);
+		item = btrfs_item_nr(slot);
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
 
 		if (found_key.objectid != key.objectid)
@@ -5454,7 +5460,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 				BTRFS_INODE_NODATASUM;
 	}
 
-	insert_inode_hash(inode);
+	btrfs_insert_inode_hash(inode);
 	inode_tree_add(inode);
 
 	trace_btrfs_inode_new(inode);
@@ -5730,7 +5736,7 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 		goto fail;
 	}
 
-	btrfs_inc_nlink(inode);
+	inc_nlink(inode);
 	inode_inc_iversion(inode);
 	inode->i_ctime = CURRENT_TIME;
 	ihold(inode);
@@ -5860,7 +5866,7 @@ static noinline int uncompress_inline(struct btrfs_path *path,
 	compress_type = btrfs_file_extent_compression(leaf, item);
 	max_size = btrfs_file_extent_ram_bytes(leaf, item);
 	inline_size = btrfs_file_extent_inline_item_len(leaf,
-					btrfs_item_nr(leaf, path->slots[0]));
+					btrfs_item_nr(path->slots[0]));
 	tmp = kmalloc(inline_size, GFP_NOFS);
 	if (!tmp)
 		return -ENOMEM;
@@ -5974,7 +5980,14 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page,
 	found_type = btrfs_key_type(&found_key);
 	if (found_key.objectid != objectid ||
 	    found_type != BTRFS_EXTENT_DATA_KEY) {
-		goto not_found;
+		/*
+		 * If we backup past the first extent we want to move forward
+		 * and see if there is an extent in front of us, otherwise we'll
+		 * say there is a hole for our whole search range which can
+		 * cause problems.
+		 */
+		extent_end = start;
+		goto next;
 	}
 
 	found_type = btrfs_file_extent_type(leaf, item);
@@ -5989,7 +6002,7 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page,
 		size = btrfs_file_extent_inline_len(leaf, item);
 		extent_end = ALIGN(extent_start + size, root->sectorsize);
 	}
-
+next:
 	if (start >= extent_end) {
 		path->slots[0]++;
 		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
@@ -6249,7 +6262,7 @@ struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *pag
 	/* adjust the range_start to make sure it doesn't
 	 * go backwards from the start they passed in
 	 */
-	range_start = max(start,range_start);
+	range_start = max(start, range_start);
 	found = found_end - range_start;
 
 	if (found > 0) {
@@ -7053,7 +7066,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 			}
 		} else {
 			submit_len += bvec->bv_len;
-			nr_pages ++;
+			nr_pages++;
 			bvec++;
 		}
 	}
@@ -7222,7 +7235,9 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 	 * outstanding dirty pages are on disk.
 	 */
 	count = iov_length(iov, nr_segs);
-	btrfs_wait_ordered_range(inode, offset, count);
+	ret = btrfs_wait_ordered_range(inode, offset, count);
+	if (ret)
+		return ret;
 
 	if (rw & WRITE) {
 		/*
@@ -7563,7 +7578,10 @@ static int btrfs_truncate(struct inode *inode)
 	u64 mask = root->sectorsize - 1;
 	u64 min_size = btrfs_calc_trunc_metadata_size(root, 1);
 
-	btrfs_wait_ordered_range(inode, inode->i_size & (~mask), (u64)-1);
+	ret = btrfs_wait_ordered_range(inode, inode->i_size & (~mask),
+				       (u64)-1);
+	if (ret)
+		return ret;
 
 	/*
 	 * Yes ladies and gentelment, this is indeed ugly.  The fact is we have
@@ -7787,6 +7805,14 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	return inode;
 }
 
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+void btrfs_test_destroy_inode(struct inode *inode)
+{
+	btrfs_drop_extent_cache(inode, 0, (u64)-1, 0);
+	kmem_cache_free(btrfs_inode_cachep, BTRFS_I(inode));
+}
+#endif
+
 static void btrfs_i_callback(struct rcu_head *head)
 {
 	struct inode *inode = container_of(head, struct inode, i_rcu);
@@ -7857,8 +7883,7 @@ int btrfs_drop_inode(struct inode *inode)
 		return 1;
 
 	/* the snap/subvol tree is on deleting */
-	if (btrfs_root_refs(&root->root_item) == 0 &&
+	if (btrfs_root_refs(&root->root_item) == 0)
-	    root != root->fs_info->tree_root)
 		return 1;
 	else
 		return generic_drop_inode(inode);
@@ -7995,8 +8020,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		if (ret == -EEXIST) {
 			/* we shouldn't get
 			 * eexist without a new_inode */
-			if (!new_inode) {
+			if (WARN_ON(!new_inode)) {
-				WARN_ON(1);
 				return ret;
 			}
 		} else {
@@ -8144,18 +8168,24 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 static void btrfs_run_delalloc_work(struct btrfs_work *work)
 {
 	struct btrfs_delalloc_work *delalloc_work;
+	struct inode *inode;
 
 	delalloc_work = container_of(work, struct btrfs_delalloc_work,
 				     work);
-	if (delalloc_work->wait)
+	inode = delalloc_work->inode;
-		btrfs_wait_ordered_range(delalloc_work->inode, 0, (u64)-1);
+	if (delalloc_work->wait) {
-	else
+		btrfs_wait_ordered_range(inode, 0, (u64)-1);
-		filemap_flush(delalloc_work->inode->i_mapping);
+	} else {
+		filemap_flush(inode->i_mapping);
+		if (test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
+			     &BTRFS_I(inode)->runtime_flags))
+			filemap_flush(inode->i_mapping);
+	}
 
 	if (delalloc_work->delay_iput)
-		btrfs_add_delayed_iput(delalloc_work->inode);
+		btrfs_add_delayed_iput(inode);
 	else
-		iput(delalloc_work->inode);
+		iput(inode);
 	complete(&delalloc_work->completion);
 }
 
@@ -8276,8 +8306,7 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput)
 	return ret;
 }
 
-int btrfs_start_all_delalloc_inodes(struct btrfs_fs_info *fs_info,
+int btrfs_start_delalloc_roots(struct btrfs_fs_info *fs_info, int delay_iput)
-				    int delay_iput)
 {
 	struct btrfs_root *root;
 	struct list_head splice;
@@ -8337,14 +8366,14 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 	int err;
 	int drop_inode = 0;
 	u64 objectid;
-	u64 index = 0 ;
+	u64 index = 0;
 	int name_len;
 	int datasize;
 	unsigned long ptr;
 	struct btrfs_file_extent_item *ei;
 	struct extent_buffer *leaf;
 
-	name_len = strlen(symname) + 1;
+	name_len = strlen(symname);
 	if (name_len > BTRFS_MAX_INLINE_DATA_SIZE(root))
 		return -ENAMETOOLONG;
 
@@ -8432,7 +8461,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 	inode->i_mapping->a_ops = &btrfs_symlink_aops;
 	inode->i_mapping->backing_dev_info = &root->fs_info->bdi;
 	inode_set_bytes(inode, name_len);
-	btrfs_i_size_write(inode, name_len - 1);
+	btrfs_i_size_write(inode, name_len);
 	err = btrfs_update_inode(trans, root, inode);
 	if (err)
 		drop_inode = 1;
@@ -8491,6 +8520,8 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 						  ins.offset, 0, 0, 0,
 						  BTRFS_FILE_EXTENT_PREALLOC);
 		if (ret) {
+			btrfs_free_reserved_extent(root, ins.objectid,
+						   ins.offset);
 			btrfs_abort_transaction(trans, root, ret);
 			if (own_trans)
 				btrfs_end_transaction(trans, root);

fs/btrfs/ioctl.c

@@ -44,7 +44,6 @@
 #include <linux/uuid.h>
 #include <linux/btrfs.h>
 #include <linux/uaccess.h>
-#include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -369,9 +368,8 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
 
 int btrfs_is_empty_uuid(u8 *uuid)
 {
-	static char empty_uuid[BTRFS_UUID_SIZE] = {0};
+	BUILD_BUG_ON(BTRFS_UUID_SIZE > PAGE_SIZE);
-
+	return !memcmp(uuid, empty_zero_page, BTRFS_UUID_SIZE);
-	return !memcmp(uuid, empty_uuid, BTRFS_UUID_SIZE);
 }
 
 static noinline int create_subvol(struct inode *dir,
@@ -436,7 +434,7 @@ static noinline int create_subvol(struct inode *dir,
 	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
 	btrfs_set_header_owner(leaf, objectid);
 
-	write_extent_buffer(leaf, root->fs_info->fsid, btrfs_header_fsid(leaf),
+	write_extent_buffer(leaf, root->fs_info->fsid, btrfs_header_fsid(),
 			    BTRFS_FSID_SIZE);
 	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
 			    btrfs_header_chunk_tree_uuid(leaf),
@@ -574,7 +572,7 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	if (ret)
 		return ret;
 
-	btrfs_wait_ordered_extents(root);
+	btrfs_wait_ordered_extents(root, -1);
 
 	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
 	if (!pending_snapshot)
@@ -688,7 +686,7 @@ static inline int btrfs_check_sticky(struct inode *dir, struct inode *inode)
  *     nfs_async_unlink().
  */
 
-static int btrfs_may_delete(struct inode *dir,struct dentry *victim,int isdir)
+static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir)
 {
 	int error;
 
@@ -842,7 +840,6 @@ static int find_new_extents(struct btrfs_root *root,
 {
 	struct btrfs_path *path;
 	struct btrfs_key min_key;
-	struct btrfs_key max_key;
 	struct extent_buffer *leaf;
 	struct btrfs_file_extent_item *extent;
 	int type;
@@ -857,15 +854,10 @@ static int find_new_extents(struct btrfs_root *root,
 	min_key.type = BTRFS_EXTENT_DATA_KEY;
 	min_key.offset = *off;
 
-	max_key.objectid = ino;
-	max_key.type = (u8)-1;
-	max_key.offset = (u64)-1;
-
 	path->keep_locks = 1;
 
-	while(1) {
+	while (1) {
-		ret = btrfs_search_forward(root, &min_key, &max_key,
+		ret = btrfs_search_forward(root, &min_key, path, newer_than);
-					   path, newer_than);
 		if (ret != 0)
 			goto none;
 		if (min_key.objectid != ino)
@@ -1206,7 +1198,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 		ra = &file->f_ra;
 	}
 
-	pages = kmalloc(sizeof(struct page *) * max_cluster,
+	pages = kmalloc_array(max_cluster, sizeof(struct page *),
 			GFP_NOFS);
 	if (!pages) {
 		ret = -ENOMEM;
@@ -1893,7 +1885,6 @@ static noinline int search_ioctl(struct inode *inode,
 {
 	struct btrfs_root *root;
 	struct btrfs_key key;
-	struct btrfs_key max_key;
 	struct btrfs_path *path;
 	struct btrfs_ioctl_search_key *sk = &args->key;
 	struct btrfs_fs_info *info = BTRFS_I(inode)->root->fs_info;
@@ -1925,15 +1916,10 @@ static noinline int search_ioctl(struct inode *inode,
 	key.type = sk->min_type;
 	key.offset = sk->min_offset;
 
-	max_key.objectid = sk->max_objectid;
-	max_key.type = sk->max_type;
-	max_key.offset = sk->max_offset;
-
 	path->keep_locks = 1;
 
-	while(1) {
+	while (1) {
-		ret = btrfs_search_forward(root, &key, &max_key, path,
+		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
-					   sk->min_transid);
 		if (ret != 0) {
 			if (ret > 0)
 				ret = 0;
@@ -2018,7 +2004,7 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 	key.type = BTRFS_INODE_REF_KEY;
 	key.offset = (u64)-1;
 
-	while(1) {
+	while (1) {
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 		if (ret < 0)
 			goto out;
@@ -2047,7 +2033,7 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 		}
 
 		*(ptr + len) = '/';
-		read_extent_buffer(l, ptr,(unsigned long)(iref + 1), len);
+		read_extent_buffer(l, ptr, (unsigned long)(iref + 1), len);
 
 		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
 			break;
@@ -2058,7 +2044,7 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 		dirid = key.objectid;
 	}
 	memmove(name, ptr, total_len);
-	name[total_len]='\0';
+	name[total_len] = '\0';
 	ret = 0;
 out:
 	btrfs_free_path(path);
@@ -2144,7 +2130,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 
 	inode = dentry->d_inode;
 	dest = BTRFS_I(inode)->root;
-	if (!capable(CAP_SYS_ADMIN)){
+	if (!capable(CAP_SYS_ADMIN)) {
 		/*
 		 * Regular user.  Only allow this with a special mount
 		 * option, when the user has write+exec access to the
@@ -2727,15 +2713,10 @@ static long btrfs_ioctl_file_extent_same(struct file *file,
 	size = sizeof(tmp) +
 		tmp.dest_count * sizeof(struct btrfs_ioctl_same_extent_info);
 
-	same = kmalloc(size, GFP_NOFS);
+	same = memdup_user((struct btrfs_ioctl_same_args __user *)argp, size);
-	if (!same) {
-		ret = -EFAULT;
-		goto out;
-	}
 
-	if (copy_from_user(same,
+	if (IS_ERR(same)) {
-			   (struct btrfs_ioctl_same_args __user *)argp, size)) {
+		ret = PTR_ERR(same);
-		ret = -EFAULT;
 		goto out;
 	}
 
