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

Pull btrfs updates from Chris Mason:
 "This is against 3.11-rc7, but was pulled and tested against your tree
  as of yesterday.  We do have two small incrementals queued up, but I
  wanted to get this bunch out the door before I hop on an airplane.

  This is a fairly large batch of fixes, performance improvements, and
  cleanups from the usual Btrfs suspects.

  We've included Stefan Behren's work to index subvolume UUIDs, which is
  targeted at speeding up send/receive with many subvolumes or snapshots
  in place.  It closes a long standing performance issue that was built
  in to the disk format.

  Mark Fasheh's offline dedup work is also here.  In this case offline
  means the FS is mounted and active, but the dedup work is not done
  inline during file IO.  This is a building block where utilities are
  able to ask the FS to dedup a series of extents.  The kernel takes
  care of verifying the data involved really is the same.  Today this
  involves reading both extents, but we'll continue to evolve the
  patches"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (118 commits)
  Btrfs: optimize key searches in btrfs_search_slot
  Btrfs: don't use an async starter for most of our workers
  Btrfs: only update disk_i_size as we remove extents
  Btrfs: fix deadlock in uuid scan kthread
  Btrfs: stop refusing the relocation of chunk 0
  Btrfs: fix memory leak of uuid_root in free_fs_info
  btrfs: reuse kbasename helper
  btrfs: return btrfs error code for dev excl ops err
  Btrfs: allow partial ordered extent completion
  Btrfs: convert all bug_ons in free-space-cache.c
  Btrfs: add support for asserts
  Btrfs: adjust the fs_devices->missing count on unmount
  Btrf: cleanup: don't check for root_refs == 0 twice
  Btrfs: fix for patch "cleanup: don't check the same thing twice"
  Btrfs: get rid of one BUG() in write_all_supers()
  Btrfs: allocate prelim_ref with a slab allocater
  Btrfs: pass gfp_t to __add_prelim_ref() to avoid always using GFP_ATOMIC
  Btrfs: fix race conditions in BTRFS_IOC_FS_INFO ioctl
  Btrfs: fix race between removing a dev and writing sbs
  Btrfs: remove ourselves from the cluster list under lock
  ...

fs/btrfs/Kconfig

@@ -72,3 +72,12 @@ config BTRFS_DEBUG
 	  performance, or export extra information via sysfs.
 
 	  If unsure, say N.
+
+config BTRFS_ASSERT
+	bool "Btrfs assert support"
+	depends on BTRFS_FS
+	help
+	  Enable run-time assertion checking.  This will result in panics if
+	  any of the assertions trip.  This is meant for btrfs developers only.
+
+	  If unsure, say N.

fs/btrfs/Makefile

@@ -8,7 +8,10 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \
 	   extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.o \
 	   export.o tree-log.o free-space-cache.o zlib.o lzo.o \
 	   compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o \
-	   reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o
+	   reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o \
+	   uuid-tree.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

fs/btrfs/backref.c

@@ -119,6 +119,26 @@ struct __prelim_ref {
 	u64 wanted_disk_byte;
 };
 
+static struct kmem_cache *btrfs_prelim_ref_cache;
+
+int __init btrfs_prelim_ref_init(void)
+{
+	btrfs_prelim_ref_cache = kmem_cache_create("btrfs_prelim_ref",
+					sizeof(struct __prelim_ref),
+					0,
+					SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD,
+					NULL);
+	if (!btrfs_prelim_ref_cache)
+		return -ENOMEM;
+	return 0;
+}
+
+void btrfs_prelim_ref_exit(void)
+{
+	if (btrfs_prelim_ref_cache)
+		kmem_cache_destroy(btrfs_prelim_ref_cache);
+}
+
 /*
  * the rules for all callers of this function are:
  * - obtaining the parent is the goal
@@ -160,12 +180,12 @@ struct __prelim_ref {
 
 static int __add_prelim_ref(struct list_head *head, u64 root_id,
 			    struct btrfs_key *key, int level,
-			    u64 parent, u64 wanted_disk_byte, int count)
+			    u64 parent, u64 wanted_disk_byte, int count,
+			    gfp_t gfp_mask)
 {
 	struct __prelim_ref *ref;
 
-	/* in case we're adding delayed refs, we're holding the refs spinlock */
-	ref = kmalloc(sizeof(*ref), GFP_ATOMIC);
+	ref = kmem_cache_alloc(btrfs_prelim_ref_cache, gfp_mask);
 	if (!ref)
 		return -ENOMEM;
 
@@ -295,10 +315,9 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
 	ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq);
 	pr_debug("search slot in root %llu (level %d, ref count %d) returned "
 		 "%d for key (%llu %u %llu)\n",
-		 (unsigned long long)ref->root_id, level, ref->count, ret,
-		 (unsigned long long)ref->key_for_search.objectid,
-		 ref->key_for_search.type,
-		 (unsigned long long)ref->key_for_search.offset);
+		 ref->root_id, level, ref->count, ret,
+		 ref->key_for_search.objectid, ref->key_for_search.type,
+		 ref->key_for_search.offset);
 	if (ret < 0)
 		goto out;
 
@@ -365,11 +384,12 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 		node = ulist_next(parents, &uiter);
 		ref->parent = node ? node->val : 0;
 		ref->inode_list = node ?
-			(struct extent_inode_elem *)(uintptr_t)node->aux : 0;
+			(struct extent_inode_elem *)(uintptr_t)node->aux : NULL;
 
 		/* additional parents require new refs being added here */
 		while ((node = ulist_next(parents, &uiter))) {
-			new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS);
+			new_ref = kmem_cache_alloc(btrfs_prelim_ref_cache,
+						   GFP_NOFS);
 			if (!new_ref) {
 				ret = -ENOMEM;
 				goto out;
@@ -493,7 +513,7 @@ static void __merge_refs(struct list_head *head, int mode)
 			ref1->count += ref2->count;
 
 			list_del(&ref2->list);
-			kfree(ref2);
+			kmem_cache_free(btrfs_prelim_ref_cache, ref2);
 		}
 
 	}
@@ -548,7 +568,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			ref = btrfs_delayed_node_to_tree_ref(node);
 			ret = __add_prelim_ref(prefs, ref->root, &op_key,
 					       ref->level + 1, 0, node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		case BTRFS_SHARED_BLOCK_REF_KEY: {
@@ -558,7 +578,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			ret = __add_prelim_ref(prefs, ref->root, NULL,
 					       ref->level + 1, ref->parent,
 					       node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		case BTRFS_EXTENT_DATA_REF_KEY: {
@@ -570,7 +590,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			key.offset = ref->offset;
 			ret = __add_prelim_ref(prefs, ref->root, &key, 0, 0,
 					       node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		case BTRFS_SHARED_DATA_REF_KEY: {
@@ -583,7 +603,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			key.offset = ref->offset;
 			ret = __add_prelim_ref(prefs, ref->root, &key, 0,
 					       ref->parent, node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		default:
@@ -657,7 +677,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 		case BTRFS_SHARED_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, 0, NULL,
 						*info_level + 1, offset,
-						bytenr, 1);
+						bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_SHARED_DATA_REF_KEY: {
 			struct btrfs_shared_data_ref *sdref;
@@ -666,13 +686,13 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 			sdref = (struct btrfs_shared_data_ref *)(iref + 1);
 			count = btrfs_shared_data_ref_count(leaf, sdref);
 			ret = __add_prelim_ref(prefs, 0, NULL, 0, offset,
-					       bytenr, count);
+					       bytenr, count, GFP_NOFS);
 			break;
 		}
 		case BTRFS_TREE_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, offset, NULL,
 					       *info_level + 1, 0,
-					       bytenr, 1);
+					       bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_EXTENT_DATA_REF_KEY: {
 			struct btrfs_extent_data_ref *dref;
@@ -687,7 +707,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
 			root = btrfs_extent_data_ref_root(leaf, dref);
 			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
-					       bytenr, count);
+					       bytenr, count, GFP_NOFS);
 			break;
 		}
 		default:
@@ -738,7 +758,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 		case BTRFS_SHARED_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, 0, NULL,
 						info_level + 1, key.offset,
-						bytenr, 1);
+						bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_SHARED_DATA_REF_KEY: {
 			struct btrfs_shared_data_ref *sdref;
@@ -748,13 +768,13 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 					      struct btrfs_shared_data_ref);
 			count = btrfs_shared_data_ref_count(leaf, sdref);
 			ret = __add_prelim_ref(prefs, 0, NULL, 0, key.offset,
-						bytenr, count);
+						bytenr, count, GFP_NOFS);
 			break;
 		}
 		case BTRFS_TREE_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, key.offset, NULL,
 					       info_level + 1, 0,
-					       bytenr, 1);
+					       bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_EXTENT_DATA_REF_KEY: {
 			struct btrfs_extent_data_ref *dref;
@@ -770,7 +790,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
 			root = btrfs_extent_data_ref_root(leaf, dref);
 			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
-					       bytenr, count);
+					       bytenr, count, GFP_NOFS);
 			break;
 		}
 		default:
@@ -911,7 +931,6 @@ again:
 
 	while (!list_empty(&prefs)) {
 		ref = list_first_entry(&prefs, struct __prelim_ref, list);
-		list_del(&ref->list);
 		WARN_ON(ref->count < 0);
 		if (ref->count && ref->root_id && ref->parent == 0) {
 			/* no parent == root of tree */
@@ -935,8 +954,10 @@ again:
 				}
 				ret = find_extent_in_eb(eb, bytenr,
 							*extent_item_pos, &eie);
-				ref->inode_list = eie;
 				free_extent_buffer(eb);
+				if (ret < 0)
+					goto out;
+				ref->inode_list = eie;
 			}
 			ret = ulist_add_merge(refs, ref->parent,
 					      (uintptr_t)ref->inode_list,
@@ -954,7 +975,8 @@ again:
 				eie->next = ref->inode_list;
 			}
 		}
-		kfree(ref);
+		list_del(&ref->list);
+		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
 
 out:
@@ -962,13 +984,13 @@ out:
 	while (!list_empty(&prefs)) {
 		ref = list_first_entry(&prefs, struct __prelim_ref, list);
 		list_del(&ref->list);
-		kfree(ref);
+		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
 	while (!list_empty(&prefs_delayed)) {
 		ref = list_first_entry(&prefs_delayed, struct __prelim_ref,
 				       list);
 		list_del(&ref->list);
-		kfree(ref);
+		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
 
 	return ret;
@@ -1326,8 +1348,7 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 	     found_key->type != BTRFS_METADATA_ITEM_KEY) ||
 	    found_key->objectid > logical ||
 	    found_key->objectid + size <= logical) {
-		pr_debug("logical %llu is not within any extent\n",
-			 (unsigned long long)logical);
+		pr_debug("logical %llu is not within any extent\n", logical);
 		return -ENOENT;
 	}
 
@@ -1340,11 +1361,8 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 
 	pr_debug("logical %llu is at position %llu within the extent (%llu "
 		 "EXTENT_ITEM %llu) flags %#llx size %u\n",
-		 (unsigned long long)logical,
-		 (unsigned long long)(logical - found_key->objectid),
-		 (unsigned long long)found_key->objectid,
-		 (unsigned long long)found_key->offset,
-		 (unsigned long long)flags, item_size);
+		 logical, logical - found_key->objectid, found_key->objectid,
+		 found_key->offset, flags, item_size);
 
 	WARN_ON(!flags_ret);
 	if (flags_ret) {
@@ -1516,7 +1534,7 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
 		while (!ret && (root_node = ulist_next(roots, &root_uiter))) {
 			pr_debug("root %llu references leaf %llu, data list "
 				 "%#llx\n", root_node->val, ref_node->val,
-				 (long long)ref_node->aux);
+				 ref_node->aux);
 			ret = iterate_leaf_refs((struct extent_inode_elem *)
 						(uintptr_t)ref_node->aux,
 						root_node->val,
@@ -1608,9 +1626,8 @@ static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
 			name_len = btrfs_inode_ref_name_len(eb, iref);
 			/* path must be released before calling iterate()! */
 			pr_debug("following ref at offset %u for inode %llu in "
-				 "tree %llu\n", cur,
-				 (unsigned long long)found_key.objectid,
-				 (unsigned long long)fs_root->objectid);
+				 "tree %llu\n", cur, found_key.objectid,
+				 fs_root->objectid);
 			ret = iterate(parent, name_len,
 				      (unsigned long)(iref + 1), eb, ctx);
 			if (ret)

fs/btrfs/backref.h

@@ -72,4 +72,6 @@ int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
 			  struct btrfs_inode_extref **ret_extref,
 			  u64 *found_off);
 
+int __init btrfs_prelim_ref_init(void);
+void btrfs_prelim_ref_exit(void);
 #endif

fs/btrfs/btrfs_inode.h

@@ -218,6 +218,27 @@ static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
 	return 0;
 }
 
+struct btrfs_dio_private {
+	struct inode *inode;
+	u64 logical_offset;
+	u64 disk_bytenr;
+	u64 bytes;
+	void *private;
+
+	/* number of bios pending for this dio */
+	atomic_t pending_bios;
+
+	/* IO errors */
+	int errors;
+
+	/* orig_bio is our btrfs_io_bio */
+	struct bio *orig_bio;
+
+	/* dio_bio came from fs/direct-io.c */
+	struct bio *dio_bio;
+	u8 csum[0];
+};
+
 /*
  * Disable DIO read nolock optimization, so new dio readers will be forced
  * to grab i_mutex. It is used to avoid the endless truncate due to

fs/btrfs/check-integrity.c

@@ -701,15 +701,13 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
 			next_bytenr = btrfs_super_root(selected_super);
 			if (state->print_mask &
 			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
-				printk(KERN_INFO "root@%llu\n",
-				       (unsigned long long)next_bytenr);
+				printk(KERN_INFO "root@%llu\n", next_bytenr);
 			break;
 		case 1:
 			next_bytenr = btrfs_super_chunk_root(selected_super);
 			if (state->print_mask &
 			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
-				printk(KERN_INFO "chunk@%llu\n",
-				       (unsigned long long)next_bytenr);
+				printk(KERN_INFO "chunk@%llu\n", next_bytenr);
 			break;
 		case 2:
 			next_bytenr = btrfs_super_log_root(selected_super);
@@ -717,8 +715,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
 				continue;
 			if (state->print_mask &
 			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
-				printk(KERN_INFO "log@%llu\n",
-				       (unsigned long long)next_bytenr);
+				printk(KERN_INFO "log@%llu\n", next_bytenr);
 			break;
 		}
 
@@ -727,7 +724,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
 				     next_bytenr, state->metablock_size);
 		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
 			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
-			       (unsigned long long)next_bytenr, num_copies);
+			       next_bytenr, num_copies);
 
 		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
 			struct btrfsic_block *next_block;
@@ -742,8 +739,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
 				printk(KERN_INFO "btrfsic:"
 				       " btrfsic_map_block(root @%llu,"
 				       " mirror %d) failed!\n",
-				       (unsigned long long)next_bytenr,
-				       mirror_num);
+				       next_bytenr, mirror_num);
 				kfree(selected_super);
 				return -1;
 			}
@@ -767,7 +763,6 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
 			if (ret < (int)PAGE_CACHE_SIZE) {
 				printk(KERN_INFO
 				       "btrfsic: read @logical %llu failed!\n",
-				       (unsigned long long)
 				       tmp_next_block_ctx.start);
 				btrfsic_release_block_ctx(&tmp_next_block_ctx);
 				kfree(selected_super);
@@ -813,7 +808,7 @@ static int btrfsic_process_superblock_dev_mirror(
 	    (bh->b_data + (dev_bytenr & 4095));
 
 	if (btrfs_super_bytenr(super_tmp) != dev_bytenr ||
-	    super_tmp->magic != cpu_to_le64(BTRFS_MAGIC) ||
+	    btrfs_super_magic(super_tmp) != BTRFS_MAGIC ||
 	    memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE) ||
 	    btrfs_super_nodesize(super_tmp) != state->metablock_size ||
 	    btrfs_super_leafsize(super_tmp) != state->metablock_size ||
@@ -847,10 +842,8 @@ static int btrfsic_process_superblock_dev_mirror(
 			printk_in_rcu(KERN_INFO "New initial S-block (bdev %p, %s)"
 				     " @%llu (%s/%llu/%d)\n",
 				     superblock_bdev,
-				     rcu_str_deref(device->name),
-				     (unsigned long long)dev_bytenr,
-				     dev_state->name,
-				     (unsigned long long)dev_bytenr,
+				     rcu_str_deref(device->name), dev_bytenr,
+				     dev_state->name, dev_bytenr,
 				     superblock_mirror_num);
 		list_add(&superblock_tmp->all_blocks_node,
 			 &state->all_blocks_list);
@@ -880,20 +873,20 @@ static int btrfsic_process_superblock_dev_mirror(
 		tmp_disk_key.offset = 0;
 		switch (pass) {
 		case 0:
-			tmp_disk_key.objectid =
-			    cpu_to_le64(BTRFS_ROOT_TREE_OBJECTID);
+			btrfs_set_disk_key_objectid(&tmp_disk_key,
+						    BTRFS_ROOT_TREE_OBJECTID);
 			additional_string = "initial root ";
 			next_bytenr = btrfs_super_root(super_tmp);
 			break;
 		case 1:
-			tmp_disk_key.objectid =
-			    cpu_to_le64(BTRFS_CHUNK_TREE_OBJECTID);
+			btrfs_set_disk_key_objectid(&tmp_disk_key,
+						    BTRFS_CHUNK_TREE_OBJECTID);
 			additional_string = "initial chunk ";
 			next_bytenr = btrfs_super_chunk_root(super_tmp);
 			break;
 		case 2:
-			tmp_disk_key.objectid =
-			    cpu_to_le64(BTRFS_TREE_LOG_OBJECTID);
+			btrfs_set_disk_key_objectid(&tmp_disk_key,
+						    BTRFS_TREE_LOG_OBJECTID);
 			additional_string = "initial log ";
 			next_bytenr = btrfs_super_log_root(super_tmp);
 			if (0 == next_bytenr)
@@ -906,7 +899,7 @@ static int btrfsic_process_superblock_dev_mirror(
 				     next_bytenr, state->metablock_size);
 		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
 			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
-			       (unsigned long long)next_bytenr, num_copies);
+			       next_bytenr, num_copies);
 		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
 			struct btrfsic_block *next_block;
 			struct btrfsic_block_data_ctx tmp_next_block_ctx;
@@ -918,8 +911,7 @@ static int btrfsic_process_superblock_dev_mirror(
 					      mirror_num)) {
 				printk(KERN_INFO "btrfsic: btrfsic_map_block("
 				       "bytenr @%llu, mirror %d) failed!\n",
-				       (unsigned long long)next_bytenr,
-				       mirror_num);
+				       next_bytenr, mirror_num);
 				brelse(bh);
 				return -1;
 			}
@@ -1003,19 +995,17 @@ continue_with_new_stack_frame:
 		    (struct btrfs_leaf *)sf->hdr;
 
 		if (-1 == sf->i) {
-			sf->nr = le32_to_cpu(leafhdr->header.nritems);
+			sf->nr = btrfs_stack_header_nritems(&leafhdr->header);
 
 			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
 				printk(KERN_INFO
 				       "leaf %llu items %d generation %llu"
 				       " owner %llu\n",
-				       (unsigned long long)
-				       sf->block_ctx->start,
-				       sf->nr,
-				       (unsigned long long)
-				       le64_to_cpu(leafhdr->header.generation),
-				       (unsigned long long)
-				       le64_to_cpu(leafhdr->header.owner));
+				       sf->block_ctx->start, sf->nr,
+				       btrfs_stack_header_generation(
+					       &leafhdr->header),
+				       btrfs_stack_header_owner(
+					       &leafhdr->header));
 		}
 
 continue_with_current_leaf_stack_frame:
@@ -1047,10 +1037,10 @@ leaf_item_out_of_bounce_error:
 						     &disk_item,
 						     disk_item_offset,
 						     sizeof(struct btrfs_item));
-			item_offset = le32_to_cpu(disk_item.offset);
-			item_size = le32_to_cpu(disk_item.size);
+			item_offset = btrfs_stack_item_offset(&disk_item);
+			item_size = btrfs_stack_item_offset(&disk_item);
 			disk_key = &disk_item.key;
-			type = disk_key->type;
+			type = btrfs_disk_key_type(disk_key);
 
 			if (BTRFS_ROOT_ITEM_KEY == type) {
 				struct btrfs_root_item root_item;
@@ -1066,7 +1056,7 @@ leaf_item_out_of_bounce_error:
 					sf->block_ctx, &root_item,
 					root_item_offset,
 					item_size);
-				next_bytenr = le64_to_cpu(root_item.bytenr);
+				next_bytenr = btrfs_root_bytenr(&root_item);
 
 				sf->error =
 				    btrfsic_create_link_to_next_block(
@@ -1081,8 +1071,8 @@ leaf_item_out_of_bounce_error:
 						&sf->num_copies,
 						&sf->mirror_num,
 						disk_key,
-						le64_to_cpu(root_item.
-						generation));
+						btrfs_root_generation(
+						&root_item));
 				if (sf->error)
 					goto one_stack_frame_backwards;
 
@@ -1130,18 +1120,17 @@ leaf_item_out_of_bounce_error:
 		struct btrfs_node *const nodehdr = (struct btrfs_node *)sf->hdr;
 
 		if (-1 == sf->i) {
-			sf->nr = le32_to_cpu(nodehdr->header.nritems);
+			sf->nr = btrfs_stack_header_nritems(&nodehdr->header);
 
 			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
 				printk(KERN_INFO "node %llu level %d items %d"
 				       " generation %llu owner %llu\n",
-				       (unsigned long long)
 				       sf->block_ctx->start,
 				       nodehdr->header.level, sf->nr,
-				       (unsigned long long)
-				       le64_to_cpu(nodehdr->header.generation),
-				       (unsigned long long)
-				       le64_to_cpu(nodehdr->header.owner));
+				       btrfs_stack_header_generation(
+				       &nodehdr->header),
+				       btrfs_stack_header_owner(
+				       &nodehdr->header));
 		}
 
 continue_with_current_node_stack_frame:
@@ -1168,7 +1157,7 @@ continue_with_current_node_stack_frame:
 			btrfsic_read_from_block_data(
 				sf->block_ctx, &key_ptr, key_ptr_offset,
 				sizeof(struct btrfs_key_ptr));
-			next_bytenr = le64_to_cpu(key_ptr.blockptr);
+			next_bytenr = btrfs_stack_key_blockptr(&key_ptr);
 
 			sf->error = btrfsic_create_link_to_next_block(
 					state,
@@ -1182,7 +1171,7 @@ continue_with_current_node_stack_frame:
 					&sf->num_copies,
 					&sf->mirror_num,
 					&key_ptr.key,
-					le64_to_cpu(key_ptr.generation));
+					btrfs_stack_key_generation(&key_ptr));
 			if (sf->error)
 				goto one_stack_frame_backwards;
 
@@ -1247,8 +1236,7 @@ static void btrfsic_read_from_block_data(
 	unsigned long i = (start_offset + offset) >> PAGE_CACHE_SHIFT;
 
 	WARN_ON(offset + len > block_ctx->len);
-	offset_in_page = (start_offset + offset) &
-			 ((unsigned long)PAGE_CACHE_SIZE - 1);
+	offset_in_page = (start_offset + offset) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
 		cur = min(len, ((size_t)PAGE_CACHE_SIZE - offset_in_page));
@@ -1290,7 +1278,7 @@ static int btrfsic_create_link_to_next_block(
 				     next_bytenr, state->metablock_size);
 		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
 			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
-			       (unsigned long long)next_bytenr, *num_copiesp);
+			       next_bytenr, *num_copiesp);
 		*mirror_nump = 1;
 	}
 
@@ -1307,7 +1295,7 @@ static int btrfsic_create_link_to_next_block(
 	if (ret) {
 		printk(KERN_INFO
 		       "btrfsic: btrfsic_map_block(@%llu, mirror=%d) failed!\n",
-		       (unsigned long long)next_bytenr, *mirror_nump);
+		       next_bytenr, *mirror_nump);
 		btrfsic_release_block_ctx(next_block_ctx);
 		*next_blockp = NULL;
 		return -1;
@@ -1335,20 +1323,16 @@ static int btrfsic_create_link_to_next_block(
 			       "Referenced block @%llu (%s/%llu/%d)"
 			       " found in hash table, %c,"
 			       " bytenr mismatch (!= stored %llu).\n",
-			       (unsigned long long)next_bytenr,
-			       next_block_ctx->dev->name,
-			       (unsigned long long)next_block_ctx->dev_bytenr,
-			       *mirror_nump,
+			       next_bytenr, next_block_ctx->dev->name,
+			       next_block_ctx->dev_bytenr, *mirror_nump,
 			       btrfsic_get_block_type(state, next_block),
-			       (unsigned long long)next_block->logical_bytenr);
+			       next_block->logical_bytenr);
 		} else if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
 			printk(KERN_INFO
 			       "Referenced block @%llu (%s/%llu/%d)"
 			       " found in hash table, %c.\n",
-			       (unsigned long long)next_bytenr,
-			       next_block_ctx->dev->name,
-			       (unsigned long long)next_block_ctx->dev_bytenr,
-			       *mirror_nump,
+			       next_bytenr, next_block_ctx->dev->name,
+			       next_block_ctx->dev_bytenr, *mirror_nump,
 			       btrfsic_get_block_type(state, next_block));
 		next_block->logical_bytenr = next_bytenr;
 
@@ -1400,7 +1384,7 @@ static int btrfsic_create_link_to_next_block(
 		if (ret < (int)next_block_ctx->len) {
 			printk(KERN_INFO
 			       "btrfsic: read block @logical %llu failed!\n",
-			       (unsigned long long)next_bytenr);
+			       next_bytenr);
 			btrfsic_release_block_ctx(next_block_ctx);
 			*next_blockp = NULL;
 			return -1;
@@ -1444,12 +1428,12 @@ static int btrfsic_handle_extent_data(
 		file_extent_item_offset,
 		offsetof(struct btrfs_file_extent_item, disk_num_bytes));
 	if (BTRFS_FILE_EXTENT_REG != file_extent_item.type ||
-	    ((u64)0) == le64_to_cpu(file_extent_item.disk_bytenr)) {
+	    btrfs_stack_file_extent_disk_bytenr(&file_extent_item) == 0) {
 		if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
 			printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu\n",
 			       file_extent_item.type,
-			       (unsigned long long)
-			       le64_to_cpu(file_extent_item.disk_bytenr));
+			       btrfs_stack_file_extent_disk_bytenr(
+			       &file_extent_item));
 		return 0;
 	}
 
@@ -1463,20 +1447,19 @@ static int btrfsic_handle_extent_data(
 	btrfsic_read_from_block_data(block_ctx, &file_extent_item,
 				     file_extent_item_offset,
 				     sizeof(struct btrfs_file_extent_item));
-	next_bytenr = le64_to_cpu(file_extent_item.disk_bytenr) +
-		      le64_to_cpu(file_extent_item.offset);
-	generation = le64_to_cpu(file_extent_item.generation);
-	num_bytes = le64_to_cpu(file_extent_item.num_bytes);
-	generation = le64_to_cpu(file_extent_item.generation);
+	next_bytenr = btrfs_stack_file_extent_disk_bytenr(&file_extent_item) +
+		      btrfs_stack_file_extent_offset(&file_extent_item);
+	generation = btrfs_stack_file_extent_generation(&file_extent_item);
+	num_bytes = btrfs_stack_file_extent_num_bytes(&file_extent_item);
+	generation = btrfs_stack_file_extent_generation(&file_extent_item);
 
 	if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
 		printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu,"
 		       " offset = %llu, num_bytes = %llu\n",
 		       file_extent_item.type,
-		       (unsigned long long)
-		       le64_to_cpu(file_extent_item.disk_bytenr),
-		       (unsigned long long)le64_to_cpu(file_extent_item.offset),
-		       (unsigned long long)num_bytes);
+		       btrfs_stack_file_extent_disk_bytenr(&file_extent_item),
+		       btrfs_stack_file_extent_offset(&file_extent_item),
+		       num_bytes);
 	while (num_bytes > 0) {
 		u32 chunk_len;
 		int num_copies;
@@ -1492,7 +1475,7 @@ static int btrfsic_handle_extent_data(
 				     next_bytenr, state->datablock_size);
 		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
 			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
-			       (unsigned long long)next_bytenr, num_copies);
+			       next_bytenr, num_copies);
 		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
 			struct btrfsic_block_data_ctx next_block_ctx;
 			struct btrfsic_block *next_block;
@@ -1504,8 +1487,7 @@ static int btrfsic_handle_extent_data(
 			if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
 				printk(KERN_INFO
 				       "\tdisk_bytenr = %llu, num_bytes %u\n",
-				       (unsigned long long)next_bytenr,
-				       chunk_len);
+				       next_bytenr, chunk_len);
 			ret = btrfsic_map_block(state, next_bytenr,
 						chunk_len, &next_block_ctx,
 						mirror_num);
@@ -1513,8 +1495,7 @@ static int btrfsic_handle_extent_data(
 				printk(KERN_INFO
 				       "btrfsic: btrfsic_map_block(@%llu,"
 				       " mirror=%d) failed!\n",
-				       (unsigned long long)next_bytenr,
-				       mirror_num);
+				       next_bytenr, mirror_num);
 				return -1;
 			}
 
@@ -1543,12 +1524,10 @@ static int btrfsic_handle_extent_data(
 					       " found in hash table, D,"
 					       " bytenr mismatch"
 					       " (!= stored %llu).\n",
-					       (unsigned long long)next_bytenr,
+					       next_bytenr,
 					       next_block_ctx.dev->name,
-					       (unsigned long long)
 					       next_block_ctx.dev_bytenr,
 					       mirror_num,
-					       (unsigned long long)
 					       next_block->logical_bytenr);
 				}
 				next_block->logical_bytenr = next_bytenr;
@@ -1675,7 +1654,7 @@ static int btrfsic_read_block(struct btrfsic_state *state,
 	if (block_ctx->dev_bytenr & ((u64)PAGE_CACHE_SIZE - 1)) {
 		printk(KERN_INFO
 		       "btrfsic: read_block() with unaligned bytenr %llu\n",
-		       (unsigned long long)block_ctx->dev_bytenr);
+		       block_ctx->dev_bytenr);
 		return -1;
 	}
 
@@ -1772,10 +1751,8 @@ static void btrfsic_dump_database(struct btrfsic_state *state)
 
 		printk(KERN_INFO "%c-block @%llu (%s/%llu/%d)\n",
 		       btrfsic_get_block_type(state, b_all),
-		       (unsigned long long)b_all->logical_bytenr,
-		       b_all->dev_state->name,
-		       (unsigned long long)b_all->dev_bytenr,
-		       b_all->mirror_num);
+		       b_all->logical_bytenr, b_all->dev_state->name,
+		       b_all->dev_bytenr, b_all->mirror_num);
 
 		list_for_each(elem_ref_to, &b_all->ref_to_list) {
 			const struct btrfsic_block_link *const l =
@@ -1787,16 +1764,13 @@ static void btrfsic_dump_database(struct btrfsic_state *state)
 			       " refers %u* to"
 			       " %c @%llu (%s/%llu/%d)\n",
 			       btrfsic_get_block_type(state, b_all),
-			       (unsigned long long)b_all->logical_bytenr,
-			       b_all->dev_state->name,
-			       (unsigned long long)b_all->dev_bytenr,
-			       b_all->mirror_num,
+			       b_all->logical_bytenr, b_all->dev_state->name,
+			       b_all->dev_bytenr, b_all->mirror_num,
 			       l->ref_cnt,
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->dev_bytenr,
 			       l->block_ref_to->mirror_num);
 		}
 
@@ -1810,16 +1784,12 @@ static void btrfsic_dump_database(struct btrfsic_state *state)
 			       " is ref %u* from"
 			       " %c @%llu (%s/%llu/%d)\n",
 			       btrfsic_get_block_type(state, b_all),
-			       (unsigned long long)b_all->logical_bytenr,
-			       b_all->dev_state->name,
-			       (unsigned long long)b_all->dev_bytenr,
-			       b_all->mirror_num,
+			       b_all->logical_bytenr, b_all->dev_state->name,
+			       b_all->dev_bytenr, b_all->mirror_num,
 			       l->ref_cnt,
 			       btrfsic_get_block_type(state, l->block_ref_from),
-			       (unsigned long long)
 			       l->block_ref_from->logical_bytenr,
 			       l->block_ref_from->dev_state->name,
-			       (unsigned long long)
 			       l->block_ref_from->dev_bytenr,
 			       l->block_ref_from->mirror_num);
 		}
@@ -1896,8 +1866,8 @@ again:
 		struct list_head *tmp_ref_to;
 
 		if (block->is_superblock) {
-			bytenr = le64_to_cpu(((struct btrfs_super_block *)
-					      mapped_datav[0])->bytenr);
+			bytenr = btrfs_super_bytenr((struct btrfs_super_block *)
+						    mapped_datav[0]);
 			if (num_pages * PAGE_CACHE_SIZE <
 			    BTRFS_SUPER_INFO_SIZE) {
 				printk(KERN_INFO
@@ -1923,8 +1893,9 @@ again:
 					return;
 				}
 				processed_len = state->metablock_size;
-				bytenr = le64_to_cpu(((struct btrfs_header *)
-						      mapped_datav[0])->bytenr);
+				bytenr = btrfs_stack_header_bytenr(
+						(struct btrfs_header *)
+						mapped_datav[0]);
 				btrfsic_cmp_log_and_dev_bytenr(state, bytenr,
 							       dev_state,
 							       dev_bytenr);
@@ -1935,12 +1906,9 @@ again:
 				       " found in hash table, %c,"
 				       " bytenr mismatch"
 				       " (!= stored %llu).\n",
-				       (unsigned long long)bytenr,
-				       dev_state->name,
-				       (unsigned long long)dev_bytenr,
+				       bytenr, dev_state->name, dev_bytenr,
 				       block->mirror_num,
 				       btrfsic_get_block_type(state, block),
-				       (unsigned long long)
 				       block->logical_bytenr);
 				block->logical_bytenr = bytenr;
 			} else if (state->print_mask &
@@ -1948,9 +1916,7 @@ again:
 				printk(KERN_INFO
 				       "Written block @%llu (%s/%llu/%d)"
 				       " found in hash table, %c.\n",
-				       (unsigned long long)bytenr,
-				       dev_state->name,
-				       (unsigned long long)dev_bytenr,
+				       bytenr, dev_state->name, dev_bytenr,
 				       block->mirror_num,
 				       btrfsic_get_block_type(state, block));
 		} else {
@@ -1966,9 +1932,7 @@ again:
 				printk(KERN_INFO
 				       "Written block @%llu (%s/%llu/%d)"
 				       " found in hash table, %c.\n",
-				       (unsigned long long)bytenr,
-				       dev_state->name,
-				       (unsigned long long)dev_bytenr,
+				       bytenr, dev_state->name, dev_bytenr,
 				       block->mirror_num,
 				       btrfsic_get_block_type(state, block));
 		}
@@ -1985,21 +1949,14 @@ again:
 			       " new(gen=%llu),"
 			       " which is referenced by most recent superblock"
 			       " (superblockgen=%llu)!\n",
-			       btrfsic_get_block_type(state, block),
-			       (unsigned long long)bytenr,
-			       dev_state->name,
-			       (unsigned long long)dev_bytenr,
-			       block->mirror_num,
-			       (unsigned long long)block->generation,
-			       (unsigned long long)
-			       le64_to_cpu(block->disk_key.objectid),
+			       btrfsic_get_block_type(state, block), bytenr,
+			       dev_state->name, dev_bytenr, block->mirror_num,
+			       block->generation,
+			       btrfs_disk_key_objectid(&block->disk_key),
 			       block->disk_key.type,
-			       (unsigned long long)
-			       le64_to_cpu(block->disk_key.offset),
-			       (unsigned long long)
-			       le64_to_cpu(((struct btrfs_header *)
-					    mapped_datav[0])->generation),
-			       (unsigned long long)
+			       btrfs_disk_key_offset(&block->disk_key),
+			       btrfs_stack_header_generation(
+				       (struct btrfs_header *) mapped_datav[0]),
 			       state->max_superblock_generation);
 			btrfsic_dump_tree(state);
 		}
@@ -2008,15 +1965,12 @@ again:
 			printk(KERN_INFO "btrfs: attempt to overwrite %c-block"
 			       " @%llu (%s/%llu/%d), oldgen=%llu, newgen=%llu,"
 			       " which is not yet iodone!\n",
-			       btrfsic_get_block_type(state, block),
-			       (unsigned long long)bytenr,
-			       dev_state->name,
-			       (unsigned long long)dev_bytenr,
-			       block->mirror_num,
-			       (unsigned long long)block->generation,
-			       (unsigned long long)
-			       le64_to_cpu(((struct btrfs_header *)
-					    mapped_datav[0])->generation));
+			       btrfsic_get_block_type(state, block), bytenr,
+			       dev_state->name, dev_bytenr, block->mirror_num,
+			       block->generation,
+			       btrfs_stack_header_generation(
+				       (struct btrfs_header *)
+				       mapped_datav[0]));
 			/* it would not be safe to go on */
 			btrfsic_dump_tree(state);
 			goto continue_loop;
@@ -2056,7 +2010,7 @@ again:
 		if (ret) {
 			printk(KERN_INFO
 			       "btrfsic: btrfsic_map_block(root @%llu)"
-			       " failed!\n", (unsigned long long)bytenr);
+			       " failed!\n", bytenr);
 			goto continue_loop;
 		}
 		block_ctx.datav = mapped_datav;
@@ -2140,7 +2094,7 @@ again:
 				printk(KERN_INFO
 				       "btrfsic: btrfsic_process_metablock"
 				       "(root @%llu) failed!\n",
-				       (unsigned long long)dev_bytenr);
+				       dev_bytenr);
 		} else {
 			block->is_metadata = 0;
 			block->mirror_num = 0;	/* unknown */
@@ -2168,8 +2122,7 @@ again:
 			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
 				printk(KERN_INFO "Written block (%s/%llu/?)"
 				       " !found in hash table, D.\n",
-				       dev_state->name,
-				       (unsigned long long)dev_bytenr);
+				       dev_state->name, dev_bytenr);
 			if (!state->include_extent_data) {
 				/* ignore that written D block */
 				goto continue_loop;
@@ -2184,17 +2137,16 @@ again:
 			block_ctx.pagev = NULL;
 		} else {
 			processed_len = state->metablock_size;
-			bytenr = le64_to_cpu(((struct btrfs_header *)
-					      mapped_datav[0])->bytenr);
+			bytenr = btrfs_stack_header_bytenr(
+					(struct btrfs_header *)
+					mapped_datav[0]);
 			btrfsic_cmp_log_and_dev_bytenr(state, bytenr, dev_state,
 						       dev_bytenr);
 			if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
 				printk(KERN_INFO
 				       "Written block @%llu (%s/%llu/?)"
 				       " !found in hash table, M.\n",
-				       (unsigned long long)bytenr,
-				       dev_state->name,
-				       (unsigned long long)dev_bytenr);
+				       bytenr, dev_state->name, dev_bytenr);
 
 			ret = btrfsic_map_block(state, bytenr, processed_len,
 						&block_ctx, 0);
@@ -2202,7 +2154,7 @@ again:
 				printk(KERN_INFO
 				       "btrfsic: btrfsic_map_block(root @%llu)"
 				       " failed!\n",
-				       (unsigned long long)dev_bytenr);
+				       dev_bytenr);
 				goto continue_loop;
 			}
 		}
@@ -2267,10 +2219,8 @@ again:
 			printk(KERN_INFO
 			       "New written %c-block @%llu (%s/%llu/%d)\n",
 			       is_metadata ? 'M' : 'D',
-			       (unsigned long long)block->logical_bytenr,
-			       block->dev_state->name,
-			       (unsigned long long)block->dev_bytenr,
-			       block->mirror_num);
+			       block->logical_bytenr, block->dev_state->name,
+			       block->dev_bytenr, block->mirror_num);
 		list_add(&block->all_blocks_node, &state->all_blocks_list);
 		btrfsic_block_hashtable_add(block, &state->block_hashtable);
 
@@ -2281,7 +2231,7 @@ again:
 				printk(KERN_INFO
 				       "btrfsic: process_metablock(root @%llu)"
 				       " failed!\n",
-				       (unsigned long long)dev_bytenr);
+				       dev_bytenr);
 		}
 		btrfsic_release_block_ctx(&block_ctx);
 	}
@@ -2319,10 +2269,8 @@ static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status)
 			       "bio_end_io(err=%d) for %c @%llu (%s/%llu/%d)\n",
 			       bio_error_status,
 			       btrfsic_get_block_type(dev_state->state, block),
-			       (unsigned long long)block->logical_bytenr,
-			       dev_state->name,
-			       (unsigned long long)block->dev_bytenr,
-			       block->mirror_num);
+			       block->logical_bytenr, dev_state->name,
+			       block->dev_bytenr, block->mirror_num);
 		next_block = block->next_in_same_bio;
 		block->iodone_w_error = iodone_w_error;
 		if (block->submit_bio_bh_rw & REQ_FLUSH) {
@@ -2332,7 +2280,6 @@ static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status)
 				printk(KERN_INFO
 				       "bio_end_io() new %s flush_gen=%llu\n",
 				       dev_state->name,
-				       (unsigned long long)
 				       dev_state->last_flush_gen);
 		}
 		if (block->submit_bio_bh_rw & REQ_FUA)
@@ -2358,10 +2305,8 @@ static void btrfsic_bh_end_io(struct buffer_head *bh, int uptodate)
 		       "bh_end_io(error=%d) for %c @%llu (%s/%llu/%d)\n",
 		       iodone_w_error,
 		       btrfsic_get_block_type(dev_state->state, block),
-		       (unsigned long long)block->logical_bytenr,
-		       block->dev_state->name,
-		       (unsigned long long)block->dev_bytenr,
-		       block->mirror_num);
+		       block->logical_bytenr, block->dev_state->name,
+		       block->dev_bytenr, block->mirror_num);
 
 	block->iodone_w_error = iodone_w_error;
 	if (block->submit_bio_bh_rw & REQ_FLUSH) {
@@ -2370,8 +2315,7 @@ static void btrfsic_bh_end_io(struct buffer_head *bh, int uptodate)
 		     BTRFSIC_PRINT_MASK_END_IO_BIO_BH))
 			printk(KERN_INFO
 			       "bh_end_io() new %s flush_gen=%llu\n",
-			       dev_state->name,
-			       (unsigned long long)dev_state->last_flush_gen);
+			       dev_state->name, dev_state->last_flush_gen);
 	}
 	if (block->submit_bio_bh_rw & REQ_FUA)
 		block->flush_gen = 0; /* FUA completed means block is on disk */
@@ -2396,26 +2340,20 @@ static int btrfsic_process_written_superblock(
 			printk(KERN_INFO
 			       "btrfsic: superblock @%llu (%s/%llu/%d)"
 			       " with old gen %llu <= %llu\n",
-			       (unsigned long long)superblock->logical_bytenr,
+			       superblock->logical_bytenr,
 			       superblock->dev_state->name,
-			       (unsigned long long)superblock->dev_bytenr,
-			       superblock->mirror_num,
-			       (unsigned long long)
+			       superblock->dev_bytenr, superblock->mirror_num,
 			       btrfs_super_generation(super_hdr),
-			       (unsigned long long)
 			       state->max_superblock_generation);
 	} else {
 		if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE)
 			printk(KERN_INFO
 			       "btrfsic: got new superblock @%llu (%s/%llu/%d)"
 			       " with new gen %llu > %llu\n",
-			       (unsigned long long)superblock->logical_bytenr,
+			       superblock->logical_bytenr,
 			       superblock->dev_state->name,
-			       (unsigned long long)superblock->dev_bytenr,
-			       superblock->mirror_num,
-			       (unsigned long long)
+			       superblock->dev_bytenr, superblock->mirror_num,
 			       btrfs_super_generation(super_hdr),
-			       (unsigned long long)
 			       state->max_superblock_generation);
 
 		state->max_superblock_generation =
@@ -2432,43 +2370,41 @@ static int btrfsic_process_written_superblock(
 		int num_copies;
 		int mirror_num;
 		const char *additional_string = NULL;
-		struct btrfs_disk_key tmp_disk_key;
+		struct btrfs_disk_key tmp_disk_key = {0};
 
-		tmp_disk_key.type = BTRFS_ROOT_ITEM_KEY;
-		tmp_disk_key.offset = 0;
+		btrfs_set_disk_key_objectid(&tmp_disk_key,
+					    BTRFS_ROOT_ITEM_KEY);
+		btrfs_set_disk_key_objectid(&tmp_disk_key, 0);
 
 		switch (pass) {
 		case 0:
-			tmp_disk_key.objectid =
-			    cpu_to_le64(BTRFS_ROOT_TREE_OBJECTID);
+			btrfs_set_disk_key_objectid(&tmp_disk_key,
+						    BTRFS_ROOT_TREE_OBJECTID);
 			additional_string = "root ";
 			next_bytenr = btrfs_super_root(super_hdr);
 			if (state->print_mask &
 			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
-				printk(KERN_INFO "root@%llu\n",
-				       (unsigned long long)next_bytenr);
+				printk(KERN_INFO "root@%llu\n", next_bytenr);
 			break;
 		case 1:
-			tmp_disk_key.objectid =
-			    cpu_to_le64(BTRFS_CHUNK_TREE_OBJECTID);
+			btrfs_set_disk_key_objectid(&tmp_disk_key,
+						    BTRFS_CHUNK_TREE_OBJECTID);
 			additional_string = "chunk ";
 			next_bytenr = btrfs_super_chunk_root(super_hdr);
 			if (state->print_mask &
 			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
-				printk(KERN_INFO "chunk@%llu\n",
-				       (unsigned long long)next_bytenr);
+				printk(KERN_INFO "chunk@%llu\n", next_bytenr);
 			break;
 		case 2:
-			tmp_disk_key.objectid =
-			    cpu_to_le64(BTRFS_TREE_LOG_OBJECTID);
+			btrfs_set_disk_key_objectid(&tmp_disk_key,
+						    BTRFS_TREE_LOG_OBJECTID);
 			additional_string = "log ";
 			next_bytenr = btrfs_super_log_root(super_hdr);
 			if (0 == next_bytenr)
 				continue;
 			if (state->print_mask &
 			    BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION)
-				printk(KERN_INFO "log@%llu\n",
-				       (unsigned long long)next_bytenr);
+				printk(KERN_INFO "log@%llu\n", next_bytenr);
 			break;
 		}
 
@@ -2477,7 +2413,7 @@ static int btrfsic_process_written_superblock(
 				     next_bytenr, BTRFS_SUPER_INFO_SIZE);
 		if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
 			printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
-			       (unsigned long long)next_bytenr, num_copies);
+			       next_bytenr, num_copies);
 		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
 			int was_created;
 
@@ -2493,8 +2429,7 @@ static int btrfsic_process_written_superblock(
 				printk(KERN_INFO
 				       "btrfsic: btrfsic_map_block(@%llu,"
 				       " mirror=%d) failed!\n",
-				       (unsigned long long)next_bytenr,
-				       mirror_num);
+				       next_bytenr, mirror_num);
 				return -1;
 			}
 
@@ -2579,26 +2514,22 @@ static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
 			       " %u* refers to %c @%llu (%s/%llu/%d)\n",
 			       recursion_level,
 			       btrfsic_get_block_type(state, block),
-			       (unsigned long long)block->logical_bytenr,
-			       block->dev_state->name,
-			       (unsigned long long)block->dev_bytenr,
-			       block->mirror_num,
+			       block->logical_bytenr, block->dev_state->name,
+			       block->dev_bytenr, block->mirror_num,
 			       l->ref_cnt,
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->dev_bytenr,
 			       l->block_ref_to->mirror_num);
 		if (l->block_ref_to->never_written) {
 			printk(KERN_INFO "btrfs: attempt to write superblock"
 			       " which references block %c @%llu (%s/%llu/%d)"
 			       " which is never written!\n",
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->dev_bytenr,
 			       l->block_ref_to->mirror_num);
 			ret = -1;
 		} else if (!l->block_ref_to->is_iodone) {
@@ -2606,10 +2537,9 @@ static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
 			       " which references block %c @%llu (%s/%llu/%d)"
 			       " which is not yet iodone!\n",
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->dev_bytenr,
 			       l->block_ref_to->mirror_num);
 			ret = -1;
 		} else if (l->block_ref_to->iodone_w_error) {
@@ -2617,10 +2547,9 @@ static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
 			       " which references block %c @%llu (%s/%llu/%d)"
 			       " which has write error!\n",
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->dev_bytenr,
 			       l->block_ref_to->mirror_num);
 			ret = -1;
 		} else if (l->parent_generation !=
@@ -2634,13 +2563,12 @@ static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
 			       " with generation %llu !="
 			       " parent generation %llu!\n",
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->dev_bytenr,
 			       l->block_ref_to->mirror_num,
-			       (unsigned long long)l->block_ref_to->generation,
-			       (unsigned long long)l->parent_generation);
+			       l->block_ref_to->generation,
+			       l->parent_generation);
 			ret = -1;
 		} else if (l->block_ref_to->flush_gen >
 			   l->block_ref_to->dev_state->last_flush_gen) {
@@ -2650,13 +2578,10 @@ static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state,
 			       " (block flush_gen=%llu,"
 			       " dev->flush_gen=%llu)!\n",
 			       btrfsic_get_block_type(state, l->block_ref_to),
-			       (unsigned long long)
 			       l->block_ref_to->logical_bytenr,
 			       l->block_ref_to->dev_state->name,
-			       (unsigned long long)l->block_ref_to->dev_bytenr,
-			       l->block_ref_to->mirror_num,
-			       (unsigned long long)block->flush_gen,
-			       (unsigned long long)
+			       l->block_ref_to->dev_bytenr,
+			       l->block_ref_to->mirror_num, block->flush_gen,
 			       l->block_ref_to->dev_state->last_flush_gen);
 			ret = -1;
 		} else if (-1 == btrfsic_check_all_ref_blocks(state,
@@ -2701,16 +2626,12 @@ static int btrfsic_is_block_ref_by_superblock(
 			       " is ref %u* from %c @%llu (%s/%llu/%d)\n",
 			       recursion_level,
 			       btrfsic_get_block_type(state, block),
-			       (unsigned long long)block->logical_bytenr,
-			       block->dev_state->name,
-			       (unsigned long long)block->dev_bytenr,
-			       block->mirror_num,
+			       block->logical_bytenr, block->dev_state->name,
+			       block->dev_bytenr, block->mirror_num,
 			       l->ref_cnt,
 			       btrfsic_get_block_type(state, l->block_ref_from),
-			       (unsigned long long)
 			       l->block_ref_from->logical_bytenr,
 			       l->block_ref_from->dev_state->name,
-			       (unsigned long long)
 			       l->block_ref_from->dev_bytenr,
 			       l->block_ref_from->mirror_num);
 		if (l->block_ref_from->is_superblock &&
@@ -2737,14 +2658,12 @@ static void btrfsic_print_add_link(const struct btrfsic_state *state,
 	       " to %c @%llu (%s/%llu/%d).\n",
 	       l->ref_cnt,
 	       btrfsic_get_block_type(state, l->block_ref_from),
-	       (unsigned long long)l->block_ref_from->logical_bytenr,
+	       l->block_ref_from->logical_bytenr,
 	       l->block_ref_from->dev_state->name,
-	       (unsigned long long)l->block_ref_from->dev_bytenr,
-	       l->block_ref_from->mirror_num,
+	       l->block_ref_from->dev_bytenr, l->block_ref_from->mirror_num,
 	       btrfsic_get_block_type(state, l->block_ref_to),
-	       (unsigned long long)l->block_ref_to->logical_bytenr,
-	       l->block_ref_to->dev_state->name,
-	       (unsigned long long)l->block_ref_to->dev_bytenr,
+	       l->block_ref_to->logical_bytenr,
+	       l->block_ref_to->dev_state->name, l->block_ref_to->dev_bytenr,
 	       l->block_ref_to->mirror_num);
 }
 