@@ -3679,9 +3660,10 @@ static long btrfs_ioctl_dev_replace(struct btrfs_root *root, void __user *arg)
 
 	switch (p->cmd) {
 	case BTRFS_IOCTL_DEV_REPLACE_CMD_START:
-		if (root->fs_info->sb->s_flags & MS_RDONLY)
+		if (root->fs_info->sb->s_flags & MS_RDONLY) {
-			return -EROFS;
+			ret = -EROFS;
-
+			goto out;
+		}
 		if (atomic_xchg(
 			&root->fs_info->mutually_exclusive_operation_running,
 			1)) {
@@ -3707,7 +3689,7 @@ static long btrfs_ioctl_dev_replace(struct btrfs_root *root, void __user *arg)
 
 	if (copy_to_user(arg, p, sizeof(*p)))
 		ret = -EFAULT;
-
+out:
 	kfree(p);
 	return ret;
 }
@@ -4557,9 +4539,15 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_logical_to_ino(root, argp);
 	case BTRFS_IOC_SPACE_INFO:
 		return btrfs_ioctl_space_info(root, argp);
-	case BTRFS_IOC_SYNC:
+	case BTRFS_IOC_SYNC: {
-		btrfs_sync_fs(file->f_dentry->d_sb, 1);
+		int ret;
-		return 0;
+
+		ret = btrfs_start_delalloc_roots(root->fs_info, 0);
+		if (ret)
+			return ret;
+		ret = btrfs_sync_fs(file->f_dentry->d_sb, 1);
+		return ret;
+	}
 	case BTRFS_IOC_START_SYNC:
 		return btrfs_ioctl_start_sync(root, argp);
 	case BTRFS_IOC_WAIT_SYNC:

fs/btrfs/ordered-data.c

@@ -537,7 +537,9 @@ void btrfs_remove_ordered_extent(struct inode *inode,
 	 */
 	if (RB_EMPTY_ROOT(&tree->tree) &&
 	    !mapping_tagged(inode->i_mapping, PAGECACHE_TAG_DIRTY)) {
+		spin_lock(&root->fs_info->ordered_root_lock);
 		list_del_init(&BTRFS_I(inode)->ordered_operations);
+		spin_unlock(&root->fs_info->ordered_root_lock);
 	}
 
 	if (!root->nr_ordered_extents) {
@@ -563,10 +565,11 @@ static void btrfs_run_ordered_extent_work(struct btrfs_work *work)
  * wait for all the ordered extents in a root.  This is done when balancing
  * space between drives.
  */
-void btrfs_wait_ordered_extents(struct btrfs_root *root)
+int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr)
 {
 	struct list_head splice, works;
 	struct btrfs_ordered_extent *ordered, *next;
+	int count = 0;
 
 	INIT_LIST_HEAD(&splice);
 	INIT_LIST_HEAD(&works);
@@ -574,7 +577,7 @@ void btrfs_wait_ordered_extents(struct btrfs_root *root)
 	mutex_lock(&root->fs_info->ordered_operations_mutex);
 	spin_lock(&root->ordered_extent_lock);
 	list_splice_init(&root->ordered_extents, &splice);
-	while (!list_empty(&splice)) {
+	while (!list_empty(&splice) && nr) {
 		ordered = list_first_entry(&splice, struct btrfs_ordered_extent,
 					   root_extent_list);
 		list_move_tail(&ordered->root_extent_list,
@@ -589,7 +592,11 @@ void btrfs_wait_ordered_extents(struct btrfs_root *root)
 
 		cond_resched();
 		spin_lock(&root->ordered_extent_lock);
+		if (nr != -1)
+			nr--;
+		count++;
 	}
+	list_splice_tail(&splice, &root->ordered_extents);
 	spin_unlock(&root->ordered_extent_lock);
 
 	list_for_each_entry_safe(ordered, next, &works, work_list) {
@@ -599,18 +606,21 @@ void btrfs_wait_ordered_extents(struct btrfs_root *root)
 		cond_resched();
 	}
 	mutex_unlock(&root->fs_info->ordered_operations_mutex);
+
+	return count;
 }
 
-void btrfs_wait_all_ordered_extents(struct btrfs_fs_info *fs_info)
+void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr)
 {
 	struct btrfs_root *root;
 	struct list_head splice;
+	int done;
 
 	INIT_LIST_HEAD(&splice);
 
 	spin_lock(&fs_info->ordered_root_lock);
 	list_splice_init(&fs_info->ordered_roots, &splice);
-	while (!list_empty(&splice)) {
+	while (!list_empty(&splice) && nr) {
 		root = list_first_entry(&splice, struct btrfs_root,
 					ordered_root);
 		root = btrfs_grab_fs_root(root);
@@ -619,10 +629,14 @@ void btrfs_wait_all_ordered_extents(struct btrfs_fs_info *fs_info)
 			       &fs_info->ordered_roots);
 		spin_unlock(&fs_info->ordered_root_lock);
 
-		btrfs_wait_ordered_extents(root);
+		done = btrfs_wait_ordered_extents(root, nr);
 		btrfs_put_fs_root(root);
 
 		spin_lock(&fs_info->ordered_root_lock);
+		if (nr != -1) {
+			nr -= done;
+			WARN_ON(nr < 0);
+		}
 	}
 	spin_unlock(&fs_info->ordered_root_lock);
 }
@@ -734,8 +748,9 @@ void btrfs_start_ordered_extent(struct inode *inode,
 /*
  * Used to wait on ordered extents across a large range of bytes.
  */
-void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
+int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 {
+	int ret = 0;
 	u64 end;
 	u64 orig_end;
 	struct btrfs_ordered_extent *ordered;
@@ -751,8 +766,9 @@ void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 	/* start IO across the range first to instantiate any delalloc
 	 * extents
 	 */
-	filemap_fdatawrite_range(inode->i_mapping, start, orig_end);
+	ret = filemap_fdatawrite_range(inode->i_mapping, start, orig_end);
-
+	if (ret)
+		return ret;
 	/*
 	 * So with compression we will find and lock a dirty page and clear the
 	 * first one as dirty, setup an async extent, and immediately return
@@ -768,10 +784,15 @@ void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 	 * right and you are wrong.
 	 */
 	if (test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
-		     &BTRFS_I(inode)->runtime_flags))
+		     &BTRFS_I(inode)->runtime_flags)) {
-		filemap_fdatawrite_range(inode->i_mapping, start, orig_end);
+		ret = filemap_fdatawrite_range(inode->i_mapping, start,
-
+					       orig_end);
-	filemap_fdatawait_range(inode->i_mapping, start, orig_end);
+		if (ret)
+			return ret;
+	}
+	ret = filemap_fdatawait_range(inode->i_mapping, start, orig_end);
+	if (ret)
+		return ret;
 
 	end = orig_end;
 	while (1) {
@@ -788,11 +809,14 @@ void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 		}
 		btrfs_start_ordered_extent(inode, ordered, 1);
 		end = ordered->file_offset;
+		if (test_bit(BTRFS_ORDERED_IOERR, &ordered->flags))
+			ret = -EIO;
 		btrfs_put_ordered_extent(ordered);
-		if (end == 0 || end == start)
+		if (ret || end == 0 || end == start)
 			break;
 		end--;
 	}
+	return ret;
 }
 
 /*
@@ -1076,7 +1100,7 @@ void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
 	 * if this file hasn't been changed since the last transaction
 	 * commit, we can safely return without doing anything
 	 */
-	if (last_mod < root->fs_info->last_trans_committed)
+	if (last_mod <= root->fs_info->last_trans_committed)
 		return;
 
 	spin_lock(&root->fs_info->ordered_root_lock);

fs/btrfs/ordered-data.h

@@ -180,7 +180,7 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_extent(struct inode *inode,
 							 u64 file_offset);
 void btrfs_start_ordered_extent(struct inode *inode,
 				struct btrfs_ordered_extent *entry, int wait);
-void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len);
+int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len);
 struct btrfs_ordered_extent *
 btrfs_lookup_first_ordered_extent(struct inode * inode, u64 file_offset);
 struct btrfs_ordered_extent *btrfs_lookup_ordered_range(struct inode *inode,
@@ -195,8 +195,8 @@ int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans,
 void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root,
 				 struct inode *inode);
-void btrfs_wait_ordered_extents(struct btrfs_root *root);
+int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr);
-void btrfs_wait_all_ordered_extents(struct btrfs_fs_info *fs_info);
+void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr);
 void btrfs_get_logged_extents(struct btrfs_root *log, struct inode *inode);
 void btrfs_wait_logged_extents(struct btrfs_root *log, u64 transid);
 void btrfs_free_logged_extents(struct btrfs_root *log, u64 transid);

fs/btrfs/print-tree.c

@@ -193,7 +193,7 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 	btrfs_info(root->fs_info, "leaf %llu total ptrs %d free space %d",
 		   btrfs_header_bytenr(l), nr, btrfs_leaf_free_space(root, l));
 	for (i = 0 ; i < nr ; i++) {
-		item = btrfs_item_nr(l, i);
+		item = btrfs_item_nr(i);
 		btrfs_item_key_to_cpu(l, &key, i);
 		type = btrfs_key_type(&key);
 		printk(KERN_INFO "\titem %d key (%llu %u %llu) itemoff %d "

fs/btrfs/raid56.c

@@ -33,7 +33,6 @@
 #include <linux/raid/xor.h>
 #include <linux/vmalloc.h>
 #include <asm/div64.h>
-#include "compat.h"
 #include "ctree.h"
 #include "extent_map.h"
 #include "disk-io.h"

fs/btrfs/relocation.c

@@ -1383,6 +1383,7 @@ int btrfs_init_reloc_root(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_root *reloc_root;
 	struct reloc_control *rc = root->fs_info->reloc_ctl;
+	struct btrfs_block_rsv *rsv;
 	int clear_rsv = 0;
 	int ret;
 
@@ -1396,13 +1397,14 @@ int btrfs_init_reloc_root(struct btrfs_trans_handle *trans,
 	    root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)
 		return 0;
 
-	if (!trans->block_rsv) {
+	if (!trans->reloc_reserved) {
+		rsv = trans->block_rsv;
 		trans->block_rsv = rc->block_rsv;
 		clear_rsv = 1;
 	}
 	reloc_root = create_reloc_root(trans, root, root->root_key.objectid);
 	if (clear_rsv)
-		trans->block_rsv = NULL;
+		trans->block_rsv = rsv;
 
 	ret = __add_reloc_root(reloc_root);
 	BUG_ON(ret < 0);
@@ -1775,8 +1777,7 @@ int replace_path(struct btrfs_trans_handle *trans,
 			new_ptr_gen = 0;
 		}
 
-		if (new_bytenr > 0 && new_bytenr == old_bytenr) {
+		if (WARN_ON(new_bytenr > 0 && new_bytenr == old_bytenr)) {
-			WARN_ON(1);
 			ret = level;
 			break;
 		}
@@ -2058,7 +2059,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 	LIST_HEAD(inode_list);
 	struct btrfs_key key;
 	struct btrfs_key next_key;
-	struct btrfs_trans_handle *trans;
+	struct btrfs_trans_handle *trans = NULL;
 	struct btrfs_root *reloc_root;
 	struct btrfs_root_item *root_item;
 	struct btrfs_path *path;
@@ -2107,18 +2108,19 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 	memset(&next_key, 0, sizeof(next_key));
 
 	while (1) {
-		trans = btrfs_start_transaction(root, 0);
-		BUG_ON(IS_ERR(trans));
-		trans->block_rsv = rc->block_rsv;
-
 		ret = btrfs_block_rsv_refill(root, rc->block_rsv, min_reserved,
 					     BTRFS_RESERVE_FLUSH_ALL);
 		if (ret) {
-			BUG_ON(ret != -EAGAIN);
+			err = ret;
-			ret = btrfs_commit_transaction(trans, root);
+			goto out;
-			BUG_ON(ret);
-			continue;
 		}
+		trans = btrfs_start_transaction(root, 0);
+		if (IS_ERR(trans)) {
+			err = PTR_ERR(trans);
+			trans = NULL;
+			goto out;
+		}
+		trans->block_rsv = rc->block_rsv;
 
 		replaced = 0;
 		max_level = level;
@@ -2164,6 +2166,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 		root_item->drop_level = level;
 
 		btrfs_end_transaction_throttle(trans, root);
+		trans = NULL;
 
 		btrfs_btree_balance_dirty(root);
 
@@ -2192,7 +2195,8 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 		btrfs_update_reloc_root(trans, root);
 	}
 
-	btrfs_end_transaction_throttle(trans, root);
+	if (trans)
+		btrfs_end_transaction_throttle(trans, root);
 
 	btrfs_btree_balance_dirty(root);
 
@@ -3258,7 +3262,7 @@ static int add_tree_block(struct reloc_control *rc,
 	struct rb_node *rb_node;
 	u32 item_size;
 	int level = -1;
-	int generation;
+	u64 generation;
 
 	eb =  path->nodes[0];
 	item_size = btrfs_item_size_nr(eb, path->slots[0]);
@@ -3407,7 +3411,6 @@ static int delete_block_group_cache(struct btrfs_fs_info *fs_info,
 				    struct inode *inode, u64 ino)
 {
 	struct btrfs_key key;
-	struct btrfs_path *path;
 	struct btrfs_root *root = fs_info->tree_root;
 	struct btrfs_trans_handle *trans;
 	int ret = 0;
@@ -3432,22 +3435,14 @@ static int delete_block_group_cache(struct btrfs_fs_info *fs_info,
 	if (ret)
 		goto out;
 
-	path = btrfs_alloc_path();
-	if (!path) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
 	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans)) {
-		btrfs_free_path(path);
 		ret = PTR_ERR(trans);
 		goto out;
 	}
 
-	ret = btrfs_truncate_free_space_cache(root, trans, path, inode);
+	ret = btrfs_truncate_free_space_cache(root, trans, inode);
 
-	btrfs_free_path(path);
 	btrfs_end_transaction(trans, root);
 	btrfs_btree_balance_dirty(root);
 out:
@@ -3549,10 +3544,8 @@ static int find_data_references(struct reloc_control *rc,
 				err = ret;
 				goto out;
 			}
-			if (ret > 0) {
+			if (WARN_ON(ret > 0))
-				WARN_ON(1);
 				goto out;
-			}
 
 			leaf = path->nodes[0];
 			nritems = btrfs_header_nritems(leaf);
@@ -3572,11 +3565,9 @@ static int find_data_references(struct reloc_control *rc,
 		}
 
 		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
-		if (key.objectid != ref_objectid ||
+		if (WARN_ON(key.objectid != ref_objectid ||
-		    key.type != BTRFS_EXTENT_DATA_KEY) {
+		    key.type != BTRFS_EXTENT_DATA_KEY))
-			WARN_ON(1);
 			break;
-		}
 
 		fi = btrfs_item_ptr(leaf, path->slots[0],
 				    struct btrfs_file_extent_item);
@@ -4001,16 +3992,6 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 			}
 		}
 
-		ret = btrfs_block_rsv_check(rc->extent_root, rc->block_rsv, 5);
-		if (ret < 0) {
-			if (ret != -ENOSPC) {
-				err = ret;
-				WARN_ON(1);
-				break;
-			}
-			rc->commit_transaction = 1;
-		}
-
 		if (rc->commit_transaction) {
 			rc->commit_transaction = 0;
 			ret = btrfs_commit_transaction(trans, rc->extent_root);
@@ -4241,12 +4222,12 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 	printk(KERN_INFO "btrfs: relocating block group %llu flags %llu\n",
 	       rc->block_group->key.objectid, rc->block_group->flags);
 
-	ret = btrfs_start_all_delalloc_inodes(fs_info, 0);
+	ret = btrfs_start_delalloc_roots(fs_info, 0);
 	if (ret < 0) {
 		err = ret;
 		goto out;
 	}
-	btrfs_wait_all_ordered_extents(fs_info);
+	btrfs_wait_ordered_roots(fs_info, -1);
 
 	while (1) {
 		mutex_lock(&fs_info->cleaner_mutex);
@@ -4264,7 +4245,12 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 			rc->extents_found);
 
 		if (rc->stage == MOVE_DATA_EXTENTS && rc->found_file_extent) {
-			btrfs_wait_ordered_range(rc->data_inode, 0, (u64)-1);
+			ret = btrfs_wait_ordered_range(rc->data_inode, 0,
+						       (u64)-1);
+			if (ret) {
+				err = ret;
+				goto out;
+			}
 			invalidate_mapping_pages(rc->data_inode->i_mapping,
 						 0, -1);
 			rc->stage = UPDATE_DATA_PTRS;
@@ -4481,6 +4467,7 @@ int btrfs_reloc_clone_csums(struct inode *inode, u64 file_pos, u64 len)
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int ret;
 	u64 disk_bytenr;
+	u64 new_bytenr;
 	LIST_HEAD(list);
 
 	ordered = btrfs_lookup_ordered_extent(inode, file_pos);
@@ -4492,13 +4479,24 @@ int btrfs_reloc_clone_csums(struct inode *inode, u64 file_pos, u64 len)
 	if (ret)
 		goto out;
 
-	disk_bytenr = ordered->start;
 	while (!list_empty(&list)) {
 		sums = list_entry(list.next, struct btrfs_ordered_sum, list);
 		list_del_init(&sums->list);
 
-		sums->bytenr = disk_bytenr;
+		/*
-		disk_bytenr += sums->len;
+		 * We need to offset the new_bytenr based on where the csum is.
+		 * We need to do this because we will read in entire prealloc
+		 * extents but we may have written to say the middle of the
+		 * prealloc extent, so we need to make sure the csum goes with
+		 * the right disk offset.
+		 *
+		 * We can do this because the data reloc inode refers strictly
+		 * to the on disk bytes, so we don't have to worry about
+		 * disk_len vs real len like with real inodes since it's all
+		 * disk length.
+		 */
+		new_bytenr = ordered->start + (sums->bytenr - disk_bytenr);
+		sums->bytenr = new_bytenr;
 
 		btrfs_add_ordered_sum(inode, ordered, sums);
 	}

fs/btrfs/scrub.c

@@ -2717,8 +2717,6 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		mutex_unlock(&fs_info->scrub_lock);
 		wake_up(&fs_info->scrub_pause_wait);
 
-		dev_replace->cursor_left = dev_replace->cursor_right;
-		dev_replace->item_needs_writeback = 1;
 		btrfs_put_block_group(cache);
 		if (ret)
 			break;
@@ -2732,6 +2730,9 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 			break;
 		}
 
+		dev_replace->cursor_left = dev_replace->cursor_right;
+		dev_replace->item_needs_writeback = 1;
+
 		key.offset = found_key.offset + length;
 		btrfs_release_path(path);
 	}
@@ -2783,7 +2784,6 @@ static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info,
 {
 	int ret = 0;
 
-	mutex_lock(&fs_info->scrub_lock);
 	if (fs_info->scrub_workers_refcnt == 0) {
 		if (is_dev_replace)
 			btrfs_init_workers(&fs_info->scrub_workers, "scrub", 1,
@@ -2813,21 +2813,17 @@ static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info,
 	}
 	++fs_info->scrub_workers_refcnt;
 out:
-	mutex_unlock(&fs_info->scrub_lock);
-
 	return ret;
 }
 
 static noinline_for_stack void scrub_workers_put(struct btrfs_fs_info *fs_info)
 {
-	mutex_lock(&fs_info->scrub_lock);
 	if (--fs_info->scrub_workers_refcnt == 0) {
 		btrfs_stop_workers(&fs_info->scrub_workers);
 		btrfs_stop_workers(&fs_info->scrub_wr_completion_workers);
 		btrfs_stop_workers(&fs_info->scrub_nocow_workers);
 	}
 	WARN_ON(fs_info->scrub_workers_refcnt < 0);
-	mutex_unlock(&fs_info->scrub_lock);
 }
 
 int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
@@ -2888,23 +2884,18 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		return -EINVAL;
 	}
 
-	ret = scrub_workers_get(fs_info, is_dev_replace);
-	if (ret)
-		return ret;
 
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
 	dev = btrfs_find_device(fs_info, devid, NULL, NULL);
 	if (!dev || (dev->missing && !is_dev_replace)) {
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-		scrub_workers_put(fs_info);
 		return -ENODEV;
 	}
-	mutex_lock(&fs_info->scrub_lock);
 
+	mutex_lock(&fs_info->scrub_lock);
 	if (!dev->in_fs_metadata || dev->is_tgtdev_for_dev_replace) {
 		mutex_unlock(&fs_info->scrub_lock);
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-		scrub_workers_put(fs_info);
 		return -EIO;
 	}
 
@@ -2915,10 +2906,17 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		btrfs_dev_replace_unlock(&fs_info->dev_replace);
 		mutex_unlock(&fs_info->scrub_lock);
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-		scrub_workers_put(fs_info);
 		return -EINPROGRESS;
 	}
 	btrfs_dev_replace_unlock(&fs_info->dev_replace);
+
+	ret = scrub_workers_get(fs_info, is_dev_replace);
+	if (ret) {
+		mutex_unlock(&fs_info->scrub_lock);
+		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
+		return ret;
+	}
+
 	sctx = scrub_setup_ctx(dev, is_dev_replace);
 	if (IS_ERR(sctx)) {
 		mutex_unlock(&fs_info->scrub_lock);
@@ -2931,13 +2929,15 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 
 	atomic_inc(&fs_info->scrubs_running);
 	mutex_unlock(&fs_info->scrub_lock);
-	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 
 	if (!is_dev_replace) {
-		down_read(&fs_info->scrub_super_lock);
+		/*
+		 * by holding device list mutex, we can
+		 * kick off writing super in log tree sync.
+		 */
 		ret = scrub_supers(sctx, dev);
-		up_read(&fs_info->scrub_super_lock);
 	}
+	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 
 	if (!ret)
 		ret = scrub_enumerate_chunks(sctx, dev, start, end,
@@ -2954,10 +2954,10 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 
 	mutex_lock(&fs_info->scrub_lock);
 	dev->scrub_device = NULL;
+	scrub_workers_put(fs_info);
 	mutex_unlock(&fs_info->scrub_lock);
 
 	scrub_free_ctx(sctx);
-	scrub_workers_put(fs_info);
 
 	return ret;
 }
@@ -2987,16 +2987,6 @@ void btrfs_scrub_continue(struct btrfs_root *root)
 	wake_up(&fs_info->scrub_pause_wait);
 }
 
-void btrfs_scrub_pause_super(struct btrfs_root *root)
-{
-	down_write(&root->fs_info->scrub_super_lock);
-}
-
-void btrfs_scrub_continue_super(struct btrfs_root *root)
-{
-	up_write(&root->fs_info->scrub_super_lock);
-}
-
 int btrfs_scrub_cancel(struct btrfs_fs_info *fs_info)
 {
 	mutex_lock(&fs_info->scrub_lock);

fs/btrfs/send.c

@@ -121,7 +121,6 @@ struct send_ctx {
 	struct list_head name_cache_list;
 	int name_cache_size;
 
-	struct file *cur_inode_filp;
 	char *read_buf;
 };
 
@@ -565,10 +564,8 @@ static int begin_cmd(struct send_ctx *sctx, int cmd)
 {
 	struct btrfs_cmd_header *hdr;
 
-	if (!sctx->send_buf) {
+	if (WARN_ON(!sctx->send_buf))
-		WARN_ON(1);
 		return -EINVAL;
-	}
 
 	BUG_ON(sctx->send_size);
 
@@ -791,7 +788,7 @@ static int iterate_inode_ref(struct btrfs_root *root, struct btrfs_path *path,
 	if (found_key->type == BTRFS_INODE_REF_KEY) {
 		ptr = (unsigned long)btrfs_item_ptr(eb, slot,
 						    struct btrfs_inode_ref);
-		item = btrfs_item_nr(eb, slot);
+		item = btrfs_item_nr(slot);
 		total = btrfs_item_size(eb, item);
 		elem_size = sizeof(*iref);
 	} else {
@@ -905,7 +902,7 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 
 	eb = path->nodes[0];
 	slot = path->slots[0];
-	item = btrfs_item_nr(eb, slot);
+	item = btrfs_item_nr(slot);
 	di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
 	cur = 0;
 	len = 0;
@@ -2119,77 +2116,6 @@ static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen,
 	return ret;
 }
 
-/*
- * Called for regular files when sending extents data. Opens a struct file
- * to read from the file.
- */
-static int open_cur_inode_file(struct send_ctx *sctx)
-{
-	int ret = 0;
-	struct btrfs_key key;
-	struct path path;
-	struct inode *inode;
-	struct dentry *dentry;
-	struct file *filp;
-	int new = 0;
-
-	if (sctx->cur_inode_filp)
-		goto out;
-
-	key.objectid = sctx->cur_ino;
-	key.type = BTRFS_INODE_ITEM_KEY;
-	key.offset = 0;
-
-	inode = btrfs_iget(sctx->send_root->fs_info->sb, &key, sctx->send_root,
-			&new);
-	if (IS_ERR(inode)) {
-		ret = PTR_ERR(inode);
-		goto out;
-	}
-
-	dentry = d_obtain_alias(inode);
-	inode = NULL;
-	if (IS_ERR(dentry)) {
-		ret = PTR_ERR(dentry);
-		goto out;
-	}
-
-	path.mnt = sctx->mnt;
-	path.dentry = dentry;
-	filp = dentry_open(&path, O_RDONLY | O_LARGEFILE, current_cred());
-	dput(dentry);
-	dentry = NULL;
-	if (IS_ERR(filp)) {
-		ret = PTR_ERR(filp);
-		goto out;
-	}
-	sctx->cur_inode_filp = filp;
-
-out:
-	/*
-	 * no xxxput required here as every vfs op
-	 * does it by itself on failure
-	 */
-	return ret;
-}
-
-/*
- * Closes the struct file that was created in open_cur_inode_file
- */
-static int close_cur_inode_file(struct send_ctx *sctx)
-{
-	int ret = 0;
-
-	if (!sctx->cur_inode_filp)
-		goto out;
-
-	ret = filp_close(sctx->cur_inode_filp, NULL);
-	sctx->cur_inode_filp = NULL;
-
-out:
-	return ret;
-}
-
 /*
  * Sends a BTRFS_SEND_C_SUBVOL command/item to userspace
  */
@@ -3622,6 +3548,72 @@ static int process_all_new_xattrs(struct send_ctx *sctx)
 	return ret;
 }
 
+static ssize_t fill_read_buf(struct send_ctx *sctx, u64 offset, u32 len)
+{
+	struct btrfs_root *root = sctx->send_root;
+	struct btrfs_fs_info *fs_info = root->fs_info;
+	struct inode *inode;
+	struct page *page;
+	char *addr;
+	struct btrfs_key key;
+	pgoff_t index = offset >> PAGE_CACHE_SHIFT;
+	pgoff_t last_index;
+	unsigned pg_offset = offset & ~PAGE_CACHE_MASK;
+	ssize_t ret = 0;
+
+	key.objectid = sctx->cur_ino;
+	key.type = BTRFS_INODE_ITEM_KEY;
+	key.offset = 0;
+
+	inode = btrfs_iget(fs_info->sb, &key, root, NULL);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	if (offset + len > i_size_read(inode)) {
+		if (offset > i_size_read(inode))
+			len = 0;
+		else
+			len = offset - i_size_read(inode);
+	}
+	if (len == 0)
+		goto out;
+
+	last_index = (offset + len - 1) >> PAGE_CACHE_SHIFT;
+	while (index <= last_index) {
+		unsigned cur_len = min_t(unsigned, len,
+					 PAGE_CACHE_SIZE - pg_offset);
+		page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
+		if (!page) {
+			ret = -ENOMEM;
+			break;
+		}
+
+		if (!PageUptodate(page)) {
+			btrfs_readpage(NULL, page);
+			lock_page(page);
+			if (!PageUptodate(page)) {
+				unlock_page(page);
+				page_cache_release(page);
+				ret = -EIO;
+				break;
+			}
+		}
+
+		addr = kmap(page);
+		memcpy(sctx->read_buf + ret, addr + pg_offset, cur_len);
+		kunmap(page);
+		unlock_page(page);
+		page_cache_release(page);
+		index++;
+		pg_offset = 0;
+		len -= cur_len;
+		ret += cur_len;
+	}
+out:
+	iput(inode);
+	return ret;
+}
+
 /*
  * Read some bytes from the current inode/file and send a write command to
  * user space.
@@ -3630,35 +3622,20 @@ static int send_write(struct send_ctx *sctx, u64 offset, u32 len)
 {
 	int ret = 0;
 	struct fs_path *p;
-	loff_t pos = offset;
+	ssize_t num_read = 0;
-	int num_read = 0;
-	mm_segment_t old_fs;
 
 	p = fs_path_alloc();
 	if (!p)
 		return -ENOMEM;
 
-	/*
-	 * vfs normally only accepts user space buffers for security reasons.
-	 * we only read from the file and also only provide the read_buf buffer
-	 * to vfs. As this buffer does not come from a user space call, it's
-	 * ok to temporary allow kernel space buffers.
-	 */
-	old_fs = get_fs();
-	set_fs(KERNEL_DS);
-
 verbose_printk("btrfs: send_write offset=%llu, len=%d\n", offset, len);
 
-	ret = open_cur_inode_file(sctx);
+	num_read = fill_read_buf(sctx, offset, len);
-	if (ret < 0)
+	if (num_read <= 0) {
-		goto out;
+		if (num_read < 0)
-
+			ret = num_read;
-	ret = vfs_read(sctx->cur_inode_filp, sctx->read_buf, len, &pos);
-	if (ret < 0)
-		goto out;
-	num_read = ret;
-	if (!num_read)
 		goto out;
+	}
 