@@ -2756,14 +2675,12 @@ static void btrfsic_print_rem_link(const struct btrfsic_state *state,
 	       " to %c @%llu (%s/%llu/%d).\n",
 	       l->ref_cnt,
 	       btrfsic_get_block_type(state, l->block_ref_from),
-	       (unsigned long long)l->block_ref_from->logical_bytenr,
+	       l->block_ref_from->logical_bytenr,
 	       l->block_ref_from->dev_state->name,
-	       (unsigned long long)l->block_ref_from->dev_bytenr,
-	       l->block_ref_from->mirror_num,
+	       l->block_ref_from->dev_bytenr, l->block_ref_from->mirror_num,
 	       btrfsic_get_block_type(state, l->block_ref_to),
-	       (unsigned long long)l->block_ref_to->logical_bytenr,
-	       l->block_ref_to->dev_state->name,
-	       (unsigned long long)l->block_ref_to->dev_bytenr,
+	       l->block_ref_to->logical_bytenr,
+	       l->block_ref_to->dev_state->name, l->block_ref_to->dev_bytenr,
 	       l->block_ref_to->mirror_num);
 }
 
@@ -2807,10 +2724,8 @@ static void btrfsic_dump_tree_sub(const struct btrfsic_state *state,
 	 */
 	indent_add = sprintf(buf, "%c-%llu(%s/%llu/%d)",
 			     btrfsic_get_block_type(state, block),
-			     (unsigned long long)block->logical_bytenr,
-			     block->dev_state->name,
-			     (unsigned long long)block->dev_bytenr,
-			     block->mirror_num);
+			     block->logical_bytenr, block->dev_state->name,
+			     block->dev_bytenr, block->mirror_num);
 	if (indent_level + indent_add > BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL) {
 		printk("[...]\n");
 		return;
@@ -2943,10 +2858,8 @@ static struct btrfsic_block *btrfsic_block_lookup_or_add(
 			       "New %s%c-block @%llu (%s/%llu/%d)\n",
 			       additional_string,
 			       btrfsic_get_block_type(state, block),
-			       (unsigned long long)block->logical_bytenr,
-			       dev_state->name,
-			       (unsigned long long)block->dev_bytenr,
-			       mirror_num);
+			       block->logical_bytenr, dev_state->name,
+			       block->dev_bytenr, mirror_num);
 		list_add(&block->all_blocks_node, &state->all_blocks_list);
 		btrfsic_block_hashtable_add(block, &state->block_hashtable);
 		if (NULL != was_created)
@@ -2980,7 +2893,7 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
 			printk(KERN_INFO "btrfsic:"
 			       " btrfsic_map_block(logical @%llu,"
 			       " mirror %d) failed!\n",
-			       (unsigned long long)bytenr, mirror_num);
+			       bytenr, mirror_num);
 			continue;
 		}
 
@@ -2997,8 +2910,7 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
 		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",
-		       (unsigned long long)bytenr, dev_state->name,
-		       (unsigned long long)dev_bytenr);
+		       bytenr, dev_state->name, dev_bytenr);
 		for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
 			ret = btrfsic_map_block(state, bytenr,
 						state->metablock_size,
@@ -3008,10 +2920,8 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
 
 			printk(KERN_INFO "Read logical bytenr @%llu maps to"
 			       " (%s/%llu/%d)\n",
-			       (unsigned long long)bytenr,
-			       block_ctx.dev->name,
-			       (unsigned long long)block_ctx.dev_bytenr,
-			       mirror_num);
+			       bytenr, block_ctx.dev->name,
+			       block_ctx.dev_bytenr, mirror_num);
 		}
 		WARN_ON(1);
 	}
@@ -3048,12 +2958,10 @@ int btrfsic_submit_bh(int rw, struct buffer_head *bh)
 		if (dev_state->state->print_mask &
 		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
 			printk(KERN_INFO
-			       "submit_bh(rw=0x%x, blocknr=%lu (bytenr %llu),"
-			       " size=%lu, data=%p, bdev=%p)\n",
-			       rw, (unsigned long)bh->b_blocknr,
-			       (unsigned long long)dev_bytenr,
-			       (unsigned long)bh->b_size, bh->b_data,
-			       bh->b_bdev);
+			       "submit_bh(rw=0x%x, blocknr=%llu (bytenr %llu),"
+			       " size=%zu, data=%p, bdev=%p)\n",
+			       rw, (unsigned long long)bh->b_blocknr,
+			       dev_bytenr, bh->b_size, bh->b_data, bh->b_bdev);
 		btrfsic_process_written_block(dev_state, dev_bytenr,
 					      &bh->b_data, 1, NULL,
 					      NULL, bh, rw);
@@ -3118,9 +3026,9 @@ void btrfsic_submit_bio(int rw, struct bio *bio)
 		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
 			printk(KERN_INFO
 			       "submit_bio(rw=0x%x, bi_vcnt=%u,"
-			       " bi_sector=%lu (bytenr %llu), bi_bdev=%p)\n",
-			       rw, bio->bi_vcnt, (unsigned long)bio->bi_sector,
-			       (unsigned long long)dev_bytenr,
+			       " bi_sector=%llu (bytenr %llu), bi_bdev=%p)\n",
+			       rw, bio->bi_vcnt,
+			       (unsigned long long)bio->bi_sector, dev_bytenr,
 			       bio->bi_bdev);
 
 		mapped_datav = kmalloc(sizeof(*mapped_datav) * bio->bi_vcnt,
@@ -3213,19 +3121,19 @@ int btrfsic_mount(struct btrfs_root *root,
 	if (root->nodesize & ((u64)PAGE_CACHE_SIZE - 1)) {
 		printk(KERN_INFO
 		       "btrfsic: cannot handle nodesize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
-		       root->nodesize, (unsigned long)PAGE_CACHE_SIZE);
+		       root->nodesize, PAGE_CACHE_SIZE);
 		return -1;
 	}
 	if (root->leafsize & ((u64)PAGE_CACHE_SIZE - 1)) {
 		printk(KERN_INFO
 		       "btrfsic: cannot handle leafsize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
-		       root->leafsize, (unsigned long)PAGE_CACHE_SIZE);
+		       root->leafsize, PAGE_CACHE_SIZE);
 		return -1;
 	}
 	if (root->sectorsize & ((u64)PAGE_CACHE_SIZE - 1)) {
 		printk(KERN_INFO
 		       "btrfsic: cannot handle sectorsize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
-		       root->sectorsize, (unsigned long)PAGE_CACHE_SIZE);
+		       root->sectorsize, PAGE_CACHE_SIZE);
 		return -1;
 	}
 	state = kzalloc(sizeof(*state), GFP_NOFS);
@@ -3369,10 +3277,8 @@ void btrfsic_unmount(struct btrfs_root *root,
 			       " @%llu (%s/%llu/%d) on umount which is"
 			       " not yet iodone!\n",
 			       btrfsic_get_block_type(state, b_all),
-			       (unsigned long long)b_all->logical_bytenr,
-			       b_all->dev_state->name,
-			       (unsigned long long)b_all->dev_bytenr,
-			       b_all->mirror_num);
+			       b_all->logical_bytenr, b_all->dev_state->name,
+			       b_all->dev_bytenr, b_all->mirror_num);
 	}
 
 	mutex_unlock(&btrfsic_mutex);

fs/btrfs/compression.c

@@ -132,9 +132,8 @@ static int check_compressed_csum(struct inode *inode,
 			printk(KERN_INFO "btrfs csum failed ino %llu "
 			       "extent %llu csum %u "
 			       "wanted %u mirror %d\n",
-			       (unsigned long long)btrfs_ino(inode),
-			       (unsigned long long)disk_start,
-			       csum, *cb_sum, cb->mirror_num);
+			       btrfs_ino(inode), disk_start, csum, *cb_sum,
+			       cb->mirror_num);
 			ret = -EIO;
 			goto fail;
 		}
@@ -639,7 +638,11 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	faili = nr_pages - 1;
 	cb->nr_pages = nr_pages;
 
-	add_ra_bio_pages(inode, em_start + em_len, cb);
+	/* In the parent-locked case, we only locked the range we are
+	 * interested in.  In all other cases, we can opportunistically
+	 * cache decompressed data that goes beyond the requested range. */
+	if (!(bio_flags & EXTENT_BIO_PARENT_LOCKED))
+		add_ra_bio_pages(inode, em_start + em_len, cb);
 
 	/* include any pages we added in add_ra-bio_pages */
 	uncompressed_len = bio->bi_vcnt * PAGE_CACHE_SIZE;

fs/btrfs/ctree.c

@@ -274,8 +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,
-			    (unsigned long)btrfs_header_fsid(cow),
+	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(cow),
 			    BTRFS_FSID_SIZE);
 
 	WARN_ON(btrfs_header_generation(buf) > trans->transid);
@@ -484,8 +483,27 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 	struct rb_node **new;
 	struct rb_node *parent = NULL;
 	struct tree_mod_elem *cur;
+	int ret = 0;
+
+	BUG_ON(!tm);
+
+	tree_mod_log_write_lock(fs_info);
+	if (list_empty(&fs_info->tree_mod_seq_list)) {
+		tree_mod_log_write_unlock(fs_info);
+		/*
+		 * Ok we no longer care about logging modifications, free up tm
+		 * and return 0.  Any callers shouldn't be using tm after
+		 * calling tree_mod_log_insert, but if they do we can just
+		 * change this to return a special error code to let the callers
+		 * do their own thing.
+		 */
+		kfree(tm);
+		return 0;
+	}
 
-	BUG_ON(!tm || !tm->seq);
+	spin_lock(&fs_info->tree_mod_seq_lock);
+	tm->seq = btrfs_inc_tree_mod_seq_minor(fs_info);
+	spin_unlock(&fs_info->tree_mod_seq_lock);
 
 	tm_root = &fs_info->tree_mod_log;
 	new = &tm_root->rb_node;
@@ -501,14 +519,17 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 		else if (cur->seq > tm->seq)
 			new = &((*new)->rb_right);
 		else {
+			ret = -EEXIST;
 			kfree(tm);
-			return -EEXIST;
+			goto out;
 		}
 	}
 
 	rb_link_node(&tm->node, parent, new);
 	rb_insert_color(&tm->node, tm_root);
-	return 0;
+out:
+	tree_mod_log_write_unlock(fs_info);
+	return ret;
 }
 
 /*
@@ -524,57 +545,19 @@ static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,
 		return 1;
 	if (eb && btrfs_header_level(eb) == 0)
 		return 1;
-
-	tree_mod_log_write_lock(fs_info);
-	if (list_empty(&fs_info->tree_mod_seq_list)) {
-		/*
-		 * someone emptied the list while we were waiting for the lock.
-		 * we must not add to the list when no blocker exists.
-		 */
-		tree_mod_log_write_unlock(fs_info);
-		return 1;
-	}
-
 	return 0;
 }
 
-/*
- * This allocates memory and gets a tree modification sequence number.
- *
- * Returns <0 on error.
- * Returns >0 (the added sequence number) on success.
- */
-static inline int tree_mod_alloc(struct btrfs_fs_info *fs_info, gfp_t flags,
-				 struct tree_mod_elem **tm_ret)
-{
-	struct tree_mod_elem *tm;
-
-	/*
-	 * once we switch from spin locks to something different, we should
-	 * honor the flags parameter here.
-	 */
-	tm = *tm_ret = kzalloc(sizeof(*tm), GFP_ATOMIC);
-	if (!tm)
-		return -ENOMEM;
-
-	spin_lock(&fs_info->tree_mod_seq_lock);
-	tm->seq = btrfs_inc_tree_mod_seq_minor(fs_info);
-	spin_unlock(&fs_info->tree_mod_seq_lock);
-
-	return tm->seq;
-}
-
 static inline int
 __tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 			  struct extent_buffer *eb, int slot,
 			  enum mod_log_op op, gfp_t flags)
 {
-	int ret;
 	struct tree_mod_elem *tm;
 
-	ret = tree_mod_alloc(fs_info, flags, &tm);
-	if (ret < 0)
-		return ret;
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm)
+		return -ENOMEM;
 
 	tm->index = eb->start >> PAGE_CACHE_SHIFT;
 	if (op != MOD_LOG_KEY_ADD) {
@@ -589,34 +572,14 @@ __tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 }
 
 static noinline int
-tree_mod_log_insert_key_mask(struct btrfs_fs_info *fs_info,
-			     struct extent_buffer *eb, int slot,
-			     enum mod_log_op op, gfp_t flags)
+tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
+			struct extent_buffer *eb, int slot,
+			enum mod_log_op op, gfp_t flags)
 {
-	int ret;
-
 	if (tree_mod_dont_log(fs_info, eb))
 		return 0;
 
-	ret = __tree_mod_log_insert_key(fs_info, eb, slot, op, flags);
-
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
-}
-
-static noinline int
-tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
-			int slot, enum mod_log_op op)
-{
-	return tree_mod_log_insert_key_mask(fs_info, eb, slot, op, GFP_NOFS);
-}
-
-static noinline int
-tree_mod_log_insert_key_locked(struct btrfs_fs_info *fs_info,
-			     struct extent_buffer *eb, int slot,
-			     enum mod_log_op op)
-{
-	return __tree_mod_log_insert_key(fs_info, eb, slot, op, GFP_NOFS);
+	return __tree_mod_log_insert_key(fs_info, eb, slot, op, flags);
 }
 
 static noinline int
@@ -637,14 +600,14 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
 	 * buffer, i.e. dst_slot < src_slot.
 	 */
 	for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
-		ret = tree_mod_log_insert_key_locked(fs_info, eb, i + dst_slot,
-					      MOD_LOG_KEY_REMOVE_WHILE_MOVING);
+		ret = __tree_mod_log_insert_key(fs_info, eb, i + dst_slot,
+				MOD_LOG_KEY_REMOVE_WHILE_MOVING, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 
-	ret = tree_mod_alloc(fs_info, flags, &tm);
-	if (ret < 0)
-		goto out;
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm)
+		return -ENOMEM;
 
 	tm->index = eb->start >> PAGE_CACHE_SHIFT;
 	tm->slot = src_slot;
@@ -652,10 +615,7 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
 	tm->move.nr_items = nr_items;
 	tm->op = MOD_LOG_MOVE_KEYS;
 
-	ret = __tree_mod_log_insert(fs_info, tm);
-out:
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
+	return __tree_mod_log_insert(fs_info, tm);
 }
 
 static inline void
@@ -670,8 +630,8 @@ __tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 
 	nritems = btrfs_header_nritems(eb);
 	for (i = nritems - 1; i >= 0; i--) {
-		ret = tree_mod_log_insert_key_locked(fs_info, eb, i,
-					      MOD_LOG_KEY_REMOVE_WHILE_FREEING);
+		ret = __tree_mod_log_insert_key(fs_info, eb, i,
+				MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 }
@@ -683,7 +643,6 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 			 int log_removal)
 {
 	struct tree_mod_elem *tm;
-	int ret;
 
 	if (tree_mod_dont_log(fs_info, NULL))
 		return 0;
@@ -691,9 +650,9 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 	if (log_removal)
 		__tree_mod_log_free_eb(fs_info, old_root);
 
-	ret = tree_mod_alloc(fs_info, flags, &tm);
-	if (ret < 0)
-		goto out;
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm)
+		return -ENOMEM;
 
 	tm->index = new_root->start >> PAGE_CACHE_SHIFT;
 	tm->old_root.logical = old_root->start;
@@ -701,10 +660,7 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 	tm->generation = btrfs_header_generation(old_root);
 	tm->op = MOD_LOG_ROOT_REPLACE;
 
-	ret = __tree_mod_log_insert(fs_info, tm);
-out:
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
+	return __tree_mod_log_insert(fs_info, tm);
 }
 
 static struct tree_mod_elem *
@@ -784,23 +740,20 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
 	if (tree_mod_dont_log(fs_info, NULL))
 		return;
 
-	if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0) {
-		tree_mod_log_write_unlock(fs_info);
+	if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0)
 		return;
-	}
 
 	for (i = 0; i < nr_items; i++) {
-		ret = tree_mod_log_insert_key_locked(fs_info, src,
+		ret = __tree_mod_log_insert_key(fs_info, src,
 						i + src_offset,
-						MOD_LOG_KEY_REMOVE);
+						MOD_LOG_KEY_REMOVE, GFP_NOFS);
 		BUG_ON(ret < 0);
-		ret = tree_mod_log_insert_key_locked(fs_info, dst,
+		ret = __tree_mod_log_insert_key(fs_info, dst,
 						     i + dst_offset,
-						     MOD_LOG_KEY_ADD);
+						     MOD_LOG_KEY_ADD,
+						     GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
-
-	tree_mod_log_write_unlock(fs_info);
 }
 
 static inline void
@@ -819,9 +772,9 @@ tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info,
 {
 	int ret;
 
-	ret = tree_mod_log_insert_key_mask(fs_info, eb, slot,
-					   MOD_LOG_KEY_REPLACE,
-					   atomic ? GFP_ATOMIC : GFP_NOFS);
+	ret = __tree_mod_log_insert_key(fs_info, eb, slot,
+					MOD_LOG_KEY_REPLACE,
+					atomic ? GFP_ATOMIC : GFP_NOFS);
 	BUG_ON(ret < 0);
 }
 
@@ -830,10 +783,7 @@ tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 {
 	if (tree_mod_dont_log(fs_info, eb))
 		return;
-
 	__tree_mod_log_free_eb(fs_info, eb);
-
-	tree_mod_log_write_unlock(fs_info);
 }
 
 static noinline void
@@ -1046,8 +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,
-			    (unsigned long)btrfs_header_fsid(cow),
+	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(cow),
 			    BTRFS_FSID_SIZE);
 
 	ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
@@ -1083,7 +1032,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 
 		WARN_ON(trans->transid != btrfs_header_generation(parent));
 		tree_mod_log_insert_key(root->fs_info, parent, parent_slot,
-					MOD_LOG_KEY_REPLACE);
+					MOD_LOG_KEY_REPLACE, GFP_NOFS);
 		btrfs_set_node_blockptr(parent, parent_slot,
 					cow->start);
 		btrfs_set_node_ptr_generation(parent, parent_slot,
@@ -1116,7 +1065,7 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
 	int looped = 0;
 
 	if (!time_seq)
-		return 0;
+		return NULL;
 
 	/*
 	 * the very last operation that's logged for a root is the replacement
@@ -1127,7 +1076,7 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
 		tm = tree_mod_log_search_oldest(fs_info, root_logical,
 						time_seq);
 		if (!looped && !tm)
-			return 0;
+			return NULL;
 		/*
 		 * if there are no tree operation for the oldest root, we simply
 		 * return it. this should only happen if that (old) root is at
@@ -1240,8 +1189,8 @@ __tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
  * is freed (its refcount is decremented).
  */
 static struct extent_buffer *
-tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
-		    u64 time_seq)
+tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
+		    struct extent_buffer *eb, u64 time_seq)
 {
 	struct extent_buffer *eb_rewin;
 	struct tree_mod_elem *tm;
@@ -1256,11 +1205,18 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 	if (!tm)
 		return eb;
 
+	btrfs_set_path_blocking(path);
+	btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+
 	if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
 		BUG_ON(tm->slot != 0);
 		eb_rewin = alloc_dummy_extent_buffer(eb->start,
 						fs_info->tree_root->nodesize);
-		BUG_ON(!eb_rewin);
+		if (!eb_rewin) {
+			btrfs_tree_read_unlock_blocking(eb);
+			free_extent_buffer(eb);
+			return NULL;
+		}
 		btrfs_set_header_bytenr(eb_rewin, eb->start);
 		btrfs_set_header_backref_rev(eb_rewin,
 					     btrfs_header_backref_rev(eb));
@@ -1268,10 +1224,15 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 		btrfs_set_header_level(eb_rewin, btrfs_header_level(eb));
 	} else {
 		eb_rewin = btrfs_clone_extent_buffer(eb);
-		BUG_ON(!eb_rewin);
+		if (!eb_rewin) {
+			btrfs_tree_read_unlock_blocking(eb);
+			free_extent_buffer(eb);
+			return NULL;
+		}
 	}
 
-	btrfs_tree_read_unlock(eb);
+	btrfs_clear_path_blocking(path, NULL, BTRFS_READ_LOCK);
+	btrfs_tree_read_unlock_blocking(eb);
 	free_extent_buffer(eb);
 
 	extent_buffer_get(eb_rewin);
@@ -1335,8 +1296,9 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 		free_extent_buffer(eb_root);
 		eb = alloc_dummy_extent_buffer(logical, root->nodesize);
 	} else {
+		btrfs_set_lock_blocking_rw(eb_root, BTRFS_READ_LOCK);
 		eb = btrfs_clone_extent_buffer(eb_root);
-		btrfs_tree_read_unlock(eb_root);
+		btrfs_tree_read_unlock_blocking(eb_root);
 		free_extent_buffer(eb_root);
 	}
 
@@ -1419,14 +1381,12 @@ noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
 
 	if (trans->transaction != root->fs_info->running_transaction)
 		WARN(1, KERN_CRIT "trans %llu running %llu\n",
-		       (unsigned long long)trans->transid,
-		       (unsigned long long)
+		       trans->transid,
 		       root->fs_info->running_transaction->transid);
 
 	if (trans->transid != root->fs_info->generation)
 		WARN(1, KERN_CRIT "trans %llu running %llu\n",
-		       (unsigned long long)trans->transid,
-		       (unsigned long long)root->fs_info->generation);
+		       trans->transid, root->fs_info->generation);
 
 	if (!should_cow_block(trans, root, buf)) {
 		*cow_ret = buf;
@@ -2466,6 +2426,40 @@ done:
 	return ret;
 }
 
+static void key_search_validate(struct extent_buffer *b,
+				struct btrfs_key *key,
+				int level)
+{
+#ifdef CONFIG_BTRFS_ASSERT
+	struct btrfs_disk_key disk_key;
+
+	btrfs_cpu_key_to_disk(&disk_key, key);
+
+	if (level == 0)
+		ASSERT(!memcmp_extent_buffer(b, &disk_key,
+		    offsetof(struct btrfs_leaf, items[0].key),
+		    sizeof(disk_key)));
+	else
+		ASSERT(!memcmp_extent_buffer(b, &disk_key,
+		    offsetof(struct btrfs_node, ptrs[0].key),
+		    sizeof(disk_key)));
+#endif
+}
+
+static int key_search(struct extent_buffer *b, struct btrfs_key *key,
+		      int level, int *prev_cmp, int *slot)
+{
+	if (*prev_cmp != 0) {
+		*prev_cmp = bin_search(b, key, level, slot);
+		return *prev_cmp;
+	}
+
+	key_search_validate(b, key, level);
+	*slot = 0;
+
+	return 0;
+}
+
 /*
  * look for key in the tree.  path is filled in with nodes along the way
  * if key is found, we return zero and you can find the item in the leaf
@@ -2494,6 +2488,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 	int write_lock_level = 0;
 	u8 lowest_level = 0;
 	int min_write_lock_level;
+	int prev_cmp;
 
 	lowest_level = p->lowest_level;
 	WARN_ON(lowest_level && ins_len > 0);
@@ -2524,6 +2519,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 	min_write_lock_level = write_lock_level;
 
 again:
+	prev_cmp = -1;
 	/*
 	 * we try very hard to do read locks on the root
 	 */
@@ -2624,7 +2620,7 @@ cow_done:
 		if (!cow)
 			btrfs_unlock_up_safe(p, level + 1);
 
-		ret = bin_search(b, key, level, &slot);
+		ret = key_search(b, key, level, &prev_cmp, &slot);
 
 		if (level != 0) {
 			int dec = 0;
@@ -2759,6 +2755,7 @@ 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;
 
 	lowest_level = p->lowest_level;
 	WARN_ON(p->nodes[0] != NULL);
@@ -2769,6 +2766,7 @@ 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;
@@ -2786,7 +2784,7 @@ again:
 		 */
 		btrfs_unlock_up_safe(p, level + 1);
 
-		ret = bin_search(b, key, level, &slot);
+		ret = key_search(b, key, level, &prev_cmp, &slot);
 
 		if (level != 0) {
 			int dec = 0;
@@ -2820,7 +2818,11 @@ again:
 				btrfs_clear_path_blocking(p, b,
 							  BTRFS_READ_LOCK);
 			}
-			b = tree_mod_log_rewind(root->fs_info, b, time_seq);
+			b = tree_mod_log_rewind(root->fs_info, p, b, time_seq);
+			if (!b) {
+				ret = -ENOMEM;
+				goto done;
+			}
 			p->locks[level] = BTRFS_READ_LOCK;
 			p->nodes[level] = b;
 		} else {
@@ -3143,13 +3145,11 @@ 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,
-			    (unsigned long)btrfs_header_fsid(c),
+	write_extent_buffer(c, root->fs_info->fsid, btrfs_header_fsid(c),
 			    BTRFS_FSID_SIZE);
 
 	write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(c),
-			    BTRFS_UUID_SIZE);
+			    btrfs_header_chunk_tree_uuid(c), BTRFS_UUID_SIZE);
 
 	btrfs_set_node_key(c, &lower_key, 0);
 	btrfs_set_node_blockptr(c, 0, lower->start);
@@ -3208,7 +3208,7 @@ static void insert_ptr(struct btrfs_trans_handle *trans,
 	}
 	if (level) {
 		ret = tree_mod_log_insert_key(root->fs_info, lower, slot,
-					      MOD_LOG_KEY_ADD);
+					      MOD_LOG_KEY_ADD, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 	btrfs_set_node_key(lower, key, slot);
@@ -3284,10 +3284,9 @@ 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,
-			    (unsigned long)btrfs_header_fsid(split),
-			    BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(split), BTRFS_FSID_SIZE);
 	write_extent_buffer(split, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(split),
+			    btrfs_header_chunk_tree_uuid(split),
 			    BTRFS_UUID_SIZE);
 
 	tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid);
@@ -4040,11 +4039,10 @@ again:
 	btrfs_set_header_owner(right, root->root_key.objectid);
 	btrfs_set_header_level(right, 0);
 	write_extent_buffer(right, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(right),
-			    BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(right), BTRFS_FSID_SIZE);
 
 	write_extent_buffer(right, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(right),
+			    btrfs_header_chunk_tree_uuid(right),
 			    BTRFS_UUID_SIZE);
 
 	if (split == 0) {
@@ -4642,7 +4640,7 @@ static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
 			      (nritems - slot - 1));
 	} else if (level) {
 		ret = tree_mod_log_insert_key(root->fs_info, parent, slot,
-					      MOD_LOG_KEY_REMOVE);
+					      MOD_LOG_KEY_REMOVE, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 
@@ -4814,7 +4812,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
  * This may release the path, and so you may lose any locks held at the
  * time you call it.
  */
-int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
+static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
 {
 	struct btrfs_key key;
 	struct btrfs_disk_key found_key;
@@ -5329,19 +5327,20 @@ int btrfs_compare_trees(struct btrfs_root *left_root,
 					goto out;
 				advance_right = ADVANCE;
 			} else {
+				enum btrfs_compare_tree_result cmp;
+
 				WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
 				ret = tree_compare_item(left_root, left_path,
 						right_path, tmp_buf);
-				if (ret) {
-					WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
-					ret = changed_cb(left_root, right_root,
-						left_path, right_path,
-						&left_key,
-						BTRFS_COMPARE_TREE_CHANGED,
-						ctx);
-					if (ret < 0)
-						goto out;
-				}
+				if (ret)
+					cmp = BTRFS_COMPARE_TREE_CHANGED;
+				else
+					cmp = BTRFS_COMPARE_TREE_SAME;
+				ret = changed_cb(left_root, right_root,
+						 left_path, right_path,
+						 &left_key, cmp, ctx);
+				if (ret < 0)
+					goto out;
 				advance_left = ADVANCE;
 				advance_right = ADVANCE;
 			}

fs/btrfs/ctree.h

@@ -23,6 +23,7 @@
 #include <linux/highmem.h>
 #include <linux/fs.h>
 #include <linux/rwsem.h>
+#include <linux/semaphore.h>
 #include <linux/completion.h>
 #include <linux/backing-dev.h>
 #include <linux/wait.h>
@@ -91,6 +92,9 @@ struct btrfs_ordered_sum;
 /* holds quota configuration and tracking */
 #define BTRFS_QUOTA_TREE_OBJECTID 8ULL
 
+/* for storing items that use the BTRFS_UUID_KEY* types */
+#define BTRFS_UUID_TREE_OBJECTID 9ULL
+
 /* for storing balance parameters in the root tree */
 #define BTRFS_BALANCE_OBJECTID -4ULL
 
@@ -142,7 +146,7 @@ struct btrfs_ordered_sum;
 
 #define BTRFS_EMPTY_SUBVOL_DIR_OBJECTID 2
 
-#define BTRFS_DEV_REPLACE_DEVID 0
+#define BTRFS_DEV_REPLACE_DEVID 0ULL
 
 /*
  * the max metadata block size.  This limit is somewhat artificial,
@@ -478,9 +482,10 @@ struct btrfs_super_block {
 	char label[BTRFS_LABEL_SIZE];
 
 	__le64 cache_generation;
+	__le64 uuid_tree_generation;
 
 	/* future expansion */
-	__le64 reserved[31];
+	__le64 reserved[30];
 	u8 sys_chunk_array[BTRFS_SYSTEM_CHUNK_ARRAY_SIZE];
 	struct btrfs_root_backup super_roots[BTRFS_NUM_BACKUP_ROOTS];
 } __attribute__ ((__packed__));
@@ -1188,6 +1193,7 @@ enum btrfs_caching_type {
 	BTRFS_CACHE_STARTED	= 1,
 	BTRFS_CACHE_FAST	= 2,
 	BTRFS_CACHE_FINISHED	= 3,
+	BTRFS_CACHE_ERROR	= 4,
 };
 
 enum btrfs_disk_cache_state {
@@ -1302,6 +1308,7 @@ struct btrfs_fs_info {
 	struct btrfs_root *fs_root;
 	struct btrfs_root *csum_root;
 	struct btrfs_root *quota_root;
+	struct btrfs_root *uuid_root;
 
 	/* the log root tree is a directory of all the other log roots */
 	struct btrfs_root *log_root_tree;
@@ -1350,6 +1357,7 @@ struct btrfs_fs_info {
 	u64 last_trans_log_full_commit;
 	unsigned long mount_opt;
 	unsigned long compress_type:4;
+	int commit_interval;
 	/*
 	 * It is a suggestive number, the read side is safe even it gets a
 	 * wrong number because we will write out the data into a regular
@@ -1411,6 +1419,13 @@ struct btrfs_fs_info {
 	 * before jumping into the main commit.
 	 */
 	struct mutex ordered_operations_mutex;
+
+	/*
+	 * Same as ordered_operations_mutex except this is for ordered extents
+	 * and not the operations.
+	 */
+	struct mutex ordered_extent_flush_mutex;
+
 	struct rw_semaphore extent_commit_sem;
 
 	struct rw_semaphore cleanup_work_sem;
@@ -1641,6 +1656,9 @@ struct btrfs_fs_info {
 	struct btrfs_dev_replace dev_replace;
 
 	atomic_t mutually_exclusive_operation_running;
+
+	struct semaphore uuid_tree_rescan_sem;
+	unsigned int update_uuid_tree_gen:1;
 };
 
 /*
@@ -1933,6 +1951,19 @@ struct btrfs_ioctl_defrag_range_args {
  */
 #define BTRFS_DEV_REPLACE_KEY	250
 
+/*
+ * Stores items that allow to quickly map UUIDs to something else.
+ * These items are part of the filesystem UUID tree.
+ * The key is built like this:
+ * (UUID_upper_64_bits, BTRFS_UUID_KEY*, UUID_lower_64_bits).
+ */
+#if BTRFS_UUID_SIZE != 16
+#error "UUID items require BTRFS_UUID_SIZE == 16!"
+#endif
+#define BTRFS_UUID_KEY_SUBVOL	251	/* for UUIDs assigned to subvols */
+#define BTRFS_UUID_KEY_RECEIVED_SUBVOL	252	/* for UUIDs assigned to
+						 * received subvols */
+
 /*
  * string items are for debugging.  They just store a short string of
  * data in the FS
@@ -1967,6 +1998,9 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_MOUNT_CHECK_INTEGRITY	(1 << 20)
 #define BTRFS_MOUNT_CHECK_INTEGRITY_INCLUDING_EXTENT_DATA (1 << 21)
 #define BTRFS_MOUNT_PANIC_ON_FATAL_ERROR	(1 << 22)
+#define BTRFS_MOUNT_RESCAN_UUID_TREE	(1 << 23)
+
+#define BTRFS_DEFAULT_COMMIT_INTERVAL	(30)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2130,14 +2164,14 @@ BTRFS_SETGET_STACK_FUNCS(stack_device_bandwidth, struct btrfs_dev_item,
 BTRFS_SETGET_STACK_FUNCS(stack_device_generation, struct btrfs_dev_item,
 			 generation, 64);
 
-static inline char *btrfs_device_uuid(struct btrfs_dev_item *d)
+static inline unsigned long btrfs_device_uuid(struct btrfs_dev_item *d)
 {
-	return (char *)d + offsetof(struct btrfs_dev_item, uuid);
+	return (unsigned long)d + offsetof(struct btrfs_dev_item, uuid);
 }
 
-static inline char *btrfs_device_fsid(struct btrfs_dev_item *d)
+static inline unsigned long btrfs_device_fsid(struct btrfs_dev_item *d)
 {
-	return (char *)d + offsetof(struct btrfs_dev_item, fsid);
+	return (unsigned long)d + offsetof(struct btrfs_dev_item, fsid);
 }
 
 BTRFS_SETGET_FUNCS(chunk_length, struct btrfs_chunk, length, 64);
@@ -2240,6 +2274,23 @@ BTRFS_SETGET_FUNCS(inode_gid, struct btrfs_inode_item, gid, 32);
 BTRFS_SETGET_FUNCS(inode_mode, struct btrfs_inode_item, mode, 32);
 BTRFS_SETGET_FUNCS(inode_rdev, struct btrfs_inode_item, rdev, 64);
 BTRFS_SETGET_FUNCS(inode_flags, struct btrfs_inode_item, flags, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_generation, struct btrfs_inode_item,
+			 generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_sequence, struct btrfs_inode_item,
+			 sequence, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_transid, struct btrfs_inode_item,
+			 transid, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_size, struct btrfs_inode_item, size, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_nbytes, struct btrfs_inode_item,
+			 nbytes, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_block_group, struct btrfs_inode_item,
+			 block_group, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_nlink, struct btrfs_inode_item, nlink, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_uid, struct btrfs_inode_item, uid, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_gid, struct btrfs_inode_item, gid, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_mode, struct btrfs_inode_item, mode, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_rdev, struct btrfs_inode_item, rdev, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_flags, struct btrfs_inode_item, flags, 64);
 
 static inline struct btrfs_timespec *
 btrfs_inode_atime(struct btrfs_inode_item *inode_item)
@@ -2267,6 +2318,8 @@ btrfs_inode_ctime(struct btrfs_inode_item *inode_item)
 
 BTRFS_SETGET_FUNCS(timespec_sec, struct btrfs_timespec, sec, 64);
 BTRFS_SETGET_FUNCS(timespec_nsec, struct btrfs_timespec, nsec, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_timespec_sec, struct btrfs_timespec, sec, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_timespec_nsec, struct btrfs_timespec, nsec, 32);
 
 /* struct btrfs_dev_extent */
 BTRFS_SETGET_FUNCS(dev_extent_chunk_tree, struct btrfs_dev_extent,
@@ -2277,10 +2330,10 @@ BTRFS_SETGET_FUNCS(dev_extent_chunk_offset, struct btrfs_dev_extent,
 		   chunk_offset, 64);
 BTRFS_SETGET_FUNCS(dev_extent_length, struct btrfs_dev_extent, length, 64);
 
-static inline u8 *btrfs_dev_extent_chunk_tree_uuid(struct btrfs_dev_extent *dev)
+static inline unsigned long btrfs_dev_extent_chunk_tree_uuid(struct btrfs_dev_extent *dev)
 {
 	unsigned long ptr = offsetof(struct btrfs_dev_extent, chunk_tree_uuid);
-	return (u8 *)((unsigned long)dev + ptr);
+	return (unsigned long)dev + ptr;
 }
 
 BTRFS_SETGET_FUNCS(extent_refs, struct btrfs_extent_item, refs, 64);
@@ -2348,6 +2401,10 @@ BTRFS_SETGET_FUNCS(ref_count_v0, struct btrfs_extent_ref_v0, count, 32);
 /* struct btrfs_node */
 BTRFS_SETGET_FUNCS(key_blockptr, struct btrfs_key_ptr, blockptr, 64);
 BTRFS_SETGET_FUNCS(key_generation, struct btrfs_key_ptr, generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_key_blockptr, struct btrfs_key_ptr,
+			 blockptr, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_key_generation, struct btrfs_key_ptr,
+			 generation, 64);
 
 static inline u64 btrfs_node_blockptr(struct extent_buffer *eb, int nr)
 {
@@ -2404,6 +2461,8 @@ static inline void btrfs_set_node_key(struct extent_buffer *eb,
 /* struct btrfs_item */
 BTRFS_SETGET_FUNCS(item_offset, struct btrfs_item, offset, 32);
 BTRFS_SETGET_FUNCS(item_size, struct btrfs_item, size, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_item_offset, struct btrfs_item, offset, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_item_size, struct btrfs_item, size, 32);
 
 static inline unsigned long btrfs_item_nr_offset(int nr)
 {
@@ -2466,6 +2525,13 @@ BTRFS_SETGET_FUNCS(dir_data_len, struct btrfs_dir_item, data_len, 16);
 BTRFS_SETGET_FUNCS(dir_type, struct btrfs_dir_item, type, 8);
 BTRFS_SETGET_FUNCS(dir_name_len, struct btrfs_dir_item, name_len, 16);
 BTRFS_SETGET_FUNCS(dir_transid, struct btrfs_dir_item, transid, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_type, struct btrfs_dir_item, type, 8);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_data_len, struct btrfs_dir_item,
+			 data_len, 16);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_name_len, struct btrfs_dir_item,
+			 name_len, 16);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_transid, struct btrfs_dir_item,
+			 transid, 64);
 
 static inline void btrfs_dir_item_key(struct extent_buffer *eb,
 				      struct btrfs_dir_item *item,
@@ -2568,6 +2634,12 @@ BTRFS_SETGET_HEADER_FUNCS(header_owner, struct btrfs_header, owner, 64);
 BTRFS_SETGET_HEADER_FUNCS(header_nritems, struct btrfs_header, nritems, 32);
 BTRFS_SETGET_HEADER_FUNCS(header_flags, struct btrfs_header, flags, 64);
 BTRFS_SETGET_HEADER_FUNCS(header_level, struct btrfs_header, level, 8);
+BTRFS_SETGET_STACK_FUNCS(stack_header_generation, struct btrfs_header,
+			 generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_header_owner, struct btrfs_header, owner, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_header_nritems, struct btrfs_header,
+			 nritems, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_header_bytenr, struct btrfs_header, bytenr, 64);
 
 static inline int btrfs_header_flag(struct extent_buffer *eb, u64 flag)
 {
@@ -2603,16 +2675,14 @@ static inline void btrfs_set_header_backref_rev(struct extent_buffer *eb,
 	btrfs_set_header_flags(eb, flags);
 }
 
-static inline u8 *btrfs_header_fsid(struct extent_buffer *eb)
+static inline unsigned long btrfs_header_fsid(struct extent_buffer *eb)
 {
-	unsigned long ptr = offsetof(struct btrfs_header, fsid);
-	return (u8 *)ptr;
+	return offsetof(struct btrfs_header, fsid);
 }
 
-static inline u8 *btrfs_header_chunk_tree_uuid(struct extent_buffer *eb)
+static inline unsigned long btrfs_header_chunk_tree_uuid(struct extent_buffer *eb)
 {
-	unsigned long ptr = offsetof(struct btrfs_header, chunk_tree_uuid);
-	return (u8 *)ptr;
+	return offsetof(struct btrfs_header, chunk_tree_uuid);
 }
 
 static inline int btrfs_is_leaf(struct extent_buffer *eb)
@@ -2830,6 +2900,9 @@ BTRFS_SETGET_STACK_FUNCS(super_csum_type, struct btrfs_super_block,
 			 csum_type, 16);
 BTRFS_SETGET_STACK_FUNCS(super_cache_generation, struct btrfs_super_block,
 			 cache_generation, 64);
+BTRFS_SETGET_STACK_FUNCS(super_magic, struct btrfs_super_block, magic, 64);
+BTRFS_SETGET_STACK_FUNCS(super_uuid_tree_generation, struct btrfs_super_block,
+			 uuid_tree_generation, 64);
 
 static inline int btrfs_super_csum_size(struct btrfs_super_block *s)
 {
@@ -2847,6 +2920,14 @@ static inline unsigned long btrfs_leaf_data(struct extent_buffer *l)
 
 /* struct btrfs_file_extent_item */
 BTRFS_SETGET_FUNCS(file_extent_type, struct btrfs_file_extent_item, type, 8);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_disk_bytenr,
+			 struct btrfs_file_extent_item, disk_bytenr, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_offset,
+			 struct btrfs_file_extent_item, offset, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_generation,
+			 struct btrfs_file_extent_item, generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_num_bytes,
+			 struct btrfs_file_extent_item, num_bytes, 64);
 
 static inline unsigned long
 btrfs_file_extent_inline_start(struct btrfs_file_extent_item *e)
@@ -3107,11 +3188,9 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   u64 root_objectid, u64 owner, u64 offset,
 				   struct btrfs_key *ins);
-int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
-				  struct btrfs_root *root,
-				  u64 num_bytes, u64 min_alloc_size,
-				  u64 empty_size, u64 hint_byte,
-				  struct btrfs_key *ins, int is_data);
+int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes,
+			 u64 min_alloc_size, u64 empty_size, u64 hint_byte,
+			 struct btrfs_key *ins, int is_data);
 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		  struct extent_buffer *buf, int full_backref, int for_cow);
 int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
@@ -3175,7 +3254,7 @@ void btrfs_orphan_release_metadata(struct inode *inode);
 int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 				     struct btrfs_block_rsv *rsv,
 				     int nitems,
-				     u64 *qgroup_reserved);
+				     u64 *qgroup_reserved, bool use_global_rsv);
 void btrfs_subvolume_release_metadata(struct btrfs_root *root,
 				      struct btrfs_block_rsv *rsv,
 				      u64 qgroup_reserved);
@@ -3245,6 +3324,7 @@ enum btrfs_compare_tree_result {
 	BTRFS_COMPARE_TREE_NEW,
 	BTRFS_COMPARE_TREE_DELETED,
 	BTRFS_COMPARE_TREE_CHANGED,
+	BTRFS_COMPARE_TREE_SAME,
 };
 typedef int (*btrfs_changed_cb_t)(struct btrfs_root *left_root,
 				  struct btrfs_root *right_root,
@@ -3380,6 +3460,7 @@ static inline void free_fs_info(struct btrfs_fs_info *fs_info)
 	kfree(fs_info->dev_root);
 	kfree(fs_info->csum_root);
 	kfree(fs_info->quota_root);
+	kfree(fs_info->uuid_root);
 	kfree(fs_info->super_copy);
 	kfree(fs_info->super_for_commit);
 	kfree(fs_info);
@@ -3414,8 +3495,6 @@ int __must_check btrfs_update_root(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   struct btrfs_key *key,
 				   struct btrfs_root_item *item);
-void btrfs_read_root_item(struct extent_buffer *eb, int slot,
-			  struct btrfs_root_item *item);
 int btrfs_find_root(struct btrfs_root *root, struct btrfs_key *search_key,
 		    struct btrfs_path *path, struct btrfs_root_item *root_item,
 		    struct btrfs_key *root_key);
@@ -3426,6 +3505,17 @@ void btrfs_check_and_init_root_item(struct btrfs_root_item *item);
 void btrfs_update_root_times(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root);
 
+/* uuid-tree.c */
+int btrfs_uuid_tree_add(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid);
+int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid);
+int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
+			    int (*check_func)(struct btrfs_fs_info *, u8 *, u8,
+					      u64));
+
 /* dir-item.c */
 int btrfs_check_dir_item_collision(struct btrfs_root *root, u64 dir,
 			  const char *name, int name_len);
@@ -3509,12 +3599,14 @@ int btrfs_find_name_in_ext_backref(struct btrfs_path *path,
 				   struct btrfs_inode_extref **extref_ret);
 
 /* file-item.c */
+struct btrfs_dio_private;
 int btrfs_del_csums(struct btrfs_trans_handle *trans,
 		    struct btrfs_root *root, u64 bytenr, u64 len);
 int btrfs_lookup_bio_sums(struct btrfs_root *root, struct inode *inode,
 			  struct bio *bio, u32 *dst);
 int btrfs_lookup_bio_sums_dio(struct btrfs_root *root, struct inode *inode,
-			      struct bio *bio, u64 logical_offset);
+			      struct btrfs_dio_private *dip, struct bio *bio,
+			      u64 logical_offset);
 int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root,
 			     u64 objectid, u64 pos,
@@ -3552,8 +3644,7 @@ void btrfs_wait_and_free_delalloc_work(struct btrfs_delalloc_work *work);
 struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *page,
 					   size_t pg_offset, u64 start, u64 len,
 					   int create);
-noinline int can_nocow_extent(struct btrfs_trans_handle *trans,
-			      struct inode *inode, u64 offset, u64 *len,
+noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 			      u64 *orig_start, u64 *orig_block_len,
 			      u64 *ram_bytes);
 
@@ -3643,11 +3734,15 @@ extern const struct dentry_operations btrfs_dentry_operations;
 long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
 void btrfs_update_iflags(struct inode *inode);
 void btrfs_inherit_iflags(struct inode *inode, struct inode *dir);
+int btrfs_is_empty_uuid(u8 *uuid);
 int btrfs_defrag_file(struct inode *inode, struct file *file,
 		      struct btrfs_ioctl_defrag_range_args *range,
 		      u64 newer_than, unsigned long max_pages);
 void btrfs_get_block_group_info(struct list_head *groups_list,
 				struct btrfs_ioctl_space_info *space);
+void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
+			       struct btrfs_ioctl_balance_args *bargs);
+
 
 /* file.c */
 int btrfs_auto_defrag_init(void);
@@ -3720,6 +3815,22 @@ void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 #define btrfs_debug(fs_info, fmt, args...) \
 	btrfs_printk(fs_info, KERN_DEBUG fmt, ##args)
 
+#ifdef CONFIG_BTRFS_ASSERT
+
+static inline void assfail(char *expr, char *file, int line)
+{
+	printk(KERN_ERR "BTRFS assertion failed: %s, file: %s, line: %d",
+	       expr, file, line);
+	BUG();
+}
+
+#define ASSERT(expr)	\
+	(likely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__))
+#else
+#define ASSERT(expr)	((void)0)
+#endif
+
+#define btrfs_assert()
 __printf(5, 6)
 void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		     unsigned int line, int errno, const char *fmt, ...);

fs/btrfs/delayed-inode.c

@@ -21,6 +21,7 @@
 #include "delayed-inode.h"
 #include "disk-io.h"
 #include "transaction.h"
+#include "ctree.h"
 