 	ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE);
 	if (ret < 0)
@@ -3677,7 +3654,6 @@ verbose_printk("btrfs: send_write offset=%llu, len=%d\n", offset, len);
 tlv_put_failure:
 out:
 	fs_path_free(p);
-	set_fs(old_fs);
 	if (ret < 0)
 		return ret;
 	return num_read;
@@ -3926,16 +3902,16 @@ static int is_extent_unchanged(struct send_ctx *sctx,
 	while (key.offset < ekey->offset + left_len) {
 		ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
 		right_type = btrfs_file_extent_type(eb, ei);
-		right_disknr = btrfs_file_extent_disk_bytenr(eb, ei);
-		right_len = btrfs_file_extent_num_bytes(eb, ei);
-		right_offset = btrfs_file_extent_offset(eb, ei);
-		right_gen = btrfs_file_extent_generation(eb, ei);
-
 		if (right_type != BTRFS_FILE_EXTENT_REG) {
 			ret = 0;
 			goto out;
 		}
 
+		right_disknr = btrfs_file_extent_disk_bytenr(eb, ei);
+		right_len = btrfs_file_extent_num_bytes(eb, ei);
+		right_offset = btrfs_file_extent_offset(eb, ei);
+		right_gen = btrfs_file_extent_generation(eb, ei);
+
 		/*
 		 * Are we at extent 8? If yes, we know the extent is changed.
 		 * This may only happen on the first iteration.
@@ -4222,10 +4198,6 @@ static int changed_inode(struct send_ctx *sctx,
 	u64 left_gen = 0;
 	u64 right_gen = 0;
 
-	ret = close_cur_inode_file(sctx);
-	if (ret < 0)
-		goto out;
-
 	sctx->cur_ino = key->objectid;
 	sctx->cur_inode_new_gen = 0;
 
@@ -4686,11 +4658,6 @@ static int send_subvol(struct send_ctx *sctx)
 	}
 
 out:
-	if (!ret)
-		ret = close_cur_inode_file(sctx);
-	else
-		close_cur_inode_file(sctx);
-
 	free_recorded_refs(sctx);
 	return ret;
 }

fs/btrfs/super.c

@@ -42,7 +42,6 @@
 #include <linux/cleancache.h>
 #include <linux/ratelimit.h>
 #include <linux/btrfs.h>
-#include "compat.h"
 #include "delayed-inode.h"
 #include "ctree.h"
 #include "disk-io.h"
@@ -921,7 +920,7 @@ int btrfs_sync_fs(struct super_block *sb, int wait)
 		return 0;
 	}
 
-	btrfs_wait_all_ordered_extents(fs_info);
+	btrfs_wait_ordered_roots(fs_info, -1);
 
 	trans = btrfs_attach_transaction_barrier(root);
 	if (IS_ERR(trans)) {
@@ -1330,6 +1329,12 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 		 * this also happens on 'umount -rf' or on shutdown, when
 		 * the filesystem is busy.
 		 */
+
+		/* wait for the uuid_scan task to finish */
+		down(&fs_info->uuid_tree_rescan_sem);
+		/* avoid complains from lockdep et al. */
+		up(&fs_info->uuid_tree_rescan_sem);
+
 		sb->s_flags |= MS_RDONLY;
 
 		btrfs_dev_replace_suspend_for_unmount(fs_info);
@@ -1465,7 +1470,7 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 	nr_devices = fs_info->fs_devices->open_devices;
 	BUG_ON(!nr_devices);
 
-	devices_info = kmalloc(sizeof(*devices_info) * nr_devices,
+	devices_info = kmalloc_array(nr_devices, sizeof(*devices_info),
 			       GFP_NOFS);
 	if (!devices_info)
 		return -ENOMEM;
@@ -1789,7 +1794,25 @@ static void btrfs_print_info(void)
 
 static int btrfs_run_sanity_tests(void)
 {
-	return btrfs_test_free_space_cache();
+	int ret;
+
+	ret = btrfs_init_test_fs();
+	if (ret)
+		return ret;
+
+	ret = btrfs_test_free_space_cache();
+	if (ret)
+		goto out;
+	ret = btrfs_test_extent_buffer_operations();
+	if (ret)
+		goto out;
+	ret = btrfs_test_extent_io();
+	if (ret)
+		goto out;
+	ret = btrfs_test_inodes();
+out:
+	btrfs_destroy_test_fs();
+	return ret;
 }
 
 static int __init init_btrfs_fs(void)

fs/btrfs/tests/btrfs-tests.c

@@ -0,0 +1,74 @@
+/*
+ * Copyright (C) 2013 Fusion IO.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#include <linux/fs.h>
+#include <linux/mount.h>
+#include <linux/magic.h>
+#include "btrfs-tests.h"
+#include "../ctree.h"
+
+static struct vfsmount *test_mnt = NULL;
+
+static const struct super_operations btrfs_test_super_ops = {
+	.alloc_inode	= btrfs_alloc_inode,
+	.destroy_inode	= btrfs_test_destroy_inode,
+};
+
+static struct dentry *btrfs_test_mount(struct file_system_type *fs_type,
+				       int flags, const char *dev_name,
+				       void *data)
+{
+	return mount_pseudo(fs_type, "btrfs_test:", &btrfs_test_super_ops,
+			    NULL, BTRFS_TEST_MAGIC);
+}
+
+static struct file_system_type test_type = {
+	.name		= "btrfs_test_fs",
+	.mount		= btrfs_test_mount,
+	.kill_sb	= kill_anon_super,
+};
+
+struct inode *btrfs_new_test_inode(void)
+{
+	return new_inode(test_mnt->mnt_sb);
+}
+
+int btrfs_init_test_fs(void)
+{
+	int ret;
+
+	ret = register_filesystem(&test_type);
+	if (ret) {
+		printk(KERN_ERR "btrfs: cannot register test file system\n");
+		return ret;
+	}
+
+	test_mnt = kern_mount(&test_type);
+	if (IS_ERR(test_mnt)) {
+		printk(KERN_ERR "btrfs: cannot mount test file system\n");
+		unregister_filesystem(&test_type);
+		return ret;
+	}
+	return 0;
+}
+
+void btrfs_destroy_test_fs(void)
+{
+	kern_unmount(test_mnt);
+	unregister_filesystem(&test_type);
+}

fs/btrfs/tests/btrfs-tests.h

@@ -24,11 +24,36 @@
 #define test_msg(fmt, ...) pr_info("btrfs: selftest: " fmt, ##__VA_ARGS__)
 
 int btrfs_test_free_space_cache(void);
+int btrfs_test_extent_buffer_operations(void);
+int btrfs_test_extent_io(void);
+int btrfs_test_inodes(void);
+int btrfs_init_test_fs(void);
+void btrfs_destroy_test_fs(void);
+struct inode *btrfs_new_test_inode(void);
 #else
 static inline int btrfs_test_free_space_cache(void)
 {
 	return 0;
 }
+static inline int btrfs_test_extent_buffer_operations(void)
+{
+	return 0;
+}
+static inline int btrfs_init_test_fs(void)
+{
+	return 0;
+}
+static inline void btrfs_destroy_test_fs(void)
+{
+}
+static inline int btrfs_test_extent_io(void)
+{
+	return 0;
+}
+static inline int btrfs_test_inodes(void)
+{
+	return 0;
+}
 #endif
 
 #endif

fs/btrfs/tests/extent-buffer-tests.c

@@ -0,0 +1,229 @@
+/*
+ * Copyright (C) 2013 Fusion IO.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#include <linux/slab.h>
+#include "btrfs-tests.h"
+#include "../ctree.h"
+#include "../extent_io.h"
+#include "../disk-io.h"
+
+static int test_btrfs_split_item(void)
+{
+	struct btrfs_path *path;
+	struct btrfs_root *root;
+	struct extent_buffer *eb;
+	struct btrfs_item *item;
+	char *value = "mary had a little lamb";
+	char *split1 = "mary had a little";
+	char *split2 = " lamb";
+	char *split3 = "mary";
+	char *split4 = " had a little";
+	char buf[32];
+	struct btrfs_key key;
+	u32 value_len = strlen(value);
+	int ret = 0;
+
+	test_msg("Running btrfs_split_item tests\n");
+
+	root = btrfs_alloc_dummy_root();
+	if (IS_ERR(root)) {
+		test_msg("Could not allocate root\n");
+		return PTR_ERR(root);
+	}
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		test_msg("Could not allocate path\n");
+		kfree(root);
+		return -ENOMEM;
+	}
+
+	path->nodes[0] = eb = alloc_dummy_extent_buffer(0, 4096);
+	if (!eb) {
+		test_msg("Could not allocate dummy buffer\n");
+		ret = -ENOMEM;
+		goto out;
+	}
+	path->slots[0] = 0;
+
+	key.objectid = 0;
+	key.type = BTRFS_EXTENT_CSUM_KEY;
+	key.offset = 0;
+
+	setup_items_for_insert(root, path, &key, &value_len, value_len,
+			       value_len + sizeof(struct btrfs_item), 1);
+	item = btrfs_item_nr(0);
+	write_extent_buffer(eb, value, btrfs_item_ptr_offset(eb, 0),
+			    value_len);
+
+	key.offset = 3;
+
+	/*
+	 * Passing NULL trans here should be safe because we have plenty of
+	 * space in this leaf to split the item without having to split the
+	 * leaf.
+	 */
+	ret = btrfs_split_item(NULL, root, path, &key, 17);
+	if (ret) {
+		test_msg("Split item failed %d\n", ret);
+		goto out;
+	}
+
+	/*
+	 * Read the first slot, it should have the original key and contain only
+	 * 'mary had a little'
+	 */
+	btrfs_item_key_to_cpu(eb, &key, 0);
+	if (key.objectid != 0 || key.type != BTRFS_EXTENT_CSUM_KEY ||
+	    key.offset != 0) {
+		test_msg("Invalid key at slot 0\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	item = btrfs_item_nr(0);
+	if (btrfs_item_size(eb, item) != strlen(split1)) {
+		test_msg("Invalid len in the first split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	read_extent_buffer(eb, buf, btrfs_item_ptr_offset(eb, 0),
+			   strlen(split1));
+	if (memcmp(buf, split1, strlen(split1))) {
+		test_msg("Data in the buffer doesn't match what it should "
+			 "in the first split have='%.*s' want '%s'\n",
+			 (int)strlen(split1), buf, split1);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_item_key_to_cpu(eb, &key, 1);
+	if (key.objectid != 0 || key.type != BTRFS_EXTENT_CSUM_KEY ||
+	    key.offset != 3) {
+		test_msg("Invalid key at slot 1\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	item = btrfs_item_nr(1);
+	if (btrfs_item_size(eb, item) != strlen(split2)) {
+		test_msg("Invalid len in the second split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	read_extent_buffer(eb, buf, btrfs_item_ptr_offset(eb, 1),
+			   strlen(split2));
+	if (memcmp(buf, split2, strlen(split2))) {
+		test_msg("Data in the buffer doesn't match what it should "
+			 "in the second split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	key.offset = 1;
+	/* Do it again so we test memmoving the other items in the leaf */
+	ret = btrfs_split_item(NULL, root, path, &key, 4);
+	if (ret) {
+		test_msg("Second split item failed %d\n", ret);
+		goto out;
+	}
+
+	btrfs_item_key_to_cpu(eb, &key, 0);
+	if (key.objectid != 0 || key.type != BTRFS_EXTENT_CSUM_KEY ||
+	    key.offset != 0) {
+		test_msg("Invalid key at slot 0\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	item = btrfs_item_nr(0);
+	if (btrfs_item_size(eb, item) != strlen(split3)) {
+		test_msg("Invalid len in the first split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	read_extent_buffer(eb, buf, btrfs_item_ptr_offset(eb, 0),
+			   strlen(split3));
+	if (memcmp(buf, split3, strlen(split3))) {
+		test_msg("Data in the buffer doesn't match what it should "
+			 "in the third split");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_item_key_to_cpu(eb, &key, 1);
+	if (key.objectid != 0 || key.type != BTRFS_EXTENT_CSUM_KEY ||
+	    key.offset != 1) {
+		test_msg("Invalid key at slot 1\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	item = btrfs_item_nr(1);
+	if (btrfs_item_size(eb, item) != strlen(split4)) {
+		test_msg("Invalid len in the second split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	read_extent_buffer(eb, buf, btrfs_item_ptr_offset(eb, 1),
+			   strlen(split4));
+	if (memcmp(buf, split4, strlen(split4))) {
+		test_msg("Data in the buffer doesn't match what it should "
+			 "in the fourth split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_item_key_to_cpu(eb, &key, 2);
+	if (key.objectid != 0 || key.type != BTRFS_EXTENT_CSUM_KEY ||
+	    key.offset != 3) {
+		test_msg("Invalid key at slot 2\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	item = btrfs_item_nr(2);
+	if (btrfs_item_size(eb, item) != strlen(split2)) {
+		test_msg("Invalid len in the second split\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	read_extent_buffer(eb, buf, btrfs_item_ptr_offset(eb, 2),
+			   strlen(split2));
+	if (memcmp(buf, split2, strlen(split2))) {
+		test_msg("Data in the buffer doesn't match what it should "
+			 "in the last chunk\n");
+		ret = -EINVAL;
+		goto out;
+	}
+out:
+	btrfs_free_path(path);
+	kfree(root);
+	return ret;
+}
+
+int btrfs_test_extent_buffer_operations(void)
+{
+	test_msg("Running extent buffer operation tests");
+	return test_btrfs_split_item();
+}