 #define BTRFS_DELAYED_WRITEBACK		512
 #define BTRFS_DELAYED_BACKGROUND	128
@@ -1453,10 +1454,10 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
 
 	dir_item = (struct btrfs_dir_item *)delayed_item->data;
 	dir_item->location = *disk_key;
-	dir_item->transid = cpu_to_le64(trans->transid);
-	dir_item->data_len = 0;
-	dir_item->name_len = cpu_to_le16(name_len);
-	dir_item->type = type;
+	btrfs_set_stack_dir_transid(dir_item, trans->transid);
+	btrfs_set_stack_dir_data_len(dir_item, 0);
+	btrfs_set_stack_dir_name_len(dir_item, name_len);
+	btrfs_set_stack_dir_type(dir_item, type);
 	memcpy((char *)(dir_item + 1), name, name_len);
 
 	ret = btrfs_delayed_item_reserve_metadata(trans, root, delayed_item);
@@ -1470,13 +1471,11 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
 	mutex_lock(&delayed_node->mutex);
 	ret = __btrfs_add_delayed_insertion_item(delayed_node, delayed_item);
 	if (unlikely(ret)) {
-		printk(KERN_ERR "err add delayed dir index item(name: %s) into "
-				"the insertion tree of the delayed node"
+		printk(KERN_ERR "err add delayed dir index item(name: %.*s) "
+				"into the insertion tree of the delayed node"
 				"(root id: %llu, inode id: %llu, errno: %d)\n",
-				name,
-				(unsigned long long)delayed_node->root->objectid,
-				(unsigned long long)delayed_node->inode_id,
-				ret);
+				name_len, name, delayed_node->root->objectid,
+				delayed_node->inode_id, ret);
 		BUG();
 	}
 	mutex_unlock(&delayed_node->mutex);
@@ -1547,9 +1546,7 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
 		printk(KERN_ERR "err add delayed dir index item(index: %llu) "
 				"into the deletion tree of the delayed node"
 				"(root id: %llu, inode id: %llu, errno: %d)\n",
-				(unsigned long long)index,
-				(unsigned long long)node->root->objectid,
-				(unsigned long long)node->inode_id,
+				index, node->root->objectid, node->inode_id,
 				ret);
 		BUG();
 	}
@@ -1699,7 +1696,7 @@ int btrfs_readdir_delayed_dir_index(struct dir_context *ctx,
 
 		di = (struct btrfs_dir_item *)curr->data;
 		name = (char *)(di + 1);
-		name_len = le16_to_cpu(di->name_len);
+		name_len = btrfs_stack_dir_name_len(di);
 
 		d_type = btrfs_filetype_table[di->type];
 		btrfs_disk_key_to_cpu(&location, &di->location);
@@ -1716,27 +1713,6 @@ int btrfs_readdir_delayed_dir_index(struct dir_context *ctx,
 	return 0;
 }
 
-BTRFS_SETGET_STACK_FUNCS(stack_inode_generation, struct btrfs_inode_item,
-			 generation, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_sequence, struct btrfs_inode_item,
-			 sequence, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_transid, struct btrfs_inode_item,
-			 transid, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_size, struct btrfs_inode_item, size, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_nbytes, struct btrfs_inode_item,
-			 nbytes, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_block_group, struct btrfs_inode_item,
-			 block_group, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_nlink, struct btrfs_inode_item, nlink, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_uid, struct btrfs_inode_item, uid, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_gid, struct btrfs_inode_item, gid, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_mode, struct btrfs_inode_item, mode, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_rdev, struct btrfs_inode_item, rdev, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_flags, struct btrfs_inode_item, flags, 64);
-
-BTRFS_SETGET_STACK_FUNCS(stack_timespec_sec, struct btrfs_timespec, sec, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_timespec_nsec, struct btrfs_timespec, nsec, 32);
-
 static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
 				  struct btrfs_inode_item *inode_item,
 				  struct inode *inode)

fs/btrfs/delayed-ref.c

@@ -241,7 +241,7 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-static void inline drop_delayed_ref(struct btrfs_trans_handle *trans,
+static inline void drop_delayed_ref(struct btrfs_trans_handle *trans,
 				    struct btrfs_delayed_ref_root *delayed_refs,
 				    struct btrfs_delayed_ref_node *ref)
 {
@@ -600,7 +600,7 @@ static noinline void add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	INIT_LIST_HEAD(&head_ref->cluster);
 	mutex_init(&head_ref->mutex);
 
-	trace_btrfs_delayed_ref_head(ref, head_ref, action);
+	trace_add_delayed_ref_head(ref, head_ref, action);
 
 	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
 
@@ -661,7 +661,7 @@ static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 		ref->type = BTRFS_TREE_BLOCK_REF_KEY;
 	full_ref->level = level;
 
-	trace_btrfs_delayed_tree_ref(ref, full_ref, action);
+	trace_add_delayed_tree_ref(ref, full_ref, action);
 
 	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
 
@@ -722,7 +722,7 @@ static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	full_ref->objectid = owner;
 	full_ref->offset = offset;
 
-	trace_btrfs_delayed_data_ref(ref, full_ref, action);
+	trace_add_delayed_data_ref(ref, full_ref, action);
 
 	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
 

fs/btrfs/dev-replace.c

@@ -148,13 +148,13 @@ no_valid_dev_replace_entry_found:
 		    !btrfs_test_opt(dev_root, DEGRADED)) {
 			ret = -EIO;
 			pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "srcdev (devid %llu) is missing, need to run 'btrfs dev scan'?\n",
-				(unsigned long long)src_devid);
+				src_devid);
 		}
 		if (!dev_replace->tgtdev &&
 		    !btrfs_test_opt(dev_root, DEGRADED)) {
 			ret = -EIO;
 			pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "tgtdev (devid %llu) is missing, need to run btrfs dev scan?\n",
-				(unsigned long long)BTRFS_DEV_REPLACE_DEVID);
+				BTRFS_DEV_REPLACE_DEVID);
 		}
 		if (dev_replace->tgtdev) {
 			if (dev_replace->srcdev) {

fs/btrfs/disk-io.c

@@ -31,6 +31,7 @@
 #include <linux/migrate.h>
 #include <linux/ratelimit.h>
 #include <linux/uuid.h>
+#include <linux/semaphore.h>
 #include <asm/unaligned.h>
 #include "compat.h"
 #include "ctree.h"
@@ -302,9 +303,8 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
 			printk_ratelimited(KERN_INFO "btrfs: %s checksum verify "
 				       "failed on %llu wanted %X found %X "
 				       "level %d\n",
-				       root->fs_info->sb->s_id,
-				       (unsigned long long)buf->start, val, found,
-				       btrfs_header_level(buf));
+				       root->fs_info->sb->s_id, buf->start,
+				       val, found, btrfs_header_level(buf));
 			if (result != (char *)&inline_result)
 				kfree(result);
 			return 1;
@@ -345,9 +345,7 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 	}
 	printk_ratelimited("parent transid verify failed on %llu wanted %llu "
 		       "found %llu\n",
-		       (unsigned long long)eb->start,
-		       (unsigned long long)parent_transid,
-		       (unsigned long long)btrfs_header_generation(eb));
+		       eb->start, parent_transid, btrfs_header_generation(eb));
 	ret = 1;
 	clear_extent_buffer_uptodate(eb);
 out:
@@ -497,8 +495,7 @@ static int check_tree_block_fsid(struct btrfs_root *root,
 	u8 fsid[BTRFS_UUID_SIZE];
 	int ret = 1;
 
-	read_extent_buffer(eb, fsid, (unsigned long)btrfs_header_fsid(eb),
-			   BTRFS_FSID_SIZE);
+	read_extent_buffer(eb, fsid, btrfs_header_fsid(eb), BTRFS_FSID_SIZE);
 	while (fs_devices) {
 		if (!memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE)) {
 			ret = 0;
@@ -512,8 +509,7 @@ static int check_tree_block_fsid(struct btrfs_root *root,
 #define CORRUPT(reason, eb, root, slot)				\
 	printk(KERN_CRIT "btrfs: corrupt leaf, %s: block=%llu,"	\
 	       "root=%llu, slot=%d\n", reason,			\
-	       (unsigned long long)btrfs_header_bytenr(eb),	\
-	       (unsigned long long)root->objectid, slot)
+	       btrfs_header_bytenr(eb),	root->objectid, slot)
 
 static noinline int check_leaf(struct btrfs_root *root,
 			       struct extent_buffer *leaf)
@@ -576,8 +572,9 @@ static noinline int check_leaf(struct btrfs_root *root,
 	return 0;
 }
 
-static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
-			       struct extent_state *state, int mirror)
+static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
+				      u64 phy_offset, struct page *page,
+				      u64 start, u64 end, int mirror)
 {
 	struct extent_io_tree *tree;
 	u64 found_start;
@@ -612,14 +609,13 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 	if (found_start != eb->start) {
 		printk_ratelimited(KERN_INFO "btrfs bad tree block start "
 			       "%llu %llu\n",
-			       (unsigned long long)found_start,
-			       (unsigned long long)eb->start);
+			       found_start, eb->start);
 		ret = -EIO;
 		goto err;
 	}
 	if (check_tree_block_fsid(root, eb)) {
 		printk_ratelimited(KERN_INFO "btrfs bad fsid on block %llu\n",
-			       (unsigned long long)eb->start);
+			       eb->start);
 		ret = -EIO;
 		goto err;
 	}
@@ -1148,6 +1144,10 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
 		return NULL;
 
 	ret = btree_read_extent_buffer_pages(root, buf, 0, parent_transid);
+	if (ret) {
+		free_extent_buffer(buf);
+		return NULL;
+	}
 	return buf;
 
 }
@@ -1291,11 +1291,10 @@ 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,
-			    (unsigned long)btrfs_header_fsid(leaf),
+	write_extent_buffer(leaf, fs_info->fsid, btrfs_header_fsid(leaf),
 			    BTRFS_FSID_SIZE);
 	write_extent_buffer(leaf, fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(leaf),
+			    btrfs_header_chunk_tree_uuid(leaf),
 			    BTRFS_UUID_SIZE);
 	btrfs_mark_buffer_dirty(leaf);
 
@@ -1379,8 +1378,7 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 	root->node = leaf;
 
 	write_extent_buffer(root->node, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(root->node),
-			    BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(root->node), BTRFS_FSID_SIZE);
 	btrfs_mark_buffer_dirty(root->node);
 	btrfs_tree_unlock(root->node);
 	return root;
@@ -1413,11 +1411,11 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	log_root->root_key.offset = root->root_key.objectid;
 
 	inode_item = &log_root->root_item.inode;
-	inode_item->generation = cpu_to_le64(1);
-	inode_item->size = cpu_to_le64(3);
-	inode_item->nlink = cpu_to_le32(1);
-	inode_item->nbytes = cpu_to_le64(root->leafsize);
-	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);
+	btrfs_set_stack_inode_generation(inode_item, 1);
+	btrfs_set_stack_inode_size(inode_item, 3);
+	btrfs_set_stack_inode_nlink(inode_item, 1);
+	btrfs_set_stack_inode_nbytes(inode_item, root->leafsize);
+	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
 
 	btrfs_set_root_node(&log_root->root_item, log_root->node);
 
@@ -1428,8 +1426,8 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
-					struct btrfs_key *key)
+static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
+					       struct btrfs_key *key)
 {
 	struct btrfs_root *root;
 	struct btrfs_fs_info *fs_info = tree_root->fs_info;
@@ -1529,8 +1527,8 @@ fail:
 	return ret;
 }
 
-struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
-					u64 root_id)
+static struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
+					       u64 root_id)
 {
 	struct btrfs_root *root;
 
@@ -1581,10 +1579,16 @@ struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
 	if (location->objectid == BTRFS_QUOTA_TREE_OBJECTID)
 		return fs_info->quota_root ? fs_info->quota_root :
 					     ERR_PTR(-ENOENT);
+	if (location->objectid == BTRFS_UUID_TREE_OBJECTID)
+		return fs_info->uuid_root ? fs_info->uuid_root :
+					    ERR_PTR(-ENOENT);
 again:
 	root = btrfs_lookup_fs_root(fs_info, location->objectid);
-	if (root)
+	if (root) {
+		if (btrfs_root_refs(&root->root_item) == 0)
+			return ERR_PTR(-ENOENT);
 		return root;
+	}
 
 	root = btrfs_read_fs_root(fs_info->tree_root, location);
 	if (IS_ERR(root))
@@ -1737,7 +1741,7 @@ static int transaction_kthread(void *arg)
 
 	do {
 		cannot_commit = false;
-		delay = HZ * 30;
+		delay = HZ * root->fs_info->commit_interval;
 		mutex_lock(&root->fs_info->transaction_kthread_mutex);
 
 		spin_lock(&root->fs_info->trans_lock);
@@ -1749,7 +1753,8 @@ static int transaction_kthread(void *arg)
 
 		now = get_seconds();
 		if (cur->state < TRANS_STATE_BLOCKED &&
-		    (now < cur->start_time || now - cur->start_time < 30)) {
+		    (now < cur->start_time ||
+		     now - cur->start_time < root->fs_info->commit_interval)) {
 			spin_unlock(&root->fs_info->trans_lock);
 			delay = HZ * 5;
 			goto sleep;
@@ -2038,6 +2043,12 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_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);
@@ -2098,11 +2109,14 @@ int open_ctree(struct super_block *sb,
 	struct btrfs_root *chunk_root;
 	struct btrfs_root *dev_root;
 	struct btrfs_root *quota_root;
+	struct btrfs_root *uuid_root;
 	struct btrfs_root *log_tree_root;
 	int ret;
 	int err = -EINVAL;
 	int num_backups_tried = 0;
 	int backup_index = 0;
+	bool create_uuid_tree;
+	bool check_uuid_tree;
 
 	tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info);
 	chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info);
@@ -2189,6 +2203,7 @@ int open_ctree(struct super_block *sb,
 	fs_info->defrag_inodes = RB_ROOT;
 	fs_info->free_chunk_space = 0;
 	fs_info->tree_mod_log = RB_ROOT;
+	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 
 	/* readahead state */
 	INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT);
@@ -2270,6 +2285,7 @@ int open_ctree(struct super_block *sb,
 
 
 	mutex_init(&fs_info->ordered_operations_mutex);
+	mutex_init(&fs_info->ordered_extent_flush_mutex);
 	mutex_init(&fs_info->tree_log_mutex);
 	mutex_init(&fs_info->chunk_mutex);
 	mutex_init(&fs_info->transaction_kthread_mutex);
@@ -2278,6 +2294,7 @@ int open_ctree(struct super_block *sb,
 	init_rwsem(&fs_info->extent_commit_sem);
 	init_rwsem(&fs_info->cleanup_work_sem);
 	init_rwsem(&fs_info->subvol_sem);
+	sema_init(&fs_info->uuid_tree_rescan_sem, 1);
 	fs_info->dev_replace.lock_owner = 0;
 	atomic_set(&fs_info->dev_replace.nesting_level, 0);
 	mutex_init(&fs_info->dev_replace.lock_finishing_cancel_unmount);
@@ -2383,7 +2400,7 @@ int open_ctree(struct super_block *sb,
 	if (features) {
 		printk(KERN_ERR "BTRFS: couldn't mount because of "
 		       "unsupported optional features (%Lx).\n",
-		       (unsigned long long)features);
+		       features);
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2453,7 +2470,7 @@ int open_ctree(struct super_block *sb,
 	if (!(sb->s_flags & MS_RDONLY) && features) {
 		printk(KERN_ERR "BTRFS: couldn't mount RDWR because of "
 		       "unsupported option features (%Lx).\n",
-		       (unsigned long long)features);
+		       features);
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2466,20 +2483,17 @@ int open_ctree(struct super_block *sb,
 			   &fs_info->generic_worker);
 
 	btrfs_init_workers(&fs_info->delalloc_workers, "delalloc",
-			   fs_info->thread_pool_size,
-			   &fs_info->generic_worker);
+			   fs_info->thread_pool_size, NULL);
 
 	btrfs_init_workers(&fs_info->flush_workers, "flush_delalloc",
-			   fs_info->thread_pool_size,
-			   &fs_info->generic_worker);
+			   fs_info->thread_pool_size, NULL);
 
 	btrfs_init_workers(&fs_info->submit_workers, "submit",
 			   min_t(u64, fs_devices->num_devices,
-			   fs_info->thread_pool_size),
-			   &fs_info->generic_worker);
+			   fs_info->thread_pool_size), NULL);
 
 	btrfs_init_workers(&fs_info->caching_workers, "cache",
-			   2, &fs_info->generic_worker);
+			   fs_info->thread_pool_size, NULL);
 
 	/* a higher idle thresh on the submit workers makes it much more
 	 * likely that bios will be send down in a sane order to the
@@ -2575,7 +2589,7 @@ int open_ctree(struct super_block *sb,
 	sb->s_blocksize = sectorsize;
 	sb->s_blocksize_bits = blksize_bits(sectorsize);
 
-	if (disk_super->magic != cpu_to_le64(BTRFS_MAGIC)) {
+	if (btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
 		printk(KERN_INFO "btrfs: valid FS not found on %s\n", sb->s_id);
 		goto fail_sb_buffer;
 	}
@@ -2615,8 +2629,7 @@ int open_ctree(struct super_block *sb,
 	chunk_root->commit_root = btrfs_root_node(chunk_root);
 
 	read_extent_buffer(chunk_root->node, fs_info->chunk_tree_uuid,
-	   (unsigned long)btrfs_header_chunk_tree_uuid(chunk_root->node),
-	   BTRFS_UUID_SIZE);
+	   btrfs_header_chunk_tree_uuid(chunk_root->node), BTRFS_UUID_SIZE);
 
 	ret = btrfs_read_chunk_tree(chunk_root);
 	if (ret) {
@@ -2696,6 +2709,22 @@ retry_root_backup:
 		fs_info->quota_root = quota_root;
 	}
 
+	location.objectid = BTRFS_UUID_TREE_OBJECTID;
+	uuid_root = btrfs_read_tree_root(tree_root, &location);
+	if (IS_ERR(uuid_root)) {
+		ret = PTR_ERR(uuid_root);
+		if (ret != -ENOENT)
+			goto recovery_tree_root;
+		create_uuid_tree = true;
+		check_uuid_tree = false;
+	} else {
+		uuid_root->track_dirty = 1;
+		fs_info->uuid_root = uuid_root;
+		create_uuid_tree = false;
+		check_uuid_tree =
+		    generation != btrfs_super_uuid_tree_generation(disk_super);
+	}
+
 	fs_info->generation = generation;
 	fs_info->last_trans_committed = generation;
 
@@ -2882,6 +2911,29 @@ retry_root_backup:
 
 	btrfs_qgroup_rescan_resume(fs_info);
 
+	if (create_uuid_tree) {
+		pr_info("btrfs: creating UUID tree\n");
+		ret = btrfs_create_uuid_tree(fs_info);
+		if (ret) {
+			pr_warn("btrfs: failed to create the UUID tree %d\n",
+				ret);
+			close_ctree(tree_root);
+			return ret;
+		}
+	} else if (check_uuid_tree ||
+		   btrfs_test_opt(tree_root, RESCAN_UUID_TREE)) {
+		pr_info("btrfs: checking UUID tree\n");
+		ret = btrfs_check_uuid_tree(fs_info);
+		if (ret) {
+			pr_warn("btrfs: failed to check the UUID tree %d\n",
+				ret);
+			close_ctree(tree_root);
+			return ret;
+		}
+	} else {
+		fs_info->update_uuid_tree_gen = 1;
+	}
+
 	return 0;
 
 fail_qgroup:
@@ -2983,15 +3035,17 @@ struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
 	 */
 	for (i = 0; i < 1; i++) {
 		bytenr = btrfs_sb_offset(i);
-		if (bytenr + 4096 >= i_size_read(bdev->bd_inode))
+		if (bytenr + BTRFS_SUPER_INFO_SIZE >=
+					i_size_read(bdev->bd_inode))
 			break;
-		bh = __bread(bdev, bytenr / 4096, 4096);
+		bh = __bread(bdev, bytenr / 4096,
+					BTRFS_SUPER_INFO_SIZE);
 		if (!bh)
 			continue;
 
 		super = (struct btrfs_super_block *)bh->b_data;
 		if (btrfs_super_bytenr(super) != bytenr ||
-		    super->magic != cpu_to_le64(BTRFS_MAGIC)) {
+		    btrfs_super_magic(super) != BTRFS_MAGIC) {
 			brelse(bh);
 			continue;
 		}
@@ -3311,7 +3365,6 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 	int total_errors = 0;
 	u64 flags;
 
-	max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
 	do_barriers = !btrfs_test_opt(root, NOBARRIER);
 	backup_super_roots(root->fs_info);
 
@@ -3320,6 +3373,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	head = &root->fs_info->fs_devices->devices;
+	max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
 
 	if (do_barriers) {
 		ret = barrier_all_devices(root->fs_info);
@@ -3362,8 +3416,10 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 		printk(KERN_ERR "btrfs: %d errors while writing supers\n",
 		       total_errors);
 
-		/* This shouldn't happen. FUA is masked off if unsupported */
-		BUG();
+		/* FUA is masked off if unsupported and can't be the reason */
+		btrfs_error(root->fs_info, -EIO,
+			    "%d errors while writing supers", total_errors);
+		return -EIO;
 	}
 
 	total_errors = 0;
@@ -3421,6 +3477,8 @@ static void free_fs_root(struct btrfs_root *root)
 {
 	iput(root->cache_inode);
 	WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
+	btrfs_free_block_rsv(root, root->orphan_block_rsv);
+	root->orphan_block_rsv = NULL;
 	if (root->anon_dev)
 		free_anon_bdev(root->anon_dev);
 	free_extent_buffer(root->node);
@@ -3510,6 +3568,11 @@ int close_ctree(struct btrfs_root *root)
 	fs_info->closing = 1;
 	smp_mb();
 
+	/* wait for the uuid_scan task to finish */
+	down(&fs_info->uuid_tree_rescan_sem);
+	/* avoid complains from lockdep et al., set sem back to initial state */
+	up(&fs_info->uuid_tree_rescan_sem);
+
 	/* pause restriper - we want to resume on mount */
 	btrfs_pause_balance(fs_info);
 
@@ -3573,6 +3636,9 @@ int close_ctree(struct btrfs_root *root)
 
 	btrfs_free_stripe_hash_table(fs_info);
 
+	btrfs_free_block_rsv(root, root->orphan_block_rsv);
+	root->orphan_block_rsv = NULL;
+
 	return 0;
 }
 
@@ -3608,9 +3674,7 @@ void btrfs_mark_buffer_dirty(struct extent_buffer *buf)
 	if (transid != root->fs_info->generation)
 		WARN(1, KERN_CRIT "btrfs transid mismatch buffer %llu, "
 		       "found %llu running %llu\n",
-			(unsigned long long)buf->start,
-			(unsigned long long)transid,
-			(unsigned long long)root->fs_info->generation);
+			buf->start, transid, root->fs_info->generation);
 	was_dirty = set_extent_buffer_dirty(buf);
 	if (!was_dirty)
 		__percpu_counter_add(&root->fs_info->dirty_metadata_bytes,
@@ -3744,8 +3808,8 @@ static void btrfs_destroy_all_ordered_extents(struct btrfs_fs_info *fs_info)
 	spin_unlock(&fs_info->ordered_root_lock);
 }
 
-int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
-			       struct btrfs_root *root)
+static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
+				      struct btrfs_root *root)
 {
 	struct rb_node *node;
 	struct btrfs_delayed_ref_root *delayed_refs;

fs/btrfs/extent-tree.c

@@ -113,7 +113,8 @@ static noinline int
 block_group_cache_done(struct btrfs_block_group_cache *cache)
 {
 	smp_mb();
-	return cache->cached == BTRFS_CACHE_FINISHED;
+	return cache->cached == BTRFS_CACHE_FINISHED ||
+		cache->cached == BTRFS_CACHE_ERROR;
 }
 
 static int block_group_bits(struct btrfs_block_group_cache *cache, u64 bits)
@@ -389,7 +390,7 @@ static noinline void caching_thread(struct btrfs_work *work)
 	u64 total_found = 0;
 	u64 last = 0;
 	u32 nritems;
-	int ret = 0;
+	int ret = -ENOMEM;
 
 	caching_ctl = container_of(work, struct btrfs_caching_control, work);
 	block_group = caching_ctl->block_group;
@@ -420,6 +421,7 @@ again:
 	/* need to make sure the commit_root doesn't disappear */
 	down_read(&fs_info->extent_commit_sem);
 
+next:
 	ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
 	if (ret < 0)
 		goto err;
@@ -459,6 +461,16 @@ again:
 			continue;
 		}
 
+		if (key.objectid < last) {
+			key.objectid = last;
+			key.offset = 0;
+			key.type = BTRFS_EXTENT_ITEM_KEY;
+
+			caching_ctl->progress = last;
+			btrfs_release_path(path);
+			goto next;
+		}
+
 		if (key.objectid < block_group->key.objectid) {
 			path->slots[0]++;
 			continue;
@@ -506,6 +518,12 @@ err:
 
 	mutex_unlock(&caching_ctl->mutex);
 out:
+	if (ret) {
+		spin_lock(&block_group->lock);
+		block_group->caching_ctl = NULL;
+		block_group->cached = BTRFS_CACHE_ERROR;
+		spin_unlock(&block_group->lock);
+	}
 	wake_up(&caching_ctl->wait);
 
 	put_caching_control(caching_ctl);
@@ -771,10 +789,23 @@ again:
 		goto out_free;
 
 	if (ret > 0 && metadata && key.type == BTRFS_METADATA_ITEM_KEY) {
-		key.type = BTRFS_EXTENT_ITEM_KEY;
-		key.offset = root->leafsize;
-		btrfs_release_path(path);
-		goto again;
+		metadata = 0;
+		if (path->slots[0]) {
+			path->slots[0]--;
+			btrfs_item_key_to_cpu(path->nodes[0], &key,
+					      path->slots[0]);
+			if (key.objectid == bytenr &&
+			    key.type == BTRFS_EXTENT_ITEM_KEY &&
+			    key.offset == root->leafsize)
+				ret = 0;
+		}
+		if (ret) {
+			key.objectid = bytenr;
+			key.type = BTRFS_EXTENT_ITEM_KEY;
+			key.offset = root->leafsize;
+			btrfs_release_path(path);
+			goto again;
+		}
 	}
 
 	if (ret == 0) {
@@ -2011,6 +2042,8 @@ static int run_delayed_data_ref(struct btrfs_trans_handle *trans,
 	ins.type = BTRFS_EXTENT_ITEM_KEY;
 
 	ref = btrfs_delayed_node_to_data_ref(node);
+	trace_run_delayed_data_ref(node, ref, node->action);
+
 	if (node->type == BTRFS_SHARED_DATA_REF_KEY)
 		parent = ref->parent;
 	else
@@ -2154,6 +2187,8 @@ static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
 						 SKINNY_METADATA);
 
 	ref = btrfs_delayed_node_to_tree_ref(node);
+	trace_run_delayed_tree_ref(node, ref, node->action);
+
 	if (node->type == BTRFS_SHARED_BLOCK_REF_KEY)
 		parent = ref->parent;
 	else
@@ -2212,6 +2247,8 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 		 */
 		BUG_ON(extent_op);
 		head = btrfs_delayed_node_to_head(node);
+		trace_run_delayed_ref_head(node, head, node->action);
+
 		if (insert_reserved) {
 			btrfs_pin_extent(root, node->bytenr,
 					 node->num_bytes, 1);
@@ -2403,6 +2440,8 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 			default:
 				WARN_ON(1);
 			}
+		} else {
+			list_del_init(&locked_ref->cluster);
 		}
 		spin_unlock(&delayed_refs->lock);
 
@@ -2425,7 +2464,6 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 		 * list before we release it.
 		 */
 		if (btrfs_delayed_ref_is_head(ref)) {
-			list_del_init(&locked_ref->cluster);
 			btrfs_delayed_ref_unlock(locked_ref);
 			locked_ref = NULL;
 		}
@@ -3799,8 +3837,12 @@ again:
 	if (force < space_info->force_alloc)
 		force = space_info->force_alloc;
 	if (space_info->full) {
+		if (should_alloc_chunk(extent_root, space_info, force))
+			ret = -ENOSPC;
+		else
+			ret = 0;
 		spin_unlock(&space_info->lock);
-		return 0;
+		return ret;
 	}
 
 	if (!should_alloc_chunk(extent_root, space_info, force)) {
@@ -4320,6 +4362,9 @@ static struct btrfs_block_rsv *get_block_rsv(
 	if (root == root->fs_info->csum_root && trans->adding_csums)
 		block_rsv = trans->block_rsv;
 
+	if (root == root->fs_info->uuid_root)
+		block_rsv = trans->block_rsv;
+
 	if (!block_rsv)
 		block_rsv = root->block_rsv;
 
@@ -4729,10 +4774,12 @@ void btrfs_orphan_release_metadata(struct inode *inode)
 int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 				     struct btrfs_block_rsv *rsv,
 				     int items,
-				     u64 *qgroup_reserved)
+				     u64 *qgroup_reserved,
+				     bool use_global_rsv)
 {
 	u64 num_bytes;
 	int ret;
+	struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
 
 	if (root->fs_info->quota_enabled) {
 		/* One for parent inode, two for dir entries */
@@ -4751,6 +4798,10 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 					    BTRFS_BLOCK_GROUP_METADATA);
 	ret = btrfs_block_rsv_add(root, rsv, num_bytes,
 				  BTRFS_RESERVE_FLUSH_ALL);
+
+	if (ret == -ENOSPC && use_global_rsv)
+		ret = btrfs_block_rsv_migrate(global_rsv, rsv, num_bytes);
+
 	if (ret) {
 		if (*qgroup_reserved)
 			btrfs_qgroup_free(root, *qgroup_reserved);
@@ -5668,7 +5719,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 
 			if (ret) {
 				btrfs_err(info, "umm, got %d back from search, was looking for %llu",
-					ret, (unsigned long long)bytenr);
+					ret, bytenr);
 				if (ret > 0)
 					btrfs_print_leaf(extent_root,
 							 path->nodes[0]);
@@ -5684,11 +5735,8 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 		WARN_ON(1);
 		btrfs_err(info,
 			"unable to find ref byte nr %llu parent %llu root %llu  owner %llu offset %llu",
-			(unsigned long long)bytenr,
-			(unsigned long long)parent,
-			(unsigned long long)root_objectid,
-			(unsigned long long)owner_objectid,
-			(unsigned long long)owner_offset);
+			bytenr, parent, root_objectid, owner_objectid,
+			owner_offset);
 	} else {
 		btrfs_abort_transaction(trans, extent_root, ret);
 		goto out;
@@ -5717,7 +5765,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 					-1, 1);
 		if (ret) {
 			btrfs_err(info, "umm, got %d back from search, was looking for %llu",
-				ret, (unsigned long long)bytenr);
+				ret, bytenr);
 			btrfs_print_leaf(extent_root, path->nodes[0]);
 		}
 		if (ret < 0) {
@@ -5999,8 +6047,11 @@ static u64 stripe_align(struct btrfs_root *root,
  * for our min num_bytes.  Another option is to have it go ahead
  * and look in the rbtree for a free extent of a given size, but this
  * is a good start.
+ *
+ * Callers of this must check if cache->cached == BTRFS_CACHE_ERROR before using
+ * any of the information in this block group.
  */
-static noinline int
+static noinline void
 wait_block_group_cache_progress(struct btrfs_block_group_cache *cache,
 				u64 num_bytes)
 {
@@ -6008,28 +6059,29 @@ wait_block_group_cache_progress(struct btrfs_block_group_cache *cache,
 
 	caching_ctl = get_caching_control(cache);
 	if (!caching_ctl)
-		return 0;
+		return;
 
 	wait_event(caching_ctl->wait, block_group_cache_done(cache) ||
 		   (cache->free_space_ctl->free_space >= num_bytes));
 
 	put_caching_control(caching_ctl);
-	return 0;
 }
 
 static noinline int
 wait_block_group_cache_done(struct btrfs_block_group_cache *cache)
 {
 	struct btrfs_caching_control *caching_ctl;
+	int ret = 0;
 
 	caching_ctl = get_caching_control(cache);
 	if (!caching_ctl)
-		return 0;
+		return (cache->cached == BTRFS_CACHE_ERROR) ? -EIO : 0;
 
 	wait_event(caching_ctl->wait, block_group_cache_done(cache));
-
+	if (cache->cached == BTRFS_CACHE_ERROR)
+		ret = -EIO;
 	put_caching_control(caching_ctl);
-	return 0;
+	return ret;
 }
 
 int __get_raid_index(u64 flags)
@@ -6070,8 +6122,7 @@ enum btrfs_loop_type {
  * ins->offset == number of blocks
  * Any available blocks before search_start are skipped.
  */
-static noinline int find_free_extent(struct btrfs_trans_handle *trans,
-				     struct btrfs_root *orig_root,
+static noinline int find_free_extent(struct btrfs_root *orig_root,
 				     u64 num_bytes, u64 empty_size,
 				     u64 hint_byte, struct btrfs_key *ins,
 				     u64 flags)
@@ -6212,6 +6263,8 @@ have_block_group:
 			ret = 0;
 		}
 
+		if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
+			goto loop;
 		if (unlikely(block_group->ro))
 			goto loop;
 
@@ -6292,10 +6345,10 @@ refill_cluster:
 					      block_group->full_stripe_len);
 
 			/* allocate a cluster in this block group */
-			ret = btrfs_find_space_cluster(trans, root,
-					       block_group, last_ptr,
-					       search_start, num_bytes,
-					       aligned_cluster);
+			ret = btrfs_find_space_cluster(root, block_group,
+						       last_ptr, search_start,
+						       num_bytes,
+						       aligned_cluster);
 			if (ret == 0) {
 				/*
 				 * now pull our allocation out of this
@@ -6426,17 +6479,28 @@ loop:
 		index = 0;
 		loop++;
 		if (loop == LOOP_ALLOC_CHUNK) {
+			struct btrfs_trans_handle *trans;
+
+			trans = btrfs_join_transaction(root);
+			if (IS_ERR(trans)) {
+				ret = PTR_ERR(trans);
+				goto out;
+			}
+
 			ret = do_chunk_alloc(trans, root, flags,
 					     CHUNK_ALLOC_FORCE);
 			/*
 			 * Do not bail out on ENOSPC since we
 			 * can do more things.
 			 */
-			if (ret < 0 && ret != -ENOSPC) {
+			if (ret < 0 && ret != -ENOSPC)
 				btrfs_abort_transaction(trans,
 							root, ret);
+			else
+				ret = 0;
+			btrfs_end_transaction(trans, root);
+			if (ret)
 				goto out;
-			}
 		}
 
 		if (loop == LOOP_NO_EMPTY_SIZE) {
@@ -6463,19 +6527,15 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
 
 	spin_lock(&info->lock);
 	printk(KERN_INFO "space_info %llu has %llu free, is %sfull\n",
-	       (unsigned long long)info->flags,
-	       (unsigned long long)(info->total_bytes - info->bytes_used -
-				    info->bytes_pinned - info->bytes_reserved -
-				    info->bytes_readonly),
+	       info->flags,
+	       info->total_bytes - info->bytes_used - info->bytes_pinned -
+	       info->bytes_reserved - info->bytes_readonly,
 	       (info->full) ? "" : "not ");
 	printk(KERN_INFO "space_info total=%llu, used=%llu, pinned=%llu, "
 	       "reserved=%llu, may_use=%llu, readonly=%llu\n",
-	       (unsigned long long)info->total_bytes,
-	       (unsigned long long)info->bytes_used,
-	       (unsigned long long)info->bytes_pinned,
-	       (unsigned long long)info->bytes_reserved,
-	       (unsigned long long)info->bytes_may_use,
-	       (unsigned long long)info->bytes_readonly);
+	       info->total_bytes, info->bytes_used, info->bytes_pinned,
+	       info->bytes_reserved, info->bytes_may_use,
+	       info->bytes_readonly);
 	spin_unlock(&info->lock);
 
 	if (!dump_block_groups)
@@ -6486,12 +6546,9 @@ again:
 	list_for_each_entry(cache, &info->block_groups[index], list) {
 		spin_lock(&cache->lock);
 		printk(KERN_INFO "block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %s\n",
-		       (unsigned long long)cache->key.objectid,
-		       (unsigned long long)cache->key.offset,
-		       (unsigned long long)btrfs_block_group_used(&cache->item),
-		       (unsigned long long)cache->pinned,
-		       (unsigned long long)cache->reserved,
-		       cache->ro ? "[readonly]" : "");
+		       cache->key.objectid, cache->key.offset,
+		       btrfs_block_group_used(&cache->item), cache->pinned,
+		       cache->reserved, cache->ro ? "[readonly]" : "");
 		btrfs_dump_free_space(cache, bytes);
 		spin_unlock(&cache->lock);
 	}
@@ -6500,8 +6557,7 @@ again:
 	up_read(&info->groups_sem);
 }
 
-int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
-			 struct btrfs_root *root,
+int btrfs_reserve_extent(struct btrfs_root *root,
 			 u64 num_bytes, u64 min_alloc_size,
 			 u64 empty_size, u64 hint_byte,
 			 struct btrfs_key *ins, int is_data)
@@ -6513,8 +6569,8 @@ int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
 	flags = btrfs_get_alloc_profile(root, is_data);
 again:
 	WARN_ON(num_bytes < root->sectorsize);
-	ret = find_free_extent(trans, root, num_bytes, empty_size,
-			       hint_byte, ins, flags);
+	ret = find_free_extent(root, num_bytes, empty_size, hint_byte, ins,
+			       flags);
 
 	if (ret == -ENOSPC) {
 		if (!final_tried) {
@@ -6529,8 +6585,7 @@ again:
 
 			sinfo = __find_space_info(root->fs_info, flags);
 			btrfs_err(root->fs_info, "allocation failed flags %llu, wanted %llu",
-				(unsigned long long)flags,
-				(unsigned long long)num_bytes);
+				flags, num_bytes);
 			if (sinfo)
 				dump_space_info(sinfo, num_bytes, 1);
 		}
@@ -6550,7 +6605,7 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root,
 	cache = btrfs_lookup_block_group(root->fs_info, start);
 	if (!cache) {
 		btrfs_err(root->fs_info, "Unable to find block group for %llu",
-			(unsigned long long)start);
+			start);
 		return -ENOSPC;
 	}
 
@@ -6646,8 +6701,7 @@ static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 	ret = update_block_group(root, ins->objectid, ins->offset, 1);
 	if (ret) { /* -ENOENT, logic error */
 		btrfs_err(fs_info, "update block group failed for %llu %llu",
-			(unsigned long long)ins->objectid,
-			(unsigned long long)ins->offset);
+			ins->objectid, ins->offset);
 		BUG();
 	}
 	return ret;
@@ -6719,8 +6773,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 	ret = update_block_group(root, ins->objectid, root->leafsize, 1);
 	if (ret) { /* -ENOENT, logic error */
 		btrfs_err(fs_info, "update block group failed for %llu %llu",
-			(unsigned long long)ins->objectid,
-			(unsigned long long)ins->offset);
+			ins->objectid, ins->offset);
 		BUG();
 	}
 	return ret;
@@ -6902,7 +6955,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
 	if (IS_ERR(block_rsv))
 		return ERR_CAST(block_rsv);
 
-	ret = btrfs_reserve_extent(trans, root, blocksize, blocksize,
+	ret = btrfs_reserve_extent(root, blocksize, blocksize,
 				   empty_size, hint, &ins, 0);
 	if (ret) {
 		unuse_block_rsv(root->fs_info, block_rsv, blocksize);
@@ -7173,6 +7226,8 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 		next = btrfs_find_create_tree_block(root, bytenr, blocksize);
 		if (!next)
 			return -ENOMEM;
+		btrfs_set_buffer_lockdep_class(root->root_key.objectid, next,
+					       level - 1);
 		reada = 1;
 	}
 	btrfs_tree_lock(next);
@@ -7658,7 +7713,7 @@ out:
 	 * don't have it in the radix (like when we recover after a power fail
 	 * or unmount) so we don't leak memory.
 	 */
-	if (root_dropped == false)
+	if (!for_reloc && root_dropped == false)
 		btrfs_add_dead_root(root);
 	if (err)
 		btrfs_std_error(root->fs_info, err);
@@ -8192,7 +8247,8 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 		 * We haven't cached this block group, which means we could
 		 * possibly have excluded extents on this block group.
 		 */
-		if (block_group->cached == BTRFS_CACHE_NO)
+		if (block_group->cached == BTRFS_CACHE_NO ||
+		    block_group->cached == BTRFS_CACHE_ERROR)
 			free_excluded_extents(info->extent_root, block_group);
 
 		btrfs_remove_free_space_cache(block_group);
@@ -8409,9 +8465,13 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		 * avoid allocating from un-mirrored block group if there are
 		 * mirrored block groups.
 		 */
-		list_for_each_entry(cache, &space_info->block_groups[3], list)
+		list_for_each_entry(cache,
+				&space_info->block_groups[BTRFS_RAID_RAID0],
+				list)
 			set_block_group_ro(cache, 1);
-		list_for_each_entry(cache, &space_info->block_groups[4], list)
+		list_for_each_entry(cache,
+				&space_info->block_groups[BTRFS_RAID_SINGLE],
+				list)
 			set_block_group_ro(cache, 1);
 	}
 

fs/btrfs/extent_io.c

@@ -61,9 +61,8 @@ void btrfs_leak_debug_check(void)
 		state = list_entry(states.next, struct extent_state, leak_list);
 		printk(KERN_ERR "btrfs state leak: start %llu end %llu "
 		       "state %lu in tree %p refs %d\n",
-		       (unsigned long long)state->start,
-		       (unsigned long long)state->end,
-		       state->state, state->tree, atomic_read(&state->refs));
+		       state->start, state->end, state->state, state->tree,
+		       atomic_read(&state->refs));
 		list_del(&state->leak_list);
 		kmem_cache_free(extent_state_cache, state);
 	}
@@ -71,8 +70,8 @@ void btrfs_leak_debug_check(void)
 	while (!list_empty(&buffers)) {
 		eb = list_entry(buffers.next, struct extent_buffer, leak_list);
 		printk(KERN_ERR "btrfs buffer leak start %llu len %lu "
-		       "refs %d\n", (unsigned long long)eb->start,
-		       eb->len, atomic_read(&eb->refs));
+		       "refs %d\n",
+		       eb->start, eb->len, atomic_read(&eb->refs));
 		list_del(&eb->leak_list);
 		kmem_cache_free(extent_buffer_cache, eb);
 	}
@@ -88,11 +87,7 @@ static inline void __btrfs_debug_check_extent_io_range(const char *caller,
 	if (end >= PAGE_SIZE && (end % 2) == 0 && end != isize - 1) {
 		printk_ratelimited(KERN_DEBUG
 		    "btrfs: %s: ino %llu isize %llu odd range [%llu,%llu]\n",
-				caller,
-				(unsigned long long)btrfs_ino(inode),
-				(unsigned long long)isize,
-				(unsigned long long)start,
-				(unsigned long long)end);
+				caller, btrfs_ino(inode), isize, start, end);
 	}
 }
 #else
@@ -388,8 +383,7 @@ static int insert_state(struct extent_io_tree *tree,
 
 	if (end < start)
 		WARN(1, KERN_ERR "btrfs end < start %llu %llu\n",
-		       (unsigned long long)end,
-		       (unsigned long long)start);
+		       end, start);
 	state->start = start;
 	state->end = end;
 
@@ -400,9 +394,8 @@ static int insert_state(struct extent_io_tree *tree,
 		struct extent_state *found;
 		found = rb_entry(node, struct extent_state, rb_node);
 		printk(KERN_ERR "btrfs found node %llu %llu on insert of "
-		       "%llu %llu\n", (unsigned long long)found->start,
-		       (unsigned long long)found->end,
-		       (unsigned long long)start, (unsigned long long)end);
+		       "%llu %llu\n",
+		       found->start, found->end, start, end);
 		return -EEXIST;
 	}
 	state->tree = tree;
@@ -762,15 +755,6 @@ static void cache_state(struct extent_state *state,
 	}
 }
 
-static void uncache_state(struct extent_state **cached_ptr)
-{
-	if (cached_ptr && (*cached_ptr)) {
-		struct extent_state *state = *cached_ptr;
-		*cached_ptr = NULL;
-		free_extent_state(state);
-	}
-}
-
 /*
  * set some bits on a range in the tree.  This may require allocations or
  * sleeping, so the gfp mask is used to indicate what is allowed.
@@ -1687,31 +1671,21 @@ out_failed:
 	return found;
 }
 
-int extent_clear_unlock_delalloc(struct inode *inode,
-				struct extent_io_tree *tree,
-				u64 start, u64 end, struct page *locked_page,
-				unsigned long op)
+int extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end,
+				 struct page *locked_page,
+				 unsigned long clear_bits,
+				 unsigned long page_ops)
 {
+	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;
 	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;
-	unsigned long clear_bits = 0;
-
-	if (op & EXTENT_CLEAR_UNLOCK)
-		clear_bits |= EXTENT_LOCKED;
-	if (op & EXTENT_CLEAR_DIRTY)
-		clear_bits |= EXTENT_DIRTY;
-
-	if (op & EXTENT_CLEAR_DELALLOC)
-		clear_bits |= EXTENT_DELALLOC;
 
 	clear_extent_bit(tree, start, end, clear_bits, 1, 0, NULL, GFP_NOFS);
-	if (!(op & (EXTENT_CLEAR_UNLOCK_PAGE | EXTENT_CLEAR_DIRTY |
-		    EXTENT_SET_WRITEBACK | EXTENT_END_WRITEBACK |
-		    EXTENT_SET_PRIVATE2)))
+	if (page_ops == 0)
 		return 0;
 
 	while (nr_pages > 0) {
@@ -1720,20 +1694,20 @@ int extent_clear_unlock_delalloc(struct inode *inode,
 				     nr_pages, ARRAY_SIZE(pages)), pages);
 		for (i = 0; i < ret; i++) {
 
-			if (op & EXTENT_SET_PRIVATE2)
+			if (page_ops & PAGE_SET_PRIVATE2)
 				SetPagePrivate2(pages[i]);
 
 			if (pages[i] == locked_page) {
 				page_cache_release(pages[i]);
 				continue;
 			}
-			if (op & EXTENT_CLEAR_DIRTY)
+			if (page_ops & PAGE_CLEAR_DIRTY)
 				clear_page_dirty_for_io(pages[i]);
-			if (op & EXTENT_SET_WRITEBACK)
+			if (page_ops & PAGE_SET_WRITEBACK)
 				set_page_writeback(pages[i]);
-			if (op & EXTENT_END_WRITEBACK)
+			if (page_ops & PAGE_END_WRITEBACK)
 				end_page_writeback(pages[i]);
-			if (op & EXTENT_CLEAR_UNLOCK_PAGE)
+			if (page_ops & PAGE_UNLOCK)
 				unlock_page(pages[i]);
 			page_cache_release(pages[i]);
 		}
@@ -1810,7 +1784,7 @@ out:
  * set the private field for a given byte offset in the tree.  If there isn't
  * an extent_state there already, this does nothing.
  */
-int set_state_private(struct extent_io_tree *tree, u64 start, u64 private)
+static int set_state_private(struct extent_io_tree *tree, u64 start, u64 private)
 {
 	struct rb_node *node;
 	struct extent_state *state;
@@ -1837,64 +1811,6 @@ out:
 	return ret;
 }
 
-void extent_cache_csums_dio(struct extent_io_tree *tree, u64 start, u32 csums[],
-			    int count)
-{
-	struct rb_node *node;
-	struct extent_state *state;
-
-	spin_lock(&tree->lock);
-	/*
-	 * this search will find all the extents that end after
-	 * our range starts.
-	 */
-	node = tree_search(tree, start);
-	BUG_ON(!node);
-
-	state = rb_entry(node, struct extent_state, rb_node);
-	BUG_ON(state->start != start);
-
-	while (count) {
-		state->private = *csums++;
-		count--;
-		state = next_state(state);
-	}
-	spin_unlock(&tree->lock);
-}
-
-static inline u64 __btrfs_get_bio_offset(struct bio *bio, int bio_index)
-{
-	struct bio_vec *bvec = bio->bi_io_vec + bio_index;
-
-	return page_offset(bvec->bv_page) + bvec->bv_offset;
-}
-
-void extent_cache_csums(struct extent_io_tree *tree, struct bio *bio, int bio_index,
-			u32 csums[], int count)
-{
-	struct rb_node *node;
-	struct extent_state *state = NULL;
-	u64 start;
-
-	spin_lock(&tree->lock);
-	do {
-		start = __btrfs_get_bio_offset(bio, bio_index);
-		if (state == NULL || state->start != start) {
-			node = tree_search(tree, start);
-			BUG_ON(!node);
-
-			state = rb_entry(node, struct extent_state, rb_node);
-			BUG_ON(state->start != start);
-		}
-		state->private = *csums++;
-		count--;
-		bio_index++;
-
-		state = next_state(state);
-	} while (count);
-	spin_unlock(&tree->lock);
-}
-
 int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private)
 {
 	struct rb_node *node;
@@ -2173,7 +2089,8 @@ static int clean_io_failure(u64 start, struct page *page)
 					    EXTENT_LOCKED);
 	spin_unlock(&BTRFS_I(inode)->io_tree.lock);
 
-	if (state && state->start == failrec->start) {
+	if (state && state->start <= failrec->start &&
+	    state->end >= failrec->start + failrec->len - 1) {
 		fs_info = BTRFS_I(inode)->root->fs_info;
 		num_copies = btrfs_num_copies(fs_info, failrec->logical,
 					      failrec->len);
@@ -2201,9 +2118,9 @@ out:
  * needed
  */
 
-static int bio_readpage_error(struct bio *failed_bio, struct page *page,
-				u64 start, u64 end, int failed_mirror,
-				struct extent_state *state)
+static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
+			      struct page *page, u64 start, u64 end,
+			      int failed_mirror)
 {
 	struct io_failure_record *failrec = NULL;
 	u64 private;
@@ -2213,6 +2130,8 @@ static int bio_readpage_error(struct bio *failed_bio, struct page *page,
 	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;
 	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
 	struct bio *bio;
+	struct btrfs_io_bio *btrfs_failed_bio;
+	struct btrfs_io_bio *btrfs_bio;
 	int num_copies;
 	int ret;
 	int read_mode;
@@ -2296,23 +2215,12 @@ static int bio_readpage_error(struct bio *failed_bio, struct page *page,
 		 * all the retry and error correction code that follows. no
 		 * matter what the error is, it is very likely to persist.
 		 */
-		pr_debug("bio_readpage_error: cannot repair, num_copies == 1. "
-			 "state=%p, num_copies=%d, next_mirror %d, "
-			 "failed_mirror %d\n", state, num_copies,
-			 failrec->this_mirror, failed_mirror);
+		pr_debug("bio_readpage_error: cannot repair, num_copies=%d, next_mirror %d, failed_mirror %d\n",
+			 num_copies, failrec->this_mirror, failed_mirror);
 		free_io_failure(inode, failrec, 0);
 		return -EIO;
 	}
 