fs/btrfs/tests/extent-io-tests.c

@@ -0,0 +1,276 @@
+/*
+ * Copyright (C) 2013 Fusion IO.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#include <linux/pagemap.h>
+#include <linux/sched.h>
+#include "btrfs-tests.h"
+#include "../extent_io.h"
+
+#define PROCESS_UNLOCK		(1 << 0)
+#define PROCESS_RELEASE		(1 << 1)
+#define PROCESS_TEST_LOCKED	(1 << 2)
+
+static noinline int process_page_range(struct inode *inode, u64 start, u64 end,
+				       unsigned long flags)
+{
+	int ret;
+	struct page *pages[16];
+	unsigned long index = start >> PAGE_CACHE_SHIFT;
+	unsigned long end_index = end >> PAGE_CACHE_SHIFT;
+	unsigned long nr_pages = end_index - index + 1;
+	int i;
+	int count = 0;
+	int loops = 0;
+
+	while (nr_pages > 0) {
+		ret = find_get_pages_contig(inode->i_mapping, index,
+				     min_t(unsigned long, nr_pages,
+				     ARRAY_SIZE(pages)), pages);
+		for (i = 0; i < ret; i++) {
+			if (flags & PROCESS_TEST_LOCKED &&
+			    !PageLocked(pages[i]))
+				count++;
+			if (flags & PROCESS_UNLOCK && PageLocked(pages[i]))
+				unlock_page(pages[i]);
+			page_cache_release(pages[i]);
+			if (flags & PROCESS_RELEASE)
+				page_cache_release(pages[i]);
+		}
+		nr_pages -= ret;
+		index += ret;
+		cond_resched();
+		loops++;
+		if (loops > 100000) {
+			printk(KERN_ERR "stuck in a loop, start %Lu, end %Lu, nr_pages %lu, ret %d\n", start, end, nr_pages, ret);
+			break;
+		}
+	}
+	return count;
+}
+
+static int test_find_delalloc(void)
+{
+	struct inode *inode;
+	struct extent_io_tree tmp;
+	struct page *page;
+	struct page *locked_page = NULL;
+	unsigned long index = 0;
+	u64 total_dirty = 256 * 1024 * 1024;
+	u64 max_bytes = 128 * 1024 * 1024;
+	u64 start, end, test_start;
+	u64 found;
+	int ret = -EINVAL;
+
+	inode = btrfs_new_test_inode();
+	if (!inode) {
+		test_msg("Failed to allocate test inode\n");
+		return -ENOMEM;
+	}
+
+	extent_io_tree_init(&tmp, &inode->i_data);
+
+	/*
+	 * First go through and create and mark all of our pages dirty, we pin
+	 * everything to make sure our pages don't get evicted and screw up our
+	 * test.
+	 */
+	for (index = 0; index < (total_dirty >> PAGE_CACHE_SHIFT); index++) {
+		page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
+		if (!page) {
+			test_msg("Failed to allocate test page\n");
+			ret = -ENOMEM;
+			goto out;
+		}
+		SetPageDirty(page);
+		if (index) {
+			unlock_page(page);
+		} else {
+			page_cache_get(page);
+			locked_page = page;
+		}
+	}
+
+	/* Test this scenario
+	 * |--- delalloc ---|
+	 * |---  search  ---|
+	 */
+	set_extent_delalloc(&tmp, 0, 4095, NULL, GFP_NOFS);
+	start = 0;
+	end = 0;
+	found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
+					 &end, max_bytes);
+	if (!found) {
+		test_msg("Should have found at least one delalloc\n");
+		goto out_bits;
+	}
+	if (start != 0 || end != 4095) {
+		test_msg("Expected start 0 end 4095, got start %Lu end %Lu\n",
+			 start, end);
+		goto out_bits;
+	}
+	unlock_extent(&tmp, start, end);
+	unlock_page(locked_page);
+	page_cache_release(locked_page);
+
+	/*
+	 * Test this scenario
+	 *
+	 * |--- delalloc ---|
+	 *           |--- search ---|
+	 */
+	test_start = 64 * 1024 * 1024;
+	locked_page = find_lock_page(inode->i_mapping,
+				     test_start >> PAGE_CACHE_SHIFT);
+	if (!locked_page) {
+		test_msg("Couldn't find the locked page\n");
+		goto out_bits;
+	}
+	set_extent_delalloc(&tmp, 4096, max_bytes - 1, NULL, GFP_NOFS);
+	start = test_start;
+	end = 0;
+	found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
+					 &end, max_bytes);
+	if (!found) {
+		test_msg("Couldn't find delalloc in our range\n");
+		goto out_bits;
+	}
+	if (start != test_start || end != max_bytes - 1) {
+		test_msg("Expected start %Lu end %Lu, got start %Lu, end "
+			 "%Lu\n", test_start, max_bytes - 1, start, end);
+		goto out_bits;
+	}
+	if (process_page_range(inode, start, end,
+			       PROCESS_TEST_LOCKED | PROCESS_UNLOCK)) {
+		test_msg("There were unlocked pages in the range\n");
+		goto out_bits;
+	}
+	unlock_extent(&tmp, start, end);
+	/* locked_page was unlocked above */
+	page_cache_release(locked_page);
+
+	/*
+	 * Test this scenario
+	 * |--- delalloc ---|
+	 *                    |--- search ---|
+	 */
+	test_start = max_bytes + 4096;
+	locked_page = find_lock_page(inode->i_mapping, test_start >>
+				     PAGE_CACHE_SHIFT);
+	if (!locked_page) {
+		test_msg("Could'nt find the locked page\n");
+		goto out_bits;
+	}
+	start = test_start;
+	end = 0;
+	found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
+					 &end, max_bytes);
+	if (found) {
+		test_msg("Found range when we shouldn't have\n");
+		goto out_bits;
+	}
+	if (end != (u64)-1) {
+		test_msg("Did not return the proper end offset\n");
+		goto out_bits;
+	}
+
+	/*
+	 * Test this scenario
+	 * [------- delalloc -------|
+	 * [max_bytes]|-- search--|
+	 *
+	 * We are re-using our test_start from above since it works out well.
+	 */
+	set_extent_delalloc(&tmp, max_bytes, total_dirty - 1, NULL, GFP_NOFS);
+	start = test_start;
+	end = 0;
+	found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
+					 &end, max_bytes);
+	if (!found) {
+		test_msg("Didn't find our range\n");
+		goto out_bits;
+	}
+	if (start != test_start || end != total_dirty - 1) {
+		test_msg("Expected start %Lu end %Lu, got start %Lu end %Lu\n",
+			 test_start, total_dirty - 1, start, end);
+		goto out_bits;
+	}
+	if (process_page_range(inode, start, end,
+			       PROCESS_TEST_LOCKED | PROCESS_UNLOCK)) {
+		test_msg("Pages in range were not all locked\n");
+		goto out_bits;
+	}
+	unlock_extent(&tmp, start, end);
+
+	/*
+	 * Now to test where we run into a page that is no longer dirty in the
+	 * range we want to find.
+	 */
+	page = find_get_page(inode->i_mapping, (max_bytes + (1 * 1024 * 1024))
+			     >> PAGE_CACHE_SHIFT);
+	if (!page) {
+		test_msg("Couldn't find our page\n");
+		goto out_bits;
+	}
+	ClearPageDirty(page);
+	page_cache_release(page);
+
+	/* We unlocked it in the previous test */
+	lock_page(locked_page);
+	start = test_start;
+	end = 0;
+	/*
+	 * Currently if we fail to find dirty pages in the delalloc range we
+	 * will adjust max_bytes down to PAGE_CACHE_SIZE and then re-search.  If
+	 * this changes at any point in the future we will need to fix this
+	 * tests expected behavior.
+	 */
+	found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
+					 &end, max_bytes);
+	if (!found) {
+		test_msg("Didn't find our range\n");
+		goto out_bits;
+	}
+	if (start != test_start && end != test_start + PAGE_CACHE_SIZE - 1) {
+		test_msg("Expected start %Lu end %Lu, got start %Lu end %Lu\n",
+			 test_start, test_start + PAGE_CACHE_SIZE - 1, start,
+			 end);
+		goto out_bits;
+	}
+	if (process_page_range(inode, start, end, PROCESS_TEST_LOCKED |
+			       PROCESS_UNLOCK)) {
+		test_msg("Pages in range were not all locked\n");
+		goto out_bits;
+	}
+	ret = 0;
+out_bits:
+	clear_extent_bits(&tmp, 0, total_dirty - 1,
+			  (unsigned long)-1, GFP_NOFS);
+out:
+	if (locked_page)
+		page_cache_release(locked_page);
+	process_page_range(inode, 0, total_dirty - 1,
+			   PROCESS_UNLOCK | PROCESS_RELEASE);
+	iput(inode);
+	return ret;
+}
+
+int btrfs_test_extent_io(void)
+{
+	test_msg("Running find delalloc tests\n");
+	return test_find_delalloc();
+}