-	if (!state) {
-		spin_lock(&tree->lock);
-		state = find_first_extent_bit_state(tree, failrec->start,
-						    EXTENT_LOCKED);
-		if (state && state->start != failrec->start)
-			state = NULL;
-		spin_unlock(&tree->lock);
-	}
-
 	/*
 	 * there are two premises:
 	 *	a) deliver good data to the caller
@@ -2349,9 +2257,8 @@ static int bio_readpage_error(struct bio *failed_bio, struct page *page,
 		read_mode = READ_SYNC;
 	}
 
-	if (!state || failrec->this_mirror > num_copies) {
-		pr_debug("bio_readpage_error: (fail) state=%p, num_copies=%d, "
-			 "next_mirror %d, failed_mirror %d\n", state,
+	if (failrec->this_mirror > num_copies) {
+		pr_debug("bio_readpage_error: (fail) num_copies=%d, next_mirror %d, failed_mirror %d\n",
 			 num_copies, failrec->this_mirror, failed_mirror);
 		free_io_failure(inode, failrec, 0);
 		return -EIO;
@@ -2362,12 +2269,24 @@ static int bio_readpage_error(struct bio *failed_bio, struct page *page,
 		free_io_failure(inode, failrec, 0);
 		return -EIO;
 	}
-	bio->bi_private = state;
 	bio->bi_end_io = failed_bio->bi_end_io;
 	bio->bi_sector = failrec->logical >> 9;
 	bio->bi_bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
 	bio->bi_size = 0;
 
+	btrfs_failed_bio = btrfs_io_bio(failed_bio);
+	if (btrfs_failed_bio->csum) {
+		struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
+		u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
+
+		btrfs_bio = btrfs_io_bio(bio);
+		btrfs_bio->csum = btrfs_bio->csum_inline;
+		phy_offset >>= inode->i_sb->s_blocksize_bits;
+		phy_offset *= csum_size;
+		memcpy(btrfs_bio->csum, btrfs_failed_bio->csum + phy_offset,
+		       csum_size);
+	}
+
 	bio_add_page(bio, page, failrec->len, start - page_offset(page));
 
 	pr_debug("bio_readpage_error: submitting new read[%#x] to "
@@ -2450,6 +2369,18 @@ static void end_bio_extent_writepage(struct bio *bio, int err)
 	bio_put(bio);
 }
 
+static void
+endio_readpage_release_extent(struct extent_io_tree *tree, u64 start, u64 len,
+			      int uptodate)
+{
+	struct extent_state *cached = NULL;
+	u64 end = start + len - 1;
+
+	if (uptodate && tree->track_uptodate)
+		set_extent_uptodate(tree, start, end, &cached, GFP_ATOMIC);
+	unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC);
+}
+
 /*
  * after a readpage IO is done, we need to:
  * clear the uptodate bits on error
@@ -2466,9 +2397,14 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
 	struct bio_vec *bvec_end = bio->bi_io_vec + bio->bi_vcnt - 1;
 	struct bio_vec *bvec = bio->bi_io_vec;
+	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
 	struct extent_io_tree *tree;
+	u64 offset = 0;
 	u64 start;
 	u64 end;
+	u64 len;
+	u64 extent_start = 0;
+	u64 extent_len = 0;
 	int mirror;
 	int ret;
 
@@ -2477,9 +2413,6 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 
 	do {
 		struct page *page = bvec->bv_page;
-		struct extent_state *cached = NULL;
-		struct extent_state *state;
-		struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
 		struct inode *inode = page->mapping->host;
 
 		pr_debug("end_bio_extent_readpage: bi_sector=%llu, err=%d, "
@@ -2500,37 +2433,32 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 
 		start = page_offset(page);
 		end = start + bvec->bv_offset + bvec->bv_len - 1;
+		len = bvec->bv_len;
 
 		if (++bvec <= bvec_end)
 			prefetchw(&bvec->bv_page->flags);
 
-		spin_lock(&tree->lock);
-		state = find_first_extent_bit_state(tree, start, EXTENT_LOCKED);
-		if (state && state->start == start) {
-			/*
-			 * take a reference on the state, unlock will drop
-			 * the ref
-			 */
-			cache_state(state, &cached);
-		}
-		spin_unlock(&tree->lock);
-
 		mirror = io_bio->mirror_num;
-		if (uptodate && tree->ops && tree->ops->readpage_end_io_hook) {
-			ret = tree->ops->readpage_end_io_hook(page, start, end,
-							      state, mirror);
+		if (likely(uptodate && tree->ops &&
+			   tree->ops->readpage_end_io_hook)) {
+			ret = tree->ops->readpage_end_io_hook(io_bio, offset,
+							      page, start, end,
+							      mirror);
 			if (ret)
 				uptodate = 0;
 			else
 				clean_io_failure(start, page);
 		}
 
-		if (!uptodate && tree->ops && tree->ops->readpage_io_failed_hook) {
+		if (likely(uptodate))
+			goto readpage_ok;
+
+		if (tree->ops && tree->ops->readpage_io_failed_hook) {
 			ret = tree->ops->readpage_io_failed_hook(page, mirror);
 			if (!ret && !err &&
 			    test_bit(BIO_UPTODATE, &bio->bi_flags))
 				uptodate = 1;
-		} else if (!uptodate) {
+		} else {
 			/*
 			 * The generic bio_readpage_error handles errors the
 			 * following way: If possible, new read requests are
@@ -2541,24 +2469,18 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 			 * can't handle the error it will return -EIO and we
 			 * remain responsible for that page.
 			 */
-			ret = bio_readpage_error(bio, page, start, end, mirror, NULL);
+			ret = bio_readpage_error(bio, offset, page, start, end,
+						 mirror);
 			if (ret == 0) {
 				uptodate =
 					test_bit(BIO_UPTODATE, &bio->bi_flags);
 				if (err)
 					uptodate = 0;
-				uncache_state(&cached);
 				continue;
 			}
 		}
-
-		if (uptodate && tree->track_uptodate) {
-			set_extent_uptodate(tree, start, end, &cached,
-					    GFP_ATOMIC);
-		}
-		unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC);
-
-		if (uptodate) {
+readpage_ok:
+		if (likely(uptodate)) {
 			loff_t i_size = i_size_read(inode);
 			pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
 			unsigned offset;
@@ -2573,8 +2495,36 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 			SetPageError(page);
 		}
 		unlock_page(page);
+		offset += len;
+
+		if (unlikely(!uptodate)) {
+			if (extent_len) {
+				endio_readpage_release_extent(tree,
+							      extent_start,
+							      extent_len, 1);
+				extent_start = 0;
+				extent_len = 0;
+			}
+			endio_readpage_release_extent(tree, start,
+						      end - start + 1, 0);
+		} else if (!extent_len) {
+			extent_start = start;
+			extent_len = end + 1 - start;
+		} else if (extent_start + extent_len == start) {
+			extent_len += end + 1 - start;
+		} else {
+			endio_readpage_release_extent(tree, extent_start,
+						      extent_len, uptodate);
+			extent_start = start;
+			extent_len = end + 1 - start;
+		}
 	} while (bvec <= bvec_end);
 
+	if (extent_len)
+		endio_readpage_release_extent(tree, extent_start, extent_len,
+					      uptodate);
+	if (io_bio->end_io)
+		io_bio->end_io(io_bio, err);
 	bio_put(bio);
 }
 
@@ -2586,6 +2536,7 @@ struct bio *
 btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
 		gfp_t gfp_flags)
 {
+	struct btrfs_io_bio *btrfs_bio;
 	struct bio *bio;
 
 	bio = bio_alloc_bioset(gfp_flags, nr_vecs, btrfs_bioset);
@@ -2601,6 +2552,10 @@ btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
 		bio->bi_size = 0;
 		bio->bi_bdev = bdev;
 		bio->bi_sector = first_sector;
+		btrfs_bio = btrfs_io_bio(bio);
+		btrfs_bio->csum = NULL;
+		btrfs_bio->csum_allocated = NULL;
+		btrfs_bio->end_io = NULL;
 	}
 	return bio;
 }
@@ -2614,7 +2569,17 @@ struct bio *btrfs_bio_clone(struct bio *bio, gfp_t gfp_mask)
 /* this also allocates from the btrfs_bioset */
 struct bio *btrfs_io_bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs)
 {
-	return bio_alloc_bioset(gfp_mask, nr_iovecs, btrfs_bioset);
+	struct btrfs_io_bio *btrfs_bio;
+	struct bio *bio;
+
+	bio = bio_alloc_bioset(gfp_mask, nr_iovecs, btrfs_bioset);
+	if (bio) {
+		btrfs_bio = btrfs_io_bio(bio);
+		btrfs_bio->csum = NULL;
+		btrfs_bio->csum_allocated = NULL;
+		btrfs_bio->end_io = NULL;
+	}
+	return bio;
 }
 
 
@@ -2738,17 +2703,45 @@ void set_page_extent_mapped(struct page *page)
 	}
 }
 
+static struct extent_map *
+__get_extent_map(struct inode *inode, struct page *page, size_t pg_offset,
+		 u64 start, u64 len, get_extent_t *get_extent,
+		 struct extent_map **em_cached)
+{
+	struct extent_map *em;
+
+	if (em_cached && *em_cached) {
+		em = *em_cached;
+		if (em->in_tree && start >= em->start &&
+		    start < extent_map_end(em)) {
+			atomic_inc(&em->refs);
+			return em;
+		}
+
+		free_extent_map(em);
+		*em_cached = NULL;
+	}
+
+	em = get_extent(inode, page, pg_offset, start, len, 0);
+	if (em_cached && !IS_ERR_OR_NULL(em)) {
+		BUG_ON(*em_cached);
+		atomic_inc(&em->refs);
+		*em_cached = em;
+	}
+	return em;
+}
 /*
  * basic readpage implementation.  Locked extent state structs are inserted
  * into the tree that are removed when the IO is done (by the end_io
  * handlers)
  * XXX JDM: This needs looking at to ensure proper page locking
  */
-static int __extent_read_full_page(struct extent_io_tree *tree,
-				   struct page *page,
-				   get_extent_t *get_extent,
-				   struct bio **bio, int mirror_num,
-				   unsigned long *bio_flags, int rw)
+static int __do_readpage(struct extent_io_tree *tree,
+			 struct page *page,
+			 get_extent_t *get_extent,
+			 struct extent_map **em_cached,
+			 struct bio **bio, int mirror_num,
+			 unsigned long *bio_flags, int rw)
 {
 	struct inode *inode = page->mapping->host;
 	u64 start = page_offset(page);
@@ -2762,35 +2755,26 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 	sector_t sector;
 	struct extent_map *em;
 	struct block_device *bdev;
-	struct btrfs_ordered_extent *ordered;
 	int ret;
 	int nr = 0;
+	int parent_locked = *bio_flags & EXTENT_BIO_PARENT_LOCKED;
 	size_t pg_offset = 0;
 	size_t iosize;
 	size_t disk_io_size;
 	size_t blocksize = inode->i_sb->s_blocksize;
-	unsigned long this_bio_flag = 0;
+	unsigned long this_bio_flag = *bio_flags & EXTENT_BIO_PARENT_LOCKED;
 
 	set_page_extent_mapped(page);
 
+	end = page_end;
 	if (!PageUptodate(page)) {
 		if (cleancache_get_page(page) == 0) {
 			BUG_ON(blocksize != PAGE_SIZE);
+			unlock_extent(tree, start, end);
 			goto out;
 		}
 	}
 
-	end = page_end;
-	while (1) {
-		lock_extent(tree, start, end);
-		ordered = btrfs_lookup_ordered_extent(inode, start);
-		if (!ordered)
-			break;
-		unlock_extent(tree, start, end);
-		btrfs_start_ordered_extent(inode, ordered, 1);
-		btrfs_put_ordered_extent(ordered);
-	}
-
 	if (page->index == last_byte >> PAGE_CACHE_SHIFT) {
 		char *userpage;
 		size_t zero_offset = last_byte & (PAGE_CACHE_SIZE - 1);
@@ -2817,15 +2801,18 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 			kunmap_atomic(userpage);
 			set_extent_uptodate(tree, cur, cur + iosize - 1,
 					    &cached, GFP_NOFS);
-			unlock_extent_cached(tree, cur, cur + iosize - 1,
-					     &cached, GFP_NOFS);
+			if (!parent_locked)
+				unlock_extent_cached(tree, cur,
+						     cur + iosize - 1,
+						     &cached, GFP_NOFS);
 			break;
 		}
-		em = get_extent(inode, page, pg_offset, cur,
-				end - cur + 1, 0);
+		em = __get_extent_map(inode, page, pg_offset, cur,
+				      end - cur + 1, get_extent, em_cached);
 		if (IS_ERR_OR_NULL(em)) {
 			SetPageError(page);
-			unlock_extent(tree, cur, end);
+			if (!parent_locked)
+				unlock_extent(tree, cur, end);
 			break;
 		}
 		extent_offset = cur - em->start;
@@ -2833,7 +2820,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 		BUG_ON(end < cur);
 
 		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) {
-			this_bio_flag = EXTENT_BIO_COMPRESSED;
+			this_bio_flag |= EXTENT_BIO_COMPRESSED;
 			extent_set_compress_type(&this_bio_flag,
 						 em->compress_type);
 		}
@@ -2877,7 +2864,8 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 		if (test_range_bit(tree, cur, cur_end,
 				   EXTENT_UPTODATE, 1, NULL)) {
 			check_page_uptodate(tree, page);
-			unlock_extent(tree, cur, cur + iosize - 1);
+			if (!parent_locked)
+				unlock_extent(tree, cur, cur + iosize - 1);
 			cur = cur + iosize;
 			pg_offset += iosize;
 			continue;
@@ -2887,7 +2875,8 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 		 */
 		if (block_start == EXTENT_MAP_INLINE) {
 			SetPageError(page);
-			unlock_extent(tree, cur, cur + iosize - 1);
+			if (!parent_locked)
+				unlock_extent(tree, cur, cur + iosize - 1);
 			cur = cur + iosize;
 			pg_offset += iosize;
 			continue;
@@ -2905,7 +2894,8 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 			*bio_flags = this_bio_flag;
 		} else {
 			SetPageError(page);
-			unlock_extent(tree, cur, cur + iosize - 1);
+			if (!parent_locked)
+				unlock_extent(tree, cur, cur + iosize - 1);
 		}
 		cur = cur + iosize;
 		pg_offset += iosize;
@@ -2919,6 +2909,104 @@ out:
 	return 0;
 }
 
+static inline void __do_contiguous_readpages(struct extent_io_tree *tree,
+					     struct page *pages[], int nr_pages,
+					     u64 start, u64 end,
+					     get_extent_t *get_extent,
+					     struct extent_map **em_cached,
+					     struct bio **bio, int mirror_num,
+					     unsigned long *bio_flags, int rw)
+{
+	struct inode *inode;
+	struct btrfs_ordered_extent *ordered;
+	int index;
+
+	inode = pages[0]->mapping->host;
+	while (1) {
+		lock_extent(tree, start, end);
+		ordered = btrfs_lookup_ordered_range(inode, start,
+						     end - start + 1);
+		if (!ordered)
+			break;
+		unlock_extent(tree, start, end);
+		btrfs_start_ordered_extent(inode, ordered, 1);
+		btrfs_put_ordered_extent(ordered);
+	}
+
+	for (index = 0; index < nr_pages; index++) {
+		__do_readpage(tree, pages[index], get_extent, em_cached, bio,
+			      mirror_num, bio_flags, rw);
+		page_cache_release(pages[index]);
+	}
+}
+
+static void __extent_readpages(struct extent_io_tree *tree,
+			       struct page *pages[],
+			       int nr_pages, get_extent_t *get_extent,
+			       struct extent_map **em_cached,
+			       struct bio **bio, int mirror_num,
+			       unsigned long *bio_flags, int rw)
+{
+	u64 start = 0;
+	u64 end = 0;
+	u64 page_start;
+	int index;
+	int first_index = 0;
+
+	for (index = 0; index < nr_pages; index++) {
+		page_start = page_offset(pages[index]);
+		if (!end) {
+			start = page_start;
+			end = start + PAGE_CACHE_SIZE - 1;
+			first_index = index;
+		} else if (end + 1 == page_start) {
+			end += PAGE_CACHE_SIZE;
+		} else {
+			__do_contiguous_readpages(tree, &pages[first_index],
+						  index - first_index, start,
+						  end, get_extent, em_cached,
+						  bio, mirror_num, bio_flags,
+						  rw);
+			start = page_start;
+			end = start + PAGE_CACHE_SIZE - 1;
+			first_index = index;
+		}
+	}
+
+	if (end)
+		__do_contiguous_readpages(tree, &pages[first_index],
+					  index - first_index, start,
+					  end, get_extent, em_cached, bio,
+					  mirror_num, bio_flags, rw);
+}
+
+static int __extent_read_full_page(struct extent_io_tree *tree,
+				   struct page *page,
+				   get_extent_t *get_extent,
+				   struct bio **bio, int mirror_num,
+				   unsigned long *bio_flags, int rw)
+{
+	struct inode *inode = page->mapping->host;
+	struct btrfs_ordered_extent *ordered;
+	u64 start = page_offset(page);
+	u64 end = start + PAGE_CACHE_SIZE - 1;
+	int ret;
+
+	while (1) {
+		lock_extent(tree, start, end);
+		ordered = btrfs_lookup_ordered_extent(inode, start);
+		if (!ordered)
+			break;
+		unlock_extent(tree, start, end);
+		btrfs_start_ordered_extent(inode, ordered, 1);
+		btrfs_put_ordered_extent(ordered);
+	}
+
+	ret = __do_readpage(tree, page, get_extent, NULL, bio, mirror_num,
+			    bio_flags, rw);
+	return ret;
+}
+
 int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 			    get_extent_t *get_extent, int mirror_num)
 {
@@ -2933,6 +3021,20 @@ int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 	return ret;
 }
 
+int extent_read_full_page_nolock(struct extent_io_tree *tree, struct page *page,
+				 get_extent_t *get_extent, int mirror_num)
+{
+	struct bio *bio = NULL;
+	unsigned long bio_flags = EXTENT_BIO_PARENT_LOCKED;
+	int ret;
+
+	ret = __do_readpage(tree, page, get_extent, NULL, &bio, mirror_num,
+				      &bio_flags, READ);
+	if (bio)
+		ret = submit_one_bio(READ, bio, mirror_num, bio_flags);
+	return ret;
+}
+
 static noinline void update_nr_written(struct page *page,
 				      struct writeback_control *wbc,
 				      unsigned long nr_written)
@@ -3189,8 +3291,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc,
 			if (!PageWriteback(page)) {
 				printk(KERN_ERR "btrfs warning page %lu not "
 				       "writeback, cur %llu end %llu\n",
-				       page->index, (unsigned long long)cur,
-				       (unsigned long long)end);
+				       page->index, cur, end);
 			}
 
 			ret = submit_extent_page(write_flags, tree, page,
@@ -3769,7 +3870,7 @@ int extent_readpages(struct extent_io_tree *tree,
 	unsigned long bio_flags = 0;
 	struct page *pagepool[16];
 	struct page *page;
-	int i = 0;
+	struct extent_map *em_cached = NULL;
 	int nr = 0;
 
 	for (page_idx = 0; page_idx < nr_pages; page_idx++) {
@@ -3786,18 +3887,16 @@ int extent_readpages(struct extent_io_tree *tree,
 		pagepool[nr++] = page;
 		if (nr < ARRAY_SIZE(pagepool))
 			continue;
-		for (i = 0; i < nr; i++) {
-			__extent_read_full_page(tree, pagepool[i], get_extent,
-					&bio, 0, &bio_flags, READ);
-			page_cache_release(pagepool[i]);
-		}
+		__extent_readpages(tree, pagepool, nr, get_extent, &em_cached,
+				   &bio, 0, &bio_flags, READ);
 		nr = 0;
 	}
-	for (i = 0; i < nr; i++) {
-		__extent_read_full_page(tree, pagepool[i], get_extent,
-					&bio, 0, &bio_flags, READ);
-		page_cache_release(pagepool[i]);
-	}
+	if (nr)
+		__extent_readpages(tree, pagepool, nr, get_extent, &em_cached,
+				   &bio, 0, &bio_flags, READ);
+
+	if (em_cached)
+		free_extent_map(em_cached);
 
 	BUG_ON(!list_empty(pages));
 	if (bio)
@@ -4136,6 +4235,76 @@ static void __free_extent_buffer(struct extent_buffer *eb)
 	kmem_cache_free(extent_buffer_cache, eb);
 }
 
+static int extent_buffer_under_io(struct extent_buffer *eb)
+{
+	return (atomic_read(&eb->io_pages) ||
+		test_bit(EXTENT_BUFFER_WRITEBACK, &eb->bflags) ||
+		test_bit(EXTENT_BUFFER_DIRTY, &eb->bflags));
+}
+
+/*
+ * Helper for releasing extent buffer page.
+ */
+static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
+						unsigned long start_idx)
+{
+	unsigned long index;
+	unsigned long num_pages;
+	struct page *page;
+	int mapped = !test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags);
+
+	BUG_ON(extent_buffer_under_io(eb));
+
+	num_pages = num_extent_pages(eb->start, eb->len);
+	index = start_idx + num_pages;
+	if (start_idx >= index)
+		return;
+
+	do {
+		index--;
+		page = extent_buffer_page(eb, index);
+		if (page && mapped) {
+			spin_lock(&page->mapping->private_lock);
+			/*
+			 * We do this since we'll remove the pages after we've
+			 * removed the eb from the radix tree, so we could race
+			 * and have this page now attached to the new eb.  So
+			 * only clear page_private if it's still connected to
+			 * this eb.
+			 */
+			if (PagePrivate(page) &&
+			    page->private == (unsigned long)eb) {
+				BUG_ON(test_bit(EXTENT_BUFFER_DIRTY, &eb->bflags));
+				BUG_ON(PageDirty(page));
+				BUG_ON(PageWriteback(page));
+				/*
+				 * We need to make sure we haven't be attached
+				 * to a new eb.
+				 */
+				ClearPagePrivate(page);
+				set_page_private(page, 0);
+				/* One for the page private */
+				page_cache_release(page);
+			}
+			spin_unlock(&page->mapping->private_lock);
+
+		}
+		if (page) {
+			/* One for when we alloced the page */
+			page_cache_release(page);
+		}
+	} while (index != start_idx);
+}
+
+/*
+ * Helper for releasing the extent buffer.
+ */
+static inline void btrfs_release_extent_buffer(struct extent_buffer *eb)
+{
+	btrfs_release_extent_buffer_page(eb, 0);
+	__free_extent_buffer(eb);
+}
+
 static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
 						   u64 start,
 						   unsigned long len,
@@ -4184,13 +4353,16 @@ struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src)
 	struct extent_buffer *new;
 	unsigned long num_pages = num_extent_pages(src->start, src->len);
 
-	new = __alloc_extent_buffer(NULL, src->start, src->len, GFP_ATOMIC);
+	new = __alloc_extent_buffer(NULL, src->start, src->len, GFP_NOFS);
 	if (new == NULL)
 		return NULL;
 
 	for (i = 0; i < num_pages; i++) {
-		p = alloc_page(GFP_ATOMIC);
-		BUG_ON(!p);
+		p = alloc_page(GFP_NOFS);
+		if (!p) {
+			btrfs_release_extent_buffer(new);
+			return NULL;
+		}
 		attach_extent_buffer_page(new, p);
 		WARN_ON(PageDirty(p));
 		SetPageUptodate(p);
@@ -4210,12 +4382,12 @@ struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len)
 	unsigned long num_pages = num_extent_pages(0, len);
 	unsigned long i;
 
-	eb = __alloc_extent_buffer(NULL, start, len, GFP_ATOMIC);
+	eb = __alloc_extent_buffer(NULL, start, len, GFP_NOFS);
 	if (!eb)
 		return NULL;
 
 	for (i = 0; i < num_pages; i++) {
-		eb->pages[i] = alloc_page(GFP_ATOMIC);
+		eb->pages[i] = alloc_page(GFP_NOFS);
 		if (!eb->pages[i])
 			goto err;
 	}
@@ -4231,76 +4403,6 @@ err:
 	return NULL;
 }
 
-static int extent_buffer_under_io(struct extent_buffer *eb)
-{
-	return (atomic_read(&eb->io_pages) ||
-		test_bit(EXTENT_BUFFER_WRITEBACK, &eb->bflags) ||
-		test_bit(EXTENT_BUFFER_DIRTY, &eb->bflags));
-}
-
-/*
- * Helper for releasing extent buffer page.
- */
-static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
-						unsigned long start_idx)
-{
-	unsigned long index;
-	unsigned long num_pages;
-	struct page *page;
-	int mapped = !test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags);
-
-	BUG_ON(extent_buffer_under_io(eb));
-
-	num_pages = num_extent_pages(eb->start, eb->len);
-	index = start_idx + num_pages;
-	if (start_idx >= index)
-		return;
-
-	do {
-		index--;
-		page = extent_buffer_page(eb, index);
-		if (page && mapped) {
-			spin_lock(&page->mapping->private_lock);
-			/*
-			 * We do this since we'll remove the pages after we've
-			 * removed the eb from the radix tree, so we could race
-			 * and have this page now attached to the new eb.  So
-			 * only clear page_private if it's still connected to
-			 * this eb.
-			 */
-			if (PagePrivate(page) &&
-			    page->private == (unsigned long)eb) {
-				BUG_ON(test_bit(EXTENT_BUFFER_DIRTY, &eb->bflags));
-				BUG_ON(PageDirty(page));
-				BUG_ON(PageWriteback(page));
-				/*
-				 * We need to make sure we haven't be attached
-				 * to a new eb.
-				 */
-				ClearPagePrivate(page);
-				set_page_private(page, 0);
-				/* One for the page private */
-				page_cache_release(page);
-			}
-			spin_unlock(&page->mapping->private_lock);
-
-		}
-		if (page) {
-			/* One for when we alloced the page */
-			page_cache_release(page);
-		}
-	} while (index != start_idx);
-}
-
-/*
- * Helper for releasing the extent buffer.
- */
-static inline void btrfs_release_extent_buffer(struct extent_buffer *eb)
-{
-	btrfs_release_extent_buffer_page(eb, 0);
-	__free_extent_buffer(eb);
-}
-
 static void check_buffer_tree_ref(struct extent_buffer *eb)
 {
 	int refs;
@@ -4771,7 +4873,7 @@ void read_extent_buffer(struct extent_buffer *eb, void *dstv,
 	WARN_ON(start > eb->len);
 	WARN_ON(start + len > eb->start + eb->len);
 
-	offset = (start_offset + start) & ((unsigned long)PAGE_CACHE_SIZE - 1);
+	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
 		page = extent_buffer_page(eb, i);
@@ -4813,8 +4915,8 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long start,
 
 	if (start + min_len > eb->len) {
 		WARN(1, KERN_ERR "btrfs bad mapping eb start %llu len %lu, "
-		       "wanted %lu %lu\n", (unsigned long long)eb->start,
-		       eb->len, start, min_len);
+		       "wanted %lu %lu\n",
+		       eb->start, eb->len, start, min_len);
 		return -EINVAL;
 	}
 
@@ -4841,7 +4943,7 @@ int memcmp_extent_buffer(struct extent_buffer *eb, const void *ptrv,
 	WARN_ON(start > eb->len);
 	WARN_ON(start + len > eb->start + eb->len);
 
-	offset = (start_offset + start) & ((unsigned long)PAGE_CACHE_SIZE - 1);
+	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
 		page = extent_buffer_page(eb, i);
@@ -4875,7 +4977,7 @@ void write_extent_buffer(struct extent_buffer *eb, const void *srcv,
 	WARN_ON(start > eb->len);
 	WARN_ON(start + len > eb->start + eb->len);
 
-	offset = (start_offset + start) & ((unsigned long)PAGE_CACHE_SIZE - 1);
+	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
 		page = extent_buffer_page(eb, i);
@@ -4905,7 +5007,7 @@ void memset_extent_buffer(struct extent_buffer *eb, char c,
 	WARN_ON(start > eb->len);
 	WARN_ON(start + len > eb->start + eb->len);
 
-	offset = (start_offset + start) & ((unsigned long)PAGE_CACHE_SIZE - 1);
+	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
 		page = extent_buffer_page(eb, i);
@@ -4936,7 +5038,7 @@ void copy_extent_buffer(struct extent_buffer *dst, struct extent_buffer *src,
 	WARN_ON(src->len != dst_len);
 
 	offset = (start_offset + dst_offset) &
-		((unsigned long)PAGE_CACHE_SIZE - 1);
+		(PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
 		page = extent_buffer_page(dst, i);
@@ -5022,9 +5124,9 @@ void memcpy_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 
 	while (len > 0) {
 		dst_off_in_page = (start_offset + dst_offset) &
-			((unsigned long)PAGE_CACHE_SIZE - 1);
+			(PAGE_CACHE_SIZE - 1);
 		src_off_in_page = (start_offset + src_offset) &
-			((unsigned long)PAGE_CACHE_SIZE - 1);
+			(PAGE_CACHE_SIZE - 1);
 
 		dst_i = (start_offset + dst_offset) >> PAGE_CACHE_SHIFT;
 		src_i = (start_offset + src_offset) >> PAGE_CACHE_SHIFT;
@@ -5075,9 +5177,9 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 		src_i = (start_offset + src_end) >> PAGE_CACHE_SHIFT;
 
 		dst_off_in_page = (start_offset + dst_end) &
-			((unsigned long)PAGE_CACHE_SIZE - 1);
+			(PAGE_CACHE_SIZE - 1);
 		src_off_in_page = (start_offset + src_end) &
-			((unsigned long)PAGE_CACHE_SIZE - 1);
+			(PAGE_CACHE_SIZE - 1);
 
 		cur = min_t(unsigned long, len, src_off_in_page + 1);
 		cur = min(cur, dst_off_in_page + 1);

fs/btrfs/extent_io.h

@@ -29,6 +29,7 @@
  */
 #define EXTENT_BIO_COMPRESSED 1
 #define EXTENT_BIO_TREE_LOG 2
+#define EXTENT_BIO_PARENT_LOCKED 4
 #define EXTENT_BIO_FLAG_SHIFT 16
 
 /* these are bit numbers for test/set bit */
@@ -44,14 +45,11 @@
 #define EXTENT_BUFFER_DUMMY 9
 
 /* these are flags for extent_clear_unlock_delalloc */
-#define EXTENT_CLEAR_UNLOCK_PAGE 0x1
-#define EXTENT_CLEAR_UNLOCK	 0x2
-#define EXTENT_CLEAR_DELALLOC	 0x4
-#define EXTENT_CLEAR_DIRTY	 0x8
-#define EXTENT_SET_WRITEBACK	 0x10
-#define EXTENT_END_WRITEBACK	 0x20
-#define EXTENT_SET_PRIVATE2	 0x40
-#define EXTENT_CLEAR_ACCOUNTING  0x80
+#define PAGE_UNLOCK		(1 << 0)
+#define PAGE_CLEAR_DIRTY	(1 << 1)
+#define PAGE_SET_WRITEBACK	(1 << 2)
+#define PAGE_END_WRITEBACK	(1 << 3)
+#define PAGE_SET_PRIVATE2	(1 << 4)
 
 /*
  * page->private values.  Every page that is controlled by the extent
@@ -62,6 +60,7 @@
 
 struct extent_state;
 struct btrfs_root;
+struct btrfs_io_bio;
 
 typedef	int (extent_submit_bio_hook_t)(struct inode *inode, int rw,
 				       struct bio *bio, int mirror_num,
@@ -77,8 +76,9 @@ struct extent_io_ops {
 			      size_t size, struct bio *bio,
 			      unsigned long bio_flags);
 	int (*readpage_io_failed_hook)(struct page *page, int failed_mirror);
-	int (*readpage_end_io_hook)(struct page *page, u64 start, u64 end,
-				    struct extent_state *state, int mirror);
+	int (*readpage_end_io_hook)(struct btrfs_io_bio *io_bio, u64 phy_offset,
+				    struct page *page, u64 start, u64 end,
+				    int mirror);
 	int (*writepage_end_io_hook)(struct page *page, u64 start, u64 end,
 				      struct extent_state *state, int uptodate);
 	void (*set_bit_hook)(struct inode *inode, struct extent_state *state,
@@ -200,6 +200,8 @@ int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end,
 int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 			  get_extent_t *get_extent, int mirror_num);
+int extent_read_full_page_nolock(struct extent_io_tree *tree, struct page *page,
+				 get_extent_t *get_extent, int mirror_num);
 int __init extent_io_init(void);
 void extent_io_exit(void);
 
@@ -261,11 +263,6 @@ int extent_readpages(struct extent_io_tree *tree,
 		     get_extent_t get_extent);
 int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len, get_extent_t *get_extent);
-int set_state_private(struct extent_io_tree *tree, u64 start, u64 private);
-void extent_cache_csums_dio(struct extent_io_tree *tree, u64 start, u32 csums[],
-			    int count);
-void extent_cache_csums(struct extent_io_tree *tree, struct bio *bio,
-			int bvec_index, u32 csums[], int count);
 int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private);
 void set_page_extent_mapped(struct page *page);
 
@@ -330,10 +327,10 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long offset,
 		      unsigned long *map_len);
 int extent_range_clear_dirty_for_io(struct inode *inode, u64 start, u64 end);
 int extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end);
-int extent_clear_unlock_delalloc(struct inode *inode,
-				struct extent_io_tree *tree,
-				u64 start, u64 end, struct page *locked_page,
-				unsigned long op);
+int extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end,
+				 struct page *locked_page,
+				 unsigned long bits_to_clear,
+				 unsigned long page_ops);
 struct bio *
 btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
 		gfp_t gfp_flags);

fs/btrfs/file-item.c

@@ -23,6 +23,7 @@
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
+#include "volumes.h"
 #include "print-tree.h"
 
 #define __MAX_CSUM_ITEMS(r, size) ((unsigned long)(((BTRFS_LEAF_DATA_SIZE(r) - \
@@ -152,28 +153,54 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+static void btrfs_io_bio_endio_readpage(struct btrfs_io_bio *bio, int err)
+{
+	kfree(bio->csum_allocated);
+}
+
 static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 				   struct inode *inode, struct bio *bio,
 				   u64 logical_offset, u32 *dst, int dio)
 {
-	u32 sum[16];
-	int len;
 	struct bio_vec *bvec = bio->bi_io_vec;
-	int bio_index = 0;
+	struct btrfs_io_bio *btrfs_bio = btrfs_io_bio(bio);
+	struct btrfs_csum_item *item = NULL;
+	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	struct btrfs_path *path;
+	u8 *csum;
 	u64 offset = 0;
 	u64 item_start_offset = 0;
 	u64 item_last_offset = 0;
 	u64 disk_bytenr;
 	u32 diff;
-	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
+	int nblocks;
+	int bio_index = 0;
 	int count;
-	struct btrfs_path *path;
-	struct btrfs_csum_item *item = NULL;
-	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
 
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+
+	nblocks = bio->bi_size >> inode->i_sb->s_blocksize_bits;
+	if (!dst) {
+		if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
+			btrfs_bio->csum_allocated = kmalloc(nblocks * csum_size,
+							    GFP_NOFS);
+			if (!btrfs_bio->csum_allocated) {
+				btrfs_free_path(path);
+				return -ENOMEM;
+			}
+			btrfs_bio->csum = btrfs_bio->csum_allocated;
+			btrfs_bio->end_io = btrfs_io_bio_endio_readpage;
+		} else {
+			btrfs_bio->csum = btrfs_bio->csum_inline;
+		}
+		csum = btrfs_bio->csum;
+	} else {
+		csum = (u8 *)dst;
+	}
+
 	if (bio->bi_size > PAGE_CACHE_SIZE * 8)
 		path->reada = 2;
 
@@ -194,11 +221,10 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 	if (dio)
 		offset = logical_offset;
 	while (bio_index < bio->bi_vcnt) {
-		len = min_t(int, ARRAY_SIZE(sum), bio->bi_vcnt - bio_index);
 		if (!dio)
 			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
-		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr, sum,
-					       len);
+		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr,
+					       (u32 *)csum, nblocks);
 		if (count)
 			goto found;
 
@@ -213,7 +239,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 						 path, disk_bytenr, 0);
 			if (IS_ERR(item)) {
 				count = 1;
-				sum[0] = 0;
+				memset(csum, 0, csum_size);
 				if (BTRFS_I(inode)->root->root_key.objectid ==
 				    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 					set_extent_bits(io_tree, offset,
@@ -222,9 +248,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 				} else {
 					printk(KERN_INFO "btrfs no csum found "
 					       "for inode %llu start %llu\n",
-					       (unsigned long long)
-					       btrfs_ino(inode),
-					       (unsigned long long)offset);
+					       btrfs_ino(inode), offset);
 				}
 				item = NULL;
 				btrfs_release_path(path);
@@ -249,23 +273,14 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 		diff = disk_bytenr - item_start_offset;
 		diff = diff / root->sectorsize;
 		diff = diff * csum_size;
-		count = min_t(int, len, (item_last_offset - disk_bytenr) >>
-					inode->i_sb->s_blocksize_bits);
-		read_extent_buffer(path->nodes[0], sum,
+		count = min_t(int, nblocks, (item_last_offset - disk_bytenr) >>
+					    inode->i_sb->s_blocksize_bits);
+		read_extent_buffer(path->nodes[0], csum,
 				   ((unsigned long)item) + diff,
 				   csum_size * count);
 found:
-		if (dst) {
-			memcpy(dst, sum, count * csum_size);
-			dst += count;
-		} else {
-			if (dio)
-				extent_cache_csums_dio(io_tree, offset, sum,
-						       count);
-			else
-				extent_cache_csums(io_tree, bio, bio_index, sum,
-					    count);
-		}
+		csum += count * csum_size;
+		nblocks -= count;
 		while (count--) {
 			disk_bytenr += bvec->bv_len;
 			offset += bvec->bv_len;
@@ -284,9 +299,19 @@ int btrfs_lookup_bio_sums(struct btrfs_root *root, struct inode *inode,
 }
 
 int btrfs_lookup_bio_sums_dio(struct btrfs_root *root, struct inode *inode,
-			      struct bio *bio, u64 offset)
+			      struct btrfs_dio_private *dip, struct bio *bio,
+			      u64 offset)
 {
-	return __btrfs_lookup_bio_sums(root, inode, bio, offset, NULL, 1);
+	int len = (bio->bi_sector << 9) - dip->disk_bytenr;
+	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
+	int ret;
+
+	len >>= inode->i_sb->s_blocksize_bits;
+	len *= csum_size;
+
+	ret = __btrfs_lookup_bio_sums(root, inode, bio, offset,
+				      (u32 *)(dip->csum + len), 1);
+	return ret;
 }
 
 int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,

fs/btrfs/file.c

@@ -1334,7 +1334,6 @@ fail:
 static noinline int check_can_nocow(struct inode *inode, loff_t pos,
 				    size_t *write_bytes)
 {
-	struct btrfs_trans_handle *trans;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_ordered_extent *ordered;
 	u64 lockstart, lockend;
@@ -1356,16 +1355,8 @@ static noinline int check_can_nocow(struct inode *inode, loff_t pos,
 		btrfs_put_ordered_extent(ordered);
 	}
 
-	trans = btrfs_join_transaction(root);
-	if (IS_ERR(trans)) {
-		unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
-		return PTR_ERR(trans);
-	}
-
 	num_bytes = lockend - lockstart + 1;
-	ret = can_nocow_extent(trans, inode, lockstart, &num_bytes, NULL, NULL,
-			       NULL);
-	btrfs_end_transaction(trans, root);
+	ret = can_nocow_extent(inode, lockstart, &num_bytes, NULL, NULL, NULL);
 	if (ret <= 0) {
 		ret = 0;
 	} else {

fs/btrfs/free-space-cache.c

@@ -308,7 +308,7 @@ static void io_ctl_unmap_page(struct io_ctl *io_ctl)
 
 static void io_ctl_map_page(struct io_ctl *io_ctl, int clear)
 {
-	BUG_ON(io_ctl->index >= io_ctl->num_pages);
+	ASSERT(io_ctl->index < io_ctl->num_pages);
 	io_ctl->page = io_ctl->pages[io_ctl->index++];
 	io_ctl->cur = kmap(io_ctl->page);
 	io_ctl->orig = io_ctl->cur;
@@ -673,8 +673,7 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 		btrfs_err(root->fs_info,
 			"free space inode generation (%llu) "
 			"did not match free space cache generation (%llu)",
-			(unsigned long long)BTRFS_I(inode)->generation,
-			(unsigned long long)generation);
+			BTRFS_I(inode)->generation, generation);
 		return 0;
 	}
 
@@ -729,7 +728,7 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 				goto free_cache;
 			}
 		} else {
-			BUG_ON(!num_bitmaps);
+			ASSERT(num_bitmaps);
 			num_bitmaps--;
 			e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS);
 			if (!e->bitmap) {
@@ -1029,7 +1028,7 @@ static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 	leaf = path->nodes[0];
 	if (ret > 0) {
 		struct btrfs_key found_key;
-		BUG_ON(!path->slots[0]);
+		ASSERT(path->slots[0]);
 		path->slots[0]--;
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 		if (found_key.objectid != BTRFS_FREE_SPACE_OBJECTID ||
@@ -1117,7 +1116,7 @@ int btrfs_write_out_cache(struct btrfs_root *root,
 static inline unsigned long offset_to_bit(u64 bitmap_start, u32 unit,
 					  u64 offset)
 {
-	BUG_ON(offset < bitmap_start);
+	ASSERT(offset >= bitmap_start);
 	offset -= bitmap_start;
 	return (unsigned long)(div_u64(offset, unit));
 }
@@ -1272,7 +1271,7 @@ tree_search_offset(struct btrfs_free_space_ctl *ctl,
 		if (n) {
 			entry = rb_entry(n, struct btrfs_free_space,
 					offset_index);
-			BUG_ON(entry->offset > offset);
+			ASSERT(entry->offset <= offset);
 		} else {
 			if (fuzzy)
 				return entry;
@@ -1336,7 +1335,7 @@ static int link_free_space(struct btrfs_free_space_ctl *ctl,
 {
 	int ret = 0;
 
-	BUG_ON(!info->bitmap && !info->bytes);
+	ASSERT(info->bytes || info->bitmap);
 	ret = tree_insert_offset(&ctl->free_space_offset, info->offset,
 				 &info->offset_index, (info->bitmap != NULL));
 	if (ret)
@@ -1359,7 +1358,7 @@ static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl)
 
 	max_bitmaps = max(max_bitmaps, 1);
 
-	BUG_ON(ctl->total_bitmaps > max_bitmaps);
+	ASSERT(ctl->total_bitmaps <= max_bitmaps);
 
 	/*
 	 * The goal is to keep the total amount of memory used per 1gb of space
@@ -1403,7 +1402,7 @@ static inline void __bitmap_clear_bits(struct btrfs_free_space_ctl *ctl,
 
 	start = offset_to_bit(info->offset, ctl->unit, offset);
 	count = bytes_to_bits(bytes, ctl->unit);
-	BUG_ON(start + count > BITS_PER_BITMAP);
+	ASSERT(start + count <= BITS_PER_BITMAP);
 
 	bitmap_clear(info->bitmap, start, count);
 
@@ -1426,7 +1425,7 @@ static void bitmap_set_bits(struct btrfs_free_space_ctl *ctl,
 
 	start = offset_to_bit(info->offset, ctl->unit, offset);
 	count = bytes_to_bits(bytes, ctl->unit);
-	BUG_ON(start + count > BITS_PER_BITMAP);
+	ASSERT(start + count <= BITS_PER_BITMAP);
 
 	bitmap_set(info->bitmap, start, count);
 
@@ -1742,7 +1741,7 @@ no_cluster_bitmap:
 	bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
 					 1, 0);
 	if (!bitmap_info) {
-		BUG_ON(added);
+		ASSERT(added == 0);
 		goto new_bitmap;
 	}
 
@@ -1882,7 +1881,7 @@ out:
 
 	if (ret) {
 		printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret);
-		BUG_ON(ret == -EEXIST);
+		ASSERT(ret != -EEXIST);
 	}
 
 	return ret;
@@ -1991,8 +1990,7 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 		if (info->bytes >= bytes && !block_group->ro)
 			count++;
 		printk(KERN_CRIT "entry offset %llu, bytes %llu, bitmap %s\n",
-		       (unsigned long long)info->offset,
-		       (unsigned long long)info->bytes,
+		       info->offset, info->bytes,
 		       (info->bitmap) ? "yes" : "no");
 	}
 	printk(KERN_INFO "block group has cluster?: %s\n",
@@ -2371,7 +2369,7 @@ again:
 	rb_erase(&entry->offset_index, &ctl->free_space_offset);
 	ret = tree_insert_offset(&cluster->root, entry->offset,
 				 &entry->offset_index, 1);
-	BUG_ON(ret); /* -EEXIST; Logic error */
+	ASSERT(!ret); /* -EEXIST; Logic error */
 
 	trace_btrfs_setup_cluster(block_group, cluster,
 				  total_found * ctl->unit, 1);
@@ -2464,7 +2462,7 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
 		ret = tree_insert_offset(&cluster->root, entry->offset,
 					 &entry->offset_index, 0);
 		total_size += entry->bytes;
-		BUG_ON(ret); /* -EEXIST; Logic error */
+		ASSERT(!ret); /* -EEXIST; Logic error */
 	} while (node && entry != last);
 
 	cluster->max_size = max_extent;
@@ -2525,8 +2523,7 @@ setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
  * returns zero and sets up cluster if things worked out, otherwise
  * it returns -enospc
  */
-int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *root,
+int btrfs_find_space_cluster(struct btrfs_root *root,
 			     struct btrfs_block_group_cache *block_group,
 			     struct btrfs_free_cluster *cluster,
 			     u64 offset, u64 bytes, u64 empty_size)
@@ -2856,7 +2853,7 @@ u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root)
 
 		ret = search_bitmap(ctl, entry, &offset, &count);
 		/* Logic error; Should be empty if it can't find anything */
-		BUG_ON(ret);
+		ASSERT(!ret);
 
 		ino = offset;
 		bitmap_clear_bits(ctl, entry, offset, 1);
@@ -2973,33 +2970,68 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 }
 
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-static struct btrfs_block_group_cache *init_test_block_group(void)
+/*
+ * Use this if you need to make a bitmap or extent entry specifically, it
+ * doesn't do any of the merging that add_free_space does, this acts a lot like
+ * how the free space cache loading stuff works, so you can get really weird
+ * configurations.
+ */
+int test_add_free_space_entry(struct btrfs_block_group_cache *cache,
+			      u64 offset, u64 bytes, bool bitmap)
 {
-	struct btrfs_block_group_cache *cache;
+	struct btrfs_free_space_ctl *ctl = cache->free_space_ctl;
+	struct btrfs_free_space *info = NULL, *bitmap_info;
+	void *map = NULL;
+	u64 bytes_added;
+	int ret;
 
-	cache = kzalloc(sizeof(*cache), GFP_NOFS);
-	if (!cache)
-		return NULL;
-	cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
-					GFP_NOFS);
-	if (!cache->free_space_ctl) {
-		kfree(cache);
-		return NULL;
+again:
+	if (!info) {
+		info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS);
+		if (!info)
+			return -ENOMEM;
 	}
 
-	cache->key.objectid = 0;
-	cache->key.offset = 1024 * 1024 * 1024;
-	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
-	cache->sectorsize = 4096;
+	if (!bitmap) {
+		spin_lock(&ctl->tree_lock);
+		info->offset = offset;
+		info->bytes = bytes;
+		ret = link_free_space(ctl, info);
+		spin_unlock(&ctl->tree_lock);
+		if (ret)
+			kmem_cache_free(btrfs_free_space_cachep, info);
+		return ret;
+	}
+
+	if (!map) {
+		map = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS);
+		if (!map) {
+			kmem_cache_free(btrfs_free_space_cachep, info);
+			return -ENOMEM;
+		}
+	}
+
+	spin_lock(&ctl->tree_lock);
+	bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
+					 1, 0);
+	if (!bitmap_info) {
+		info->bitmap = map;
+		map = NULL;
+		add_new_bitmap(ctl, info, offset);
+		bitmap_info = info;
+	}
 
-	spin_lock_init(&cache->lock);
-	INIT_LIST_HEAD(&cache->list);
-	INIT_LIST_HEAD(&cache->cluster_list);
-	INIT_LIST_HEAD(&cache->new_bg_list);
+	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
+	bytes -= bytes_added;
+	offset += bytes_added;
+	spin_unlock(&ctl->tree_lock);
 
-	btrfs_init_free_space_ctl(cache);
+	if (bytes)
+		goto again;
 
-	return cache;
+	if (map)
+		kfree(map);
+	return 0;
 }
 
 /*
@@ -3007,8 +3039,8 @@ static struct btrfs_block_group_cache *init_test_block_group(void)
  * just used to check the absence of space, so if there is free space in the
  * range at all we will return 1.
  */
-static int check_exists(struct btrfs_block_group_cache *cache, u64 offset,
-			u64 bytes)
+int test_check_exists(struct btrfs_block_group_cache *cache,
+		      u64 offset, u64 bytes)
 {
 	struct btrfs_free_space_ctl *ctl = cache->free_space_ctl;
 	struct btrfs_free_space *info;
@@ -3085,411 +3117,4 @@ out:
 	spin_unlock(&ctl->tree_lock);
 	return ret;
 }
-
-/*
- * Use this if you need to make a bitmap or extent entry specifically, it
- * doesn't do any of the merging that add_free_space does, this acts a lot like
- * how the free space cache loading stuff works, so you can get really weird
- * configurations.
- */
-static int add_free_space_entry(struct btrfs_block_group_cache *cache,
-				u64 offset, u64 bytes, bool bitmap)
-{
-	struct btrfs_free_space_ctl *ctl = cache->free_space_ctl;
-	struct btrfs_free_space *info = NULL, *bitmap_info;
-	void *map = NULL;
-	u64 bytes_added;
-	int ret;
-
-again:
-	if (!info) {
-		info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS);
-		if (!info)
-			return -ENOMEM;
-	}
-
-	if (!bitmap) {
-		spin_lock(&ctl->tree_lock);
-		info->offset = offset;
-		info->bytes = bytes;
-		ret = link_free_space(ctl, info);
-		spin_unlock(&ctl->tree_lock);
-		if (ret)
-			kmem_cache_free(btrfs_free_space_cachep, info);
-		return ret;
-	}
-
-	if (!map) {
-		map = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS);
-		if (!map) {
-			kmem_cache_free(btrfs_free_space_cachep, info);
-			return -ENOMEM;
-		}
-	}
-
-	spin_lock(&ctl->tree_lock);
-	bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
-					 1, 0);
-	if (!bitmap_info) {
-		info->bitmap = map;
-		map = NULL;
-		add_new_bitmap(ctl, info, offset);
-		bitmap_info = info;
-	}
-
-	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
-	bytes -= bytes_added;
-	offset += bytes_added;
-	spin_unlock(&ctl->tree_lock);
-
-	if (bytes)
-		goto again;
-
-	if (map)
-		kfree(map);
-	return 0;
-}
-
-#define test_msg(fmt, ...) printk(KERN_INFO "btrfs: selftest: " fmt, ##__VA_ARGS__)
-
-/*
- * This test just does basic sanity checking, making sure we can add an exten
- * entry and remove space from either end and the middle, and make sure we can
- * remove space that covers adjacent extent entries.
- */
-static int test_extents(struct btrfs_block_group_cache *cache)
-{
-	int ret = 0;
-
-	test_msg("Running extent only tests\n");
-
-	/* First just make sure we can remove an entire entry */
-	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error adding initial extents %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing extent %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 4 * 1024 * 1024)) {
-		test_msg("Full remove left some lingering space\n");
-		return -1;
-	}
-
-	/* Ok edge and middle cases now */
-	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error adding half extent %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing tail end %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing front end %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 2 * 1024 * 1024, 4096);
-	if (ret) {
-		test_msg("Error removing middle piece %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 1 * 1024 * 1024)) {
-		test_msg("Still have space at the front\n");
-		return -1;
-	}
-
-	if (check_exists(cache, 2 * 1024 * 1024, 4096)) {
-		test_msg("Still have space in the middle\n");
-		return -1;
-	}
-
-	if (check_exists(cache, 3 * 1024 * 1024, 1 * 1024 * 1024)) {
-		test_msg("Still have space at the end\n");
-		return -1;
-	}
-
-	/* Cleanup */
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-
-	return 0;
-}
-
-static int test_bitmaps(struct btrfs_block_group_cache *cache)
-{
-	u64 next_bitmap_offset;
-	int ret;
-
-	test_msg("Running bitmap only tests\n");
-
-	ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't create a bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing bitmap full range %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 4 * 1024 * 1024)) {
-		test_msg("Left some space in bitmap\n");
-		return -1;
-	}
-
-	ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add to our bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 2 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove middle chunk %d\n", ret);
-		return ret;
-	}
-
-	/*
-	 * The first bitmap we have starts at offset 0 so the next one is just
-	 * at the end of the first bitmap.
-	 */
-	next_bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
-
-	/* Test a bit straddling two bitmaps */
-	ret = add_free_space_entry(cache, next_bitmap_offset -
-				   (2 * 1024 * 1024), 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add space that straddles two bitmaps %d\n",
-				ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, next_bitmap_offset -
-				      (1 * 1024 * 1024), 2 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove overlapping space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, next_bitmap_offset - (1 * 1024 * 1024),
-			 2 * 1024 * 1024)) {
-		test_msg("Left some space when removing overlapping\n");
-		return -1;
-	}
-
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-
-	return 0;
-}
-
-/* This is the high grade jackassery */
-static int test_bitmaps_and_extents(struct btrfs_block_group_cache *cache)
-{
-	u64 bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
-	int ret;
-
-	test_msg("Running bitmap and extent tests\n");
-
-	/*
-	 * First let's do something simple, an extent at the same offset as the
-	 * bitmap, but the free space completely in the extent and then
-	 * completely in the bitmap.
-	 */
-	ret = add_free_space_entry(cache, 4 * 1024 * 1024, 1 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't create bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove extent entry %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 1 * 1024 * 1024)) {
-		test_msg("Left remnants after our remove\n");
-		return -1;
-	}
-
-	/* Now to add back the extent entry and remove from the bitmap */
-	ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't re-add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 4 * 1024 * 1024, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove from bitmap %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 4 * 1024 * 1024, 1 * 1024 * 1024)) {
-		test_msg("Left remnants in the bitmap\n");
-		return -1;
-	}
-
-	/*
-	 * Ok so a little more evil, extent entry and bitmap at the same offset,
-	 * removing an overlapping chunk.
-	 */
-	ret = add_free_space_entry(cache, 1 * 1024 * 1024, 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add to a bitmap %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 512 * 1024, 3 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove overlapping space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 512 * 1024, 3 * 1024 * 1024)) {
-		test_msg("Left over peices after removing overlapping\n");
-		return -1;
-	}
-
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-
-	/* Now with the extent entry offset into the bitmap */
-	ret = add_free_space_entry(cache, 4 * 1024 * 1024, 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add space to the bitmap %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, 2 * 1024 * 1024, 2 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent to the cache %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Problem removing overlapping space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 3 * 1024 * 1024, 4 * 1024 * 1024)) {
-		test_msg("Left something behind when removing space");
-		return -1;
-	}
-
-	/*
-	 * This has blown up in the past, the extent entry starts before the
-	 * bitmap entry, but we're trying to remove an offset that falls
-	 * completely within the bitmap range and is in both the extent entry
-	 * and the bitmap entry, looks like this
-	 *
-	 *   [ extent ]
-	 *      [ bitmap ]
-	 *        [ del ]
-	 */
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-	ret = add_free_space_entry(cache, bitmap_offset + 4 * 1024 * 1024,
-				   4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add bitmap %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, bitmap_offset - 1 * 1024 * 1024,
-				   5 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, bitmap_offset + 1 * 1024 * 1024,
-				      5 * 1024 * 1024);
-	if (ret) {
-		test_msg("Failed to free our space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, bitmap_offset + 1 * 1024 * 1024,
-			 5 * 1024 * 1024)) {
-		test_msg("Left stuff over\n");
-		return -1;
-	}
-
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-
-	/*
-	 * This blew up before, we have part of the free space in a bitmap and
-	 * then the entirety of the rest of the space in an extent.  This used
-	 * to return -EAGAIN back from btrfs_remove_extent, make sure this
-	 * doesn't happen.
-	 */
-	ret = add_free_space_entry(cache, 1 * 1024 * 1024, 2 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, 3 * 1024 * 1024, 1 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 3 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing bitmap and extent overlapping %d\n", ret);
-		return ret;
-	}
-
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-	return 0;
-}
-
-void btrfs_test_free_space_cache(void)
-{
-	struct btrfs_block_group_cache *cache;
-
-	test_msg("Running btrfs free space cache tests\n");
-
-	cache = init_test_block_group();
-	if (!cache) {
-		test_msg("Couldn't run the tests\n");
-		return;
-	}
-
-	if (test_extents(cache))
-		goto out;
-	if (test_bitmaps(cache))
-		goto out;
-	if (test_bitmaps_and_extents(cache))
-		goto out;
-out:
-	__btrfs_remove_free_space_cache(cache->free_space_ctl);
-	kfree(cache->free_space_ctl);
-	kfree(cache);
-	test_msg("Free space cache tests finished\n");
-}
-#undef test_msg
-#else /* !CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
-void btrfs_test_free_space_cache(void) {}
-#endif /* !CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
+#endif /* CONFIG_BTRFS_FS_RUN_SANITY_TESTS */

fs/btrfs/free-space-cache.h

@@ -98,8 +98,7 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group,
 u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root);
 void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 			   u64 bytes);
-int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *root,
+int btrfs_find_space_cluster(struct btrfs_root *root,
 			     struct btrfs_block_group_cache *block_group,
 			     struct btrfs_free_cluster *cluster,
 			     u64 offset, u64 bytes, u64 empty_size);
@@ -113,6 +112,12 @@ int btrfs_return_cluster_to_free_space(
 int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group,
 			   u64 *trimmed, u64 start, u64 end, u64 minlen);
 
-void btrfs_test_free_space_cache(void);
+/* Support functions for runnint our sanity tests */
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+int test_add_free_space_entry(struct btrfs_block_group_cache *cache,
+			      u64 offset, u64 bytes, bool bitmap);
+int test_check_exists(struct btrfs_block_group_cache *cache,
+		      u64 offset, u64 bytes);
+#endif
 
 #endif

fs/btrfs/inode.c

@@ -230,12 +230,13 @@ fail:
  * does the checks required to make sure the data is small enough
  * to fit as an inline extent.
  */
-static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
-				 struct btrfs_root *root,
-				 struct inode *inode, u64 start, u64 end,
-				 size_t compressed_size, int compress_type,
-				 struct page **compressed_pages)
+static noinline int cow_file_range_inline(struct btrfs_root *root,
+					  struct inode *inode, u64 start,
+					  u64 end, size_t compressed_size,
+					  int compress_type,
+					  struct page **compressed_pages)
 {
+	struct btrfs_trans_handle *trans;
 	u64 isize = i_size_read(inode);
 	u64 actual_end = min(end + 1, isize);
 	u64 inline_len = actual_end - start;
@@ -256,9 +257,16 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
 		return 1;
 	}
 
+	trans = btrfs_join_transaction(root);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
+
 	ret = btrfs_drop_extents(trans, root, inode, start, aligned_end, 1);
-	if (ret)
-		return ret;
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto out;
+	}
 
 	if (isize > actual_end)
 		inline_len = min_t(u64, isize, actual_end);
@@ -267,15 +275,18 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
 				   compress_type, compressed_pages);
 	if (ret && ret != -ENOSPC) {
 		btrfs_abort_transaction(trans, root, ret);
-		return ret;
+		goto out;
 	} else if (ret == -ENOSPC) {
-		return 1;
+		ret = 1;
+		goto out;
 	}
 
 	set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &BTRFS_I(inode)->runtime_flags);
 	btrfs_delalloc_release_metadata(inode, end + 1 - start);
 	btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
-	return 0;
+out:
+	btrfs_end_transaction(trans, root);
+	return ret;
 }
 
 struct async_extent {
@@ -343,7 +354,6 @@ static noinline int compress_file_range(struct inode *inode,
 					int *num_added)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct btrfs_trans_handle *trans;
 	u64 num_bytes;
 	u64 blocksize = root->sectorsize;
 	u64 actual_end;
@@ -461,45 +471,36 @@ again:
 	}
 cont:
 	if (start == 0) {
-		trans = btrfs_join_transaction(root);
-		if (IS_ERR(trans)) {
-			ret = PTR_ERR(trans);
-			trans = NULL;
-			goto cleanup_and_out;
-		}
-		trans->block_rsv = &root->fs_info->delalloc_block_rsv;
-
 		/* lets try to make an inline extent */
 		if (ret || total_in < (actual_end - start)) {
 			/* we didn't compress the entire range, try
 			 * to make an uncompressed inline extent.
 			 */
-			ret = cow_file_range_inline(trans, root, inode,
-						    start, end, 0, 0, NULL);
+			ret = cow_file_range_inline(root, inode, start, end,
+						    0, 0, NULL);
 		} else {
 			/* try making a compressed inline extent */
-			ret = cow_file_range_inline(trans, root, inode,
-						    start, end,
+			ret = cow_file_range_inline(root, inode, start, end,
 						    total_compressed,
 						    compress_type, pages);
 		}
 		if (ret <= 0) {
+			unsigned long clear_flags = EXTENT_DELALLOC |
+				EXTENT_DEFRAG;
+			clear_flags |= (ret < 0) ? EXTENT_DO_ACCOUNTING : 0;
+
 			/*
 			 * inline extent creation worked or returned error,
 			 * we don't need to create any more async work items.
 			 * Unlock and free up our temp pages.
 			 */
-			extent_clear_unlock_delalloc(inode,
-			     &BTRFS_I(inode)->io_tree,
-			     start, end, NULL,
-			     EXTENT_CLEAR_UNLOCK_PAGE | EXTENT_CLEAR_DIRTY |
-			     EXTENT_CLEAR_DELALLOC |
-			     EXTENT_SET_WRITEBACK | EXTENT_END_WRITEBACK);
-
-			btrfs_end_transaction(trans, root);
+			extent_clear_unlock_delalloc(inode, start, end, NULL,
+						     clear_flags, PAGE_UNLOCK |
+						     PAGE_CLEAR_DIRTY |
+						     PAGE_SET_WRITEBACK |
+						     PAGE_END_WRITEBACK);
 			goto free_pages_out;
 		}
-		btrfs_end_transaction(trans, root);
 	}
 
 	if (will_compress) {
@@ -590,20 +591,6 @@ free_pages_out:
 	kfree(pages);
 
 	goto out;
-
-cleanup_and_out:
-	extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
-				     start, end, NULL,
-				     EXTENT_CLEAR_UNLOCK_PAGE |
-				     EXTENT_CLEAR_DIRTY |
-				     EXTENT_CLEAR_DELALLOC |
-				     EXTENT_SET_WRITEBACK |
-				     EXTENT_END_WRITEBACK);
-	if (!trans || IS_ERR(trans))
-		btrfs_error(root->fs_info, ret, "Failed to join transaction");
-	else
-		btrfs_abort_transaction(trans, root, ret);
-	goto free_pages_out;
 }
 
 /*
@@ -617,7 +604,6 @@ static noinline int submit_compressed_extents(struct inode *inode,
 {
 	struct async_extent *async_extent;
 	u64 alloc_hint = 0;
-	struct btrfs_trans_handle *trans;
 	struct btrfs_key ins;
 	struct extent_map *em;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
@@ -678,20 +664,10 @@ retry:
 		lock_extent(io_tree, async_extent->start,
 			    async_extent->start + async_extent->ram_size - 1);
 
-		trans = btrfs_join_transaction(root);
-		if (IS_ERR(trans)) {
-			ret = PTR_ERR(trans);
-		} else {
-			trans->block_rsv = &root->fs_info->delalloc_block_rsv;
-			ret = btrfs_reserve_extent(trans, root,
+		ret = btrfs_reserve_extent(root,
 					   async_extent->compressed_size,
 					   async_extent->compressed_size,
 					   0, alloc_hint, &ins, 1);
-			if (ret && ret != -ENOSPC)
-				btrfs_abort_transaction(trans, root, ret);
-			btrfs_end_transaction(trans, root);
-		}
-
 		if (ret) {
 			int i;
 
@@ -770,16 +746,12 @@ retry:
 		/*
 		 * clear dirty, set writeback and unlock the pages.
 		 */
-		extent_clear_unlock_delalloc(inode,
-				&BTRFS_I(inode)->io_tree,
-				async_extent->start,
+		extent_clear_unlock_delalloc(inode, async_extent->start,
 				async_extent->start +
 				async_extent->ram_size - 1,
-				NULL, EXTENT_CLEAR_UNLOCK_PAGE |
-				EXTENT_CLEAR_UNLOCK |
-				EXTENT_CLEAR_DELALLOC |
-				EXTENT_CLEAR_DIRTY | EXTENT_SET_WRITEBACK);
-
+				NULL, EXTENT_LOCKED | EXTENT_DELALLOC,
+				PAGE_UNLOCK | PAGE_CLEAR_DIRTY |
+				PAGE_SET_WRITEBACK);
 		ret = btrfs_submit_compressed_write(inode,
 				    async_extent->start,
 				    async_extent->ram_size,
@@ -798,16 +770,13 @@ out:
 out_free_reserve:
 	btrfs_free_reserved_extent(root, ins.objectid, ins.offset);
 out_free:
-	extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
-				     async_extent->start,
+	extent_clear_unlock_delalloc(inode, async_extent->start,
 				     async_extent->start +
 				     async_extent->ram_size - 1,
-				     NULL, EXTENT_CLEAR_UNLOCK_PAGE |
-				     EXTENT_CLEAR_UNLOCK |
-				     EXTENT_CLEAR_DELALLOC |
-				     EXTENT_CLEAR_DIRTY |
-				     EXTENT_SET_WRITEBACK |
-				     EXTENT_END_WRITEBACK);
+				     NULL, EXTENT_LOCKED | EXTENT_DELALLOC |
+				     EXTENT_DEFRAG | EXTENT_DO_ACCOUNTING,
+				     PAGE_UNLOCK | PAGE_CLEAR_DIRTY |
+				     PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK);
 	kfree(async_extent);
 	goto again;
 }
@@ -857,14 +826,13 @@ static u64 get_extent_allocation_hint(struct inode *inode, u64 start,
  * required to start IO on it.  It may be clean and already done with
  * IO when we return.
  */
-static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
-				     struct inode *inode,
-				     struct btrfs_root *root,
-				     struct page *locked_page,
-				     u64 start, u64 end, int *page_started,
-				     unsigned long *nr_written,
-				     int unlock)
+static noinline int cow_file_range(struct inode *inode,
+				   struct page *locked_page,
+				   u64 start, u64 end, int *page_started,
+				   unsigned long *nr_written,
+				   int unlock)
 {
+	struct btrfs_root *root = BTRFS_I(inode)->root;
 	u64 alloc_hint = 0;
 	u64 num_bytes;
 	unsigned long ram_size;
@@ -885,29 +853,24 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 	/* if this is a small write inside eof, kick off defrag */
 	if (num_bytes < 64 * 1024 &&
 	    (start > 0 || end + 1 < BTRFS_I(inode)->disk_i_size))
-		btrfs_add_inode_defrag(trans, inode);
+		btrfs_add_inode_defrag(NULL, inode);
 
 	if (start == 0) {
 		/* lets try to make an inline extent */
-		ret = cow_file_range_inline(trans, root, inode,
-					    start, end, 0, 0, NULL);
+		ret = cow_file_range_inline(root, inode, start, end, 0, 0,
+					    NULL);
 		if (ret == 0) {
-			extent_clear_unlock_delalloc(inode,
-				     &BTRFS_I(inode)->io_tree,
-				     start, end, NULL,
-				     EXTENT_CLEAR_UNLOCK_PAGE |
-				     EXTENT_CLEAR_UNLOCK |
-				     EXTENT_CLEAR_DELALLOC |
-				     EXTENT_CLEAR_DIRTY |
-				     EXTENT_SET_WRITEBACK |
-				     EXTENT_END_WRITEBACK);
+			extent_clear_unlock_delalloc(inode, start, end, NULL,
+				     EXTENT_LOCKED | EXTENT_DELALLOC |
+				     EXTENT_DEFRAG, PAGE_UNLOCK |
+				     PAGE_CLEAR_DIRTY | PAGE_SET_WRITEBACK |
+				     PAGE_END_WRITEBACK);
 
 			*nr_written = *nr_written +
 			     (end - start + PAGE_CACHE_SIZE) / PAGE_CACHE_SIZE;
 			*page_started = 1;
 			goto out;
 		} else if (ret < 0) {
-			btrfs_abort_transaction(trans, root, ret);
 			goto out_unlock;
 		}
 	}
@@ -922,13 +885,11 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 		unsigned long op;
 
 		cur_alloc_size = disk_num_bytes;
-		ret = btrfs_reserve_extent(trans, root, cur_alloc_size,
+		ret = btrfs_reserve_extent(root, cur_alloc_size,
 					   root->sectorsize, 0, alloc_hint,
 					   &ins, 1);
-		if (ret < 0) {
-			btrfs_abort_transaction(trans, root, ret);
+		if (ret < 0)
 			goto out_unlock;
-		}
 
 		em = alloc_extent_map();
 		if (!em) {
@@ -974,10 +935,8 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 		    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 			ret = btrfs_reloc_clone_csums(inode, start,
 						      cur_alloc_size);
-			if (ret) {
-				btrfs_abort_transaction(trans, root, ret);
+			if (ret)
 				goto out_reserve;
-			}
 		}
 
 		if (disk_num_bytes < cur_alloc_size)
@@ -990,13 +949,13 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 		 * Do set the Private2 bit so we know this page was properly
 		 * setup for writepage
 		 */
-		op = unlock ? EXTENT_CLEAR_UNLOCK_PAGE : 0;
-		op |= EXTENT_CLEAR_UNLOCK | EXTENT_CLEAR_DELALLOC |
-			EXTENT_SET_PRIVATE2;
+		op = unlock ? PAGE_UNLOCK : 0;
+		op |= PAGE_SET_PRIVATE2;
 
-		extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
-					     start, start + ram_size - 1,
-					     locked_page, op);
+		extent_clear_unlock_delalloc(inode, start,
+					     start + ram_size - 1, locked_page,
+					     EXTENT_LOCKED | EXTENT_DELALLOC,
+					     op);
 		disk_num_bytes -= cur_alloc_size;
 		num_bytes -= cur_alloc_size;
 		alloc_hint = ins.objectid + ins.offset;
@@ -1008,52 +967,14 @@ out:
 out_reserve:
 	btrfs_free_reserved_extent(root, ins.objectid, ins.offset);
 out_unlock:
-	extent_clear_unlock_delalloc(inode,
-		     &BTRFS_I(inode)->io_tree,
-		     start, end, locked_page,
-		     EXTENT_CLEAR_UNLOCK_PAGE |
-		     EXTENT_CLEAR_UNLOCK |
-		     EXTENT_CLEAR_DELALLOC |
-		     EXTENT_CLEAR_DIRTY |
-		     EXTENT_SET_WRITEBACK |
-		     EXTENT_END_WRITEBACK);
-
+	extent_clear_unlock_delalloc(inode, start, end, locked_page,
+				     EXTENT_LOCKED | EXTENT_DO_ACCOUNTING |
+				     EXTENT_DELALLOC | EXTENT_DEFRAG,
+				     PAGE_UNLOCK | PAGE_CLEAR_DIRTY |
+				     PAGE_SET_WRITEBACK | PAGE_END_WRITEBACK);
 	goto out;
 }
 
-static noinline int cow_file_range(struct inode *inode,
-				   struct page *locked_page,
-				   u64 start, u64 end, int *page_started,
-				   unsigned long *nr_written,
-				   int unlock)
-{
-	struct btrfs_trans_handle *trans;
-	struct btrfs_root *root = BTRFS_I(inode)->root;
-	int ret;
-
-	trans = btrfs_join_transaction(root);
-	if (IS_ERR(trans)) {
-		extent_clear_unlock_delalloc(inode,
-			     &BTRFS_I(inode)->io_tree,
-			     start, end, locked_page,
-			     EXTENT_CLEAR_UNLOCK_PAGE |
-			     EXTENT_CLEAR_UNLOCK |
-			     EXTENT_CLEAR_DELALLOC |
-			     EXTENT_CLEAR_DIRTY |
-			     EXTENT_SET_WRITEBACK |
-			     EXTENT_END_WRITEBACK);
-		return PTR_ERR(trans);
-	}
-	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
-
-	ret = __cow_file_range(trans, inode, root, locked_page, start, end,
-			       page_started, nr_written, unlock);
-
-	btrfs_end_transaction(trans, root);
-
-	return ret;
-}
-
 /*
  * work queue call back to started compression on a file and pages
  */
@@ -1221,15 +1142,13 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 
 	path = btrfs_alloc_path();
 	if (!path) {
-		extent_clear_unlock_delalloc(inode,
-			     &BTRFS_I(inode)->io_tree,
-			     start, end, locked_page,
-			     EXTENT_CLEAR_UNLOCK_PAGE |
-			     EXTENT_CLEAR_UNLOCK |
-			     EXTENT_CLEAR_DELALLOC |
-			     EXTENT_CLEAR_DIRTY |
-			     EXTENT_SET_WRITEBACK |
-			     EXTENT_END_WRITEBACK);
+		extent_clear_unlock_delalloc(inode, start, end, locked_page,
+					     EXTENT_LOCKED | EXTENT_DELALLOC |
+					     EXTENT_DO_ACCOUNTING |
+					     EXTENT_DEFRAG, PAGE_UNLOCK |
+					     PAGE_CLEAR_DIRTY |
+					     PAGE_SET_WRITEBACK |
+					     PAGE_END_WRITEBACK);
 		return -ENOMEM;
 	}
 
@@ -1241,15 +1160,13 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 		trans = btrfs_join_transaction(root);
 
 	if (IS_ERR(trans)) {
-		extent_clear_unlock_delalloc(inode,
-			     &BTRFS_I(inode)->io_tree,
-			     start, end, locked_page,
-			     EXTENT_CLEAR_UNLOCK_PAGE |
-			     EXTENT_CLEAR_UNLOCK |
-			     EXTENT_CLEAR_DELALLOC |
-			     EXTENT_CLEAR_DIRTY |
-			     EXTENT_SET_WRITEBACK |
-			     EXTENT_END_WRITEBACK);
+		extent_clear_unlock_delalloc(inode, start, end, locked_page,
+					     EXTENT_LOCKED | EXTENT_DELALLOC |
+					     EXTENT_DO_ACCOUNTING |
+					     EXTENT_DEFRAG, PAGE_UNLOCK |
+					     PAGE_CLEAR_DIRTY |
+					     PAGE_SET_WRITEBACK |
+					     PAGE_END_WRITEBACK);
 		btrfs_free_path(path);
 		return PTR_ERR(trans);
 	}
@@ -1369,9 +1286,9 @@ out_check:
 
 		btrfs_release_path(path);
 		if (cow_start != (u64)-1) {
-			ret = __cow_file_range(trans, inode, root, locked_page,
-					       cow_start, found_key.offset - 1,
-					       page_started, nr_written, 1);
+			ret = cow_file_range(inode, locked_page,
+					     cow_start, found_key.offset - 1,
+					     page_started, nr_written, 1);
 			if (ret) {
 				btrfs_abort_transaction(trans, root, ret);
 				goto error;
@@ -1428,11 +1345,11 @@ out_check:
 			}
 		}
 
-		extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
-				cur_offset, cur_offset + num_bytes - 1,
-				locked_page, EXTENT_CLEAR_UNLOCK_PAGE |
-				EXTENT_CLEAR_UNLOCK | EXTENT_CLEAR_DELALLOC |
-				EXTENT_SET_PRIVATE2);
+		extent_clear_unlock_delalloc(inode, cur_offset,
+					     cur_offset + num_bytes - 1,
+					     locked_page, EXTENT_LOCKED |
+					     EXTENT_DELALLOC, PAGE_UNLOCK |
+					     PAGE_SET_PRIVATE2);
 		cur_offset = extent_end;
 		if (cur_offset > end)
 			break;
@@ -1445,9 +1362,8 @@ out_check:
 	}
 
 	if (cow_start != (u64)-1) {
-		ret = __cow_file_range(trans, inode, root, locked_page,
-				       cow_start, end,
-				       page_started, nr_written, 1);
+		ret = cow_file_range(inode, locked_page, cow_start, end,
+				     page_started, nr_written, 1);
 		if (ret) {
 			btrfs_abort_transaction(trans, root, ret);
 			goto error;
@@ -1460,16 +1376,13 @@ error:
 		ret = err;
 
 	if (ret && cur_offset < end)
-		extent_clear_unlock_delalloc(inode,
-			     &BTRFS_I(inode)->io_tree,
-			     cur_offset, end, locked_page,
-			     EXTENT_CLEAR_UNLOCK_PAGE |
-			     EXTENT_CLEAR_UNLOCK |
-			     EXTENT_CLEAR_DELALLOC |
-			     EXTENT_CLEAR_DIRTY |
-			     EXTENT_SET_WRITEBACK |
-			     EXTENT_END_WRITEBACK);
-
+		extent_clear_unlock_delalloc(inode, cur_offset, end,
+					     locked_page, EXTENT_LOCKED |
+					     EXTENT_DELALLOC | EXTENT_DEFRAG |
+					     EXTENT_DO_ACCOUNTING, PAGE_UNLOCK |
+					     PAGE_CLEAR_DIRTY |
+					     PAGE_SET_WRITEBACK |
+					     PAGE_END_WRITEBACK);
 	btrfs_free_path(path);
 	return ret;
 }
@@ -2132,6 +2045,7 @@ static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
 		WARN_ON(1);
 		return ret;
 	}
+	ret = 0;
 
 	while (1) {
 		cond_resched();
@@ -2181,8 +2095,6 @@ static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
 		    old->len || extent_offset + num_bytes <=
 		    old->extent_offset + old->offset)
 			continue;
-
-		ret = 0;
 		break;
 	}
 
@@ -2238,16 +2150,18 @@ static noinline bool record_extent_backrefs(struct btrfs_path *path,
 
 static int relink_is_mergable(struct extent_buffer *leaf,
 			      struct btrfs_file_extent_item *fi,
-			      u64 disk_bytenr)
+			      struct new_sa_defrag_extent *new)
 {
-	if (btrfs_file_extent_disk_bytenr(leaf, fi) != disk_bytenr)
+	if (btrfs_file_extent_disk_bytenr(leaf, fi) != new->bytenr)
 		return 0;
 
 	if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG)
 		return 0;
 
-	if (btrfs_file_extent_compression(leaf, fi) ||
-	    btrfs_file_extent_encryption(leaf, fi) ||
+	if (btrfs_file_extent_compression(leaf, fi) != new->compress_type)
+		return 0;
+
+	if (btrfs_file_extent_encryption(leaf, fi) ||
 	    btrfs_file_extent_other_encoding(leaf, fi))
 		return 0;
 
@@ -2391,8 +2305,8 @@ again:
 				    struct btrfs_file_extent_item);
 		extent_len = btrfs_file_extent_num_bytes(leaf, fi);
 
-		if (relink_is_mergable(leaf, fi, new->bytenr) &&
-		    extent_len + found_key.offset == start) {
+		if (extent_len + found_key.offset == start &&
+		    relink_is_mergable(leaf, fi, new)) {
 			btrfs_set_file_extent_num_bytes(leaf, fi,
 							extent_len + len);
 			btrfs_mark_buffer_dirty(leaf);
@@ -2648,8 +2562,10 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	struct extent_state *cached_state = NULL;
 	struct new_sa_defrag_extent *new = NULL;
 	int compress_type = 0;
-	int ret;
+	int ret = 0;
+	u64 logical_len = ordered_extent->len;
 	bool nolock;
+	bool truncated = false;
 
 	nolock = btrfs_is_free_space_inode(inode);
 
@@ -2658,6 +2574,14 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 		goto out;
 	}
 
+	if (test_bit(BTRFS_ORDERED_TRUNCATED, &ordered_extent->flags)) {
+		truncated = true;
+		logical_len = ordered_extent->truncated_len;
+		/* Truncated the entire extent, don't bother adding */
+		if (!logical_len)
+			goto out;
+	}
+
 	if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
 		BUG_ON(!list_empty(&ordered_extent->list)); /* Logic error */
 		btrfs_ordered_update_i_size(inode, 0, ordered_extent);
@@ -2713,15 +2637,14 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 		ret = btrfs_mark_extent_written(trans, inode,
 						ordered_extent->file_offset,
 						ordered_extent->file_offset +
-						ordered_extent->len);
+						logical_len);
 	} else {
 		BUG_ON(root == root->fs_info->tree_root);
 		ret = insert_reserved_file_extent(trans, inode,
 						ordered_extent->file_offset,
 						ordered_extent->start,
 						ordered_extent->disk_len,
-						ordered_extent->len,
-						ordered_extent->len,
+						logical_len, logical_len,
 						compress_type, 0, 0,
 						BTRFS_FILE_EXTENT_REG);
 	}
@@ -2753,17 +2676,27 @@ out:
 	if (trans)
 		btrfs_end_transaction(trans, root);
 
-	if (ret) {
-		clear_extent_uptodate(io_tree, ordered_extent->file_offset,
-				      ordered_extent->file_offset +
-				      ordered_extent->len - 1, NULL, GFP_NOFS);
+	if (ret || truncated) {
+		u64 start, end;
+
+		if (truncated)
+			start = ordered_extent->file_offset + logical_len;
+		else
+			start = ordered_extent->file_offset;
+		end = ordered_extent->file_offset + ordered_extent->len - 1;
+		clear_extent_uptodate(io_tree, start, end, NULL, GFP_NOFS);
+
+		/* Drop the cache for the part of the extent we didn't write. */
+		btrfs_drop_extent_cache(inode, start, end, 0);
 
 		/*
 		 * If the ordered extent had an IOERR or something else went
 		 * wrong we need to return the space for this ordered extent
-		 * back to the allocator.
+		 * back to the allocator.  We only free the extent in the
+		 * truncated case if we didn't write out the extent at all.
 		 */
-		if (!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) &&
+		if ((ret || !logical_len) &&
+		    !test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) &&
 		    !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags))
 			btrfs_free_reserved_extent(root, ordered_extent->start,
 						   ordered_extent->disk_len);
@@ -2827,16 +2760,16 @@ static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
  * if there's a match, we allow the bio to finish.  If not, the code in
  * extent_io.c will try to find good copies for us.
  */
-static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
-			       struct extent_state *state, int mirror)
+static int btrfs_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
+				      u64 phy_offset, struct page *page,
+				      u64 start, u64 end, int mirror)
 {
 	size_t offset = start - page_offset(page);
 	struct inode *inode = page->mapping->host;
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	char *kaddr;
-	u64 private = ~(u32)0;
-	int ret;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
+	u32 csum_expected;
 	u32 csum = ~(u32)0;
 	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
 	                              DEFAULT_RATELIMIT_BURST);
@@ -2856,19 +2789,13 @@ static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 		return 0;
 	}
 
-	if (state && state->start == start) {
-		private = state->private;
-		ret = 0;
-	} else {
-		ret = get_state_private(io_tree, start, &private);
-	}
-	kaddr = kmap_atomic(page);
-	if (ret)
-		goto zeroit;
+	phy_offset >>= inode->i_sb->s_blocksize_bits;
+	csum_expected = *(((u32 *)io_bio->csum) + phy_offset);
 
+	kaddr = kmap_atomic(page);
 	csum = btrfs_csum_data(kaddr + offset, csum,  end - start + 1);
 	btrfs_csum_final(csum, (char *)&csum);
-	if (csum != private)
+	if (csum != csum_expected)
 		goto zeroit;
 
 	kunmap_atomic(kaddr);
@@ -2877,14 +2804,12 @@ good:
 
 zeroit:
 	if (__ratelimit(&_rs))
-		btrfs_info(root->fs_info, "csum failed ino %llu off %llu csum %u private %llu",
-			(unsigned long long)btrfs_ino(page->mapping->host),
-			(unsigned long long)start, csum,
-			(unsigned long long)private);
+		btrfs_info(root->fs_info, "csum failed ino %llu off %llu csum %u expected csum %u",
+			btrfs_ino(page->mapping->host), start, csum, csum_expected);
 	memset(kaddr + offset, 1, end - start + 1);
 	flush_dcache_page(page);
 	kunmap_atomic(kaddr);
-	if (private == 0)
+	if (csum_expected == 0)
 		return 0;
 	return -EIO;
 }
@@ -2971,8 +2896,10 @@ void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans,
 	    btrfs_root_refs(&root->root_item) > 0) {
 		ret = btrfs_del_orphan_item(trans, root->fs_info->tree_root,
 					    root->root_key.objectid);
-		BUG_ON(ret);
-		root->orphan_item_inserted = 0;
+		if (ret)
+			btrfs_abort_transaction(trans, root, ret);
+		else
+			root->orphan_item_inserted = 0;
 	}
 
 	if (block_rsv) {
@@ -3041,11 +2968,18 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
 	/* insert an orphan item to track this unlinked/truncated file */
 	if (insert >= 1) {
 		ret = btrfs_insert_orphan_item(trans, root, btrfs_ino(inode));
-		if (ret && ret != -EEXIST) {
-			clear_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
-				  &BTRFS_I(inode)->runtime_flags);
-			btrfs_abort_transaction(trans, root, ret);
-			return ret;
+		if (ret) {
+			if (reserve) {
+				clear_bit(BTRFS_INODE_ORPHAN_META_RESERVED,
+					  &BTRFS_I(inode)->runtime_flags);
+				btrfs_orphan_release_metadata(inode);
+			}
+			if (ret != -EEXIST) {
+				clear_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
+					  &BTRFS_I(inode)->runtime_flags);
+				btrfs_abort_transaction(trans, root, ret);
+				return ret;
+			}
 		}
 		ret = 0;
 	}
@@ -3084,17 +3018,15 @@ static int btrfs_orphan_del(struct btrfs_trans_handle *trans,
 		release_rsv = 1;
 	spin_unlock(&root->orphan_lock);
 
-	if (trans && delete_item) {
+	if (trans && delete_item)
 		ret = btrfs_del_orphan_item(trans, root, btrfs_ino(inode));
-		BUG_ON(ret); /* -ENOMEM or corruption (JDM: Recheck) */
-	}
 
 	if (release_rsv) {
 		btrfs_orphan_release_metadata(inode);
 		atomic_dec(&root->orphan_inodes);
 	}
 
-	return 0;
+	return ret;
 }
 
 /*
@@ -3224,8 +3156,9 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 				found_key.objectid);
 			ret = btrfs_del_orphan_item(trans, root,
 						    found_key.objectid);
-			BUG_ON(ret); /* -ENOMEM or corruption (JDM: Recheck) */
 			btrfs_end_transaction(trans, root);
+			if (ret)
+				goto out;
 			continue;
 		}
 
@@ -3657,8 +3590,7 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans,
 	if (ret) {
 		btrfs_info(root->fs_info,
 			"failed to delete reference to %.*s, inode %llu parent %llu",
-			name_len, name,
-			(unsigned long long)ino, (unsigned long long)dir_ino);
+			name_len, name, ino, dir_ino);
 		btrfs_abort_transaction(trans, root, ret);
 		goto err;
 	}
@@ -3929,6 +3861,7 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,
 	u64 extent_num_bytes = 0;
 	u64 extent_offset = 0;
 	u64 item_end = 0;
+	u64 last_size = (u64)-1;
 	u32 found_type = (u8)-1;
 	int found_extent;
 	int del_item;
@@ -4026,6 +3959,11 @@ search_again:
 		if (found_type != BTRFS_EXTENT_DATA_KEY)
 			goto delete;
 
+		if (del_item)
+			last_size = found_key.offset;
+		else
+			last_size = new_size;
+
 		if (extent_type != BTRFS_FILE_EXTENT_INLINE) {
 			u64 num_dec;
 			extent_start = btrfs_file_extent_disk_bytenr(leaf, fi);
@@ -4137,6 +4075,8 @@ out:
 			btrfs_abort_transaction(trans, root, ret);
 	}
 error:
+	if (last_size != (u64)-1)
+		btrfs_ordered_update_i_size(inode, last_size, NULL);
 	btrfs_free_path(path);
 	return err;
 }
@@ -4465,8 +4405,26 @@ static int btrfs_setsize(struct inode *inode, struct iattr *attr)
 		btrfs_inode_resume_unlocked_dio(inode);
 
 		ret = btrfs_truncate(inode);
-		if (ret && inode->i_nlink)
-			btrfs_orphan_del(NULL, inode);
+		if (ret && inode->i_nlink) {
+			int err;
+
+			/*
+			 * failed to truncate, disk_i_size is only adjusted down
+			 * as we remove extents, so it should represent the true
+			 * size of the inode, so reset the in memory size and
+			 * delete our orphan entry.
+			 */
+			trans = btrfs_join_transaction(root);
+			if (IS_ERR(trans)) {
+				btrfs_orphan_del(NULL, inode);
+				return ret;
+			}
+			i_size_write(inode, BTRFS_I(inode)->disk_i_size);
+			err = btrfs_orphan_del(trans, inode);
+			if (err)
+				btrfs_abort_transaction(trans, root, err);
+			btrfs_end_transaction(trans, root);
+		}
 	}
 
 	return ret;
@@ -4601,10 +4559,15 @@ void btrfs_evict_inode(struct inode *inode)
 
 	btrfs_free_block_rsv(root, rsv);
 
+	/*
+	 * Errors here aren't a big deal, it just means we leave orphan items
+	 * in the tree.  They will be cleaned up on the next mount.
+	 */
 	if (ret == 0) {
 		trans->block_rsv = root->orphan_block_rsv;
-		ret = btrfs_orphan_del(trans, inode);
-		BUG_ON(ret);
+		btrfs_orphan_del(trans, inode);
+	} else {
+		btrfs_orphan_del(NULL, inode);
 	}
 
 	trans->block_rsv = &root->fs_info->trans_block_rsv;
@@ -6161,10 +6124,7 @@ insert:
 	btrfs_release_path(path);
 	if (em->start > start || extent_map_end(em) <= start) {
 		btrfs_err(root->fs_info, "bad extent! em: [%llu %llu] passed [%llu %llu]",
-			(unsigned long long)em->start,
-			(unsigned long long)em->len,
-			(unsigned long long)start,
-			(unsigned long long)len);
+			em->start, em->len, start, len);
 		err = -EIO;
 		goto out;
 	}
@@ -6362,39 +6322,32 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode,
 						  u64 start, u64 len)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct btrfs_trans_handle *trans;
 	struct extent_map *em;
 	struct btrfs_key ins;
 	u64 alloc_hint;
 	int ret;
 
-	trans = btrfs_join_transaction(root);
-	if (IS_ERR(trans))
-		return ERR_CAST(trans);
-
-	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
-
 	alloc_hint = get_extent_allocation_hint(inode, start, len);
-	ret = btrfs_reserve_extent(trans, root, len, root->sectorsize, 0,
+	ret = btrfs_reserve_extent(root, len, root->sectorsize, 0,
 				   alloc_hint, &ins, 1);
-	if (ret) {
-		em = ERR_PTR(ret);
-		goto out;
-	}
+	if (ret)
+		return ERR_PTR(ret);
 
 	em = create_pinned_em(inode, start, ins.offset, start, ins.objectid,
 			      ins.offset, ins.offset, ins.offset, 0);
-	if (IS_ERR(em))
-		goto out;
+	if (IS_ERR(em)) {
+		btrfs_free_reserved_extent(root, ins.objectid, ins.offset);
+		return em;
+	}
 
 	ret = btrfs_add_ordered_extent_dio(inode, start, ins.objectid,
 					   ins.offset, ins.offset, 0);
 	if (ret) {
 		btrfs_free_reserved_extent(root, ins.objectid, ins.offset);
-		em = ERR_PTR(ret);
+		free_extent_map(em);
+		return ERR_PTR(ret);
 	}
-out:
-	btrfs_end_transaction(trans, root);
+
 	return em;
 }
 
@@ -6402,11 +6355,11 @@ out:
  * returns 1 when the nocow is safe, < 1 on error, 0 if the
  * block must be cow'd
  */
-noinline int can_nocow_extent(struct btrfs_trans_handle *trans,
-			      struct inode *inode, u64 offset, u64 *len,
+noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 			      u64 *orig_start, u64 *orig_block_len,
 			      u64 *ram_bytes)
 {
+	struct btrfs_trans_handle *trans;
 	struct btrfs_path *path;
 	int ret;
 	struct extent_buffer *leaf;
@@ -6424,7 +6377,7 @@ noinline int can_nocow_extent(struct btrfs_trans_handle *trans,
 	if (!path)
 		return -ENOMEM;
 
-	ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
+	ret = btrfs_lookup_file_extent(NULL, root, path, btrfs_ino(inode),
 				       offset, 0);
 	if (ret < 0)
 		goto out;
@@ -6489,9 +6442,19 @@ noinline int can_nocow_extent(struct btrfs_trans_handle *trans,
 	 * look for other files referencing this extent, if we
 	 * find any we must cow
 	 */
-	if (btrfs_cross_ref_exist(trans, root, btrfs_ino(inode),
-				  key.offset - backref_offset, disk_bytenr))
+	trans = btrfs_join_transaction(root);
+	if (IS_ERR(trans)) {
+		ret = 0;
 		goto out;
+	}
+
+	ret = btrfs_cross_ref_exist(trans, root, btrfs_ino(inode),
+				    key.offset - backref_offset, disk_bytenr);
+	btrfs_end_transaction(trans, root);
+	if (ret) {
+		ret = 0;
+		goto out;
+	}
 
 	/*
 	 * adjust disk_bytenr and num_bytes to cover just the bytes
@@ -6633,7 +6596,6 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 	u64 start = iblock << inode->i_blkbits;
 	u64 lockstart, lockend;
 	u64 len = bh_result->b_size;
-	struct btrfs_trans_handle *trans;
 	int unlock_bits = EXTENT_LOCKED;
 	int ret = 0;
 
@@ -6715,16 +6677,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 		len = min(len, em->len - (start - em->start));
 		block_start = em->block_start + (start - em->start);
 
-		/*
-		 * we're not going to log anything, but we do need
-		 * to make sure the current transaction stays open
-		 * while we look for nocow cross refs
-		 */
-		trans = btrfs_join_transaction(root);
-		if (IS_ERR(trans))
-			goto must_cow;
-
-		if (can_nocow_extent(trans, inode, start, &len, &orig_start,
+		if (can_nocow_extent(inode, start, &len, &orig_start,
 				     &orig_block_len, &ram_bytes) == 1) {
 			if (type == BTRFS_ORDERED_PREALLOC) {
 				free_extent_map(em);
@@ -6733,24 +6686,20 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 						       block_start, len,
 						       orig_block_len,
 						       ram_bytes, type);
-				if (IS_ERR(em)) {
-					btrfs_end_transaction(trans, root);
+				if (IS_ERR(em))
 					goto unlock_err;
-				}
 			}
 
 			ret = btrfs_add_ordered_extent_dio(inode, start,
 					   block_start, len, len, type);
-			btrfs_end_transaction(trans, root);
 			if (ret) {
 				free_extent_map(em);
 				goto unlock_err;
 			}
 			goto unlock;
 		}
-		btrfs_end_transaction(trans, root);
 	}
-must_cow:
+
 	/*
 	 * this will cow the extent, reset the len in case we changed
 	 * it above
@@ -6813,26 +6762,6 @@ unlock_err:
 	return ret;
 }
 
-struct btrfs_dio_private {
-	struct inode *inode;
-	u64 logical_offset;
-	u64 disk_bytenr;
-	u64 bytes;
-	void *private;
-
-	/* number of bios pending for this dio */
-	atomic_t pending_bios;
-
-	/* IO errors */
-	int errors;
-
-	/* orig_bio is our btrfs_io_bio */
-	struct bio *orig_bio;
-
-	/* dio_bio came from fs/direct-io.c */
-	struct bio *dio_bio;
-};
-
 static void btrfs_endio_direct_read(struct bio *bio, int err)
 {
 	struct btrfs_dio_private *dip = bio->bi_private;
@@ -6841,6 +6770,8 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 	struct inode *inode = dip->inode;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct bio *dio_bio;
+	u32 *csums = (u32 *)dip->csum;
+	int index = 0;
 	u64 start;
 
 	start = dip->logical_offset;
@@ -6849,12 +6780,8 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 			struct page *page = bvec->bv_page;
 			char *kaddr;
 			u32 csum = ~(u32)0;
-			u64 private = ~(u32)0;
 			unsigned long flags;
 
-			if (get_state_private(&BTRFS_I(inode)->io_tree,
-					      start, &private))
-				goto failed;
 			local_irq_save(flags);
 			kaddr = kmap_atomic(page);
 			csum = btrfs_csum_data(kaddr + bvec->bv_offset,
@@ -6864,18 +6791,17 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 			local_irq_restore(flags);
 
 			flush_dcache_page(bvec->bv_page);
-			if (csum != private) {
-failed:
-				btrfs_err(root->fs_info, "csum failed ino %llu off %llu csum %u private %u",
-					(unsigned long long)btrfs_ino(inode),
-					(unsigned long long)start,
-					csum, (unsigned)private);
+			if (csum != csums[index]) {
+				btrfs_err(root->fs_info, "csum failed ino %llu off %llu csum %u expected csum %u",
+					  btrfs_ino(inode), start, csum,
+					  csums[index]);
 				err = -EIO;
 			}
 		}
 
 		start += bvec->bv_len;
 		bvec++;
+		index++;
 	} while (bvec <= bvec_end);
 
 	unlock_extent(&BTRFS_I(inode)->io_tree, dip->logical_offset,
@@ -6956,7 +6882,7 @@ static void btrfs_end_dio_bio(struct bio *bio, int err)
 	if (err) {
 		printk(KERN_ERR "btrfs direct IO failed ino %llu rw %lu "
 		      "sector %#Lx len %u err no %d\n",
-		      (unsigned long long)btrfs_ino(dip->inode), bio->bi_rw,
+		      btrfs_ino(dip->inode), bio->bi_rw,
 		      (unsigned long long)bio->bi_sector, bio->bi_size, err);
 		dip->errors = 1;
 
@@ -6992,6 +6918,7 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
 					 int rw, u64 file_offset, int skip_sum,
 					 int async_submit)
 {
+	struct btrfs_dio_private *dip = bio->bi_private;
 	int write = rw & REQ_WRITE;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int ret;
@@ -7026,7 +6953,8 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
 		if (ret)
 			goto err;
 	} else if (!skip_sum) {
-		ret = btrfs_lookup_bio_sums_dio(root, inode, bio, file_offset);
+		ret = btrfs_lookup_bio_sums_dio(root, inode, dip, bio,
+						file_offset);
 		if (ret)
 			goto err;
 	}
@@ -7061,6 +6989,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 		bio_put(orig_bio);
 		return -EIO;
 	}
+
 	if (map_length >= orig_bio->bi_size) {
 		bio = orig_bio;
 		goto submit;
@@ -7156,19 +7085,28 @@ static void btrfs_submit_direct(int rw, struct bio *dio_bio,
 	struct btrfs_dio_private *dip;
 	struct bio *io_bio;
 	int skip_sum;
+	int sum_len;
 	int write = rw & REQ_WRITE;
 	int ret = 0;
+	u16 csum_size;
 
 	skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 
 	io_bio = btrfs_bio_clone(dio_bio, GFP_NOFS);
-
 	if (!io_bio) {
 		ret = -ENOMEM;
 		goto free_ordered;
 	}
 