fs/btrfs/tests/inode-tests.c

@@ -0,0 +1,955 @@
+/*
+ * Copyright (C) 2013 Fusion IO.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#include "btrfs-tests.h"
+#include "../ctree.h"
+#include "../btrfs_inode.h"
+#include "../disk-io.h"
+#include "../extent_io.h"
+#include "../volumes.h"
+
+static struct btrfs_fs_info *alloc_dummy_fs_info(void)
+{
+	struct btrfs_fs_info *fs_info = kzalloc(sizeof(struct btrfs_fs_info),
+						GFP_NOFS);
+	if (!fs_info)
+		return fs_info;
+	fs_info->fs_devices = kzalloc(sizeof(struct btrfs_fs_devices),
+				      GFP_NOFS);
+	if (!fs_info->fs_devices) {
+		kfree(fs_info);
+		return NULL;
+	}
+	return fs_info;
+}
+static void free_dummy_root(struct btrfs_root *root)
+{
+	if (!root)
+		return;
+	if (root->fs_info) {
+		kfree(root->fs_info->fs_devices);
+		kfree(root->fs_info);
+	}
+	if (root->node)
+		free_extent_buffer(root->node);
+	kfree(root);
+}
+
+static void insert_extent(struct btrfs_root *root, u64 start, u64 len,
+			  u64 ram_bytes, u64 offset, u64 disk_bytenr,
+			  u64 disk_len, u32 type, u8 compression, int slot)
+{
+	struct btrfs_path path;
+	struct btrfs_file_extent_item *fi;
+	struct extent_buffer *leaf = root->node;
+	struct btrfs_key key;
+	u32 value_len = sizeof(struct btrfs_file_extent_item);
+
+	if (type == BTRFS_FILE_EXTENT_INLINE)
+		value_len += len;
+	memset(&path, 0, sizeof(path));
+
+	path.nodes[0] = leaf;
+	path.slots[0] = slot;
+
+	key.objectid = BTRFS_FIRST_FREE_OBJECTID;
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = start;
+
+	setup_items_for_insert(root, &path, &key, &value_len, value_len,
+			       value_len + sizeof(struct btrfs_item), 1);
+	fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
+	btrfs_set_file_extent_generation(leaf, fi, 1);
+	btrfs_set_file_extent_type(leaf, fi, type);
+	btrfs_set_file_extent_disk_bytenr(leaf, fi, disk_bytenr);
+	btrfs_set_file_extent_disk_num_bytes(leaf, fi, disk_len);
+	btrfs_set_file_extent_offset(leaf, fi, offset);
+	btrfs_set_file_extent_num_bytes(leaf, fi, len);
+	btrfs_set_file_extent_ram_bytes(leaf, fi, ram_bytes);
+	btrfs_set_file_extent_compression(leaf, fi, compression);
+	btrfs_set_file_extent_encryption(leaf, fi, 0);
+	btrfs_set_file_extent_other_encoding(leaf, fi, 0);
+}
+
+static void insert_inode_item_key(struct btrfs_root *root)
+{
+	struct btrfs_path path;
+	struct extent_buffer *leaf = root->node;
+	struct btrfs_key key;
+	u32 value_len = 0;
+
+	memset(&path, 0, sizeof(path));
+
+	path.nodes[0] = leaf;
+	path.slots[0] = 0;
+
+	key.objectid = BTRFS_INODE_ITEM_KEY;
+	key.type = BTRFS_INODE_ITEM_KEY;
+	key.offset = 0;
+
+	setup_items_for_insert(root, &path, &key, &value_len, value_len,
+			       value_len + sizeof(struct btrfs_item), 1);
+}
+
+/*
+ * Build the most complicated map of extents the earth has ever seen.  We want
+ * this so we can test all of the corner cases of btrfs_get_extent.  Here is a
+ * diagram of how the extents will look though this may not be possible we still
+ * want to make sure everything acts normally (the last number is not inclusive)
+ *
+ * [0 - 5][5 -  6][6 - 10][10 - 4096][  4096 - 8192 ][8192 - 12288]
+ * [hole ][inline][ hole ][ regular ][regular1 split][    hole    ]
+ *
+ * [ 12288 - 20480][20480 - 24576][  24576 - 28672  ][28672 - 36864][36864 - 45056]
+ * [regular1 split][   prealloc1 ][prealloc1 written][   prealloc1 ][ compressed  ]
+ *
+ * [45056 - 49152][49152-53248][53248-61440][61440-65536][     65536+81920   ]
+ * [ compressed1 ][  regular  ][compressed1][  regular  ][ hole but no extent]
+ *
+ * [81920-86016]
+ * [  regular  ]
+ */
+static void setup_file_extents(struct btrfs_root *root)
+{
+	int slot = 0;
+	u64 disk_bytenr = 1 * 1024 * 1024;
+	u64 offset = 0;
+
+	/* First we want a hole */
+	insert_extent(root, offset, 5, 5, 0, 0, 0, BTRFS_FILE_EXTENT_REG, 0,
+		      slot);
+	slot++;
+	offset += 5;
+
+	/*
+	 * Now we want an inline extent, I don't think this is possible but hey
+	 * why not?  Also keep in mind if we have an inline extent it counts as
+	 * the whole first page.  If we were to expand it we would have to cow
+	 * and we wouldn't have an inline extent anymore.
+	 */
+	insert_extent(root, offset, 1, 1, 0, 0, 0, BTRFS_FILE_EXTENT_INLINE, 0,
+		      slot);
+	slot++;
+	offset = 4096;
+
+	/* Now another hole */
+	insert_extent(root, offset, 4, 4, 0, 0, 0, BTRFS_FILE_EXTENT_REG, 0,
+		      slot);
+	slot++;
+	offset += 4;
+
+	/* Now for a regular extent */
+	insert_extent(root, offset, 4095, 4095, 0, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+	slot++;
+	disk_bytenr += 4096;
+	offset += 4095;
+
+	/*
+	 * Now for 3 extents that were split from a hole punch so we test
+	 * offsets properly.
+	 */
+	insert_extent(root, offset, 4096, 16384, 0, disk_bytenr, 16384,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+	slot++;
+	offset += 4096;
+	insert_extent(root, offset, 4096, 4096, 0, 0, 0, BTRFS_FILE_EXTENT_REG,
+		      0, slot);
+	slot++;
+	offset += 4096;
+	insert_extent(root, offset, 8192, 16384, 8192, disk_bytenr, 16384,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+	slot++;
+	offset += 8192;
+	disk_bytenr += 16384;
+
+	/* Now for a unwritten prealloc extent */
+	insert_extent(root, offset, 4096, 4096, 0, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_PREALLOC, 0, slot);
+	slot++;
+	offset += 4096;
+
+	/*
+	 * We want to jack up disk_bytenr a little more so the em stuff doesn't
+	 * merge our records.
+	 */
+	disk_bytenr += 8192;
+
+	/*
+	 * Now for a partially written prealloc extent, basically the same as
+	 * the hole punch example above.  Ram_bytes never changes when you mark
+	 * extents written btw.
+	 */
+	insert_extent(root, offset, 4096, 16384, 0, disk_bytenr, 16384,
+		      BTRFS_FILE_EXTENT_PREALLOC, 0, slot);
+	slot++;
+	offset += 4096;
+	insert_extent(root, offset, 4096, 16384, 4096, disk_bytenr, 16384,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+	slot++;
+	offset += 4096;
+	insert_extent(root, offset, 8192, 16384, 8192, disk_bytenr, 16384,
+		      BTRFS_FILE_EXTENT_PREALLOC, 0, slot);
+	slot++;
+	offset += 8192;
+	disk_bytenr += 16384;
+
+	/* Now a normal compressed extent */
+	insert_extent(root, offset, 8192, 8192, 0, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_REG, BTRFS_COMPRESS_ZLIB, slot);
+	slot++;
+	offset += 8192;
+	/* No merges */
+	disk_bytenr += 8192;
+
+	/* Now a split compressed extent */
+	insert_extent(root, offset, 4096, 16384, 0, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_REG, BTRFS_COMPRESS_ZLIB, slot);
+	slot++;
+	offset += 4096;
+	insert_extent(root, offset, 4096, 4096, 0, disk_bytenr + 4096, 4096,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+	slot++;
+	offset += 4096;
+	insert_extent(root, offset, 8192, 16384, 8192, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_REG, BTRFS_COMPRESS_ZLIB, slot);
+	slot++;
+	offset += 8192;
+	disk_bytenr += 8192;
+
+	/* Now extents that have a hole but no hole extent */
+	insert_extent(root, offset, 4096, 4096, 0, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+	slot++;
+	offset += 16384;
+	disk_bytenr += 4096;
+	insert_extent(root, offset, 4096, 4096, 0, disk_bytenr, 4096,
+		      BTRFS_FILE_EXTENT_REG, 0, slot);
+}
+
+static unsigned long prealloc_only = 0;
+static unsigned long compressed_only = 0;
+static unsigned long vacancy_only = 0;
+
+static noinline int test_btrfs_get_extent(void)
+{
+	struct inode *inode = NULL;
+	struct btrfs_root *root = NULL;
+	struct extent_map *em = NULL;
+	u64 orig_start;
+	u64 disk_bytenr;
+	u64 offset;
+	int ret = -ENOMEM;
+
+	inode = btrfs_new_test_inode();
+	if (!inode) {
+		test_msg("Couldn't allocate inode\n");
+		return ret;
+	}
+
+	BTRFS_I(inode)->location.type = BTRFS_INODE_ITEM_KEY;
+	BTRFS_I(inode)->location.objectid = BTRFS_FIRST_FREE_OBJECTID;
+	BTRFS_I(inode)->location.offset = 0;
+
+	root = btrfs_alloc_dummy_root();
+	if (IS_ERR(root)) {
+		test_msg("Couldn't allocate root\n");
+		goto out;
+	}
+
+	/*
+	 * We do this since btrfs_get_extent wants to assign em->bdev to
+	 * root->fs_info->fs_devices->latest_bdev.
+	 */
+	root->fs_info = alloc_dummy_fs_info();
+	if (!root->fs_info) {
+		test_msg("Couldn't allocate dummy fs info\n");
+		goto out;
+	}
+
+	root->node = alloc_dummy_extent_buffer(0, 4096);
+	if (!root->node) {
+		test_msg("Couldn't allocate dummy buffer\n");
+		goto out;
+	}
+
+	/*
+	 * We will just free a dummy node if it's ref count is 2 so we need an
+	 * extra ref so our searches don't accidently release our page.
+	 */
+	extent_buffer_get(root->node);
+	btrfs_set_header_nritems(root->node, 0);
+	btrfs_set_header_level(root->node, 0);
+	ret = -EINVAL;
+
+	/* First with no extents */
+	BTRFS_I(inode)->root = root;
+	em = btrfs_get_extent(inode, NULL, 0, 0, 4096, 0);
+	if (IS_ERR(em)) {
+		em = NULL;
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_HOLE) {
+		test_msg("Expected a hole, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (!test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
+		test_msg("Vacancy flag wasn't set properly\n");
+		goto out;
+	}
+	free_extent_map(em);
+	btrfs_drop_extent_cache(inode, 0, (u64)-1, 0);
+
+	/*
+	 * All of the magic numbers are based on the mapping setup in
+	 * setup_file_extents, so if you change anything there you need to
+	 * update the comment and update the expected values below.
+	 */
+	setup_file_extents(root);
+
+	em = btrfs_get_extent(inode, NULL, 0, 0, (u64)-1, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_HOLE) {
+		test_msg("Expected a hole, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != 0 || em->len != 5) {
+		test_msg("Unexpected extent wanted start 0 len 5, got start "
+			 "%llu len %llu\n", em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_INLINE) {
+		test_msg("Expected an inline, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4091) {
+		test_msg("Unexpected extent wanted start %llu len 1, got start "
+			 "%llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	/*
+	 * We don't test anything else for inline since it doesn't get set
+	 * unless we have a page for it to write into.  Maybe we should change
+	 * this?
+	 */
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_HOLE) {
+		test_msg("Expected a hole, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4) {
+		test_msg("Unexpected extent wanted start %llu len 4, got start "
+			 "%llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* Regular extent */
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4095) {
+		test_msg("Unexpected extent wanted start %llu len 4095, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* The next 3 are split extents */
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	disk_bytenr = em->block_start;
+	orig_start = em->start;
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_HOLE) {
+		test_msg("Expected a hole, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 8192) {
+		test_msg("Unexpected extent wanted start %llu len 8192, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != orig_start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n",
+			 orig_start, em->orig_start);
+		goto out;
+	}
+	disk_bytenr += (em->start - orig_start);
+	if (em->block_start != disk_bytenr) {
+		test_msg("Wrong block start, want %llu, have %llu\n",
+			 disk_bytenr, em->block_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* Prealloc extent */
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != prealloc_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 prealloc_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* The next 3 are a half written prealloc extent */
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != prealloc_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 prealloc_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	disk_bytenr = em->block_start;
+	orig_start = em->start;
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_HOLE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != orig_start) {
+		test_msg("Unexpected orig offset, wanted %llu, have %llu\n",
+			 orig_start, em->orig_start);
+		goto out;
+	}
+	if (em->block_start != (disk_bytenr + (em->start - em->orig_start))) {
+		test_msg("Unexpected block start, wanted %llu, have %llu\n",
+			 disk_bytenr + (em->start - em->orig_start),
+			 em->block_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 8192) {
+		test_msg("Unexpected extent wanted start %llu len 8192, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != prealloc_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 prealloc_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != orig_start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", orig_start,
+			 em->orig_start);
+		goto out;
+	}
+	if (em->block_start != (disk_bytenr + (em->start - em->orig_start))) {
+		test_msg("Unexpected block start, wanted %llu, have %llu\n",
+			 disk_bytenr + (em->start - em->orig_start),
+			 em->block_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* Now for the compressed extent */
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 8192) {
+		test_msg("Unexpected extent wanted start %llu len 8192, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != compressed_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 compressed_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n",
+			 em->start, em->orig_start);
+		goto out;
+	}
+	if (em->compress_type != BTRFS_COMPRESS_ZLIB) {
+		test_msg("Unexpected compress type, wanted %d, got %d\n",
+			 BTRFS_COMPRESS_ZLIB, em->compress_type);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* Split compressed extent */
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != compressed_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 compressed_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n",
+			 em->start, em->orig_start);
+		goto out;
+	}
+	if (em->compress_type != BTRFS_COMPRESS_ZLIB) {
+		test_msg("Unexpected compress type, wanted %d, got %d\n",
+			 BTRFS_COMPRESS_ZLIB, em->compress_type);
+		goto out;
+	}
+	disk_bytenr = em->block_start;
+	orig_start = em->start;
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != disk_bytenr) {
+		test_msg("Block start does not match, want %llu got %llu\n",
+			 disk_bytenr, em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 8192) {
+		test_msg("Unexpected extent wanted start %llu len 8192, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != compressed_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 compressed_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != orig_start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n",
+			 em->start, orig_start);
+		goto out;
+	}
+	if (em->compress_type != BTRFS_COMPRESS_ZLIB) {
+		test_msg("Unexpected compress type, wanted %d, got %d\n",
+			 BTRFS_COMPRESS_ZLIB, em->compress_type);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	/* A hole between regular extents but no hole extent */
+	em = btrfs_get_extent(inode, NULL, 0, offset + 6, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096 * 1024, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_HOLE) {
+		test_msg("Expected a hole extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	/*
+	 * Currently we just return a length that we requested rather than the
+	 * length of the actual hole, if this changes we'll have to change this
+	 * test.
+	 */
+	if (em->start != offset || em->len != 12288) {
+		test_msg("Unexpected extent wanted start %llu len 12288, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != vacancy_only) {
+		test_msg("Unexpected flags set, want %lu have %lu\n",
+			 vacancy_only, em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	offset = em->start + em->len;
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, offset, 4096, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != offset || em->len != 4096) {
+		test_msg("Unexpected extent wanted start %llu len 4096, got "
+			 "start %llu len %llu\n", offset, em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, want 0 have %lu\n", em->flags);
+		goto out;
+	}
+	if (em->orig_start != em->start) {
+		test_msg("Wrong orig offset, want %llu, have %llu\n", em->start,
+			 em->orig_start);
+		goto out;
+	}
+	ret = 0;
+out:
+	if (!IS_ERR(em))
+		free_extent_map(em);
+	iput(inode);
+	free_dummy_root(root);
+	return ret;
+}
+
+static int test_hole_first(void)
+{
+	struct inode *inode = NULL;
+	struct btrfs_root *root = NULL;
+	struct extent_map *em = NULL;
+	int ret = -ENOMEM;
+
+	inode = btrfs_new_test_inode();
+	if (!inode) {
+		test_msg("Couldn't allocate inode\n");
+		return ret;
+	}
+
+	BTRFS_I(inode)->location.type = BTRFS_INODE_ITEM_KEY;
+	BTRFS_I(inode)->location.objectid = BTRFS_FIRST_FREE_OBJECTID;
+	BTRFS_I(inode)->location.offset = 0;
+
+	root = btrfs_alloc_dummy_root();
+	if (IS_ERR(root)) {
+		test_msg("Couldn't allocate root\n");
+		goto out;
+	}
+
+	root->fs_info = alloc_dummy_fs_info();
+	if (!root->fs_info) {
+		test_msg("Couldn't allocate dummy fs info\n");
+		goto out;
+	}
+
+	root->node = alloc_dummy_extent_buffer(0, 4096);
+	if (!root->node) {
+		test_msg("Couldn't allocate dummy buffer\n");
+		goto out;
+	}
+
+	extent_buffer_get(root->node);
+	btrfs_set_header_nritems(root->node, 0);
+	btrfs_set_header_level(root->node, 0);
+	BTRFS_I(inode)->root = root;
+	ret = -EINVAL;
+
+	/*
+	 * Need a blank inode item here just so we don't confuse
+	 * btrfs_get_extent.
+	 */
+	insert_inode_item_key(root);
+	insert_extent(root, 4096, 4096, 4096, 0, 4096, 4096,
+		      BTRFS_FILE_EXTENT_REG, 0, 1);
+	em = btrfs_get_extent(inode, NULL, 0, 0, 8192, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != EXTENT_MAP_HOLE) {
+		test_msg("Expected a hole, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != 0 || em->len != 4096) {
+		test_msg("Unexpected extent wanted start 0 len 4096, got start "
+			 "%llu len %llu\n", em->start, em->len);
+		goto out;
+	}
+	if (em->flags != vacancy_only) {
+		test_msg("Wrong flags, wanted %lu, have %lu\n", vacancy_only,
+			 em->flags);
+		goto out;
+	}
+	free_extent_map(em);
+
+	em = btrfs_get_extent(inode, NULL, 0, 4096, 8192, 0);
+	if (IS_ERR(em)) {
+		test_msg("Got an error when we shouldn't have\n");
+		goto out;
+	}
+	if (em->block_start != 4096) {
+		test_msg("Expected a real extent, got %llu\n", em->block_start);
+		goto out;
+	}
+	if (em->start != 4096 || em->len != 4096) {
+		test_msg("Unexpected extent wanted start 4096 len 4096, got "
+			 "start %llu len %llu\n", em->start, em->len);
+		goto out;
+	}
+	if (em->flags != 0) {
+		test_msg("Unexpected flags set, wanted 0 got %lu\n",
+			 em->flags);
+		goto out;
+	}
+	ret = 0;
+out:
+	if (!IS_ERR(em))
+		free_extent_map(em);
+	iput(inode);
+	free_dummy_root(root);
+	return ret;
+}
+
+int btrfs_test_inodes(void)
+{
+	int ret;
+
+	set_bit(EXTENT_FLAG_COMPRESSED, &compressed_only);
+	set_bit(EXTENT_FLAG_VACANCY, &vacancy_only);
+	set_bit(EXTENT_FLAG_PREALLOC, &prealloc_only);
+
+	test_msg("Running btrfs_get_extent tests\n");
+	ret = test_btrfs_get_extent();
+	if (ret)
+		return ret;
+	test_msg("Running hole first btrfs_get_extent test\n");
+	return test_hole_first();
+}

fs/btrfs/transaction.c

@@ -57,7 +57,7 @@ static unsigned int btrfs_blocked_trans_types[TRANS_STATE_MAX] = {
 					   __TRANS_JOIN_NOLOCK),
 };
 
-static void put_transaction(struct btrfs_transaction *transaction)
+void btrfs_put_transaction(struct btrfs_transaction *transaction)
 {
 	WARN_ON(atomic_read(&transaction->use_count) == 0);
 	if (atomic_dec_and_test(&transaction->use_count)) {
@@ -332,7 +332,7 @@ static void wait_current_trans(struct btrfs_root *root)
 		wait_event(root->fs_info->transaction_wait,
 			   cur_trans->state >= TRANS_STATE_UNBLOCKED ||
 			   cur_trans->aborted);
-		put_transaction(cur_trans);
+		btrfs_put_transaction(cur_trans);
 	} else {
 		spin_unlock(&root->fs_info->trans_lock);
 	}
@@ -353,6 +353,17 @@ static int may_wait_transaction(struct btrfs_root *root, int type)
 	return 0;
 }
 
+static inline bool need_reserve_reloc_root(struct btrfs_root *root)
+{
+	if (!root->fs_info->reloc_ctl ||
+	    !root->ref_cows ||
+	    root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
+	    root->reloc_root)
+		return false;
+
+	return true;
+}
+
 static struct btrfs_trans_handle *
 start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 		  enum btrfs_reserve_flush_enum flush)
@@ -360,8 +371,9 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 	struct btrfs_trans_handle *h;
 	struct btrfs_transaction *cur_trans;
 	u64 num_bytes = 0;
-	int ret;
 	u64 qgroup_reserved = 0;
+	bool reloc_reserved = false;
+	int ret;
 
 	if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))
 		return ERR_PTR(-EROFS);
@@ -390,6 +402,14 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 		}
 
 		num_bytes = btrfs_calc_trans_metadata_size(root, num_items);
+		/*
+		 * Do the reservation for the relocation root creation
+		 */
+		if (unlikely(need_reserve_reloc_root(root))) {
+			num_bytes += root->nodesize;
+			reloc_reserved = true;
+		}
+
 		ret = btrfs_block_rsv_add(root,
 					  &root->fs_info->trans_block_rsv,
 					  num_bytes, flush);
@@ -451,6 +471,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 	h->delayed_ref_elem.seq = 0;
 	h->type = type;
 	h->allocating_chunk = false;
+	h->reloc_reserved = false;
 	INIT_LIST_HEAD(&h->qgroup_ref_list);
 	INIT_LIST_HEAD(&h->new_bgs);
 
@@ -466,6 +487,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 					      h->transid, num_bytes, 1);
 		h->block_rsv = &root->fs_info->trans_block_rsv;
 		h->bytes_reserved = num_bytes;
+		h->reloc_reserved = reloc_reserved;
 	}
 	h->qgroup_reserved = qgroup_reserved;
 
@@ -610,7 +632,7 @@ int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid)
 	}
 
 	wait_for_commit(root, cur_trans);
-	put_transaction(cur_trans);
+	btrfs_put_transaction(cur_trans);
 out:
 	return ret;
 }
@@ -735,7 +757,7 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	smp_mb();
 	if (waitqueue_active(&cur_trans->writer_wait))
 		wake_up(&cur_trans->writer_wait);
-	put_transaction(cur_trans);
+	btrfs_put_transaction(cur_trans);
 
 	if (current->journal_info == trans)
 		current->journal_info = NULL;
@@ -744,8 +766,10 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 		btrfs_run_delayed_iputs(root);
 
 	if (trans->aborted ||
-	    test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))
+	    test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) {
+		wake_up_process(info->transaction_kthread);
 		err = -EIO;
+	}
 	assert_qgroups_uptodate(trans);
 
 	kmem_cache_free(btrfs_trans_handle_cachep, trans);
@@ -948,16 +972,19 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans,
 		return ret;
 
 	ret = btrfs_run_dev_stats(trans, root->fs_info);
-	WARN_ON(ret);
+	if (ret)
+		return ret;
 	ret = btrfs_run_dev_replace(trans, root->fs_info);
-	WARN_ON(ret);
+	if (ret)
-
+		return ret;
 	ret = btrfs_run_qgroups(trans, root->fs_info);
-	BUG_ON(ret);
+	if (ret)
+		return ret;
 
 	/* run_qgroups might have added some more refs */
 	ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
-	BUG_ON(ret);
+	if (ret)
+		return ret;
 
 	while (!list_empty(&fs_info->dirty_cowonly_roots)) {
 		next = fs_info->dirty_cowonly_roots.next;
@@ -1510,7 +1537,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans,
 	if (current->journal_info == trans)
 		current->journal_info = NULL;
 
-	put_transaction(cur_trans);
+	btrfs_put_transaction(cur_trans);
 	return 0;
 }
 
@@ -1552,8 +1579,10 @@ static void cleanup_transaction(struct btrfs_trans_handle *trans,
 		root->fs_info->running_transaction = NULL;
 	spin_unlock(&root->fs_info->trans_lock);
 
-	put_transaction(cur_trans);
+	if (trans->type & __TRANS_FREEZABLE)
-	put_transaction(cur_trans);
+		sb_end_intwrite(root->fs_info->sb);
+	btrfs_put_transaction(cur_trans);
+	btrfs_put_transaction(cur_trans);
 
 	trace_btrfs_transaction_commit(root);
 
@@ -1571,15 +1600,19 @@ static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans,
 	int ret;
 
 	ret = btrfs_run_delayed_items(trans, root);
-	if (ret)
-		return ret;
-
 	/*
 	 * running the delayed items may have added new refs. account
 	 * them now so that they hinder processing of more delayed refs
 	 * as little as possible.
 	 */
-	btrfs_delayed_refs_qgroup_accounting(trans, root->fs_info);
+	if (ret) {
+		btrfs_delayed_refs_qgroup_accounting(trans, root->fs_info);
+		return ret;
+	}
+
+	ret = btrfs_delayed_refs_qgroup_accounting(trans, root->fs_info);
+	if (ret)
+		return ret;
 
 	/*
 	 * rename don't use btrfs_join_transaction, so, once we
@@ -1596,14 +1629,14 @@ static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans,
 static inline int btrfs_start_delalloc_flush(struct btrfs_fs_info *fs_info)
 {
 	if (btrfs_test_opt(fs_info->tree_root, FLUSHONCOMMIT))
-		return btrfs_start_all_delalloc_inodes(fs_info, 1);
+		return btrfs_start_delalloc_roots(fs_info, 1);
 	return 0;
 }
 
 static inline void btrfs_wait_delalloc_flush(struct btrfs_fs_info *fs_info)
 {
 	if (btrfs_test_opt(fs_info->tree_root, FLUSHONCOMMIT))
-		btrfs_wait_all_ordered_extents(fs_info);
+		btrfs_wait_ordered_roots(fs_info, -1);
 }
 
 int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
@@ -1669,7 +1702,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 		wait_for_commit(root, cur_trans);
 
-		put_transaction(cur_trans);
+		btrfs_put_transaction(cur_trans);
 
 		return ret;
 	}
@@ -1686,7 +1719,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 			wait_for_commit(root, prev_trans);
 
-			put_transaction(prev_trans);
+			btrfs_put_transaction(prev_trans);
 		} else {
 			spin_unlock(&root->fs_info->trans_lock);
 		}
@@ -1885,8 +1918,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	list_del_init(&cur_trans->list);
 	spin_unlock(&root->fs_info->trans_lock);
 
-	put_transaction(cur_trans);
+	btrfs_put_transaction(cur_trans);
-	put_transaction(cur_trans);
+	btrfs_put_transaction(cur_trans);
 
 	if (trans->type & __TRANS_FREEZABLE)
 		sb_end_intwrite(root->fs_info->sb);

fs/btrfs/transaction.h

@@ -92,6 +92,7 @@ struct btrfs_trans_handle {
 	short aborted;
 	short adding_csums;
 	bool allocating_chunk;
+	bool reloc_reserved;
 	unsigned int type;
 	/*
 	 * this root is only needed to validate that the root passed to
@@ -166,4 +167,5 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_transaction_blocked(struct btrfs_fs_info *info);
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info);
+void btrfs_put_transaction(struct btrfs_transaction *transaction);
 #endif

fs/btrfs/tree-defrag.c

@@ -37,7 +37,6 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
 	int ret = 0;
 	int wret;
 	int level;
-	int is_extent = 0;
 	int next_key_ret = 0;
 	u64 last_ret = 0;
 	u64 min_trans = 0;
@@ -50,7 +49,7 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
 		goto out;
 	}
 
-	if (root->ref_cows == 0 && !is_extent)
+	if (root->ref_cows == 0)
 		goto out;
 
 	if (btrfs_test_opt(root, SSD))
@@ -85,7 +84,7 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
 
 	path->keep_locks = 1;
 
-	ret = btrfs_search_forward(root, &key, NULL, path, min_trans);
+	ret = btrfs_search_forward(root, &key, path, min_trans);
 	if (ret < 0)
 		goto out;
 	if (ret > 0) {

fs/btrfs/tree-log.c

@@ -26,7 +26,6 @@
 #include "locking.h"
 #include "print-tree.h"
 #include "backref.h"
-#include "compat.h"
 #include "tree-log.h"
 #include "hash.h"
 
@@ -936,7 +935,7 @@ static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
 					    parent_objectid,
 					    victim_name,
 					    victim_name_len)) {
-				btrfs_inc_nlink(inode);
+				inc_nlink(inode);
 				btrfs_release_path(path);
 
 				ret = btrfs_unlink_inode(trans, root, dir,
@@ -1006,7 +1005,7 @@ static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
 				victim_parent = read_one_inode(root,
 							       parent_objectid);
 				if (victim_parent) {
-					btrfs_inc_nlink(inode);
+					inc_nlink(inode);
 					btrfs_release_path(path);
 
 					ret = btrfs_unlink_inode(trans, root,
@@ -1113,11 +1112,11 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 				  struct extent_buffer *eb, int slot,
 				  struct btrfs_key *key)
 {
-	struct inode *dir;
+	struct inode *dir = NULL;
-	struct inode *inode;
+	struct inode *inode = NULL;
 	unsigned long ref_ptr;
 	unsigned long ref_end;
-	char *name;
+	char *name = NULL;
 	int namelen;
 	int ret;
 	int search_done = 0;
@@ -1150,13 +1149,15 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 	 * care of the rest
 	 */
 	dir = read_one_inode(root, parent_objectid);
-	if (!dir)
+	if (!dir) {
-		return -ENOENT;
+		ret = -ENOENT;
+		goto out;
+	}
 
 	inode = read_one_inode(root, inode_objectid);
 	if (!inode) {
-		iput(dir);
+		ret = -EIO;
-		return -EIO;
+		goto out;
 	}
 
 	while (ref_ptr < ref_end) {
@@ -1169,14 +1170,16 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 			 */
 			if (!dir)
 				dir = read_one_inode(root, parent_objectid);
-			if (!dir)
+			if (!dir) {
-				return -ENOENT;
+				ret = -ENOENT;
+				goto out;
+			}
 		} else {
 			ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
 					     &ref_index);
 		}
 		if (ret)
-			return ret;
+			goto out;
 
 		/* if we already have a perfect match, we're done */
 		if (!inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode),
@@ -1196,12 +1199,11 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 						      parent_objectid,
 						      ref_index, name, namelen,
 						      &search_done);
-				if (ret == 1) {
+				if (ret) {
-					ret = 0;
+					if (ret == 1)
+						ret = 0;
 					goto out;
 				}
-				if (ret)
-					goto out;
 			}
 
 			/* insert our name */
@@ -1215,6 +1217,7 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 
 		ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen;
 		kfree(name);
+		name = NULL;
 		if (log_ref_ver) {
 			iput(dir);
 			dir = NULL;
@@ -1225,6 +1228,7 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 	ret = overwrite_item(trans, root, path, eb, slot, key);
 out:
 	btrfs_release_path(path);
+	kfree(name);
 	iput(dir);
 	iput(inode);
 	return ret;
@@ -1307,6 +1311,7 @@ static int count_inode_refs(struct btrfs_root *root,
 				break;
 			path->slots[0]--;
 		}
+process_slot:
 		btrfs_item_key_to_cpu(path->nodes[0], &key,
 				      path->slots[0]);
 		if (key.objectid != ino ||
@@ -1327,6 +1332,10 @@ static int count_inode_refs(struct btrfs_root *root,
 
 		if (key.offset == 0)
 			break;
+		if (path->slots[0] > 0) {
+			path->slots[0]--;
+			goto process_slot;
+		}
 		key.offset--;
 		btrfs_release_path(path);
 	}
@@ -1480,7 +1489,7 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
 		if (!inode->i_nlink)
 			set_nlink(inode, 1);
 		else
-			btrfs_inc_nlink(inode);
+			inc_nlink(inode);
 		ret = btrfs_update_inode(trans, root, inode);
 	} else if (ret == -EEXIST) {
 		ret = 0;
@@ -1823,7 +1832,7 @@ static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
 						     dir_key->offset,
 						     name, name_len, 0);
 		}
-		if (IS_ERR_OR_NULL(log_di)) {
+		if (!log_di || (IS_ERR(log_di) && PTR_ERR(log_di) == -ENOENT)) {
 			btrfs_dir_item_key_to_cpu(eb, di, &location);
 			btrfs_release_path(path);
 			btrfs_release_path(log_path);
@@ -1841,7 +1850,7 @@ static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
 				goto out;
 			}
 
-			btrfs_inc_nlink(inode);
+			inc_nlink(inode);
 			ret = btrfs_unlink_inode(trans, root, dir, inode,
 						 name, name_len);
 			if (!ret)
@@ -1860,6 +1869,9 @@ static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
 				goto again;
 			ret = 0;
 			goto out;
+		} else if (IS_ERR(log_di)) {
+			kfree(name);
+			return PTR_ERR(log_di);
 		}
 		btrfs_release_path(log_path);
 		kfree(name);
@@ -2118,8 +2130,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 		WARN_ON(*level >= BTRFS_MAX_LEVEL);
 		cur = path->nodes[*level];
 
-		if (btrfs_header_level(cur) != *level)
+		WARN_ON(btrfs_header_level(cur) != *level);
-			WARN_ON(1);
 
 		if (path->slots[*level] >=
 		    btrfs_header_nritems(cur))
@@ -2151,11 +2162,13 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 					return ret;
 				}
 