-	dip = kmalloc(sizeof(*dip), GFP_NOFS);
+	if (!skip_sum && !write) {
+		csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
+		sum_len = dio_bio->bi_size >> inode->i_sb->s_blocksize_bits;
+		sum_len *= csum_size;
+	} else {
+		sum_len = 0;
+	}
+
+	dip = kmalloc(sizeof(*dip) + sum_len, GFP_NOFS);
 	if (!dip) {
 		ret = -ENOMEM;
 		goto free_io_bio;
@@ -7443,10 +7381,23 @@ static void btrfs_invalidatepage(struct page *page, unsigned int offset,
 		 * whoever cleared the private bit is responsible
 		 * for the finish_ordered_io
 		 */
-		if (TestClearPagePrivate2(page) &&
-		    btrfs_dec_test_ordered_pending(inode, &ordered, page_start,
-						   PAGE_CACHE_SIZE, 1)) {
-			btrfs_finish_ordered_io(ordered);
+		if (TestClearPagePrivate2(page)) {
+			struct btrfs_ordered_inode_tree *tree;
+			u64 new_len;
+
+			tree = &BTRFS_I(inode)->ordered_tree;
+
+			spin_lock_irq(&tree->lock);
+			set_bit(BTRFS_ORDERED_TRUNCATED, &ordered->flags);
+			new_len = page_start - ordered->file_offset;
+			if (new_len < ordered->truncated_len)
+				ordered->truncated_len = new_len;
+			spin_unlock_irq(&tree->lock);
+
+			if (btrfs_dec_test_ordered_pending(inode, &ordered,
+							   page_start,
+							   PAGE_CACHE_SIZE, 1))
+				btrfs_finish_ordered_io(ordered);
 		}
 		btrfs_put_ordered_extent(ordered);
 		cached_state = NULL;
@@ -7612,7 +7563,6 @@ static int btrfs_truncate(struct inode *inode)
 	u64 min_size = btrfs_calc_trunc_metadata_size(root, 1);
 
 	btrfs_wait_ordered_range(inode, inode->i_size & (~mask), (u64)-1);
-	btrfs_ordered_update_i_size(inode, inode->i_size, NULL);
 
 	/*
 	 * Yes ladies and gentelment, this is indeed ugly.  The fact is we have
@@ -7876,7 +7826,7 @@ void btrfs_destroy_inode(struct inode *inode)
 	if (test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
 		     &BTRFS_I(inode)->runtime_flags)) {
 		btrfs_info(root->fs_info, "inode %llu still on the orphan list",
-			(unsigned long long)btrfs_ino(inode));
+			btrfs_ino(inode));
 		atomic_dec(&root->orphan_inodes);
 	}
 
@@ -7886,8 +7836,7 @@ void btrfs_destroy_inode(struct inode *inode)
 			break;
 		else {
 			btrfs_err(root->fs_info, "found ordered extent %llu %llu on inode cleanup",
-				(unsigned long long)ordered->file_offset,
-				(unsigned long long)ordered->len);
+				ordered->file_offset, ordered->len);
 			btrfs_remove_ordered_extent(inode, ordered);
 			btrfs_put_ordered_extent(ordered);
 			btrfs_put_ordered_extent(ordered);
@@ -8161,10 +8110,8 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 						 new_dentry->d_name.name,
 						 new_dentry->d_name.len);
 		}
-		if (!ret && new_inode->i_nlink == 0) {
+		if (!ret && new_inode->i_nlink == 0)
 			ret = btrfs_orphan_add(trans, new_dentry->d_inode);
-			BUG_ON(ret);
-		}
 		if (ret) {
 			btrfs_abort_transaction(trans, root, ret);
 			goto out_fail;
@@ -8525,8 +8472,8 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 
 		cur_bytes = min(num_bytes, 256ULL * 1024 * 1024);
 		cur_bytes = max(cur_bytes, min_size);
-		ret = btrfs_reserve_extent(trans, root, cur_bytes,
-					   min_size, 0, *alloc_hint, &ins, 1);
+		ret = btrfs_reserve_extent(root, cur_bytes, min_size, 0,
+					   *alloc_hint, &ins, 1);
 		if (ret) {
 			if (own_trans)
 				btrfs_end_transaction(trans, root);

fs/btrfs/ioctl.c

@@ -43,6 +43,7 @@
 #include <linux/blkdev.h>
 #include <linux/uuid.h>
 #include <linux/btrfs.h>
+#include <linux/uaccess.h>
 #include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
@@ -57,6 +58,9 @@
 #include "send.h"
 #include "dev-replace.h"
 
+static int btrfs_clone(struct inode *src, struct inode *inode,
+		       u64 off, u64 olen, u64 olen_aligned, u64 destoff);
+
 /* Mask out flags that are inappropriate for the given type of inode. */
 static inline __u32 btrfs_mask_flags(umode_t mode, __u32 flags)
 {
@@ -363,6 +367,13 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
 	return 0;
 }
 
+int btrfs_is_empty_uuid(u8 *uuid)
+{
+	static char empty_uuid[BTRFS_UUID_SIZE] = {0};
+
+	return !memcmp(uuid, empty_uuid, BTRFS_UUID_SIZE);
+}
+
 static noinline int create_subvol(struct inode *dir,
 				  struct dentry *dentry,
 				  char *name, int namelen,
@@ -396,7 +407,7 @@ static noinline int create_subvol(struct inode *dir,
 	 * of create_snapshot().
 	 */
 	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
-					       7, &qgroup_reserved);
+					       8, &qgroup_reserved, false);
 	if (ret)
 		return ret;
 
@@ -425,26 +436,25 @@ 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,
-			    (unsigned long)btrfs_header_fsid(leaf),
+	write_extent_buffer(leaf, root->fs_info->fsid, btrfs_header_fsid(leaf),
 			    BTRFS_FSID_SIZE);
 	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(leaf),
+			    btrfs_header_chunk_tree_uuid(leaf),
 			    BTRFS_UUID_SIZE);
 	btrfs_mark_buffer_dirty(leaf);
 
 	memset(&root_item, 0, sizeof(root_item));
 
 	inode_item = &root_item.inode;
-	inode_item->generation = cpu_to_le64(1);
-	inode_item->size = cpu_to_le64(3);
-	inode_item->nlink = cpu_to_le32(1);
-	inode_item->nbytes = cpu_to_le64(root->leafsize);
-	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);
+	btrfs_set_stack_inode_generation(inode_item, 1);
+	btrfs_set_stack_inode_size(inode_item, 3);
+	btrfs_set_stack_inode_nlink(inode_item, 1);
+	btrfs_set_stack_inode_nbytes(inode_item, root->leafsize);
+	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
 
-	root_item.flags = 0;
-	root_item.byte_limit = 0;
-	inode_item->flags = cpu_to_le64(BTRFS_INODE_ROOT_ITEM_INIT);
+	btrfs_set_root_flags(&root_item, 0);
+	btrfs_set_root_limit(&root_item, 0);
+	btrfs_set_stack_inode_flags(inode_item, BTRFS_INODE_ROOT_ITEM_INIT);
 
 	btrfs_set_root_bytenr(&root_item, leaf->start);
 	btrfs_set_root_generation(&root_item, trans->transid);
@@ -457,8 +467,8 @@ static noinline int create_subvol(struct inode *dir,
 			btrfs_root_generation(&root_item));
 	uuid_le_gen(&new_uuid);
 	memcpy(root_item.uuid, new_uuid.b, BTRFS_UUID_SIZE);
-	root_item.otime.sec = cpu_to_le64(cur_time.tv_sec);
-	root_item.otime.nsec = cpu_to_le32(cur_time.tv_nsec);
+	btrfs_set_stack_timespec_sec(&root_item.otime, cur_time.tv_sec);
+	btrfs_set_stack_timespec_nsec(&root_item.otime, cur_time.tv_nsec);
 	root_item.ctime = root_item.otime;
 	btrfs_set_root_ctransid(&root_item, trans->transid);
 	btrfs_set_root_otransid(&root_item, trans->transid);
@@ -518,9 +528,14 @@ static noinline int create_subvol(struct inode *dir,
 	ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
 				 objectid, root->root_key.objectid,
 				 btrfs_ino(dir), index, name, namelen);
-
 	BUG_ON(ret);
 
+	ret = btrfs_uuid_tree_add(trans, root->fs_info->uuid_root,
+				  root_item.uuid, BTRFS_UUID_KEY_SUBVOL,
+				  objectid);
+	if (ret)
+		btrfs_abort_transaction(trans, root, ret);
+
 fail:
 	trans->block_rsv = NULL;
 	trans->bytes_reserved = 0;
@@ -573,10 +588,12 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	 * 1 - root item
 	 * 2 - root ref/backref
 	 * 1 - root of snapshot
+	 * 1 - UUID item
 	 */
 	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
-					&pending_snapshot->block_rsv, 7,
-					&pending_snapshot->qgroup_reserved);
+					&pending_snapshot->block_rsv, 8,
+					&pending_snapshot->qgroup_reserved,
+					false);
 	if (ret)
 		goto out;
 
@@ -1267,9 +1284,6 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			cluster = max_cluster;
 		}
 
-		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
-			BTRFS_I(inode)->force_compress = compress_type;
-
 		if (i + cluster > ra_index) {
 			ra_index = max(i, ra_index);
 			btrfs_force_ra(inode->i_mapping, ra, file, ra_index,
@@ -1278,6 +1292,8 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 		}
 
 		mutex_lock(&inode->i_mutex);
+		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
+			BTRFS_I(inode)->force_compress = compress_type;
 		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
 		if (ret < 0) {
 			mutex_unlock(&inode->i_mutex);
@@ -1334,10 +1350,6 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			    atomic_read(&root->fs_info->async_delalloc_pages) == 0));
 		}
 		atomic_dec(&root->fs_info->async_submit_draining);
-
-		mutex_lock(&inode->i_mutex);
-		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
-		mutex_unlock(&inode->i_mutex);
 	}
 
 	if (range->compress_type == BTRFS_COMPRESS_LZO) {
@@ -1347,6 +1359,11 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	ret = defrag_count;
 
 out_ra:
+	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
+		mutex_lock(&inode->i_mutex);
+		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
+		mutex_unlock(&inode->i_mutex);
+	}
 	if (!file)
 		kfree(ra);
 	kfree(pages);
@@ -1377,9 +1394,8 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 
 	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
 			1)) {
-		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
 		mnt_drop_write_file(file);
-		return -EINVAL;
+		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 	}
 
 	mutex_lock(&root->fs_info->volume_mutex);
@@ -1403,14 +1419,13 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 			ret = -EINVAL;
 			goto out_free;
 		}
-		printk(KERN_INFO "btrfs: resizing devid %llu\n",
-		       (unsigned long long)devid);
+		printk(KERN_INFO "btrfs: resizing devid %llu\n", devid);
 	}
 
 	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
 	if (!device) {
 		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
-		       (unsigned long long)devid);
+		       devid);
 		ret = -ENODEV;
 		goto out_free;
 	}
@@ -1418,7 +1433,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 	if (!device->writeable) {
 		printk(KERN_INFO "btrfs: resizer unable to apply on "
 		       "readonly device %llu\n",
-		       (unsigned long long)devid);
+		       devid);
 		ret = -EPERM;
 		goto out_free;
 	}
@@ -1470,8 +1485,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 	new_size *= root->sectorsize;
 
 	printk_in_rcu(KERN_INFO "btrfs: new size for %s is %llu\n",
-		      rcu_str_deref(device->name),
-		      (unsigned long long)new_size);
+		      rcu_str_deref(device->name), new_size);
 
 	if (new_size > old_size) {
 		trans = btrfs_start_transaction(root, 0);
@@ -1721,13 +1735,28 @@ out:
 static noinline int may_destroy_subvol(struct btrfs_root *root)
 {
 	struct btrfs_path *path;
+	struct btrfs_dir_item *di;
 	struct btrfs_key key;
+	u64 dir_id;
 	int ret;
 
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
 
+	/* Make sure this root isn't set as the default subvol */
+	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
+	di = btrfs_lookup_dir_item(NULL, root->fs_info->tree_root, path,
+				   dir_id, "default", 7, 0);
+	if (di && !IS_ERR(di)) {
+		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &key);
+		if (key.objectid == root->root_key.objectid) {
+			ret = -ENOTEMPTY;
+			goto out;
+		}
+		btrfs_release_path(path);
+	}
+
 	key.objectid = root->root_key.objectid;
 	key.type = BTRFS_ROOT_REF_KEY;
 	key.offset = (u64)-1;
@@ -1993,25 +2022,29 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 		if (ret < 0)
 			goto out;
+		else if (ret > 0) {
+			ret = btrfs_previous_item(root, path, dirid,
+						  BTRFS_INODE_REF_KEY);
+			if (ret < 0)
+				goto out;
+			else if (ret > 0) {
+				ret = -ENOENT;
+				goto out;
+			}
+		}
 
 		l = path->nodes[0];
 		slot = path->slots[0];
-		if (ret > 0 && slot > 0)
-			slot--;
 		btrfs_item_key_to_cpu(l, &key, slot);
 
-		if (ret > 0 && (key.objectid != dirid ||
-				key.type != BTRFS_INODE_REF_KEY)) {
-			ret = -ENOENT;
-			goto out;
-		}
-
 		iref = btrfs_item_ptr(l, slot, struct btrfs_inode_ref);
 		len = btrfs_inode_ref_name_len(l, iref);
 		ptr -= len + 1;
 		total_len += len + 1;
-		if (ptr < name)
+		if (ptr < name) {
+			ret = -ENAMETOOLONG;
 			goto out;
+		}
 
 		*(ptr + len) = '/';
 		read_extent_buffer(l, ptr,(unsigned long)(iref + 1), len);
@@ -2024,8 +2057,6 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 		key.offset = (u64)-1;
 		dirid = key.objectid;
 	}
-	if (ptr < name)
-		goto out;
 	memmove(name, ptr, total_len);
 	name[total_len]='\0';
 	ret = 0;
@@ -2174,7 +2205,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 	 * ref/backref.
 	 */
 	err = btrfs_subvolume_reserve_metadata(root, &block_rsv,
-					       5, &qgroup_reserved);
+					       5, &qgroup_reserved, true);
 	if (err)
 		goto out_up_write;
 
@@ -2213,6 +2244,27 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 			goto out_end_trans;
 		}
 	}
+
+	ret = btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
+				  dest->root_item.uuid, BTRFS_UUID_KEY_SUBVOL,
+				  dest->root_key.objectid);
+	if (ret && ret != -ENOENT) {
+		btrfs_abort_transaction(trans, root, ret);
+		err = ret;
+		goto out_end_trans;
+	}
+	if (!btrfs_is_empty_uuid(dest->root_item.received_uuid)) {
+		ret = btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
+					  dest->root_item.received_uuid,
+					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+					  dest->root_key.objectid);
+		if (ret && ret != -ENOENT) {
+			btrfs_abort_transaction(trans, root, ret);
+			err = ret;
+			goto out_end_trans;
+		}
+	}
+
 out_end_trans:
 	trans->block_rsv = NULL;
 	trans->bytes_reserved = 0;
@@ -2326,8 +2378,7 @@ static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
 
 	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
 			1)) {
-		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
-		return -EINVAL;
+		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 	}
 
 	mutex_lock(&root->fs_info->volume_mutex);
@@ -2400,10 +2451,10 @@ static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
 	if (!fi_args)
 		return -ENOMEM;
 
+	mutex_lock(&fs_devices->device_list_mutex);
 	fi_args->num_devices = fs_devices->num_devices;
 	memcpy(&fi_args->fsid, root->fs_info->fsid, sizeof(fi_args->fsid));
 
-	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
 		if (device->devid > fi_args->max_id)
 			fi_args->max_id = device->devid;
@@ -2424,7 +2475,6 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 	int ret = 0;
 	char *s_uuid = NULL;
-	char empty_uuid[BTRFS_UUID_SIZE] = {0};
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -2433,7 +2483,7 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 	if (IS_ERR(di_args))
 		return PTR_ERR(di_args);
 
-	if (memcmp(empty_uuid, di_args->uuid, BTRFS_UUID_SIZE) != 0)
+	if (!btrfs_is_empty_uuid(di_args->uuid))
 		s_uuid = di_args->uuid;
 
 	mutex_lock(&fs_devices->device_list_mutex);
@@ -2469,150 +2519,336 @@ out:
 	return ret;
 }
 
-static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
-				       u64 off, u64 olen, u64 destoff)
+static struct page *extent_same_get_page(struct inode *inode, u64 off)
+{
+	struct page *page;
+	pgoff_t index;
+	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;
+
+	index = off >> PAGE_CACHE_SHIFT;
+
+	page = grab_cache_page(inode->i_mapping, index);
+	if (!page)
+		return NULL;
+
+	if (!PageUptodate(page)) {
+		if (extent_read_full_page_nolock(tree, page, btrfs_get_extent,
+						 0))
+			return NULL;
+		lock_page(page);
+		if (!PageUptodate(page)) {
+			unlock_page(page);
+			page_cache_release(page);
+			return NULL;
+		}
+	}
+	unlock_page(page);
+
+	return page;
+}
+
+static inline void lock_extent_range(struct inode *inode, u64 off, u64 len)
+{
+	/* do any pending delalloc/csum calc on src, one way or
+	   another, and lock file content */
+	while (1) {
+		struct btrfs_ordered_extent *ordered;
+		lock_extent(&BTRFS_I(inode)->io_tree, off, off + len - 1);
+		ordered = btrfs_lookup_first_ordered_extent(inode,
+							    off + len - 1);
+		if (!ordered &&
+		    !test_range_bit(&BTRFS_I(inode)->io_tree, off,
+				    off + len - 1, EXTENT_DELALLOC, 0, NULL))
+			break;
+		unlock_extent(&BTRFS_I(inode)->io_tree, off, off + len - 1);
+		if (ordered)
+			btrfs_put_ordered_extent(ordered);
+		btrfs_wait_ordered_range(inode, off, len);
+	}
+}
+
+static void btrfs_double_unlock(struct inode *inode1, u64 loff1,
+				struct inode *inode2, u64 loff2, u64 len)
+{
+	unlock_extent(&BTRFS_I(inode1)->io_tree, loff1, loff1 + len - 1);
+	unlock_extent(&BTRFS_I(inode2)->io_tree, loff2, loff2 + len - 1);
+
+	mutex_unlock(&inode1->i_mutex);
+	mutex_unlock(&inode2->i_mutex);
+}
+
+static void btrfs_double_lock(struct inode *inode1, u64 loff1,
+			      struct inode *inode2, u64 loff2, u64 len)
+{
+	if (inode1 < inode2) {
+		swap(inode1, inode2);
+		swap(loff1, loff2);
+	}
+
+	mutex_lock_nested(&inode1->i_mutex, I_MUTEX_PARENT);
+	lock_extent_range(inode1, loff1, len);
+	if (inode1 != inode2) {
+		mutex_lock_nested(&inode2->i_mutex, I_MUTEX_CHILD);
+		lock_extent_range(inode2, loff2, len);
+	}
+}
+
+static int btrfs_cmp_data(struct inode *src, u64 loff, struct inode *dst,
+			  u64 dst_loff, u64 len)
+{
+	int ret = 0;
+	struct page *src_page, *dst_page;
+	unsigned int cmp_len = PAGE_CACHE_SIZE;
+	void *addr, *dst_addr;
+
+	while (len) {
+		if (len < PAGE_CACHE_SIZE)
+			cmp_len = len;
+
+		src_page = extent_same_get_page(src, loff);
+		if (!src_page)
+			return -EINVAL;
+		dst_page = extent_same_get_page(dst, dst_loff);
+		if (!dst_page) {
+			page_cache_release(src_page);
+			return -EINVAL;
+		}
+		addr = kmap_atomic(src_page);
+		dst_addr = kmap_atomic(dst_page);
+
+		flush_dcache_page(src_page);
+		flush_dcache_page(dst_page);
+
+		if (memcmp(addr, dst_addr, cmp_len))
+			ret = BTRFS_SAME_DATA_DIFFERS;
+
+		kunmap_atomic(addr);
+		kunmap_atomic(dst_addr);
+		page_cache_release(src_page);
+		page_cache_release(dst_page);
+
+		if (ret)
+			break;
+
+		loff += cmp_len;
+		dst_loff += cmp_len;
+		len -= cmp_len;
+	}
+
+	return ret;
+}
+
+static int extent_same_check_offsets(struct inode *inode, u64 off, u64 len)
+{
+	u64 bs = BTRFS_I(inode)->root->fs_info->sb->s_blocksize;
+
+	if (off + len > inode->i_size || off + len < off)
+		return -EINVAL;
+	/* Check that we are block aligned - btrfs_clone() requires this */
+	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs))
+		return -EINVAL;
+
+	return 0;
+}
+
+static int btrfs_extent_same(struct inode *src, u64 loff, u64 len,
+			     struct inode *dst, u64 dst_loff)
 {
-	struct inode *inode = file_inode(file);
-	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct fd src_file;
-	struct inode *src;
-	struct btrfs_trans_handle *trans;
-	struct btrfs_path *path;
-	struct extent_buffer *leaf;
-	char *buf;
-	struct btrfs_key key;
-	u32 nritems;
-	int slot;
 	int ret;
-	u64 len = olen;
-	u64 bs = root->fs_info->sb->s_blocksize;
-	int same_inode = 0;
 
 	/*
-	 * TODO:
-	 * - split compressed inline extents.  annoying: we need to
-	 *   decompress into destination's address_space (the file offset
-	 *   may change, so source mapping won't do), then recompress (or
-	 *   otherwise reinsert) a subrange.
-	 * - allow ranges within the same file to be cloned (provided
-	 *   they don't overlap)?
+	 * btrfs_clone() can't handle extents in the same file
+	 * yet. Once that works, we can drop this check and replace it
+	 * with a check for the same inode, but overlapping extents.
 	 */
-
-	/* the destination must be opened for writing */
-	if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND))
+	if (src == dst)
 		return -EINVAL;
 
-	if (btrfs_root_readonly(root))
-		return -EROFS;
+	btrfs_double_lock(src, loff, dst, dst_loff, len);
+
+	ret = extent_same_check_offsets(src, loff, len);
+	if (ret)
+		goto out_unlock;
+
+	ret = extent_same_check_offsets(dst, dst_loff, len);
+	if (ret)
+		goto out_unlock;
+
+	/* don't make the dst file partly checksummed */
+	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
+	    (BTRFS_I(dst)->flags & BTRFS_INODE_NODATASUM)) {
+		ret = -EINVAL;
+		goto out_unlock;
+	}
+
+	ret = btrfs_cmp_data(src, loff, dst, dst_loff, len);
+	if (ret == 0)
+		ret = btrfs_clone(src, dst, loff, len, len, dst_loff);
+
+out_unlock:
+	btrfs_double_unlock(src, loff, dst, dst_loff, len);
+
+	return ret;
+}
+
+#define BTRFS_MAX_DEDUPE_LEN	(16 * 1024 * 1024)
+
+static long btrfs_ioctl_file_extent_same(struct file *file,
+					 void __user *argp)
+{
+	struct btrfs_ioctl_same_args *args = argp;
+	struct btrfs_ioctl_same_args same;
+	struct btrfs_ioctl_same_extent_info info;
+	struct inode *src = file->f_dentry->d_inode;
+	struct file *dst_file = NULL;
+	struct inode *dst;
+	u64 off;
+	u64 len;
+	int i;
+	int ret;
+	u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize;
+	bool is_admin = capable(CAP_SYS_ADMIN);
+
+	if (!(file->f_mode & FMODE_READ))
+		return -EINVAL;
 
 	ret = mnt_want_write_file(file);
 	if (ret)
 		return ret;
 
-	src_file = fdget(srcfd);
-	if (!src_file.file) {
-		ret = -EBADF;
-		goto out_drop_write;
+	if (copy_from_user(&same,
+			   (struct btrfs_ioctl_same_args __user *)argp,
+			   sizeof(same))) {
+		ret = -EFAULT;
+		goto out;
 	}
 
-	ret = -EXDEV;
-	if (src_file.file->f_path.mnt != file->f_path.mnt)
-		goto out_fput;
+	off = same.logical_offset;
+	len = same.length;
 
-	src = file_inode(src_file.file);
+	/*
+	 * Limit the total length we will dedupe for each operation.
+	 * This is intended to bound the total time spent in this
+	 * ioctl to something sane.
+	 */
+	if (len > BTRFS_MAX_DEDUPE_LEN)
+		len = BTRFS_MAX_DEDUPE_LEN;
 
-	ret = -EINVAL;
-	if (src == inode)
-		same_inode = 1;
+	if (WARN_ON_ONCE(bs < PAGE_CACHE_SIZE)) {
+		/*
+		 * Btrfs does not support blocksize < page_size. As a
+		 * result, btrfs_cmp_data() won't correctly handle
+		 * this situation without an update.
+		 */
+		ret = -EINVAL;
+		goto out;
+	}
 
-	/* the src must be open for reading */
-	if (!(src_file.file->f_mode & FMODE_READ))
-		goto out_fput;
+	ret = -EISDIR;
+	if (S_ISDIR(src->i_mode))
+		goto out;
 
-	/* don't make the dst file partly checksummed */
-	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
-	    (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM))
-		goto out_fput;
+	ret = -EACCES;
+	if (!S_ISREG(src->i_mode))
+		goto out;
 
-	ret = -EISDIR;
-	if (S_ISDIR(src->i_mode) || S_ISDIR(inode->i_mode))
-		goto out_fput;
+	ret = 0;
+	for (i = 0; i < same.dest_count; i++) {
+		if (copy_from_user(&info, &args->info[i], sizeof(info))) {
+			ret = -EFAULT;
+			goto out;
+		}
 
-	ret = -EXDEV;
-	if (src->i_sb != inode->i_sb)
-		goto out_fput;
+		info.bytes_deduped = 0;
 
-	ret = -ENOMEM;
-	buf = vmalloc(btrfs_level_size(root, 0));
-	if (!buf)
-		goto out_fput;
+		dst_file = fget(info.fd);
+		if (!dst_file) {
+			info.status = -EBADF;
+			goto next;
+		}
 
-	path = btrfs_alloc_path();
-	if (!path) {
-		vfree(buf);
-		goto out_fput;
-	}
-	path->reada = 2;
+		if (!(is_admin || (dst_file->f_mode & FMODE_WRITE))) {
+			info.status = -EINVAL;
+			goto next;
+		}
 
-	if (!same_inode) {
-		if (inode < src) {
-			mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
-			mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
-		} else {
-			mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
-			mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
+		info.status = -EXDEV;
+		if (file->f_path.mnt != dst_file->f_path.mnt)
+			goto next;
+
+		dst = dst_file->f_dentry->d_inode;
+		if (src->i_sb != dst->i_sb)
+			goto next;
+
+		if (S_ISDIR(dst->i_mode)) {
+			info.status = -EISDIR;
+			goto next;
 		}
-	} else {
-		mutex_lock(&src->i_mutex);
-	}
 
-	/* determine range to clone */
-	ret = -EINVAL;
-	if (off + len > src->i_size || off + len < off)
-		goto out_unlock;
-	if (len == 0)
-		olen = len = src->i_size - off;
-	/* if we extend to eof, continue to block boundary */
-	if (off + len == src->i_size)
-		len = ALIGN(src->i_size, bs) - off;
+		if (!S_ISREG(dst->i_mode)) {
+			info.status = -EACCES;
+			goto next;
+		}
 
-	/* verify the end result is block aligned */
-	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs) ||
-	    !IS_ALIGNED(destoff, bs))
-		goto out_unlock;
+		info.status = btrfs_extent_same(src, off, len, dst,
+						info.logical_offset);
+		if (info.status == 0)
+			info.bytes_deduped += len;
 
-	/* verify if ranges are overlapped within the same file */
-	if (same_inode) {
-		if (destoff + len > off && destoff < off + len)
-			goto out_unlock;
-	}
+next:
+		if (dst_file)
+			fput(dst_file);
 
-	if (destoff > inode->i_size) {
-		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
-		if (ret)
-			goto out_unlock;
+		if (__put_user_unaligned(info.status, &args->info[i].status) ||
+		    __put_user_unaligned(info.bytes_deduped,
+					 &args->info[i].bytes_deduped)) {
+			ret = -EFAULT;
+			goto out;
+		}                                                               
 	}
 
-	/* truncate page cache pages from target inode range */
-	truncate_inode_pages_range(&inode->i_data, destoff,
-				   PAGE_CACHE_ALIGN(destoff + len) - 1);
+out:
+	mnt_drop_write_file(file);
+	return ret;
+}
 
-	/* do any pending delalloc/csum calc on src, one way or
-	   another, and lock file content */
-	while (1) {
-		struct btrfs_ordered_extent *ordered;
-		lock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
-		ordered = btrfs_lookup_first_ordered_extent(src, off + len - 1);
-		if (!ordered &&
-		    !test_range_bit(&BTRFS_I(src)->io_tree, off, off + len - 1,
-				    EXTENT_DELALLOC, 0, NULL))
-			break;
-		unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
-		if (ordered)
-			btrfs_put_ordered_extent(ordered);
-		btrfs_wait_ordered_range(src, off, len);
+/**
+ * btrfs_clone() - clone a range from inode file to another
+ *
+ * @src: Inode to clone from
+ * @inode: Inode to clone to
+ * @off: Offset within source to start clone from
+ * @olen: Original length, passed by user, of range to clone
+ * @olen_aligned: Block-aligned value of olen, extent_same uses
+ *               identical values here
+ * @destoff: Offset within @inode to start clone
+ */
+static int btrfs_clone(struct inode *src, struct inode *inode,
+		       u64 off, u64 olen, u64 olen_aligned, u64 destoff)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct btrfs_path *path = NULL;
+	struct extent_buffer *leaf;
+	struct btrfs_trans_handle *trans;
+	char *buf = NULL;
+	struct btrfs_key key;
+	u32 nritems;
+	int slot;
+	int ret;
+	u64 len = olen_aligned;
+
+	ret = -ENOMEM;
+	buf = vmalloc(btrfs_level_size(root, 0));
+	if (!buf)
+		return ret;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		vfree(buf);
+		return ret;
 	}
 
+	path->reada = 2;
 	/* clone data */
 	key.objectid = btrfs_ino(src);
 	key.type = BTRFS_EXTENT_DATA_KEY;
@@ -2858,15 +3094,132 @@ next:
 		key.offset++;
 	}
 	ret = 0;
+
 out:
 	btrfs_release_path(path);
+	btrfs_free_path(path);
+	vfree(buf);
+	return ret;
+}
+
+static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
+				       u64 off, u64 olen, u64 destoff)
+{
+	struct inode *inode = fdentry(file)->d_inode;
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct fd src_file;
+	struct inode *src;
+	int ret;
+	u64 len = olen;
+	u64 bs = root->fs_info->sb->s_blocksize;
+	int same_inode = 0;
+
+	/*
+	 * TODO:
+	 * - split compressed inline extents.  annoying: we need to
+	 *   decompress into destination's address_space (the file offset
+	 *   may change, so source mapping won't do), then recompress (or
+	 *   otherwise reinsert) a subrange.
+	 * - allow ranges within the same file to be cloned (provided
+	 *   they don't overlap)?
+	 */
+
+	/* the destination must be opened for writing */
+	if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND))
+		return -EINVAL;
+
+	if (btrfs_root_readonly(root))
+		return -EROFS;
+
+	ret = mnt_want_write_file(file);
+	if (ret)
+		return ret;
+
+	src_file = fdget(srcfd);
+	if (!src_file.file) {
+		ret = -EBADF;
+		goto out_drop_write;
+	}
+
+	ret = -EXDEV;
+	if (src_file.file->f_path.mnt != file->f_path.mnt)
+		goto out_fput;
+
+	src = file_inode(src_file.file);
+
+	ret = -EINVAL;
+	if (src == inode)
+		same_inode = 1;
+
+	/* the src must be open for reading */
+	if (!(src_file.file->f_mode & FMODE_READ))
+		goto out_fput;
+
+	/* don't make the dst file partly checksummed */
+	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
+	    (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM))
+		goto out_fput;
+
+	ret = -EISDIR;
+	if (S_ISDIR(src->i_mode) || S_ISDIR(inode->i_mode))
+		goto out_fput;
+
+	ret = -EXDEV;
+	if (src->i_sb != inode->i_sb)
+		goto out_fput;
+
+	if (!same_inode) {
+		if (inode < src) {
+			mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
+			mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
+		} else {
+			mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
+			mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
+		}
+	} else {
+		mutex_lock(&src->i_mutex);
+	}
+
+	/* determine range to clone */
+	ret = -EINVAL;
+	if (off + len > src->i_size || off + len < off)
+		goto out_unlock;
+	if (len == 0)
+		olen = len = src->i_size - off;
+	/* if we extend to eof, continue to block boundary */
+	if (off + len == src->i_size)
+		len = ALIGN(src->i_size, bs) - off;
+
+	/* verify the end result is block aligned */
+	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs) ||
+	    !IS_ALIGNED(destoff, bs))
+		goto out_unlock;
+
+	/* verify if ranges are overlapped within the same file */
+	if (same_inode) {
+		if (destoff + len > off && destoff < off + len)
+			goto out_unlock;
+	}
+
+	if (destoff > inode->i_size) {
+		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
+		if (ret)
+			goto out_unlock;
+	}
+
+	/* truncate page cache pages from target inode range */
+	truncate_inode_pages_range(&inode->i_data, destoff,
+				   PAGE_CACHE_ALIGN(destoff + len) - 1);
+
+	lock_extent_range(src, off, len);
+
+	ret = btrfs_clone(src, inode, off, olen, len, destoff);
+
 	unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
 out_unlock:
 	mutex_unlock(&src->i_mutex);
 	if (!same_inode)
 		mutex_unlock(&inode->i_mutex);
-	vfree(buf);
-	btrfs_free_path(path);
 out_fput:
 	fdput(src_file);
 out_drop_write:
@@ -3312,11 +3665,13 @@ 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)
+			return -EROFS;
+
 		if (atomic_xchg(
 			&root->fs_info->mutually_exclusive_operation_running,
 			1)) {
-			pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
-			ret = -EINPROGRESS;
+			ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 		} else {
 			ret = btrfs_dev_replace_start(root, p);
 			atomic_set(
@@ -3560,8 +3915,7 @@ again:
 	} else {
 		/* this is (1) */
 		mutex_unlock(&fs_info->balance_mutex);
-		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
-		ret = -EINVAL;
+		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 		goto out;
 	}
 
@@ -3967,6 +4321,7 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 	struct btrfs_trans_handle *trans;
 	struct timespec ct = CURRENT_TIME;
 	int ret = 0;
+	int received_uuid_changed;
 
 	ret = mnt_want_write_file(file);
 	if (ret < 0)
@@ -3996,7 +4351,11 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 		goto out;
 	}
 
-	trans = btrfs_start_transaction(root, 1);
+	/*
+	 * 1 - root item
+	 * 2 - uuid items (received uuid + subvol uuid)
+	 */
+	trans = btrfs_start_transaction(root, 3);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
 		trans = NULL;
@@ -4007,24 +4366,42 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 	sa->rtime.sec = ct.tv_sec;
 	sa->rtime.nsec = ct.tv_nsec;
 
+	received_uuid_changed = memcmp(root_item->received_uuid, sa->uuid,
+				       BTRFS_UUID_SIZE);
+	if (received_uuid_changed &&
+	    !btrfs_is_empty_uuid(root_item->received_uuid))
+		btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
+				    root_item->received_uuid,
+				    BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+				    root->root_key.objectid);
 	memcpy(root_item->received_uuid, sa->uuid, BTRFS_UUID_SIZE);
 	btrfs_set_root_stransid(root_item, sa->stransid);
 	btrfs_set_root_rtransid(root_item, sa->rtransid);
-	root_item->stime.sec = cpu_to_le64(sa->stime.sec);
-	root_item->stime.nsec = cpu_to_le32(sa->stime.nsec);
-	root_item->rtime.sec = cpu_to_le64(sa->rtime.sec);
-	root_item->rtime.nsec = cpu_to_le32(sa->rtime.nsec);
+	btrfs_set_stack_timespec_sec(&root_item->stime, sa->stime.sec);
+	btrfs_set_stack_timespec_nsec(&root_item->stime, sa->stime.nsec);
+	btrfs_set_stack_timespec_sec(&root_item->rtime, sa->rtime.sec);
+	btrfs_set_stack_timespec_nsec(&root_item->rtime, sa->rtime.nsec);
 
 	ret = btrfs_update_root(trans, root->fs_info->tree_root,
 				&root->root_key, &root->root_item);
 	if (ret < 0) {
 		btrfs_end_transaction(trans, root);
-		trans = NULL;
 		goto out;
-	} else {
-		ret = btrfs_commit_transaction(trans, root);
-		if (ret < 0)
+	}
+	if (received_uuid_changed && !btrfs_is_empty_uuid(sa->uuid)) {
+		ret = btrfs_uuid_tree_add(trans, root->fs_info->uuid_root,
+					  sa->uuid,
+					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+					  root->root_key.objectid);
+		if (ret < 0 && ret != -EEXIST) {
+			btrfs_abort_transaction(trans, root, ret);
 			goto out;
+		}
+	}
+	ret = btrfs_commit_transaction(trans, root);
+	if (ret < 0) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto out;
 	}
 
 	ret = copy_to_user(arg, sa, sizeof(*sa));
@@ -4041,18 +4418,22 @@ out:
 static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
 {
 	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
-	const char *label = root->fs_info->super_copy->label;
-	size_t len = strnlen(label, BTRFS_LABEL_SIZE);
+	size_t len;
 	int ret;
+	char label[BTRFS_LABEL_SIZE];
+
+	spin_lock(&root->fs_info->super_lock);
+	memcpy(label, root->fs_info->super_copy->label, BTRFS_LABEL_SIZE);
+	spin_unlock(&root->fs_info->super_lock);
+
+	len = strnlen(label, BTRFS_LABEL_SIZE);
 
 	if (len == BTRFS_LABEL_SIZE) {
 		pr_warn("btrfs: label is too long, return the first %zu bytes\n",
 			--len);
 	}
 
-	mutex_lock(&root->fs_info->volume_mutex);
 	ret = copy_to_user(arg, label, len);
-	mutex_unlock(&root->fs_info->volume_mutex);
 
 	return ret ? -EFAULT : 0;
 }
@@ -4081,18 +4462,18 @@ static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
 	if (ret)
 		return ret;
 
-	mutex_lock(&root->fs_info->volume_mutex);
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
 		goto out_unlock;
 	}
 
+	spin_lock(&root->fs_info->super_lock);
 	strcpy(super_block->label, label);
+	spin_unlock(&root->fs_info->super_lock);
 	ret = btrfs_end_transaction(trans, root);
 
 out_unlock:
-	mutex_unlock(&root->fs_info->volume_mutex);
 	mnt_drop_write_file(file);
 	return ret;
 }
@@ -4207,6 +4588,8 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_get_fslabel(file, argp);
 	case BTRFS_IOC_SET_FSLABEL:
 		return btrfs_ioctl_set_fslabel(file, argp);
+	case BTRFS_IOC_FILE_EXTENT_SAME:
+		return btrfs_ioctl_file_extent_same(file, argp);
 	}
 
 	return -ENOTTY;

fs/btrfs/lzo.c

@@ -207,8 +207,10 @@ static int lzo_compress_pages(struct list_head *ws,
 		}
 
 		/* we're making it bigger, give up */
-		if (tot_in > 8192 && tot_in < tot_out)
+		if (tot_in > 8192 && tot_in < tot_out) {
+			ret = -1;
 			goto out;
+		}
 
 		/* we're all done */
 		if (tot_in >= len)

fs/btrfs/ordered-data.c

@@ -67,7 +67,7 @@ static void ordered_data_tree_panic(struct inode *inode, int errno,
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	btrfs_panic(fs_info, errno, "Inconsistency in ordered tree at offset "
-		    "%llu\n", (unsigned long long)offset);
+		    "%llu\n", offset);
 }
 
 /*
@@ -205,6 +205,7 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset,
 	entry->bytes_left = len;
 	entry->inode = igrab(inode);
 	entry->compress_type = compress_type;
+	entry->truncated_len = (u64)-1;
 	if (type != BTRFS_ORDERED_IO_DONE && type != BTRFS_ORDERED_COMPLETE)
 		set_bit(type, &entry->flags);
 
@@ -336,14 +337,12 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode,
 	*file_offset = dec_end;
 	if (dec_start > dec_end) {
 		printk(KERN_CRIT "bad ordering dec_start %llu end %llu\n",
-		       (unsigned long long)dec_start,
-		       (unsigned long long)dec_end);
+		       dec_start, dec_end);
 	}
 	to_dec = dec_end - dec_start;
 	if (to_dec > entry->bytes_left) {
 		printk(KERN_CRIT "bad ordered accounting left %llu size %llu\n",
-		       (unsigned long long)entry->bytes_left,
-		       (unsigned long long)to_dec);
+		       entry->bytes_left, to_dec);
 	}
 	entry->bytes_left -= to_dec;
 	if (!uptodate)
@@ -403,8 +402,7 @@ have_entry:
 
 	if (io_size > entry->bytes_left) {
 		printk(KERN_CRIT "bad ordered accounting left %llu size %llu\n",
-		       (unsigned long long)entry->bytes_left,
-		       (unsigned long long)io_size);
+		       entry->bytes_left, io_size);
 	}
 	entry->bytes_left -= io_size;
 	if (!uptodate)
@@ -671,7 +669,7 @@ int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans,
 	INIT_LIST_HEAD(&splice);
 	INIT_LIST_HEAD(&works);
 
-	mutex_lock(&root->fs_info->ordered_operations_mutex);
+	mutex_lock(&root->fs_info->ordered_extent_flush_mutex);
 	spin_lock(&root->fs_info->ordered_root_lock);
 	list_splice_init(&cur_trans->ordered_operations, &splice);
 	while (!list_empty(&splice)) {
@@ -718,7 +716,7 @@ out:
 		list_del_init(&work->list);
 		btrfs_wait_and_free_delalloc_work(work);
 	}
-	mutex_unlock(&root->fs_info->ordered_operations_mutex);
+	mutex_unlock(&root->fs_info->ordered_extent_flush_mutex);
 	return ret;
 }
 
@@ -923,12 +921,16 @@ int btrfs_ordered_update_i_size(struct inode *inode, u64 offset,
 	struct btrfs_ordered_extent *test;
 	int ret = 1;
 
-	if (ordered)
+	spin_lock_irq(&tree->lock);
+	if (ordered) {
 		offset = entry_end(ordered);
-	else
+		if (test_bit(BTRFS_ORDERED_TRUNCATED, &ordered->flags))
+			offset = min(offset,
+				     ordered->file_offset +
+				     ordered->truncated_len);
+	} else {
 		offset = ALIGN(offset, BTRFS_I(inode)->root->sectorsize);
-
-	spin_lock_irq(&tree->lock);
+	}
 	disk_i_size = BTRFS_I(inode)->disk_i_size;
 
 	/* truncate file */

fs/btrfs/ordered-data.h

@@ -69,6 +69,7 @@ struct btrfs_ordered_sum {
 				       * the isize. */
 #define BTRFS_ORDERED_LOGGED_CSUM 8 /* We've logged the csums on this ordered
 				       ordered extent */
+#define BTRFS_ORDERED_TRUNCATED 9 /* Set when we have to truncate an extent */
 
 struct btrfs_ordered_extent {
 	/* logical offset in the file */
@@ -96,6 +97,12 @@ struct btrfs_ordered_extent {
 	 */
 	u64 outstanding_isize;
 
+	/*
+	 * If we get truncated we need to adjust the file extent we enter for
+	 * this ordered extent so that we do not expose stale data.
+	 */
+	u64 truncated_len;
+
 	/* flags (described above) */
 	unsigned long flags;
 

fs/btrfs/print-tree.c

@@ -26,14 +26,12 @@ static void print_chunk(struct extent_buffer *eb, struct btrfs_chunk *chunk)
 	int i;
 	printk(KERN_INFO "\t\tchunk length %llu owner %llu type %llu "
 	       "num_stripes %d\n",
-	       (unsigned long long)btrfs_chunk_length(eb, chunk),
-	       (unsigned long long)btrfs_chunk_owner(eb, chunk),
-	       (unsigned long long)btrfs_chunk_type(eb, chunk),
-	       num_stripes);
+	       btrfs_chunk_length(eb, chunk), btrfs_chunk_owner(eb, chunk),
+	       btrfs_chunk_type(eb, chunk), num_stripes);
 	for (i = 0 ; i < num_stripes ; i++) {
 		printk(KERN_INFO "\t\t\tstripe %d devid %llu offset %llu\n", i,
-		      (unsigned long long)btrfs_stripe_devid_nr(eb, chunk, i),
-		      (unsigned long long)btrfs_stripe_offset_nr(eb, chunk, i));
+		      btrfs_stripe_devid_nr(eb, chunk, i),
+		      btrfs_stripe_offset_nr(eb, chunk, i));
 	}
 }
 static void print_dev_item(struct extent_buffer *eb,
@@ -41,18 +39,18 @@ static void print_dev_item(struct extent_buffer *eb,
 {
 	printk(KERN_INFO "\t\tdev item devid %llu "
 	       "total_bytes %llu bytes used %llu\n",
-	       (unsigned long long)btrfs_device_id(eb, dev_item),
-	       (unsigned long long)btrfs_device_total_bytes(eb, dev_item),
-	       (unsigned long long)btrfs_device_bytes_used(eb, dev_item));
+	       btrfs_device_id(eb, dev_item),
+	       btrfs_device_total_bytes(eb, dev_item),
+	       btrfs_device_bytes_used(eb, dev_item));
 }
 static void print_extent_data_ref(struct extent_buffer *eb,
 				  struct btrfs_extent_data_ref *ref)
 {
 	printk(KERN_INFO "\t\textent data backref root %llu "
 	       "objectid %llu offset %llu count %u\n",
-	       (unsigned long long)btrfs_extent_data_ref_root(eb, ref),
-	       (unsigned long long)btrfs_extent_data_ref_objectid(eb, ref),
-	       (unsigned long long)btrfs_extent_data_ref_offset(eb, ref),
+	       btrfs_extent_data_ref_root(eb, ref),
+	       btrfs_extent_data_ref_objectid(eb, ref),
+	       btrfs_extent_data_ref_offset(eb, ref),
 	       btrfs_extent_data_ref_count(eb, ref));
 }
 
@@ -87,19 +85,17 @@ static void print_extent_item(struct extent_buffer *eb, int slot)
 	flags = btrfs_extent_flags(eb, ei);
 
 	printk(KERN_INFO "\t\textent refs %llu gen %llu flags %llu\n",
-	       (unsigned long long)btrfs_extent_refs(eb, ei),
-	       (unsigned long long)btrfs_extent_generation(eb, ei),
-	       (unsigned long long)flags);
+	       btrfs_extent_refs(eb, ei), btrfs_extent_generation(eb, ei),
+	       flags);
 
 	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
 		struct btrfs_tree_block_info *info;
 		info = (struct btrfs_tree_block_info *)(ei + 1);
 		btrfs_tree_block_key(eb, info, &key);
-		printk(KERN_INFO "\t\ttree block key (%llu %x %llu) "
+		printk(KERN_INFO "\t\ttree block key (%llu %u %llu) "
 		       "level %d\n",
-		       (unsigned long long)btrfs_disk_key_objectid(&key),
-		       key.type,
-		       (unsigned long long)btrfs_disk_key_offset(&key),
+		       btrfs_disk_key_objectid(&key), key.type,
+		       btrfs_disk_key_offset(&key),
 		       btrfs_tree_block_level(eb, info));
 		iref = (struct btrfs_extent_inline_ref *)(info + 1);
 	} else {
@@ -115,11 +111,11 @@ static void print_extent_item(struct extent_buffer *eb, int slot)
 		switch (type) {
 		case BTRFS_TREE_BLOCK_REF_KEY:
 			printk(KERN_INFO "\t\ttree block backref "
-				"root %llu\n", (unsigned long long)offset);
+				"root %llu\n", offset);
 			break;
 		case BTRFS_SHARED_BLOCK_REF_KEY:
 			printk(KERN_INFO "\t\tshared block backref "
-				"parent %llu\n", (unsigned long long)offset);
+				"parent %llu\n", offset);
 			break;
 		case BTRFS_EXTENT_DATA_REF_KEY:
 			dref = (struct btrfs_extent_data_ref *)(&iref->offset);
@@ -129,8 +125,7 @@ static void print_extent_item(struct extent_buffer *eb, int slot)
 			sref = (struct btrfs_shared_data_ref *)(iref + 1);
 			printk(KERN_INFO "\t\tshared data backref "
 			       "parent %llu count %u\n",
-			       (unsigned long long)offset,
-			       btrfs_shared_data_ref_count(eb, sref));
+			       offset, btrfs_shared_data_ref_count(eb, sref));
 			break;
 		default:
 			BUG();
@@ -148,13 +143,32 @@ static void print_extent_ref_v0(struct extent_buffer *eb, int slot)
 	ref0 = btrfs_item_ptr(eb, slot, struct btrfs_extent_ref_v0);
 	printk("\t\textent back ref root %llu gen %llu "
 		"owner %llu num_refs %lu\n",
-		(unsigned long long)btrfs_ref_root_v0(eb, ref0),
-		(unsigned long long)btrfs_ref_generation_v0(eb, ref0),
-		(unsigned long long)btrfs_ref_objectid_v0(eb, ref0),
+		btrfs_ref_root_v0(eb, ref0),
+		btrfs_ref_generation_v0(eb, ref0),
+		btrfs_ref_objectid_v0(eb, ref0),
 		(unsigned long)btrfs_ref_count_v0(eb, ref0));
 }
 #endif
 
+static void print_uuid_item(struct extent_buffer *l, unsigned long offset,
+			    u32 item_size)
+{
+	if (!IS_ALIGNED(item_size, sizeof(u64))) {
+		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			(unsigned long)item_size);
+		return;
+	}
+	while (item_size) {
+		__le64 subvol_id;
+
+		read_extent_buffer(l, &subvol_id, offset, sizeof(subvol_id));
+		printk(KERN_INFO "\t\tsubvol_id %llu\n",
+		       (unsigned long long)le64_to_cpu(subvol_id));
+		item_size -= sizeof(u64);
+		offset += sizeof(u64);
+	}
+}
+
 void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 {
 	int i;
@@ -177,39 +191,34 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 	nr = btrfs_header_nritems(l);
 
 	btrfs_info(root->fs_info, "leaf %llu total ptrs %d free space %d",
-		(unsigned long long)btrfs_header_bytenr(l), nr,
-		btrfs_leaf_free_space(root, l));
+		   btrfs_header_bytenr(l), nr, btrfs_leaf_free_space(root, l));
 	for (i = 0 ; i < nr ; i++) {
 		item = btrfs_item_nr(l, i);
 		btrfs_item_key_to_cpu(l, &key, i);
 		type = btrfs_key_type(&key);
-		printk(KERN_INFO "\titem %d key (%llu %x %llu) itemoff %d "
+		printk(KERN_INFO "\titem %d key (%llu %u %llu) itemoff %d "
 		       "itemsize %d\n",
-			i,
-			(unsigned long long)key.objectid, type,
-			(unsigned long long)key.offset,
+			i, key.objectid, type, key.offset,
 			btrfs_item_offset(l, item), btrfs_item_size(l, item));
 		switch (type) {
 		case BTRFS_INODE_ITEM_KEY:
 			ii = btrfs_item_ptr(l, i, struct btrfs_inode_item);
 			printk(KERN_INFO "\t\tinode generation %llu size %llu "
 			       "mode %o\n",
-			       (unsigned long long)
 			       btrfs_inode_generation(l, ii),
-			      (unsigned long long)btrfs_inode_size(l, ii),
+			       btrfs_inode_size(l, ii),
 			       btrfs_inode_mode(l, ii));
 			break;
 		case BTRFS_DIR_ITEM_KEY:
 			di = btrfs_item_ptr(l, i, struct btrfs_dir_item);
 			btrfs_dir_item_key_to_cpu(l, di, &found_key);
 			printk(KERN_INFO "\t\tdir oid %llu type %u\n",
-				(unsigned long long)found_key.objectid,
+				found_key.objectid,
 				btrfs_dir_type(l, di));
 			break;
 		case BTRFS_ROOT_ITEM_KEY:
 			ri = btrfs_item_ptr(l, i, struct btrfs_root_item);
 			printk(KERN_INFO "\t\troot data bytenr %llu refs %u\n",
-				(unsigned long long)
 				btrfs_disk_root_bytenr(l, ri),
 				btrfs_disk_root_refs(l, ri));
 			break;
@@ -245,17 +254,12 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			}
 			printk(KERN_INFO "\t\textent data disk bytenr %llu "
 			       "nr %llu\n",
-			       (unsigned long long)
 			       btrfs_file_extent_disk_bytenr(l, fi),
-			       (unsigned long long)
 			       btrfs_file_extent_disk_num_bytes(l, fi));
 			printk(KERN_INFO "\t\textent data offset %llu "
 			       "nr %llu ram %llu\n",
-			       (unsigned long long)
 			       btrfs_file_extent_offset(l, fi),
-			       (unsigned long long)
 			       btrfs_file_extent_num_bytes(l, fi),
-			       (unsigned long long)
 			       btrfs_file_extent_ram_bytes(l, fi));
 			break;
 		case BTRFS_EXTENT_REF_V0_KEY:
@@ -269,7 +273,6 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			bi = btrfs_item_ptr(l, i,
 					    struct btrfs_block_group_item);
 			printk(KERN_INFO "\t\tblock group used %llu\n",
-			       (unsigned long long)
 			       btrfs_disk_block_group_used(l, bi));
 			break;
 		case BTRFS_CHUNK_ITEM_KEY:
@@ -286,13 +289,9 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			printk(KERN_INFO "\t\tdev extent chunk_tree %llu\n"
 			       "\t\tchunk objectid %llu chunk offset %llu "
 			       "length %llu\n",
-			       (unsigned long long)
 			       btrfs_dev_extent_chunk_tree(l, dev_extent),
-			       (unsigned long long)
 			       btrfs_dev_extent_chunk_objectid(l, dev_extent),
-			       (unsigned long long)
 			       btrfs_dev_extent_chunk_offset(l, dev_extent),
-			       (unsigned long long)
 			       btrfs_dev_extent_length(l, dev_extent));
 			break;
 		case BTRFS_DEV_STATS_KEY:
@@ -301,6 +300,11 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 		case BTRFS_DEV_REPLACE_KEY:
 			printk(KERN_INFO "\t\tdev replace\n");
 			break;
+		case BTRFS_UUID_KEY_SUBVOL:
+		case BTRFS_UUID_KEY_RECEIVED_SUBVOL:
+			print_uuid_item(l, btrfs_item_ptr_offset(l, i),
+					btrfs_item_size_nr(l, i));
+			break;
 		};
 	}
 }
@@ -320,16 +324,13 @@ void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c)
 		return;
 	}
 	btrfs_info(root->fs_info, "node %llu level %d total ptrs %d free spc %u",
-		(unsigned long long)btrfs_header_bytenr(c),
-		level, nr, (u32)BTRFS_NODEPTRS_PER_BLOCK(root) - nr);
+		btrfs_header_bytenr(c), level, nr,
+		(u32)BTRFS_NODEPTRS_PER_BLOCK(root) - nr);
 	for (i = 0; i < nr; i++) {
 		btrfs_node_key_to_cpu(c, &key, i);
 		printk(KERN_INFO "\tkey %d (%llu %u %llu) block %llu\n",
-		       i,
-		       (unsigned long long)key.objectid,
-		       key.type,
-		       (unsigned long long)key.offset,
-		       (unsigned long long)btrfs_node_blockptr(c, i));
+		       i, key.objectid, key.type, key.offset,
+		       btrfs_node_blockptr(c, i));
 	}
 	for (i = 0; i < nr; i++) {
 		struct extent_buffer *next = read_tree_block(root,

fs/btrfs/qgroup.c

@@ -157,18 +157,11 @@ static struct btrfs_qgroup *add_qgroup_rb(struct btrfs_fs_info *fs_info,
 	return qgroup;
 }
 
-/* must be called with qgroup_lock held */
-static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid)
+static void __del_qgroup_rb(struct btrfs_qgroup *qgroup)
 {
-	struct btrfs_qgroup *qgroup = find_qgroup_rb(fs_info, qgroupid);
 	struct btrfs_qgroup_list *list;
 
-	if (!qgroup)
-		return -ENOENT;
-
-	rb_erase(&qgroup->node, &fs_info->qgroup_tree);
 	list_del(&qgroup->dirty);
-
 	while (!list_empty(&qgroup->groups)) {
 		list = list_first_entry(&qgroup->groups,
 					struct btrfs_qgroup_list, next_group);
@@ -185,7 +178,18 @@ static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid)
 		kfree(list);
 	}
 	kfree(qgroup);
+}
 
+/* must be called with qgroup_lock held */
+static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid)
+{
+	struct btrfs_qgroup *qgroup = find_qgroup_rb(fs_info, qgroupid);
+
+	if (!qgroup)
+		return -ENOENT;
+
+	rb_erase(&qgroup->node, &fs_info->qgroup_tree);
+	__del_qgroup_rb(qgroup);
 	return 0;
 }
 
@@ -394,8 +398,7 @@ next1:
 		if (ret == -ENOENT) {
 			printk(KERN_WARNING
 				"btrfs: orphan qgroup relation 0x%llx->0x%llx\n",
-				(unsigned long long)found_key.objectid,
-				(unsigned long long)found_key.offset);
+				found_key.objectid, found_key.offset);
 			ret = 0;	/* ignore the error */
 		}
 		if (ret)
@@ -428,39 +431,28 @@ out:
 }
 
 /*
- * This is only called from close_ctree() or open_ctree(), both in single-
- * treaded paths. Clean up the in-memory structures. No locking needed.
+ * This is called from close_ctree() or open_ctree() or btrfs_quota_disable(),
+ * first two are in single-threaded paths.And for the third one, we have set
+ * quota_root to be null with qgroup_lock held before, so it is safe to clean
+ * up the in-memory structures without qgroup_lock held.
  */
 void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info)
 {
 	struct rb_node *n;
 	struct btrfs_qgroup *qgroup;
-	struct btrfs_qgroup_list *list;
 
 	while ((n = rb_first(&fs_info->qgroup_tree))) {
 		qgroup = rb_entry(n, struct btrfs_qgroup, node);
 		rb_erase(n, &fs_info->qgroup_tree);
-
-		while (!list_empty(&qgroup->groups)) {
-			list = list_first_entry(&qgroup->groups,
-						struct btrfs_qgroup_list,
-						next_group);
-			list_del(&list->next_group);
-			list_del(&list->next_member);
-			kfree(list);
-		}
-
-		while (!list_empty(&qgroup->members)) {
-			list = list_first_entry(&qgroup->members,
-						struct btrfs_qgroup_list,
-						next_member);
-			list_del(&list->next_group);
-			list_del(&list->next_member);
-			kfree(list);
-		}
-		kfree(qgroup);
+		__del_qgroup_rb(qgroup);
 	}
+	/*
+	 * we call btrfs_free_qgroup_config() when umounting
+	 * filesystem and disabling quota, so we set qgroup_ulit
+	 * to be null here to avoid double free.
+	 */
 	ulist_free(fs_info->qgroup_ulist);
+	fs_info->qgroup_ulist = NULL;
 }
 
 static int add_qgroup_relation_item(struct btrfs_trans_handle *trans,
@@ -946,13 +938,9 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 	fs_info->pending_quota_state = 0;
 	quota_root = fs_info->quota_root;
 	fs_info->quota_root = NULL;
-	btrfs_free_qgroup_config(fs_info);
 	spin_unlock(&fs_info->qgroup_lock);
 
-	if (!quota_root) {
-		ret = -EINVAL;
-		goto out;
-	}
+	btrfs_free_qgroup_config(fs_info);
 
 	ret = btrfs_clean_quota_tree(trans, quota_root);
 	if (ret)
@@ -1174,7 +1162,7 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans,
 	if (ret) {
 		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
 		printk(KERN_INFO "unable to update quota limit for %llu\n",
-		       (unsigned long long)qgroupid);
+		       qgroupid);
 	}
 
 	spin_lock(&fs_info->qgroup_lock);
@@ -1884,10 +1872,9 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 					 path, 1, 0);
 
 	pr_debug("current progress key (%llu %u %llu), search_slot ret %d\n",
-		 (unsigned long long)fs_info->qgroup_rescan_progress.objectid,
+		 fs_info->qgroup_rescan_progress.objectid,
 		 fs_info->qgroup_rescan_progress.type,
-		 (unsigned long long)fs_info->qgroup_rescan_progress.offset,
-		 ret);
+		 fs_info->qgroup_rescan_progress.offset, ret);
 
 	if (ret) {
 		/*

fs/btrfs/raid56.c

@@ -1540,8 +1540,10 @@ static int full_stripe_write(struct btrfs_raid_bio *rbio)
 	int ret;
 
 	ret = alloc_rbio_parity_pages(rbio);
-	if (ret)
+	if (ret) {
+		__free_raid_bio(rbio);
 		return ret;
+	}
 
 	ret = lock_stripe_add(rbio);
 	if (ret == 0)
@@ -1687,11 +1689,8 @@ int raid56_parity_write(struct btrfs_root *root, struct bio *bio,
 	struct blk_plug_cb *cb;
 
 	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
-	if (IS_ERR(rbio)) {
-		kfree(raid_map);
-		kfree(bbio);
+	if (IS_ERR(rbio))
 		return PTR_ERR(rbio);
-	}
 	bio_list_add(&rbio->bio_list, bio);
 	rbio->bio_list_bytes = bio->bi_size;
 
@@ -2041,9 +2040,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio,
 	int ret;
 
 	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
-	if (IS_ERR(rbio)) {
+	if (IS_ERR(rbio))
 		return PTR_ERR(rbio);
-	}
 
 	rbio->read_rebuild = 1;
 	bio_list_add(&rbio->bio_list, bio);
@@ -2052,6 +2050,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio,
 	rbio->faila = find_logical_bio_stripe(rbio, bio);
 	if (rbio->faila == -1) {
 		BUG();
+		kfree(raid_map);
+		kfree(bbio);
 		kfree(rbio);
 		return -EIO;
 	}

fs/btrfs/relocation.c

@@ -335,7 +335,7 @@ static void backref_tree_panic(struct rb_node *rb_node, int errno, u64 bytenr)
 	if (bnode->root)
 		fs_info = bnode->root->fs_info;
 	btrfs_panic(fs_info, errno, "Inconsistency in backref cache "
-		    "found at offset %llu\n", (unsigned long long)bytenr);
+		    "found at offset %llu\n", bytenr);
 }
 
 /*
@@ -641,6 +641,11 @@ int find_inline_backref(struct extent_buffer *leaf, int slot,
 		WARN_ON(item_size < sizeof(*ei) + sizeof(*bi));
 		return 1;
 	}
+	if (key.type == BTRFS_METADATA_ITEM_KEY &&
+	    item_size <= sizeof(*ei)) {
+		WARN_ON(item_size < sizeof(*ei));
+		return 1;
+	}
 
 	if (key.type == BTRFS_EXTENT_ITEM_KEY) {
 		bi = (struct btrfs_tree_block_info *)(ei + 1);
@@ -691,6 +696,7 @@ struct backref_node *build_backref_tree(struct reloc_control *rc,
 	int cowonly;
 	int ret;
 	int err = 0;
+	bool need_check = true;
 
 	path1 = btrfs_alloc_path();
 	path2 = btrfs_alloc_path();
@@ -914,6 +920,7 @@ again:
 			cur->bytenr);
 
 		lower = cur;
+		need_check = true;
 		for (; level < BTRFS_MAX_LEVEL; level++) {
 			if (!path2->nodes[level]) {
 				BUG_ON(btrfs_root_bytenr(&root->root_item) !=
@@ -957,14 +964,12 @@ again:
 
 				/*
 				 * add the block to pending list if we
-				 * need check its backrefs. only block
-				 * at 'cur->level + 1' is added to the
-				 * tail of pending list. this guarantees
-				 * we check backrefs from lower level
-				 * blocks to upper level blocks.
+				 * need check its backrefs, we only do this once
+				 * while walking up a tree as we will catch
+				 * anything else later on.
 				 */
-				if (!upper->checked &&
-				    level == cur->level + 1) {
+				if (!upper->checked && need_check) {
+					need_check = false;
 					list_add_tail(&edge->list[UPPER],
 						      &list);
 				} else
@@ -2314,8 +2319,13 @@ again:
 			BUG_ON(root->reloc_root != reloc_root);
 
 			ret = merge_reloc_root(rc, root);
-			if (ret)
+			if (ret) {
+				__update_reloc_root(reloc_root, 1);
+				free_extent_buffer(reloc_root->node);
+				free_extent_buffer(reloc_root->commit_root);
+				kfree(reloc_root);
 				goto out;
+			}
 		} else {
 			list_del_init(&reloc_root->root_list);
 		}
@@ -2344,9 +2354,6 @@ again:
 			if (IS_ERR(root))
 				continue;
 
-			if (btrfs_root_refs(&root->root_item) == 0)
-				continue;
-
 			trans = btrfs_join_transaction(root);
 			BUG_ON(IS_ERR(trans));
 
@@ -3628,7 +3635,7 @@ int add_data_references(struct reloc_control *rc,
 	unsigned long ptr;
 	unsigned long end;
 	u32 blocksize = btrfs_level_size(rc->extent_root, 0);
-	int ret;
+	int ret = 0;
 	int err = 0;
 
 	eb = path->nodes[0];
@@ -3655,6 +3662,10 @@ int add_data_references(struct reloc_control *rc,
 		} else {
 			BUG();
 		}
+		if (ret) {
+			err = ret;
+			goto out;
+		}
 		ptr += btrfs_extent_inline_ref_size(key.type);
 	}
 	WARN_ON(ptr > end);
@@ -3700,6 +3711,7 @@ int add_data_references(struct reloc_control *rc,
 		}
 		path->slots[0]++;
 	}
+out:
 	btrfs_release_path(path);
 	if (err)
 		free_block_list(blocks);
@@ -4219,8 +4231,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 	}
 
 	printk(KERN_INFO "btrfs: relocating block group %llu flags %llu\n",
-	       (unsigned long long)rc->block_group->key.objectid,
-	       (unsigned long long)rc->block_group->flags);
+	       rc->block_group->key.objectid, rc->block_group->flags);
 
 	ret = btrfs_start_all_delalloc_inodes(fs_info, 0);
 	if (ret < 0) {
@@ -4242,7 +4253,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 			break;
 
 		printk(KERN_INFO "btrfs: found %llu extents\n",
-			(unsigned long long)rc->extents_found);
+			rc->extents_found);
 
 		if (rc->stage == MOVE_DATA_EXTENTS && rc->found_file_extent) {
 			btrfs_wait_ordered_range(rc->data_inode, 0, (u64)-1);

fs/btrfs/root-tree.c

@@ -29,8 +29,8 @@
  * generation numbers as then we know the root was once mounted with an older
  * kernel that was not aware of the root item structure change.
  */
-void btrfs_read_root_item(struct extent_buffer *eb, int slot,
-			  struct btrfs_root_item *item)
+static void btrfs_read_root_item(struct extent_buffer *eb, int slot,
+				struct btrfs_root_item *item)
 {
 	uuid_le uuid;
 	int len;
@@ -155,8 +155,7 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root
 	if (ret != 0) {
 		btrfs_print_leaf(root, path->nodes[0]);
 		printk(KERN_CRIT "unable to update root key %llu %u %llu\n",
-		       (unsigned long long)key->objectid, key->type,
-		       (unsigned long long)key->offset);
+		       key->objectid, key->type, key->offset);
 		BUG_ON(1);
 	}
 
@@ -490,13 +489,13 @@ again:
  */
 void btrfs_check_and_init_root_item(struct btrfs_root_item *root_item)
 {
-	u64 inode_flags = le64_to_cpu(root_item->inode.flags);
+	u64 inode_flags = btrfs_stack_inode_flags(&root_item->inode);
 
 	if (!(inode_flags & BTRFS_INODE_ROOT_ITEM_INIT)) {
 		inode_flags |= BTRFS_INODE_ROOT_ITEM_INIT;
-		root_item->inode.flags = cpu_to_le64(inode_flags);
-		root_item->flags = 0;
-		root_item->byte_limit = 0;
+		btrfs_set_stack_inode_flags(&root_item->inode, inode_flags);
+		btrfs_set_root_flags(root_item, 0);
+		btrfs_set_root_limit(root_item, 0);
 	}
 }
 
@@ -507,8 +506,8 @@ void btrfs_update_root_times(struct btrfs_trans_handle *trans,
 	struct timespec ct = CURRENT_TIME;
 
 	spin_lock(&root->root_item_lock);
-	item->ctransid = cpu_to_le64(trans->transid);
-	item->ctime.sec = cpu_to_le64(ct.tv_sec);
-	item->ctime.nsec = cpu_to_le32(ct.tv_nsec);
+	btrfs_set_root_ctransid(item, trans->transid);
+	btrfs_set_stack_timespec_sec(&item->ctime, ct.tv_sec);
+	btrfs_set_stack_timespec_nsec(&item->ctime, ct.tv_nsec);
 	spin_unlock(&root->root_item_lock);
 }

fs/btrfs/scrub.c

@@ -754,8 +754,7 @@ out:
 			num_uncorrectable_read_errors);
 		printk_ratelimited_in_rcu(KERN_ERR
 			"btrfs: unable to fixup (nodatasum) error at logical %llu on dev %s\n",
-			(unsigned long long)fixup->logical,
-			rcu_str_deref(fixup->dev->name));
+			fixup->logical, rcu_str_deref(fixup->dev->name));
 	}
 
 	btrfs_free_path(path);
@@ -1154,8 +1153,7 @@ corrected_error:
 			spin_unlock(&sctx->stat_lock);
 			printk_ratelimited_in_rcu(KERN_ERR
 				"btrfs: fixed up error at logical %llu on dev %s\n",
-				(unsigned long long)logical,
-				rcu_str_deref(dev->name));
+				logical, rcu_str_deref(dev->name));
 		}
 	} else {
 did_not_correct_error:
@@ -1164,8 +1162,7 @@ did_not_correct_error:
 		spin_unlock(&sctx->stat_lock);
 		printk_ratelimited_in_rcu(KERN_ERR
 			"btrfs: unable to fixup (regular) error at logical %llu on dev %s\n",
-			(unsigned long long)logical,
-			rcu_str_deref(dev->name));
+			logical, rcu_str_deref(dev->name));
 	}
 
 out:
@@ -1345,12 +1342,12 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
 		mapped_buffer = kmap_atomic(sblock->pagev[0]->page);
 		h = (struct btrfs_header *)mapped_buffer;
 
-		if (sblock->pagev[0]->logical != le64_to_cpu(h->bytenr) ||
+		if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h) ||
 		    memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE) ||
 		    memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
 			   BTRFS_UUID_SIZE)) {
 			sblock->header_error = 1;
-		} else if (generation != le64_to_cpu(h->generation)) {
+		} else if (generation != btrfs_stack_header_generation(h)) {
 			sblock->header_error = 1;
 			sblock->generation_error = 1;
 		}
@@ -1720,10 +1717,10 @@ static int scrub_checksum_tree_block(struct scrub_block *sblock)
 	 * b) the page is already kmapped
 	 */
 
-	if (sblock->pagev[0]->logical != le64_to_cpu(h->bytenr))
+	if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h))
 		++fail;
 
-	if (sblock->pagev[0]->generation != le64_to_cpu(h->generation))
+	if (sblock->pagev[0]->generation != btrfs_stack_header_generation(h))
 		++fail;
 
 	if (memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
@@ -1786,10 +1783,10 @@ static int scrub_checksum_super(struct scrub_block *sblock)
 	s = (struct btrfs_super_block *)mapped_buffer;
 	memcpy(on_disk_csum, s->csum, sctx->csum_size);
 
-	if (sblock->pagev[0]->logical != le64_to_cpu(s->bytenr))
+	if (sblock->pagev[0]->logical != btrfs_super_bytenr(s))
 		++fail_cor;
 
-	if (sblock->pagev[0]->generation != le64_to_cpu(s->generation))
+	if (sblock->pagev[0]->generation != btrfs_super_generation(s))
 		++fail_gen;
 
 	if (memcmp(s->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
@@ -2455,8 +2452,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 				printk(KERN_ERR
 				       "btrfs scrub: tree block %llu spanning "
 				       "stripes, ignored. logical=%llu\n",
-				       (unsigned long long)key.objectid,
-				       (unsigned long long)logical);
+				       key.objectid, logical);
 				goto next;
 			}
 
@@ -2863,9 +2859,8 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	if (fs_info->chunk_root->sectorsize != PAGE_SIZE) {
 		/* not supported for data w/o checksums */
 		printk(KERN_ERR
-		       "btrfs_scrub: size assumption sectorsize != PAGE_SIZE (%d != %lld) fails\n",
-		       fs_info->chunk_root->sectorsize,
-		       (unsigned long long)PAGE_SIZE);
+		       "btrfs_scrub: size assumption sectorsize != PAGE_SIZE (%d != %lu) fails\n",
+		       fs_info->chunk_root->sectorsize, PAGE_SIZE);
 		return -EINVAL;
 	}
 
@@ -3175,11 +3170,9 @@ static void copy_nocow_pages_worker(struct btrfs_work *work)
 					  copy_nocow_pages_for_inode,
 					  nocow_ctx);
 	if (ret != 0 && ret != -ENOENT) {
-		pr_warn("iterate_inodes_from_logical() failed: log %llu, phys %llu, len %llu, mir %llu, ret %d\n",
-			(unsigned long long)logical,
-			(unsigned long long)physical_for_dev_replace,
-			(unsigned long long)len,
-			(unsigned long long)mirror_num, ret);
+		pr_warn("iterate_inodes_from_logical() failed: log %llu, phys %llu, len %llu, mir %u, ret %d\n",
+			logical, physical_for_dev_replace, len, mirror_num,
+			ret);
 		not_written = 1;
 		goto out;
 	}
@@ -3224,11 +3217,6 @@ static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root, void *ctx)
 		return PTR_ERR(local_root);
 	}
 
-	if (btrfs_root_refs(&local_root->root_item) == 0) {
-		srcu_read_unlock(&fs_info->subvol_srcu, srcu_index);
-		return -ENOENT;
-	}
-
 	key.type = BTRFS_INODE_ITEM_KEY;
 	key.objectid = inum;
 	key.offset = 0;

fs/btrfs/send.c

@@ -26,6 +26,7 @@
 #include <linux/radix-tree.h>
 #include <linux/crc32c.h>
 #include <linux/vmalloc.h>
+#include <linux/string.h>
 
 #include "send.h"
 #include "backref.h"
@@ -54,8 +55,8 @@ struct fs_path {
 
 			char *buf;
 			int buf_len;
-			int reversed:1;
-			int virtual_mem:1;
+			unsigned int reversed:1;
+			unsigned int virtual_mem:1;
 			char inline_buf[];
 		};
 		char pad[PAGE_SIZE];
@@ -1668,6 +1669,7 @@ static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen,
 			      u64 *who_ino, u64 *who_gen)
 {
 	int ret = 0;
+	u64 gen;
 	u64 other_inode = 0;
 	u8 other_type = 0;
 
@@ -1678,6 +1680,24 @@ static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen,
 	if (ret <= 0)
 		goto out;
 
+	/*
+	 * If we have a parent root we need to verify that the parent dir was
+	 * not delted and then re-created, if it was then we have no overwrite
+	 * and we can just unlink this entry.
+	 */
+	if (sctx->parent_root) {
+		ret = get_inode_info(sctx->parent_root, dir, NULL, &gen, NULL,
+				     NULL, NULL, NULL);
+		if (ret < 0 && ret != -ENOENT)
+			goto out;
+		if (ret) {
+			ret = 0;
+			goto out;
+		}
+		if (gen != dir_gen)
+			goto out;
+	}
+
 	ret = lookup_dir_item_inode(sctx->parent_root, dir, name, name_len,
 			&other_inode, &other_type);
 	if (ret < 0 && ret != -ENOENT)
@@ -2519,7 +2539,8 @@ static int did_create_dir(struct send_ctx *sctx, u64 dir)
 		di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
 		btrfs_dir_item_key_to_cpu(eb, di, &di_key);
 
-		if (di_key.objectid < sctx->send_progress) {
+		if (di_key.type != BTRFS_ROOT_ITEM_KEY &&
+		    di_key.objectid < sctx->send_progress) {
 			ret = 1;
 			goto out;
 		}
@@ -2581,7 +2602,6 @@ static int record_ref(struct list_head *head, u64 dir,
 		      u64 dir_gen, struct fs_path *path)
 {
 	struct recorded_ref *ref;
-	char *tmp;
 
 	ref = kmalloc(sizeof(*ref), GFP_NOFS);
 	if (!ref)
@@ -2591,25 +2611,35 @@ static int record_ref(struct list_head *head, u64 dir,
 	ref->dir_gen = dir_gen;
 	ref->full_path = path;
 
-	tmp = strrchr(ref->full_path->start, '/');
-	if (!tmp) {
-		ref->name_len = ref->full_path->end - ref->full_path->start;
-		ref->name = ref->full_path->start;
+	ref->name = (char *)kbasename(ref->full_path->start);
+	ref->name_len = ref->full_path->end - ref->name;
+	ref->dir_path = ref->full_path->start;
+	if (ref->name == ref->full_path->start)
 		ref->dir_path_len = 0;
-		ref->dir_path = ref->full_path->start;
-	} else {
-		tmp++;
-		ref->name_len = ref->full_path->end - tmp;
-		ref->name = tmp;
-		ref->dir_path = ref->full_path->start;
+	else
 		ref->dir_path_len = ref->full_path->end -
 				ref->full_path->start - 1 - ref->name_len;
-	}
 
 	list_add_tail(&ref->list, head);
 	return 0;
 }
 
+static int dup_ref(struct recorded_ref *ref, struct list_head *list)
+{
+	struct recorded_ref *new;
+
+	new = kmalloc(sizeof(*ref), GFP_NOFS);
+	if (!new)
+		return -ENOMEM;
+
+	new->dir = ref->dir;
+	new->dir_gen = ref->dir_gen;
+	new->full_path = NULL;
+	INIT_LIST_HEAD(&new->list);
+	list_add_tail(&new->list, list);
+	return 0;
+}
+
 static void __free_recorded_refs(struct list_head *head)
 {
 	struct recorded_ref *cur;
@@ -2724,9 +2754,7 @@ static int process_recorded_refs(struct send_ctx *sctx)
 	int ret = 0;
 	struct recorded_ref *cur;
 	struct recorded_ref *cur2;
-	struct ulist *check_dirs = NULL;
-	struct ulist_iterator uit;
-	struct ulist_node *un;
+	struct list_head check_dirs;
 	struct fs_path *valid_path = NULL;
 	u64 ow_inode = 0;
 	u64 ow_gen;
@@ -2740,6 +2768,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 	 * which is always '..'
 	 */
 	BUG_ON(sctx->cur_ino <= BTRFS_FIRST_FREE_OBJECTID);
+	INIT_LIST_HEAD(&check_dirs);
 
 	valid_path = fs_path_alloc();
 	if (!valid_path) {
@@ -2747,12 +2776,6 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 		goto out;
 	}
 
-	check_dirs = ulist_alloc(GFP_NOFS);
-	if (!check_dirs) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
 	/*
 	 * First, check if the first ref of the current inode was overwritten
 	 * before. If yes, we know that the current inode was already orphanized
@@ -2889,8 +2912,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 					goto out;
 			}
 		}
-		ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-				GFP_NOFS);
+		ret = dup_ref(cur, &check_dirs);
 		if (ret < 0)
 			goto out;
 	}
@@ -2918,8 +2940,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 		}
 
 		list_for_each_entry(cur, &sctx->deleted_refs, list) {
-			ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-					GFP_NOFS);
+			ret = dup_ref(cur, &check_dirs);
 			if (ret < 0)
 				goto out;
 		}
@@ -2930,8 +2951,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 		 */
 		cur = list_entry(sctx->deleted_refs.next, struct recorded_ref,
 				list);
-		ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-				GFP_NOFS);
+		ret = dup_ref(cur, &check_dirs);
 		if (ret < 0)
 			goto out;
 	} else if (!S_ISDIR(sctx->cur_inode_mode)) {
@@ -2951,12 +2971,10 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 				if (ret < 0)
 					goto out;
 			}
-			ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-					GFP_NOFS);
+			ret = dup_ref(cur, &check_dirs);
 			if (ret < 0)
 				goto out;
 		}
-
 		/*
 		 * If the inode is still orphan, unlink the orphan. This may
 		 * happen when a previous inode did overwrite the first ref
@@ -2978,33 +2996,32 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 	 * deletion and if it's finally possible to perform the rmdir now.
 	 * We also update the inode stats of the parent dirs here.
 	 */
-	ULIST_ITER_INIT(&uit);
-	while ((un = ulist_next(check_dirs, &uit))) {
+	list_for_each_entry(cur, &check_dirs, list) {
 		/*
 		 * In case we had refs into dirs that were not processed yet,
 		 * we don't need to do the utime and rmdir logic for these dirs.
 		 * The dir will be processed later.
 		 */
-		if (un->val > sctx->cur_ino)
+		if (cur->dir > sctx->cur_ino)
 			continue;
 
-		ret = get_cur_inode_state(sctx, un->val, un->aux);
+		ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
 		if (ret < 0)
 			goto out;
 
 		if (ret == inode_state_did_create ||
 		    ret == inode_state_no_change) {
 			/* TODO delayed utimes */
-			ret = send_utimes(sctx, un->val, un->aux);
+			ret = send_utimes(sctx, cur->dir, cur->dir_gen);
 			if (ret < 0)
 				goto out;
 		} else if (ret == inode_state_did_delete) {
-			ret = can_rmdir(sctx, un->val, sctx->cur_ino);
+			ret = can_rmdir(sctx, cur->dir, sctx->cur_ino);
 			if (ret < 0)
 				goto out;
 			if (ret) {
-				ret = get_cur_path(sctx, un->val, un->aux,
-						valid_path);
+				ret = get_cur_path(sctx, cur->dir,
+						   cur->dir_gen, valid_path);
 				if (ret < 0)
 					goto out;
 				ret = send_rmdir(sctx, valid_path);
@@ -3017,8 +3034,8 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 	ret = 0;
 
 out:
+	__free_recorded_refs(&check_dirs);
 	free_recorded_refs(sctx);
-	ulist_free(check_dirs);
 	fs_path_free(valid_path);
 	return ret;
 }
@@ -3119,6 +3136,8 @@ out:
 
 struct find_ref_ctx {
 	u64 dir;
+	u64 dir_gen;
+	struct btrfs_root *root;
 	struct fs_path *name;
 	int found_idx;
 };
@@ -3128,9 +3147,21 @@ static int __find_iref(int num, u64 dir, int index,
 		       void *ctx_)
 {
 	struct find_ref_ctx *ctx = ctx_;
+	u64 dir_gen;
+	int ret;
 
 	if (dir == ctx->dir && fs_path_len(name) == fs_path_len(ctx->name) &&
 	    strncmp(name->start, ctx->name->start, fs_path_len(name)) == 0) {
+		/*
+		 * To avoid doing extra lookups we'll only do this if everything
+		 * else matches.
+		 */
+		ret = get_inode_info(ctx->root, dir, NULL, &dir_gen, NULL,
+				     NULL, NULL, NULL);
+		if (ret)
+			return ret;
+		if (dir_gen != ctx->dir_gen)
+			return 0;
 		ctx->found_idx = num;
 		return 1;
 	}
@@ -3140,14 +3171,16 @@ static int __find_iref(int num, u64 dir, int index,
 static int find_iref(struct btrfs_root *root,
 		     struct btrfs_path *path,
 		     struct btrfs_key *key,
-		     u64 dir, struct fs_path *name)
+		     u64 dir, u64 dir_gen, struct fs_path *name)
 {
 	int ret;
 	struct find_ref_ctx ctx;
 
 	ctx.dir = dir;
 	ctx.name = name;
+	ctx.dir_gen = dir_gen;
 	ctx.found_idx = -1;
+	ctx.root = root;
 
 	ret = iterate_inode_ref(root, path, key, 0, __find_iref, &ctx);
 	if (ret < 0)
@@ -3163,11 +3196,17 @@ static int __record_changed_new_ref(int num, u64 dir, int index,
 				    struct fs_path *name,
 				    void *ctx)
 {
+	u64 dir_gen;
 	int ret;
 	struct send_ctx *sctx = ctx;
 
+	ret = get_inode_info(sctx->send_root, dir, NULL, &dir_gen, NULL,
+			     NULL, NULL, NULL);
+	if (ret)
+		return ret;
+
 	ret = find_iref(sctx->parent_root, sctx->right_path,
-			sctx->cmp_key, dir, name);
+			sctx->cmp_key, dir, dir_gen, name);
 	if (ret == -ENOENT)
 		ret = __record_new_ref(num, dir, index, name, sctx);
 	else if (ret > 0)
@@ -3180,11 +3219,17 @@ static int __record_changed_deleted_ref(int num, u64 dir, int index,
 					struct fs_path *name,
 					void *ctx)
 {
+	u64 dir_gen;
 	int ret;
 	struct send_ctx *sctx = ctx;
 
+	ret = get_inode_info(sctx->parent_root, dir, NULL, &dir_gen, NULL,
+			     NULL, NULL, NULL);
+	if (ret)
+		return ret;
+
 	ret = find_iref(sctx->send_root, sctx->left_path, sctx->cmp_key,
-			dir, name);
+			dir, dir_gen, name);
 	if (ret == -ENOENT)
 		ret = __record_deleted_ref(num, dir, index, name, sctx);
 	else if (ret > 0)
@@ -3869,7 +3914,8 @@ static int is_extent_unchanged(struct send_ctx *sctx,
 	btrfs_item_key_to_cpu(eb, &found_key, slot);
 	if (found_key.objectid != key.objectid ||
 	    found_key.type != key.type) {
-		ret = 0;
+		/* If we're a hole then just pretend nothing changed */
+		ret = (left_disknr) ? 0 : 1;
 		goto out;
 	}
 
@@ -3895,7 +3941,8 @@ static int is_extent_unchanged(struct send_ctx *sctx,
 		 * This may only happen on the first iteration.
 		 */
 		if (found_key.offset + right_len <= ekey->offset) {
-			ret = 0;
+			/* If we're a hole just pretend nothing changed */
+			ret = (left_disknr) ? 0 : 1;
 			goto out;
 		}
 
@@ -3960,8 +4007,8 @@ static int process_extent(struct send_ctx *sctx,
 			  struct btrfs_path *path,
 			  struct btrfs_key *key)
 {
-	int ret = 0;
 	struct clone_root *found_clone = NULL;
+	int ret = 0;
 
 	if (S_ISLNK(sctx->cur_inode_mode))
 		return 0;
@@ -3974,6 +4021,32 @@ static int process_extent(struct send_ctx *sctx,
 			ret = 0;
 			goto out;
 		}
+	} else {
+		struct btrfs_file_extent_item *ei;
+		u8 type;
+
+		ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				    struct btrfs_file_extent_item);
+		type = btrfs_file_extent_type(path->nodes[0], ei);
+		if (type == BTRFS_FILE_EXTENT_PREALLOC ||
+		    type == BTRFS_FILE_EXTENT_REG) {
+			/*
+			 * The send spec does not have a prealloc command yet,
+			 * so just leave a hole for prealloc'ed extents until
+			 * we have enough commands queued up to justify rev'ing
+			 * the send spec.
+			 */
+			if (type == BTRFS_FILE_EXTENT_PREALLOC) {
+				ret = 0;
+				goto out;
+			}
+
+			/* Have a hole, just skip it. */
+			if (btrfs_file_extent_disk_bytenr(path->nodes[0], ei) == 0) {
+				ret = 0;
+				goto out;
+			}
+		}
 	}
 
 	ret = find_extent_clone(sctx, path, key->objectid, key->offset,
@@ -4361,6 +4434,64 @@ static int changed_extent(struct send_ctx *sctx,
 	return ret;
 }
 
+static int dir_changed(struct send_ctx *sctx, u64 dir)
+{
+	u64 orig_gen, new_gen;
+	int ret;
+
+	ret = get_inode_info(sctx->send_root, dir, NULL, &new_gen, NULL, NULL,
+			     NULL, NULL);
+	if (ret)
+		return ret;
+
+	ret = get_inode_info(sctx->parent_root, dir, NULL, &orig_gen, NULL,
+			     NULL, NULL, NULL);
+	if (ret)
+		return ret;
+
+	return (orig_gen != new_gen) ? 1 : 0;
+}
+
+static int compare_refs(struct send_ctx *sctx, struct btrfs_path *path,
+			struct btrfs_key *key)
+{
+	struct btrfs_inode_extref *extref;
+	struct extent_buffer *leaf;
+	u64 dirid = 0, last_dirid = 0;
+	unsigned long ptr;
+	u32 item_size;
+	u32 cur_offset = 0;
+	int ref_name_len;
+	int ret = 0;
+
+	/* Easy case, just check this one dirid */
+	if (key->type == BTRFS_INODE_REF_KEY) {
+		dirid = key->offset;
+
+		ret = dir_changed(sctx, dirid);
+		goto out;
+	}
+
+	leaf = path->nodes[0];
+	item_size = btrfs_item_size_nr(leaf, path->slots[0]);
+	ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
+	while (cur_offset < item_size) {
+		extref = (struct btrfs_inode_extref *)(ptr +
+						       cur_offset);
+		dirid = btrfs_inode_extref_parent(leaf, extref);
+		ref_name_len = btrfs_inode_extref_name_len(leaf, extref);
+		cur_offset += ref_name_len + sizeof(*extref);
+		if (dirid == last_dirid)
+			continue;
+		ret = dir_changed(sctx, dirid);
+		if (ret)
+			break;
+		last_dirid = dirid;
+	}
+out:
+	return ret;
+}
+
 /*
  * Updates compare related fields in sctx and simply forwards to the actual
  * changed_xxx functions.
@@ -4376,6 +4507,19 @@ static int changed_cb(struct btrfs_root *left_root,
 	int ret = 0;
 	struct send_ctx *sctx = ctx;
 
+	if (result == BTRFS_COMPARE_TREE_SAME) {
+		if (key->type != BTRFS_INODE_REF_KEY &&
+		    key->type != BTRFS_INODE_EXTREF_KEY)
+			return 0;
+		ret = compare_refs(sctx, left_path, key);
+		if (!ret)
+			return 0;
+		if (ret < 0)
+			return ret;
+		result = BTRFS_COMPARE_TREE_CHANGED;
+		ret = 0;
+	}
+
 	sctx->left_path = left_path;
 	sctx->right_path = right_path;
 	sctx->cmp_key = key;

fs/btrfs/super.c

@@ -56,6 +56,8 @@
 #include "rcu-string.h"
 #include "dev-replace.h"
 #include "free-space-cache.h"
+#include "backref.h"
+#include "tests/btrfs-tests.h"
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/btrfs.h>
@@ -320,14 +322,15 @@ enum {
 	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_inode_cache,
 	Opt_no_space_cache, Opt_recovery, Opt_skip_balance,
 	Opt_check_integrity, Opt_check_integrity_including_extent_data,
-	Opt_check_integrity_print_mask, Opt_fatal_errors,
+	Opt_check_integrity_print_mask, Opt_fatal_errors, Opt_rescan_uuid_tree,
+	Opt_commit_interval,
 	Opt_err,
 };
 
 static match_table_t tokens = {
 	{Opt_degraded, "degraded"},
 	{Opt_subvol, "subvol=%s"},
-	{Opt_subvolid, "subvolid=%d"},
+	{Opt_subvolid, "subvolid=%s"},
 	{Opt_device, "device=%s"},
 	{Opt_nodatasum, "nodatasum"},
 	{Opt_nodatacow, "nodatacow"},
@@ -360,7 +363,9 @@ static match_table_t tokens = {
 	{Opt_check_integrity, "check_int"},
 	{Opt_check_integrity_including_extent_data, "check_int_data"},
 	{Opt_check_integrity_print_mask, "check_int_print_mask=%d"},
+	{Opt_rescan_uuid_tree, "rescan_uuid_tree"},
 	{Opt_fatal_errors, "fatal_errors=%s"},
+	{Opt_commit_interval, "commit=%d"},
 	{Opt_err, NULL},
 };
 
@@ -496,10 +501,15 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			btrfs_set_opt(info->mount_opt, NOBARRIER);
 			break;
 		case Opt_thread_pool:
-			intarg = 0;
-			match_int(&args[0], &intarg);
-			if (intarg)
+			ret = match_int(&args[0], &intarg);
+			if (ret) {
+				goto out;
+			} else if (intarg > 0) {
 				info->thread_pool_size = intarg;
+			} else {
+				ret = -EINVAL;
+				goto out;
+			}
 			break;
 		case Opt_max_inline:
 			num = match_strdup(&args[0]);
@@ -513,7 +523,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 						root->sectorsize);
 				}
 				printk(KERN_INFO "btrfs: max_inline at %llu\n",
-					(unsigned long long)info->max_inline);
+					info->max_inline);
+			} else {
+				ret = -ENOMEM;
+				goto out;
 			}
 			break;
 		case Opt_alloc_start:
@@ -525,7 +538,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				kfree(num);
 				printk(KERN_INFO
 					"btrfs: allocations start at %llu\n",
-					(unsigned long long)info->alloc_start);
+					info->alloc_start);
+			} else {
+				ret = -ENOMEM;
+				goto out;
 			}
 			break;
 		case Opt_noacl:
@@ -540,12 +556,16 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			btrfs_set_opt(info->mount_opt, FLUSHONCOMMIT);
 			break;
 		case Opt_ratio:
-			intarg = 0;
-			match_int(&args[0], &intarg);
-			if (intarg) {
+			ret = match_int(&args[0], &intarg);
+			if (ret) {
+				goto out;
+			} else if (intarg >= 0) {
 				info->metadata_ratio = intarg;
 				printk(KERN_INFO "btrfs: metadata ratio %d\n",
 				       info->metadata_ratio);
+			} else {
+				ret = -EINVAL;
+				goto out;
 			}
 			break;
 		case Opt_discard:
@@ -554,6 +574,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		case Opt_space_cache:
 			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
 			break;
+		case Opt_rescan_uuid_tree:
+			btrfs_set_opt(info->mount_opt, RESCAN_UUID_TREE);
+			break;
 		case Opt_no_space_cache:
 			printk(KERN_INFO "btrfs: disabling disk space caching\n");
 			btrfs_clear_opt(info->mount_opt, SPACE_CACHE);
@@ -596,13 +619,17 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY);
 			break;
 		case Opt_check_integrity_print_mask:
-			intarg = 0;
-			match_int(&args[0], &intarg);
-			if (intarg) {
+			ret = match_int(&args[0], &intarg);
+			if (ret) {
+				goto out;
+			} else if (intarg >= 0) {
 				info->check_integrity_print_mask = intarg;
 				printk(KERN_INFO "btrfs:"
 				       " check_integrity_print_mask 0x%x\n",
 				       info->check_integrity_print_mask);
+			} else {
+				ret = -EINVAL;
+				goto out;
 			}
 			break;
 #else
@@ -626,6 +653,29 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				goto out;
 			}
 			break;
+		case Opt_commit_interval:
+			intarg = 0;
+			ret = match_int(&args[0], &intarg);
+			if (ret < 0) {
+				printk(KERN_ERR
+					"btrfs: invalid commit interval\n");
+				ret = -EINVAL;
+				goto out;
+			}
+			if (intarg > 0) {
+				if (intarg > 300) {
+					printk(KERN_WARNING
+					    "btrfs: excessive commit interval %d\n",
+							intarg);
+				}
+				info->commit_interval = intarg;
+			} else {
+				printk(KERN_INFO
+				    "btrfs: using default commit interval %ds\n",
+				    BTRFS_DEFAULT_COMMIT_INTERVAL);
+				info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
+			}
+			break;
 		case Opt_err:
 			printk(KERN_INFO "btrfs: unrecognized mount option "
 			       "'%s'\n", p);
@@ -654,8 +704,8 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags,
 {
 	substring_t args[MAX_OPT_ARGS];
 	char *device_name, *opts, *orig, *p;
+	char *num = NULL;
 	int error = 0;
-	int intarg;
 
 	if (!options)
 		return 0;
@@ -679,17 +729,23 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags,
 		case Opt_subvol:
 			kfree(*subvol_name);
 			*subvol_name = match_strdup(&args[0]);
+			if (!*subvol_name) {
+				error = -ENOMEM;
+				goto out;
+			}
 			break;
 		case Opt_subvolid:
-			intarg = 0;
-			error = match_int(&args[0], &intarg);
-			if (!error) {
+			num = match_strdup(&args[0]);
+			if (num) {
+				*subvol_objectid = memparse(num, NULL);
+				kfree(num);
 				/* we want the original fs_tree */
-				if (!intarg)
+				if (!*subvol_objectid)
 					*subvol_objectid =
 						BTRFS_FS_TREE_OBJECTID;
-				else
-					*subvol_objectid = intarg;
+			} else {
+				error = -EINVAL;
+				goto out;
 			}
 			break;
 		case Opt_subvolrootid:
@@ -892,11 +948,9 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 	if (btrfs_test_opt(root, NOBARRIER))
 		seq_puts(seq, ",nobarrier");
 	if (info->max_inline != 8192 * 1024)
-		seq_printf(seq, ",max_inline=%llu",
-			   (unsigned long long)info->max_inline);
+		seq_printf(seq, ",max_inline=%llu", info->max_inline);
 	if (info->alloc_start != 0)
-		seq_printf(seq, ",alloc_start=%llu",
-			   (unsigned long long)info->alloc_start);
+		seq_printf(seq, ",alloc_start=%llu", info->alloc_start);
 	if (info->thread_pool_size !=  min_t(unsigned long,
 					     num_online_cpus() + 2, 8))
 		seq_printf(seq, ",thread_pool=%d", info->thread_pool_size);
@@ -928,6 +982,8 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",space_cache");
 	else
 		seq_puts(seq, ",nospace_cache");
+	if (btrfs_test_opt(root, RESCAN_UUID_TREE))
+		seq_puts(seq, ",rescan_uuid_tree");
 	if (btrfs_test_opt(root, CLEAR_CACHE))
 		seq_puts(seq, ",clear_cache");
 	if (btrfs_test_opt(root, USER_SUBVOL_RM_ALLOWED))
@@ -940,8 +996,24 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",inode_cache");
 	if (btrfs_test_opt(root, SKIP_BALANCE))
 		seq_puts(seq, ",skip_balance");
+	if (btrfs_test_opt(root, RECOVERY))
+		seq_puts(seq, ",recovery");
+#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
+	if (btrfs_test_opt(root, CHECK_INTEGRITY_INCLUDING_EXTENT_DATA))
+		seq_puts(seq, ",check_int_data");
+	else if (btrfs_test_opt(root, CHECK_INTEGRITY))
+		seq_puts(seq, ",check_int");
+	if (info->check_integrity_print_mask)
+		seq_printf(seq, ",check_int_print_mask=%d",
+				info->check_integrity_print_mask);
+#endif
+	if (info->metadata_ratio)
+		seq_printf(seq, ",metadata_ratio=%d",
+				info->metadata_ratio);
 	if (btrfs_test_opt(root, PANIC_ON_FATAL_ERROR))
 		seq_puts(seq, ",fatal_errors=panic");
+	if (info->commit_interval != BTRFS_DEFAULT_COMMIT_INTERVAL)
+		seq_printf(seq, ",commit=%d", info->commit_interval);
 	return 0;
 }
 
@@ -1696,6 +1768,11 @@ static void btrfs_print_info(void)
 			"\n");
 }
 
+static int btrfs_run_sanity_tests(void)
+{
+	return btrfs_test_free_space_cache();
+}
+
 static int __init init_btrfs_fs(void)
 {
 	int err;
@@ -1734,23 +1811,32 @@ static int __init init_btrfs_fs(void)
 	if (err)
 		goto free_auto_defrag;
 
-	err = btrfs_interface_init();
+	err = btrfs_prelim_ref_init();
 	if (err)
-		goto free_delayed_ref;
+		goto free_prelim_ref;
 