-				btrfs_tree_lock(next);
+				if (trans) {
-				btrfs_set_lock_blocking(next);
+					btrfs_tree_lock(next);
-				clean_tree_block(trans, root, next);
+					btrfs_set_lock_blocking(next);
-				btrfs_wait_tree_block_writeback(next);
+					clean_tree_block(trans, root, next);
-				btrfs_tree_unlock(next);
+					btrfs_wait_tree_block_writeback(next);
+					btrfs_tree_unlock(next);
+				}
 
 				WARN_ON(root_owner !=
 					BTRFS_TREE_LOG_OBJECTID);
@@ -2227,11 +2240,13 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
 
 				next = path->nodes[*level];
 
-				btrfs_tree_lock(next);
+				if (trans) {
-				btrfs_set_lock_blocking(next);
+					btrfs_tree_lock(next);
-				clean_tree_block(trans, root, next);
+					btrfs_set_lock_blocking(next);
-				btrfs_wait_tree_block_writeback(next);
+					clean_tree_block(trans, root, next);
-				btrfs_tree_unlock(next);
+					btrfs_wait_tree_block_writeback(next);
+					btrfs_tree_unlock(next);
+				}
 
 				WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID);
 				ret = btrfs_free_and_pin_reserved_extent(root,
@@ -2301,11 +2316,13 @@ static int walk_log_tree(struct btrfs_trans_handle *trans,
 
 			next = path->nodes[orig_level];
 
-			btrfs_tree_lock(next);
+			if (trans) {
-			btrfs_set_lock_blocking(next);
+				btrfs_tree_lock(next);
-			clean_tree_block(trans, log, next);
+				btrfs_set_lock_blocking(next);
-			btrfs_wait_tree_block_writeback(next);
+				clean_tree_block(trans, log, next);
-			btrfs_tree_unlock(next);
+				btrfs_wait_tree_block_writeback(next);
+				btrfs_tree_unlock(next);
+			}
 
 			WARN_ON(log->root_key.objectid !=
 				BTRFS_TREE_LOG_OBJECTID);
@@ -2571,9 +2588,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	 * the running transaction open, so a full commit can't hop
 	 * in and cause problems either.
 	 */
-	btrfs_scrub_pause_super(root);
 	ret = write_ctree_super(trans, root->fs_info->tree_root, 1);
-	btrfs_scrub_continue_super(root);
 	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
 		goto out_wake_log_root;
@@ -2608,13 +2623,10 @@ static void free_log_tree(struct btrfs_trans_handle *trans,
 		.process_func = process_one_buffer
 	};
 
-	if (trans) {
+	ret = walk_log_tree(trans, log, &wc);
-		ret = walk_log_tree(trans, log, &wc);
+	/* I don't think this can happen but just in case */
-
+	if (ret)
-		/* I don't think this can happen but just in case */
+		btrfs_abort_transaction(trans, log, ret);
-		if (ret)
-			btrfs_abort_transaction(trans, log, ret);
-	}
 
 	while (1) {
 		ret = find_first_extent_bit(&log->dirty_log_pages,
@@ -2867,7 +2879,6 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 			  u64 min_offset, u64 *last_offset_ret)
 {
 	struct btrfs_key min_key;
-	struct btrfs_key max_key;
 	struct btrfs_root *log = root->log_root;
 	struct extent_buffer *src;
 	int err = 0;
@@ -2879,9 +2890,6 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 	u64 ino = btrfs_ino(inode);
 
 	log = root->log_root;
-	max_key.objectid = ino;
-	max_key.offset = (u64)-1;
-	max_key.type = key_type;
 
 	min_key.objectid = ino;
 	min_key.type = key_type;
@@ -2889,8 +2897,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 
 	path->keep_locks = 1;
 
-	ret = btrfs_search_forward(root, &min_key, &max_key,
+	ret = btrfs_search_forward(root, &min_key, path, trans->transid);
-				   path, trans->transid);
 
 	/*
 	 * we didn't find anything from this transaction, see if there
@@ -2943,10 +2950,8 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 
 	/* find the first key from this transaction again */
 	ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
-	if (ret != 0) {
+	if (WARN_ON(ret != 0))
-		WARN_ON(1);
 		goto done;
-	}
 
 	/*
 	 * we have a block from this transaction, log every item in it
@@ -3172,11 +3177,10 @@ static int log_inode_item(struct btrfs_trans_handle *trans,
 			  struct inode *inode)
 {
 	struct btrfs_inode_item *inode_item;
-	struct btrfs_key key;
 	int ret;
 
-	memcpy(&key, &BTRFS_I(inode)->location, sizeof(key));
+	ret = btrfs_insert_empty_item(trans, log, path,
-	ret = btrfs_insert_empty_item(trans, log, path, &key,
+				      &BTRFS_I(inode)->location,
 				      sizeof(*inode_item));
 	if (ret && ret != -EEXIST)
 		return ret;
@@ -3375,7 +3379,7 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 		btrfs_set_token_file_extent_type(leaf, fi,
 						 BTRFS_FILE_EXTENT_REG,
 						 &token);
-		if (em->block_start == 0)
+		if (em->block_start == EXTENT_MAP_HOLE)
 			skip_csum = true;
 	}
 
@@ -3417,11 +3421,6 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 	if (skip_csum)
 		return 0;
 
-	if (em->compress_type) {
-		csum_offset = 0;
-		csum_len = block_len;
-	}
-
 	/*
 	 * First check and see if our csums are on our outstanding ordered
 	 * extents.
@@ -3505,8 +3504,13 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 	if (!mod_len || ret)
 		return ret;
 
-	csum_offset = mod_start - em->start;
+	if (em->compress_type) {
-	csum_len = mod_len;
+		csum_offset = 0;
+		csum_len = block_len;
+	} else {
+		csum_offset = mod_start - em->start;
+		csum_len = mod_len;
+	}
 
 	/* block start is already adjusted for the file extent offset. */
 	ret = btrfs_lookup_csums_range(log->fs_info->csum_root,
@@ -3719,7 +3723,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 
 	while (1) {
 		ins_nr = 0;
-		ret = btrfs_search_forward(root, &min_key, &max_key,
+		ret = btrfs_search_forward(root, &min_key,
 					   path, trans->transid);
 		if (ret != 0)
 			break;
@@ -3769,14 +3773,14 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 		}
 		btrfs_release_path(path);
 
-		if (min_key.offset < (u64)-1)
+		if (min_key.offset < (u64)-1) {
 			min_key.offset++;
-		else if (min_key.type < (u8)-1)
+		} else if (min_key.type < max_key.type) {
 			min_key.type++;
-		else if (min_key.objectid < (u64)-1)
+			min_key.offset = 0;
-			min_key.objectid++;
+		} else {
-		else
 			break;
+		}
 	}
 	if (ins_nr) {
 		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,

fs/btrfs/uuid-tree.c

@@ -260,7 +260,6 @@ int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
 {
 	struct btrfs_root *root = fs_info->uuid_root;
 	struct btrfs_key key;
-	struct btrfs_key max_key;
 	struct btrfs_path *path;
 	int ret = 0;
 	struct extent_buffer *leaf;
@@ -277,13 +276,10 @@ int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
 	key.objectid = 0;
 	key.type = 0;
 	key.offset = 0;
-	max_key.objectid = (u64)-1;
-	max_key.type = (u8)-1;
-	max_key.offset = (u64)-1;
 
 again_search_slot:
 	path->keep_locks = 1;
-	ret = btrfs_search_forward(root, &key, &max_key, path, 0);
+	ret = btrfs_search_forward(root, &key, path, 0);
 	if (ret) {
 		if (ret > 0)
 			ret = 0;

fs/btrfs/volumes.c

@@ -28,7 +28,6 @@
 #include <linux/raid/pq.h>
 #include <linux/semaphore.h>
 #include <asm/div64.h>
-#include "compat.h"
 #include "ctree.h"
 #include "extent_map.h"
 #include "disk-io.h"
@@ -666,7 +665,8 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 		if (device->bdev)
 			fs_devices->open_devices--;
 
-		if (device->writeable && !device->is_tgtdev_for_dev_replace) {
+		if (device->writeable &&
+		    device->devid != BTRFS_DEV_REPLACE_DEVID) {
 			list_del_init(&device->dev_alloc_list);
 			fs_devices->rw_devices--;
 		}
@@ -2041,6 +2041,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	device->in_fs_metadata = 1;
 	device->is_tgtdev_for_dev_replace = 0;
 	device->mode = FMODE_EXCL;
+	device->dev_stats_valid = 1;
 	set_blocksize(device->bdev, 4096);
 
 	if (seeding_dev) {
@@ -2208,6 +2209,7 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 	device->in_fs_metadata = 1;
 	device->is_tgtdev_for_dev_replace = 1;
 	device->mode = FMODE_EXCL;
+	device->dev_stats_valid = 1;
 	set_blocksize(device->bdev, 4096);
 	device->fs_devices = fs_info->fs_devices;
 	list_add(&device->dev_list, &fs_info->fs_devices->devices);
@@ -2550,8 +2552,7 @@ static int btrfs_relocate_sys_chunks(struct btrfs_root *root)
 		failed = 0;
 		retried = true;
 		goto again;
-	} else if (failed && retried) {
+	} else if (WARN_ON(failed && retried)) {
-		WARN_ON(1);
 		ret = -ENOSPC;
 	}
 error:
@@ -3423,6 +3424,9 @@ int btrfs_pause_balance(struct btrfs_fs_info *fs_info)
 
 int btrfs_cancel_balance(struct btrfs_fs_info *fs_info)
 {
+	if (fs_info->sb->s_flags & MS_RDONLY)
+		return -EROFS;
+
 	mutex_lock(&fs_info->balance_mutex);
 	if (!fs_info->balance_ctl) {
 		mutex_unlock(&fs_info->balance_mutex);
@@ -3488,7 +3492,7 @@ static int btrfs_uuid_scan_kthread(void *data)
 	path->keep_locks = 1;
 
 	while (1) {
-		ret = btrfs_search_forward(root, &key, &max_key, path, 0);
+		ret = btrfs_search_forward(root, &key, path, 0);
 		if (ret) {
 			if (ret > 0)
 				ret = 0;
@@ -4488,6 +4492,7 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len)
 		btrfs_crit(fs_info, "Invalid mapping for %Lu-%Lu, got "
 			    "%Lu-%Lu\n", logical, logical+len, em->start,
 			    em->start + em->len);
+		free_extent_map(em);
 		return 1;
 	}
 
@@ -4668,6 +4673,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 		btrfs_crit(fs_info, "found a bad mapping, wanted %Lu, "
 			   "found %Lu-%Lu\n", logical, em->start,
 			   em->start + em->len);
+		free_extent_map(em);
 		return -EINVAL;
 	}
 
@@ -4895,7 +4901,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 			num_stripes = map->num_stripes;
 			max_errors = nr_parity_stripes(map);
 
-			raid_map = kmalloc(sizeof(u64) * num_stripes,
+			raid_map = kmalloc_array(num_stripes, sizeof(u64),
 					   GFP_NOFS);
 			if (!raid_map) {
 				ret = -ENOMEM;
@@ -5395,10 +5401,8 @@ static int bio_size_ok(struct block_device *bdev, struct bio *bio,
 		.bi_rw = bio->bi_rw,
 	};
 
-	if (bio->bi_vcnt == 0) {
+	if (WARN_ON(bio->bi_vcnt == 0))
-		WARN_ON(1);
 		return 1;
-	}
 
 	prev = &bio->bi_io_vec[bio->bi_vcnt - 1];
 	if (bio_sectors(bio) > max_sectors)
@@ -5631,10 +5635,8 @@ struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
 	struct btrfs_device *dev;
 	u64 tmp;
 
-	if (!devid && !fs_info) {
+	if (WARN_ON(!devid && !fs_info))
-		WARN_ON(1);
 		return ERR_PTR(-EINVAL);
-	}
 
 	dev = __alloc_device();
 	if (IS_ERR(dev))

fs/btrfs/volumes.h

@@ -43,9 +43,8 @@ struct btrfs_device {
 	/* WRITE_SYNC bios */
 	struct btrfs_pending_bios pending_sync_bios;
 
-	int running_pending;
 	u64 generation;
-
+	int running_pending;
 	int writeable;
 	int in_fs_metadata;
 	int missing;
@@ -53,11 +52,11 @@ struct btrfs_device {
 	int is_tgtdev_for_dev_replace;
 
 	spinlock_t io_lock;
+	/* the mode sent to blkdev_get */
+	fmode_t mode;
 
 	struct block_device *bdev;
 
-	/* the mode sent to blkdev_get */
-	fmode_t mode;
 
 	struct rcu_string *name;
 
@@ -78,16 +77,21 @@ struct btrfs_device {
 
 	/* optimal io width for this device */
 	u32 io_width;
+	/* type and info about this device */
+	u64 type;
 
 	/* minimal io size for this device */
 	u32 sector_size;
 
-	/* type and info about this device */
-	u64 type;
 
 	/* physical drive uuid (or lvm uuid) */
 	u8 uuid[BTRFS_UUID_SIZE];
 
+	/* for sending down flush barriers */
+	int nobarriers;
+	struct bio *flush_bio;
+	struct completion flush_wait;
+
 	/* per-device scrub information */
 	struct scrub_ctx *scrub_device;
 
@@ -103,10 +107,6 @@ struct btrfs_device {
 	struct radix_tree_root reada_zones;
 	struct radix_tree_root reada_extents;
 
-	/* for sending down flush barriers */
-	struct bio *flush_bio;
-	struct completion flush_wait;
-	int nobarriers;
 
 	/* disk I/O failure stats. For detailed description refer to
 	 * enum btrfs_dev_stat_values in ioctl.h */
@@ -132,7 +132,9 @@ struct btrfs_fs_devices {
 
 	/* all of the devices in the FS, protected by a mutex
 	 * so we can safely walk it to write out the supers without
-	 * worrying about add/remove by the multi-device code
+	 * worrying about add/remove by the multi-device code.
+	 * Scrubbing super can kick off supers writing by holding
+	 * this mutex lock.
 	 */
 	struct mutex device_list_mutex;
 	struct list_head devices;

include/uapi/linux/magic.h

@@ -71,6 +71,6 @@
 #define USBDEVICE_SUPER_MAGIC	0x9fa2
 #define MTD_INODE_FS_MAGIC      0x11307854
 #define ANON_INODE_FS_MAGIC	0x09041934
-
+#define BTRFS_TEST_MAGIC	0x73727279
 
 #endif /* __LINUX_MAGIC_H__ */