-	err = register_filesystem(&btrfs_fs_type);
+	err = btrfs_interface_init();
 	if (err)
-		goto unregister_ioctl;
+		goto free_delayed_ref;
 
 	btrfs_init_lockdep();
 
 	btrfs_print_info();
-	btrfs_test_free_space_cache();
+
+	err = btrfs_run_sanity_tests();
+	if (err)
+		goto unregister_ioctl;
+
+	err = register_filesystem(&btrfs_fs_type);
+	if (err)
+		goto unregister_ioctl;
 
 	return 0;
 
 unregister_ioctl:
 	btrfs_interface_exit();
+free_prelim_ref:
+	btrfs_prelim_ref_exit();
 free_delayed_ref:
 	btrfs_delayed_ref_exit();
 free_auto_defrag:
@@ -1777,6 +1863,7 @@ static void __exit exit_btrfs_fs(void)
 	btrfs_delayed_ref_exit();
 	btrfs_auto_defrag_exit();
 	btrfs_delayed_inode_exit();
+	btrfs_prelim_ref_exit();
 	ordered_data_exit();
 	extent_map_exit();
 	extent_io_exit();

fs/btrfs/tests/btrfs-tests.h

@@ -0,0 +1,34 @@
+/*
+ * 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.
+ */
+
+#ifndef __BTRFS_TESTS
+#define __BTRFS_TESTS
+
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+
+#define test_msg(fmt, ...) pr_info("btrfs: selftest: " fmt, ##__VA_ARGS__)
+
+int btrfs_test_free_space_cache(void);
+#else
+static inline int btrfs_test_free_space_cache(void)
+{
+	return 0;
+}
+#endif
+
+#endif

fs/btrfs/tests/free-space-tests.c

@@ -0,0 +1,395 @@
+/*
+ * 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 "../free-space-cache.h"
+
+#define BITS_PER_BITMAP		(PAGE_CACHE_SIZE * 8)
+static struct btrfs_block_group_cache *init_test_block_group(void)
+{
+	struct btrfs_block_group_cache *cache;
+
+	cache = kzalloc(sizeof(*cache), GFP_NOFS);
+	if (!cache)
+		return NULL;
+	cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
+					GFP_NOFS);
+	if (!cache->free_space_ctl) {
+		kfree(cache);
+		return NULL;
+	}
+
+	cache->key.objectid = 0;
+	cache->key.offset = 1024 * 1024 * 1024;
+	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
+	cache->sectorsize = 4096;
+
+	spin_lock_init(&cache->lock);
+	INIT_LIST_HEAD(&cache->list);
+	INIT_LIST_HEAD(&cache->cluster_list);
+	INIT_LIST_HEAD(&cache->new_bg_list);
+
+	btrfs_init_free_space_ctl(cache);
+
+	return cache;
+}
+
+/*
+ * This test just does basic sanity checking, making sure we can add an exten
+ * entry and remove space from either end and the middle, and make sure we can
+ * remove space that covers adjacent extent entries.
+ */
+static int test_extents(struct btrfs_block_group_cache *cache)
+{
+	int ret = 0;
+
+	test_msg("Running extent only tests\n");
+
+	/* First just make sure we can remove an entire entry */
+	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error adding initial extents %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing extent %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 4 * 1024 * 1024)) {
+		test_msg("Full remove left some lingering space\n");
+		return -1;
+	}
+
+	/* Ok edge and middle cases now */
+	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error adding half extent %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing tail end %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing front end %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 2 * 1024 * 1024, 4096);
+	if (ret) {
+		test_msg("Error removing middle peice %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 1 * 1024 * 1024)) {
+		test_msg("Still have space at the front\n");
+		return -1;
+	}
+
+	if (test_check_exists(cache, 2 * 1024 * 1024, 4096)) {
+		test_msg("Still have space in the middle\n");
+		return -1;
+	}
+
+	if (test_check_exists(cache, 3 * 1024 * 1024, 1 * 1024 * 1024)) {
+		test_msg("Still have space at the end\n");
+		return -1;
+	}
+
+	/* Cleanup */
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	return 0;
+}
+
+static int test_bitmaps(struct btrfs_block_group_cache *cache)
+{
+	u64 next_bitmap_offset;
+	int ret;
+
+	test_msg("Running bitmap only tests\n");
+
+	ret = test_add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't create a bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing bitmap full range %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 4 * 1024 * 1024)) {
+		test_msg("Left some space in bitmap\n");
+		return -1;
+	}
+
+	ret = test_add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add to our bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 2 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove middle chunk %d\n", ret);
+		return ret;
+	}
+
+	/*
+	 * The first bitmap we have starts at offset 0 so the next one is just
+	 * at the end of the first bitmap.
+	 */
+	next_bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
+
+	/* Test a bit straddling two bitmaps */
+	ret = test_add_free_space_entry(cache, next_bitmap_offset -
+				   (2 * 1024 * 1024), 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add space that straddles two bitmaps %d\n",
+				ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, next_bitmap_offset -
+				      (1 * 1024 * 1024), 2 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, next_bitmap_offset - (1 * 1024 * 1024),
+			 2 * 1024 * 1024)) {
+		test_msg("Left some space when removing overlapping\n");
+		return -1;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	return 0;
+}
+
+/* This is the high grade jackassery */
+static int test_bitmaps_and_extents(struct btrfs_block_group_cache *cache)
+{
+	u64 bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
+	int ret;
+
+	test_msg("Running bitmap and extent tests\n");
+
+	/*
+	 * First let's do something simple, an extent at the same offset as the
+	 * bitmap, but the free space completely in the extent and then
+	 * completely in the bitmap.
+	 */
+	ret = test_add_free_space_entry(cache, 4 * 1024 * 1024, 1 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't create bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove extent entry %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 1 * 1024 * 1024)) {
+		test_msg("Left remnants after our remove\n");
+		return -1;
+	}
+
+	/* Now to add back the extent entry and remove from the bitmap */
+	ret = test_add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't re-add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 4 * 1024 * 1024, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove from bitmap %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 4 * 1024 * 1024, 1 * 1024 * 1024)) {
+		test_msg("Left remnants in the bitmap\n");
+		return -1;
+	}
+
+	/*
+	 * Ok so a little more evil, extent entry and bitmap at the same offset,
+	 * removing an overlapping chunk.
+	 */
+	ret = test_add_free_space_entry(cache, 1 * 1024 * 1024, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add to a bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 512 * 1024, 3 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 512 * 1024, 3 * 1024 * 1024)) {
+		test_msg("Left over peices after removing overlapping\n");
+		return -1;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	/* Now with the extent entry offset into the bitmap */
+	ret = test_add_free_space_entry(cache, 4 * 1024 * 1024, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add space to the bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, 2 * 1024 * 1024, 2 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent to the cache %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Problem removing overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 3 * 1024 * 1024, 4 * 1024 * 1024)) {
+		test_msg("Left something behind when removing space");
+		return -1;
+	}
+
+	/*
+	 * This has blown up in the past, the extent entry starts before the
+	 * bitmap entry, but we're trying to remove an offset that falls
+	 * completely within the bitmap range and is in both the extent entry
+	 * and the bitmap entry, looks like this
+	 *
+	 *   [ extent ]
+	 *      [ bitmap ]
+	 *        [ del ]
+	 */
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	ret = test_add_free_space_entry(cache, bitmap_offset + 4 * 1024 * 1024,
+				   4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, bitmap_offset - 1 * 1024 * 1024,
+				   5 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, bitmap_offset + 1 * 1024 * 1024,
+				      5 * 1024 * 1024);
+	if (ret) {
+		test_msg("Failed to free our space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, bitmap_offset + 1 * 1024 * 1024,
+			 5 * 1024 * 1024)) {
+		test_msg("Left stuff over\n");
+		return -1;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	/*
+	 * This blew up before, we have part of the free space in a bitmap and
+	 * then the entirety of the rest of the space in an extent.  This used
+	 * to return -EAGAIN back from btrfs_remove_extent, make sure this
+	 * doesn't happen.
+	 */
+	ret = test_add_free_space_entry(cache, 1 * 1024 * 1024, 2 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, 3 * 1024 * 1024, 1 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 3 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing bitmap and extent overlapping %d\n", ret);
+		return ret;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	return 0;
+}
+
+int btrfs_test_free_space_cache(void)
+{
+	struct btrfs_block_group_cache *cache;
+	int ret;
+
+	test_msg("Running btrfs free space cache tests\n");
+
+	cache = init_test_block_group();
+	if (!cache) {
+		test_msg("Couldn't run the tests\n");
+		return 0;
+	}
+
+	ret = test_extents(cache);
+	if (ret)
+		goto out;
+	ret = test_bitmaps(cache);
+	if (ret)
+		goto out;
+	ret = test_bitmaps_and_extents(cache);
+	if (ret)
+		goto out;
+out:
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	kfree(cache->free_space_ctl);
+	kfree(cache);
+	test_msg("Free space cache tests finished\n");
+	return ret;
+}

fs/btrfs/transaction.c

@@ -837,7 +837,7 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
  * them in one of two extent_io trees.  This is used to make sure all of
  * those extents are on disk for transaction or log commit
  */
-int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
+static int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark)
 {
 	int ret;
@@ -1225,8 +1225,8 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 		btrfs_set_root_stransid(new_root_item, 0);
 		btrfs_set_root_rtransid(new_root_item, 0);
 	}
-	new_root_item->otime.sec = cpu_to_le64(cur_time.tv_sec);
-	new_root_item->otime.nsec = cpu_to_le32(cur_time.tv_nsec);
+	btrfs_set_stack_timespec_sec(&new_root_item->otime, cur_time.tv_sec);
+	btrfs_set_stack_timespec_nsec(&new_root_item->otime, cur_time.tv_nsec);
 	btrfs_set_root_otransid(new_root_item, trans->transid);
 
 	old = btrfs_lock_root_node(root);
@@ -1311,8 +1311,26 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 					 dentry->d_name.len * 2);
 	parent_inode->i_mtime = parent_inode->i_ctime = CURRENT_TIME;
 	ret = btrfs_update_inode_fallback(trans, parent_root, parent_inode);
-	if (ret)
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto fail;
+	}
+	ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root, new_uuid.b,
+				  BTRFS_UUID_KEY_SUBVOL, objectid);
+	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
+		goto fail;
+	}
+	if (!btrfs_is_empty_uuid(new_root_item->received_uuid)) {
+		ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
+					  new_root_item->received_uuid,
+					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+					  objectid);
+		if (ret && ret != -EEXIST) {
+			btrfs_abort_transaction(trans, root, ret);
+			goto fail;
+		}
+	}
 fail:
 	pending->error = ret;
 dir_item_existed:
@@ -1362,6 +1380,8 @@ static void update_super_roots(struct btrfs_root *root)
 	super->root_level = root_item->level;
 	if (btrfs_test_opt(root, SPACE_CACHE))
 		super->cache_generation = root_item->generation;
+	if (root->fs_info->update_uuid_tree_gen)
+		super->uuid_tree_generation = root_item->generation;
 }
 
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info)
@@ -1928,8 +1948,7 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 	list_del_init(&root->root_list);
 	spin_unlock(&fs_info->trans_lock);
 
-	pr_debug("btrfs: cleaner removing %llu\n",
-			(unsigned long long)root->objectid);
+	pr_debug("btrfs: cleaner removing %llu\n", root->objectid);
 
 	btrfs_kill_all_delayed_nodes(root);
 
@@ -1942,6 +1961,5 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 	 * If we encounter a transaction abort during snapshot cleaning, we
 	 * don't want to crash here
 	 */
-	BUG_ON(ret < 0 && ret != -EAGAIN && ret != -EROFS);
-	return 1;
+	return (ret < 0) ? 0 : 1;
 }

fs/btrfs/transaction.h

@@ -160,8 +160,6 @@ int btrfs_should_end_transaction(struct btrfs_trans_handle *trans,
 void btrfs_throttle(struct btrfs_root *root);
 int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root);
-int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
-				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_write_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_wait_marked_extents(struct btrfs_root *root,

fs/btrfs/tree-log.c

@@ -747,7 +747,8 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
 	ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
 	if (ret)
 		goto out;
-	btrfs_run_delayed_items(trans, root);
+	else
+		ret = btrfs_run_delayed_items(trans, root);
 out:
 	kfree(name);
 	iput(inode);
@@ -923,7 +924,9 @@ again:
 				kfree(victim_name);
 				if (ret)
 					return ret;
-				btrfs_run_delayed_items(trans, root);
+				ret = btrfs_run_delayed_items(trans, root);
+				if (ret)
+					return ret;
 				*search_done = 1;
 				goto again;
 			}
@@ -990,7 +993,9 @@ again:
 								 inode,
 								 victim_name,
 								 victim_name_len);
-					btrfs_run_delayed_items(trans, root);
+					if (!ret)
+						ret = btrfs_run_delayed_items(
+								  trans, root);
 				}
 				iput(victim_parent);
 				kfree(victim_name);
@@ -1536,8 +1541,10 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 
 	name_len = btrfs_dir_name_len(eb, di);
 	name = kmalloc(name_len, GFP_NOFS);
-	if (!name)
-		return -ENOMEM;
+	if (!name) {
+		ret = -ENOMEM;
+		goto out;
+	}
 
 	log_type = btrfs_dir_type(eb, di);
 	read_extent_buffer(eb, name, (unsigned long)(di + 1),
@@ -1810,7 +1817,7 @@ again:
 			ret = btrfs_unlink_inode(trans, root, dir, inode,
 						 name, name_len);
 			if (!ret)
-				btrfs_run_delayed_items(trans, root);
+				ret = btrfs_run_delayed_items(trans, root);
 			kfree(name);
 			iput(inode);
 			if (ret)

fs/btrfs/uuid-tree.c

@@ -0,0 +1,358 @@
+/*
+ * Copyright (C) STRATO AG 2013.  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/uuid.h>
+#include <asm/unaligned.h>
+#include "ctree.h"
+#include "transaction.h"
+#include "disk-io.h"
+#include "print-tree.h"
+
+
+static void btrfs_uuid_to_key(u8 *uuid, u8 type, struct btrfs_key *key)
+{
+	key->type = type;
+	key->objectid = get_unaligned_le64(uuid);
+	key->offset = get_unaligned_le64(uuid + sizeof(u64));
+}
+
+/* return -ENOENT for !found, < 0 for errors, or 0 if an item was found */
+static int btrfs_uuid_tree_lookup(struct btrfs_root *uuid_root, u8 *uuid,
+				  u8 type, u64 subid)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct extent_buffer *eb;
+	int slot;
+	u32 item_size;
+	unsigned long offset;
+	struct btrfs_key key;
+
+	if (WARN_ON_ONCE(!uuid_root)) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, type, &key);
+	ret = btrfs_search_slot(NULL, uuid_root, &key, path, 0, 0);
+	if (ret < 0) {
+		goto out;
+	} else if (ret > 0) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	item_size = btrfs_item_size_nr(eb, slot);
+	offset = btrfs_item_ptr_offset(eb, slot);
+	ret = -ENOENT;
+
+	if (!IS_ALIGNED(item_size, sizeof(u64))) {
+		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			(unsigned long)item_size);
+		goto out;
+	}
+	while (item_size) {
+		__le64 data;
+
+		read_extent_buffer(eb, &data, offset, sizeof(data));
+		if (le64_to_cpu(data) == subid) {
+			ret = 0;
+			break;
+		}
+		offset += sizeof(data);
+		item_size -= sizeof(data);
+	}
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+int btrfs_uuid_tree_add(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid_cpu)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	int slot;
+	unsigned long offset;
+	__le64 subid_le;
+
+	ret = btrfs_uuid_tree_lookup(uuid_root, uuid, type, subid_cpu);
+	if (ret != -ENOENT)
+		return ret;
+
+	if (WARN_ON_ONCE(!uuid_root)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, type, &key);
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_insert_empty_item(trans, uuid_root, path, &key,
+				      sizeof(subid_le));
+	if (ret >= 0) {
+		/* Add an item for the type for the first time */
+		eb = path->nodes[0];
+		slot = path->slots[0];
+		offset = btrfs_item_ptr_offset(eb, slot);
+	} else if (ret == -EEXIST) {
+		/*
+		 * An item with that type already exists.
+		 * Extend the item and store the new subid at the end.
+		 */
+		btrfs_extend_item(uuid_root, path, sizeof(subid_le));
+		eb = path->nodes[0];
+		slot = path->slots[0];
+		offset = btrfs_item_ptr_offset(eb, slot);
+		offset += btrfs_item_size_nr(eb, slot) - sizeof(subid_le);
+	} else if (ret < 0) {
+		pr_warn("btrfs: insert uuid item failed %d (0x%016llx, 0x%016llx) type %u!\n",
+			ret, (unsigned long long)key.objectid,
+			(unsigned long long)key.offset, type);
+		goto out;
+	}
+
+	ret = 0;
+	subid_le = cpu_to_le64(subid_cpu);
+	write_extent_buffer(eb, &subid_le, offset, sizeof(subid_le));
+	btrfs_mark_buffer_dirty(eb);
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	int slot;
+	unsigned long offset;
+	u32 item_size;
+	unsigned long move_dst;
+	unsigned long move_src;
+	unsigned long move_len;
+
+	if (WARN_ON_ONCE(!uuid_root)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, type, &key);
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_search_slot(trans, uuid_root, &key, path, -1, 1);
+	if (ret < 0) {
+		pr_warn("btrfs: error %d while searching for uuid item!\n",
+			ret);
+		goto out;
+	}
+	if (ret > 0) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	offset = btrfs_item_ptr_offset(eb, slot);
+	item_size = btrfs_item_size_nr(eb, slot);
+	if (!IS_ALIGNED(item_size, sizeof(u64))) {
+		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			(unsigned long)item_size);
+		ret = -ENOENT;
+		goto out;
+	}
+	while (item_size) {
+		__le64 read_subid;
+
+		read_extent_buffer(eb, &read_subid, offset, sizeof(read_subid));
+		if (le64_to_cpu(read_subid) == subid)
+			break;
+		offset += sizeof(read_subid);
+		item_size -= sizeof(read_subid);
+	}
+
+	if (!item_size) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	item_size = btrfs_item_size_nr(eb, slot);
+	if (item_size == sizeof(subid)) {
+		ret = btrfs_del_item(trans, uuid_root, path);
+		goto out;
+	}
+
+	move_dst = offset;
+	move_src = offset + sizeof(subid);
+	move_len = item_size - (move_src - btrfs_item_ptr_offset(eb, slot));
+	memmove_extent_buffer(eb, move_dst, move_src, move_len);
+	btrfs_truncate_item(uuid_root, path, item_size - sizeof(subid), 1);
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+static int btrfs_uuid_iter_rem(struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			       u64 subid)
+{
+	struct btrfs_trans_handle *trans;
+	int ret;
+
+	/* 1 - for the uuid item */
+	trans = btrfs_start_transaction(uuid_root, 1);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		goto out;
+	}
+
+	ret = btrfs_uuid_tree_rem(trans, uuid_root, uuid, type, subid);
+	btrfs_end_transaction(trans, uuid_root);
+
+out:
+	return ret;
+}
+
+int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
+			    int (*check_func)(struct btrfs_fs_info *, u8 *, u8,
+					      u64))
+{
+	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;
+	int slot;
+	u32 item_size;
+	unsigned long offset;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	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);
+	if (ret) {
+		if (ret > 0)
+			ret = 0;
+		goto out;
+	}
+
+	while (1) {
+		cond_resched();
+		leaf = path->nodes[0];
+		slot = path->slots[0];
+		btrfs_item_key_to_cpu(leaf, &key, slot);
+
+		if (key.type != BTRFS_UUID_KEY_SUBVOL &&
+		    key.type != BTRFS_UUID_KEY_RECEIVED_SUBVOL)
+			goto skip;
+
+		offset = btrfs_item_ptr_offset(leaf, slot);
+		item_size = btrfs_item_size_nr(leaf, slot);
+		if (!IS_ALIGNED(item_size, sizeof(u64))) {
+			pr_warn("btrfs: uuid item with illegal size %lu!\n",
+				(unsigned long)item_size);
+			goto skip;
+		}
+		while (item_size) {
+			u8 uuid[BTRFS_UUID_SIZE];
+			__le64 subid_le;
+			u64 subid_cpu;
+
+			put_unaligned_le64(key.objectid, uuid);
+			put_unaligned_le64(key.offset, uuid + sizeof(u64));
+			read_extent_buffer(leaf, &subid_le, offset,
+					   sizeof(subid_le));
+			subid_cpu = le64_to_cpu(subid_le);
+			ret = check_func(fs_info, uuid, key.type, subid_cpu);
+			if (ret < 0)
+				goto out;
+			if (ret > 0) {
+				btrfs_release_path(path);
+				ret = btrfs_uuid_iter_rem(root, uuid, key.type,
+							  subid_cpu);
+				if (ret == 0) {
+					/*
+					 * this might look inefficient, but the
+					 * justification is that it is an
+					 * exception that check_func returns 1,
+					 * and that in the regular case only one
+					 * entry per UUID exists.
+					 */
+					goto again_search_slot;
+				}
+				if (ret < 0 && ret != -ENOENT)
+					goto out;
+			}
+			item_size -= sizeof(subid_le);
+			offset += sizeof(subid_le);
+		}
+
+skip:
+		ret = btrfs_next_item(root, path);
+		if (ret == 0)
+			continue;
+		else if (ret > 0)
+			ret = 0;
+		break;
+	}
+
+out:
+	btrfs_free_path(path);
+	if (ret)
+		pr_warn("btrfs: btrfs_uuid_tree_iterate failed %d\n", ret);
+	return 0;
+}

fs/btrfs/volumes.c

@@ -26,6 +26,7 @@
 #include <linux/ratelimit.h>
 #include <linux/kthread.h>
 #include <linux/raid/pq.h>
+#include <linux/semaphore.h>
 #include <asm/div64.h>
 #include "compat.h"
 #include "ctree.h"
@@ -62,6 +63,48 @@ static void unlock_chunks(struct btrfs_root *root)
 	mutex_unlock(&root->fs_info->chunk_mutex);
 }
 
+static struct btrfs_fs_devices *__alloc_fs_devices(void)
+{
+	struct btrfs_fs_devices *fs_devs;
+
+	fs_devs = kzalloc(sizeof(*fs_devs), GFP_NOFS);
+	if (!fs_devs)
+		return ERR_PTR(-ENOMEM);
+
+	mutex_init(&fs_devs->device_list_mutex);
+
+	INIT_LIST_HEAD(&fs_devs->devices);
+	INIT_LIST_HEAD(&fs_devs->alloc_list);
+	INIT_LIST_HEAD(&fs_devs->list);
+
+	return fs_devs;
+}
+
+/**
+ * alloc_fs_devices - allocate struct btrfs_fs_devices
+ * @fsid:	a pointer to UUID for this FS.  If NULL a new UUID is
+ *		generated.
+ *
+ * Return: a pointer to a new &struct btrfs_fs_devices on success;
+ * ERR_PTR() on error.  Returned struct is not linked onto any lists and
+ * can be destroyed with kfree() right away.
+ */
+static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid)
+{
+	struct btrfs_fs_devices *fs_devs;
+
+	fs_devs = __alloc_fs_devices();
+	if (IS_ERR(fs_devs))
+		return fs_devs;
+
+	if (fsid)
+		memcpy(fs_devs->fsid, fsid, BTRFS_FSID_SIZE);
+	else
+		generate_random_uuid(fs_devs->fsid);
+
+	return fs_devs;
+}
+
 static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
 {
 	struct btrfs_device *device;
@@ -101,6 +144,27 @@ void btrfs_cleanup_fs_uuids(void)
 	}
 }
 
+static struct btrfs_device *__alloc_device(void)
+{
+	struct btrfs_device *dev;
+
+	dev = kzalloc(sizeof(*dev), GFP_NOFS);
+	if (!dev)
+		return ERR_PTR(-ENOMEM);
+
+	INIT_LIST_HEAD(&dev->dev_list);
+	INIT_LIST_HEAD(&dev->dev_alloc_list);
+
+	spin_lock_init(&dev->io_lock);
+
+	spin_lock_init(&dev->reada_lock);
+	atomic_set(&dev->reada_in_flight, 0);
+	INIT_RADIX_TREE(&dev->reada_zones, GFP_NOFS & ~__GFP_WAIT);
+	INIT_RADIX_TREE(&dev->reada_extents, GFP_NOFS & ~__GFP_WAIT);
+
+	return dev;
+}
+
 static noinline struct btrfs_device *__find_device(struct list_head *head,
 						   u64 devid, u8 *uuid)
 {
@@ -395,16 +459,14 @@ static noinline int device_list_add(const char *path,
 
 	fs_devices = find_fsid(disk_super->fsid);
 	if (!fs_devices) {
-		fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
-		if (!fs_devices)
-			return -ENOMEM;
-		INIT_LIST_HEAD(&fs_devices->devices);
-		INIT_LIST_HEAD(&fs_devices->alloc_list);
+		fs_devices = alloc_fs_devices(disk_super->fsid);
+		if (IS_ERR(fs_devices))
+			return PTR_ERR(fs_devices);
+
 		list_add(&fs_devices->list, &fs_uuids);
-		memcpy(fs_devices->fsid, disk_super->fsid, BTRFS_FSID_SIZE);
 		fs_devices->latest_devid = devid;
 		fs_devices->latest_trans = found_transid;
-		mutex_init(&fs_devices->device_list_mutex);
+
 		device = NULL;
 	} else {
 		device = __find_device(&fs_devices->devices, devid,
@@ -414,17 +476,12 @@ static noinline int device_list_add(const char *path,
 		if (fs_devices->opened)
 			return -EBUSY;
 
-		device = kzalloc(sizeof(*device), GFP_NOFS);
-		if (!device) {
+		device = btrfs_alloc_device(NULL, &devid,
+					    disk_super->dev_item.uuid);
+		if (IS_ERR(device)) {
 			/* we can safely leave the fs_devices entry around */
-			return -ENOMEM;
+			return PTR_ERR(device);
 		}
-		device->devid = devid;
-		device->dev_stats_valid = 0;
-		device->work.func = pending_bios_fn;
-		memcpy(device->uuid, disk_super->dev_item.uuid,
-		       BTRFS_UUID_SIZE);
-		spin_lock_init(&device->io_lock);
 
 		name = rcu_string_strdup(path, GFP_NOFS);
 		if (!name) {
@@ -432,22 +489,13 @@ static noinline int device_list_add(const char *path,
 			return -ENOMEM;
 		}
 		rcu_assign_pointer(device->name, name);
-		INIT_LIST_HEAD(&device->dev_alloc_list);
-
-		/* init readahead state */
-		spin_lock_init(&device->reada_lock);
-		device->reada_curr_zone = NULL;
-		atomic_set(&device->reada_in_flight, 0);
-		device->reada_next = 0;
-		INIT_RADIX_TREE(&device->reada_zones, GFP_NOFS & ~__GFP_WAIT);
-		INIT_RADIX_TREE(&device->reada_extents, GFP_NOFS & ~__GFP_WAIT);
 
 		mutex_lock(&fs_devices->device_list_mutex);
 		list_add_rcu(&device->dev_list, &fs_devices->devices);
+		fs_devices->num_devices++;
 		mutex_unlock(&fs_devices->device_list_mutex);
 
 		device->fs_devices = fs_devices;
-		fs_devices->num_devices++;
 	} else if (!device->name || strcmp(device->name->str, path)) {
 		name = rcu_string_strdup(path, GFP_NOFS);
 		if (!name)
@@ -474,25 +522,21 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 	struct btrfs_device *device;
 	struct btrfs_device *orig_dev;
 
-	fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
-	if (!fs_devices)
-		return ERR_PTR(-ENOMEM);
+	fs_devices = alloc_fs_devices(orig->fsid);
+	if (IS_ERR(fs_devices))
+		return fs_devices;
 
-	INIT_LIST_HEAD(&fs_devices->devices);
-	INIT_LIST_HEAD(&fs_devices->alloc_list);
-	INIT_LIST_HEAD(&fs_devices->list);
-	mutex_init(&fs_devices->device_list_mutex);
 	fs_devices->latest_devid = orig->latest_devid;
 	fs_devices->latest_trans = orig->latest_trans;
 	fs_devices->total_devices = orig->total_devices;
-	memcpy(fs_devices->fsid, orig->fsid, sizeof(fs_devices->fsid));
 
 	/* We have held the volume lock, it is safe to get the devices. */
 	list_for_each_entry(orig_dev, &orig->devices, dev_list) {
 		struct rcu_string *name;
 
-		device = kzalloc(sizeof(*device), GFP_NOFS);
-		if (!device)
+		device = btrfs_alloc_device(NULL, &orig_dev->devid,
+					    orig_dev->uuid);
+		if (IS_ERR(device))
 			goto error;
 
 		/*
@@ -506,13 +550,6 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 		}
 		rcu_assign_pointer(device->name, name);
 
-		device->devid = orig_dev->devid;
-		device->work.func = pending_bios_fn;
-		memcpy(device->uuid, orig_dev->uuid, sizeof(device->uuid));
-		spin_lock_init(&device->io_lock);
-		INIT_LIST_HEAD(&device->dev_list);
-		INIT_LIST_HEAD(&device->dev_alloc_list);
-
 		list_add(&device->dev_list, &fs_devices->devices);
 		device->fs_devices = fs_devices;
 		fs_devices->num_devices++;
@@ -636,23 +673,22 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 
 		if (device->can_discard)
 			fs_devices->num_can_discard--;
+		if (device->missing)
+			fs_devices->missing_devices--;
 
-		new_device = kmalloc(sizeof(*new_device), GFP_NOFS);
-		BUG_ON(!new_device); /* -ENOMEM */
-		memcpy(new_device, device, sizeof(*new_device));
+		new_device = btrfs_alloc_device(NULL, &device->devid,
+						device->uuid);
+		BUG_ON(IS_ERR(new_device)); /* -ENOMEM */
 
 		/* Safe because we are under uuid_mutex */
 		if (device->name) {
 			name = rcu_string_strdup(device->name->str, GFP_NOFS);
-			BUG_ON(device->name && !name); /* -ENOMEM */
+			BUG_ON(!name); /* -ENOMEM */
 			rcu_assign_pointer(new_device->name, name);
 		}
-		new_device->bdev = NULL;
-		new_device->writeable = 0;
-		new_device->in_fs_metadata = 0;
-		new_device->can_discard = 0;
-		spin_lock_init(&new_device->io_lock);
+
 		list_replace_rcu(&device->dev_list, &new_device->dev_list);
+		new_device->fs_devices = device->fs_devices;
 
 		call_rcu(&device->rcu, free_device);
 	}
@@ -865,7 +901,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	disk_super = p + (bytenr & ~PAGE_CACHE_MASK);
 
 	if (btrfs_super_bytenr(disk_super) != bytenr ||
-	    disk_super->magic != cpu_to_le64(BTRFS_MAGIC))
+	    btrfs_super_magic(disk_super) != BTRFS_MAGIC)
 		goto error_unmap;
 
 	devid = btrfs_stack_device_id(&disk_super->dev_item);
@@ -880,8 +916,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 		printk(KERN_INFO "device fsid %pU ", disk_super->fsid);
 	}
 
-	printk(KERN_CONT "devid %llu transid %llu %s\n",
-	       (unsigned long long)devid, (unsigned long long)transid, path);
+	printk(KERN_CONT "devid %llu transid %llu %s\n", devid, transid, path);
 
 	ret = device_list_add(path, disk_super, devid, fs_devices_ret);
 	if (!ret && fs_devices_ret)
@@ -1278,8 +1313,7 @@ static int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
 	btrfs_set_dev_extent_chunk_offset(leaf, extent, chunk_offset);
 
 	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
-		    (unsigned long)btrfs_dev_extent_chunk_tree_uuid(extent),
-		    BTRFS_UUID_SIZE);
+		    btrfs_dev_extent_chunk_tree_uuid(extent), BTRFS_UUID_SIZE);
 
 	btrfs_set_dev_extent_length(leaf, extent, num_bytes);
 	btrfs_mark_buffer_dirty(leaf);
@@ -1307,15 +1341,14 @@ static u64 find_next_chunk(struct btrfs_fs_info *fs_info)
 	return ret;
 }
 
-static noinline int find_next_devid(struct btrfs_root *root, u64 *objectid)
+static noinline int find_next_devid(struct btrfs_fs_info *fs_info,
+				    u64 *devid_ret)
 {
 	int ret;
 	struct btrfs_key key;
 	struct btrfs_key found_key;
 	struct btrfs_path *path;
 
-	root = root->fs_info->chunk_root;
-
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -1324,20 +1357,21 @@ static noinline int find_next_devid(struct btrfs_root *root, u64 *objectid)
 	key.type = BTRFS_DEV_ITEM_KEY;
 	key.offset = (u64)-1;
 
-	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	ret = btrfs_search_slot(NULL, fs_info->chunk_root, &key, path, 0, 0);
 	if (ret < 0)
 		goto error;
 
 	BUG_ON(ret == 0); /* Corruption */
 
-	ret = btrfs_previous_item(root, path, BTRFS_DEV_ITEMS_OBJECTID,
+	ret = btrfs_previous_item(fs_info->chunk_root, path,
+				  BTRFS_DEV_ITEMS_OBJECTID,
 				  BTRFS_DEV_ITEM_KEY);
 	if (ret) {
-		*objectid = 1;
+		*devid_ret = 1;
 	} else {
 		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
 				      path->slots[0]);
-		*objectid = found_key.offset + 1;
+		*devid_ret = found_key.offset + 1;
 	}
 	ret = 0;
 error:
@@ -1391,9 +1425,9 @@ static int btrfs_add_device(struct btrfs_trans_handle *trans,
 	btrfs_set_device_bandwidth(leaf, dev_item, 0);
 	btrfs_set_device_start_offset(leaf, dev_item, 0);
 
-	ptr = (unsigned long)btrfs_device_uuid(dev_item);
+	ptr = btrfs_device_uuid(dev_item);
 	write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
-	ptr = (unsigned long)btrfs_device_fsid(dev_item);
+	ptr = btrfs_device_fsid(dev_item);
 	write_extent_buffer(leaf, root->fs_info->fsid, ptr, BTRFS_UUID_SIZE);
 	btrfs_mark_buffer_dirty(leaf);
 
@@ -1562,7 +1596,9 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 		clear_super = true;
 	}
 
+	mutex_unlock(&uuid_mutex);
 	ret = btrfs_shrink_device(device, 0);
+	mutex_lock(&uuid_mutex);
 	if (ret)
 		goto error_undo;
 
@@ -1586,7 +1622,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	/*
 	 * the device list mutex makes sure that we don't change
 	 * the device list while someone else is writing out all
-	 * the device supers.
+	 * the device supers. Whoever is writing all supers, should
+	 * lock the device list mutex before getting the number of
+	 * devices in the super block (super_copy). Conversely,
+	 * whoever updates the number of devices in the super block
+	 * (super_copy) should hold the device list mutex.
 	 */
 
 	cur_devices = device->fs_devices;
@@ -1610,10 +1650,10 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 		device->fs_devices->open_devices--;
 
 	call_rcu(&device->rcu, free_device);
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	num_devices = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
 	btrfs_set_super_num_devices(root->fs_info->super_copy, num_devices);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	if (cur_devices->open_devices == 0) {
 		struct btrfs_fs_devices *fs_devices;
@@ -1793,9 +1833,9 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	if (!fs_devices->seeding)
 		return -EINVAL;
 
-	seed_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
-	if (!seed_devices)
-		return -ENOMEM;
+	seed_devices = __alloc_fs_devices();
+	if (IS_ERR(seed_devices))
+		return PTR_ERR(seed_devices);
 
 	old_devices = clone_fs_devices(fs_devices);
 	if (IS_ERR(old_devices)) {
@@ -1814,7 +1854,6 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices,
 			      synchronize_rcu);
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list);
 	list_for_each_entry(device, &seed_devices->devices, dev_list) {
@@ -1830,6 +1869,8 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	generate_random_uuid(fs_devices->fsid);
 	memcpy(root->fs_info->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
 	memcpy(disk_super->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+
 	super_flags = btrfs_super_flags(disk_super) &
 		      ~BTRFS_SUPER_FLAG_SEEDING;
 	btrfs_set_super_flags(disk_super, super_flags);
@@ -1889,11 +1930,9 @@ next_slot:
 		dev_item = btrfs_item_ptr(leaf, path->slots[0],
 					  struct btrfs_dev_item);
 		devid = btrfs_device_id(leaf, dev_item);
-		read_extent_buffer(leaf, dev_uuid,
-				   (unsigned long)btrfs_device_uuid(dev_item),
+		read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
 				   BTRFS_UUID_SIZE);
-		read_extent_buffer(leaf, fs_uuid,
-				   (unsigned long)btrfs_device_fsid(dev_item),
+		read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
 				   BTRFS_UUID_SIZE);
 		device = btrfs_find_device(root->fs_info, devid, dev_uuid,
 					   fs_uuid);
@@ -1956,10 +1995,10 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	}
 	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
-	device = kzalloc(sizeof(*device), GFP_NOFS);
-	if (!device) {
+	device = btrfs_alloc_device(root->fs_info, NULL, NULL);
+	if (IS_ERR(device)) {
 		/* we can safely leave the fs_devices entry around */
-		ret = -ENOMEM;
+		ret = PTR_ERR(device);
 		goto error;
 	}
 
@@ -1971,13 +2010,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	}
 	rcu_assign_pointer(device->name, name);
 
-	ret = find_next_devid(root, &device->devid);
-	if (ret) {
-		rcu_string_free(device->name);
-		kfree(device);
-		goto error;
-	}
-
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		rcu_string_free(device->name);
@@ -1992,9 +2024,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	if (blk_queue_discard(q))
 		device->can_discard = 1;
 	device->writeable = 1;
-	device->work.func = pending_bios_fn;
-	generate_random_uuid(device->uuid);
-	spin_lock_init(&device->io_lock);
 	device->generation = trans->transid;
 	device->io_width = root->sectorsize;
 	device->io_align = root->sectorsize;
@@ -2121,6 +2150,7 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct list_head *devices;
 	struct rcu_string *name;
+	u64 devid = BTRFS_DEV_REPLACE_DEVID;
 	int ret = 0;
 
 	*device_out = NULL;
@@ -2142,9 +2172,9 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 		}
 	}
 
-	device = kzalloc(sizeof(*device), GFP_NOFS);
-	if (!device) {
-		ret = -ENOMEM;
+	device = btrfs_alloc_device(NULL, &devid, NULL);
+	if (IS_ERR(device)) {
+		ret = PTR_ERR(device);
 		goto error;
 	}
 
@@ -2161,10 +2191,6 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 		device->can_discard = 1;
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	device->writeable = 1;
-	device->work.func = pending_bios_fn;
-	generate_random_uuid(device->uuid);
-	device->devid = BTRFS_DEV_REPLACE_DEVID;
-	spin_lock_init(&device->io_lock);
 	device->generation = 0;
 	device->io_width = root->sectorsize;
 	device->io_align = root->sectorsize;
@@ -2971,10 +2997,6 @@ again:
 		if (found_key.objectid != key.objectid)
 			break;
 
-		/* chunk zero is special */
-		if (found_key.offset == 0)
-			break;
-
 		chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
 
 		if (!counting) {
@@ -3010,6 +3032,8 @@ again:
 			spin_unlock(&fs_info->balance_lock);
 		}
 loop:
+		if (found_key.offset == 0)
+			break;
 		key.offset = found_key.offset - 1;
 	}
 
@@ -3074,9 +3098,6 @@ static void __cancel_balance(struct btrfs_fs_info *fs_info)
 	atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
 }
 
-void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
-			       struct btrfs_ioctl_balance_args *bargs);
-
 /*
  * Should be called with both balance and volume mutexes held
  */
@@ -3139,7 +3160,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	     (bctl->data.target & ~allowed))) {
 		printk(KERN_ERR "btrfs: unable to start balance with target "
 		       "data profile %llu\n",
-		       (unsigned long long)bctl->data.target);
+		       bctl->data.target);
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3148,7 +3169,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	     (bctl->meta.target & ~allowed))) {
 		printk(KERN_ERR "btrfs: unable to start balance with target "
 		       "metadata profile %llu\n",
-		       (unsigned long long)bctl->meta.target);
+		       bctl->meta.target);
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3157,7 +3178,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	     (bctl->sys.target & ~allowed))) {
 		printk(KERN_ERR "btrfs: unable to start balance with target "
 		       "system profile %llu\n",
-		       (unsigned long long)bctl->sys.target);
+		       bctl->sys.target);
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3430,6 +3451,264 @@ int btrfs_cancel_balance(struct btrfs_fs_info *fs_info)
 	return 0;
 }
 
+static int btrfs_uuid_scan_kthread(void *data)
+{
+	struct btrfs_fs_info *fs_info = data;
+	struct btrfs_root *root = fs_info->tree_root;
+	struct btrfs_key key;
+	struct btrfs_key max_key;
+	struct btrfs_path *path = NULL;
+	int ret = 0;
+	struct extent_buffer *eb;
+	int slot;
+	struct btrfs_root_item root_item;
+	u32 item_size;
+	struct btrfs_trans_handle *trans = NULL;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	key.objectid = 0;
+	key.type = BTRFS_ROOT_ITEM_KEY;
+	key.offset = 0;
+
+	max_key.objectid = (u64)-1;
+	max_key.type = BTRFS_ROOT_ITEM_KEY;
+	max_key.offset = (u64)-1;
+
+	path->keep_locks = 1;
+
+	while (1) {
+		ret = btrfs_search_forward(root, &key, &max_key, path, 0);
+		if (ret) {
+			if (ret > 0)
+				ret = 0;
+			break;
+		}
+
+		if (key.type != BTRFS_ROOT_ITEM_KEY ||
+		    (key.objectid < BTRFS_FIRST_FREE_OBJECTID &&
+		     key.objectid != BTRFS_FS_TREE_OBJECTID) ||
+		    key.objectid > BTRFS_LAST_FREE_OBJECTID)
+			goto skip;
+
+		eb = path->nodes[0];
+		slot = path->slots[0];
+		item_size = btrfs_item_size_nr(eb, slot);
+		if (item_size < sizeof(root_item))
+			goto skip;
+
+		read_extent_buffer(eb, &root_item,
+				   btrfs_item_ptr_offset(eb, slot),
+				   (int)sizeof(root_item));
+		if (btrfs_root_refs(&root_item) == 0)
+			goto skip;
+
+		if (!btrfs_is_empty_uuid(root_item.uuid) ||
+		    !btrfs_is_empty_uuid(root_item.received_uuid)) {
+			if (trans)
+				goto update_tree;
+
+			btrfs_release_path(path);
+			/*
+			 * 1 - subvol uuid item
+			 * 1 - received_subvol uuid item
+			 */
+			trans = btrfs_start_transaction(fs_info->uuid_root, 2);
+			if (IS_ERR(trans)) {
+				ret = PTR_ERR(trans);
+				break;
+			}
+			continue;
+		} else {
+			goto skip;
+		}
+update_tree:
+		if (!btrfs_is_empty_uuid(root_item.uuid)) {
+			ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
+						  root_item.uuid,
+						  BTRFS_UUID_KEY_SUBVOL,
+						  key.objectid);
+			if (ret < 0) {
+				pr_warn("btrfs: uuid_tree_add failed %d\n",
+					ret);
+				break;
+			}
+		}
+
+		if (!btrfs_is_empty_uuid(root_item.received_uuid)) {
+			ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
+						  root_item.received_uuid,
+						 BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+						  key.objectid);
+			if (ret < 0) {
+				pr_warn("btrfs: uuid_tree_add failed %d\n",
+					ret);
+				break;
+			}
+		}
+
+skip:
+		if (trans) {
+			ret = btrfs_end_transaction(trans, fs_info->uuid_root);
+			trans = NULL;
+			if (ret)
+				break;
+		}
+
+		btrfs_release_path(path);
+		if (key.offset < (u64)-1) {
+			key.offset++;
+		} else if (key.type < BTRFS_ROOT_ITEM_KEY) {
+			key.offset = 0;
+			key.type = BTRFS_ROOT_ITEM_KEY;
+		} else if (key.objectid < (u64)-1) {
+			key.offset = 0;
+			key.type = BTRFS_ROOT_ITEM_KEY;
+			key.objectid++;
+		} else {
+			break;
+		}
+		cond_resched();
+	}
+
+out:
+	btrfs_free_path(path);
+	if (trans && !IS_ERR(trans))
+		btrfs_end_transaction(trans, fs_info->uuid_root);
+	if (ret)
+		pr_warn("btrfs: btrfs_uuid_scan_kthread failed %d\n", ret);
+	else
+		fs_info->update_uuid_tree_gen = 1;
+	up(&fs_info->uuid_tree_rescan_sem);
+	return 0;
+}
+
+/*
+ * Callback for btrfs_uuid_tree_iterate().
+ * returns:
+ * 0	check succeeded, the entry is not outdated.
+ * < 0	if an error occured.
+ * > 0	if the check failed, which means the caller shall remove the entry.
+ */
+static int btrfs_check_uuid_tree_entry(struct btrfs_fs_info *fs_info,
+				       u8 *uuid, u8 type, u64 subid)
+{
+	struct btrfs_key key;
+	int ret = 0;
+	struct btrfs_root *subvol_root;
+
+	if (type != BTRFS_UUID_KEY_SUBVOL &&
+	    type != BTRFS_UUID_KEY_RECEIVED_SUBVOL)
+		goto out;
+
+	key.objectid = subid;
+	key.type = BTRFS_ROOT_ITEM_KEY;
+	key.offset = (u64)-1;
+	subvol_root = btrfs_read_fs_root_no_name(fs_info, &key);
+	if (IS_ERR(subvol_root)) {
+		ret = PTR_ERR(subvol_root);
+		if (ret == -ENOENT)
+			ret = 1;
+		goto out;
+	}
+
+	switch (type) {
+	case BTRFS_UUID_KEY_SUBVOL:
+		if (memcmp(uuid, subvol_root->root_item.uuid, BTRFS_UUID_SIZE))
+			ret = 1;
+		break;
+	case BTRFS_UUID_KEY_RECEIVED_SUBVOL:
+		if (memcmp(uuid, subvol_root->root_item.received_uuid,
+			   BTRFS_UUID_SIZE))
+			ret = 1;
+		break;
+	}
+
+out:
+	return ret;
+}
+
+static int btrfs_uuid_rescan_kthread(void *data)
+{
+	struct btrfs_fs_info *fs_info = (struct btrfs_fs_info *)data;
+	int ret;
+
+	/*
+	 * 1st step is to iterate through the existing UUID tree and
+	 * to delete all entries that contain outdated data.
+	 * 2nd step is to add all missing entries to the UUID tree.
+	 */
+	ret = btrfs_uuid_tree_iterate(fs_info, btrfs_check_uuid_tree_entry);
+	if (ret < 0) {
+		pr_warn("btrfs: iterating uuid_tree failed %d\n", ret);
+		up(&fs_info->uuid_tree_rescan_sem);
+		return ret;
+	}
+	return btrfs_uuid_scan_kthread(data);
+}
+
+int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_trans_handle *trans;
+	struct btrfs_root *tree_root = fs_info->tree_root;
+	struct btrfs_root *uuid_root;
+	struct task_struct *task;
+	int ret;
+
+	/*
+	 * 1 - root node
+	 * 1 - root item
+	 */
+	trans = btrfs_start_transaction(tree_root, 2);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	uuid_root = btrfs_create_tree(trans, fs_info,
+				      BTRFS_UUID_TREE_OBJECTID);
+	if (IS_ERR(uuid_root)) {
+		btrfs_abort_transaction(trans, tree_root,
+					PTR_ERR(uuid_root));
+		return PTR_ERR(uuid_root);
+	}
+
+	fs_info->uuid_root = uuid_root;
+
+	ret = btrfs_commit_transaction(trans, tree_root);
+	if (ret)
+		return ret;
+
+	down(&fs_info->uuid_tree_rescan_sem);
+	task = kthread_run(btrfs_uuid_scan_kthread, fs_info, "btrfs-uuid");
+	if (IS_ERR(task)) {
+		/* fs_info->update_uuid_tree_gen remains 0 in all error case */
+		pr_warn("btrfs: failed to start uuid_scan task\n");
+		up(&fs_info->uuid_tree_rescan_sem);
+		return PTR_ERR(task);
+	}
+
+	return 0;
+}
+
+int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info)
+{
+	struct task_struct *task;
+
+	down(&fs_info->uuid_tree_rescan_sem);
+	task = kthread_run(btrfs_uuid_rescan_kthread, fs_info, "btrfs-uuid");
+	if (IS_ERR(task)) {
+		/* fs_info->update_uuid_tree_gen remains 0 in all error case */
+		pr_warn("btrfs: failed to start uuid_rescan task\n");
+		up(&fs_info->uuid_tree_rescan_sem);
+		return PTR_ERR(task);
+	}
+
+	return 0;
+}
+
 /*
  * shrinking a device means finding all of the device extents past
  * the new size, and then following the back refs to the chunks.
@@ -4194,13 +4473,13 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len)
 	 * and exit, so return 1 so the callers don't try to use other copies.
 	 */
 	if (!em) {
-		btrfs_emerg(fs_info, "No mapping for %Lu-%Lu\n", logical,
+		btrfs_crit(fs_info, "No mapping for %Lu-%Lu\n", logical,
 			    logical+len);
 		return 1;
 	}
 
 	if (em->start > logical || em->start + em->len < logical) {
-		btrfs_emerg(fs_info, "Invalid mapping for %Lu-%Lu, got "
+		btrfs_crit(fs_info, "Invalid mapping for %Lu-%Lu, got "
 			    "%Lu-%Lu\n", logical, logical+len, em->start,
 			    em->start + em->len);
 		return 1;
@@ -4375,8 +4654,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 
 	if (!em) {
 		btrfs_crit(fs_info, "unable to find logical %llu len %llu",
-			(unsigned long long)logical,
-			(unsigned long long)*length);
+			logical, *length);
 		return -EINVAL;
 	}
 
@@ -4671,6 +4949,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 	}
 	bbio = kzalloc(btrfs_bio_size(num_alloc_stripes), GFP_NOFS);
 	if (!bbio) {
+		kfree(raid_map);
 		ret = -ENOMEM;
 		goto out;
 	}
@@ -5246,9 +5525,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
 
 	if (map_length < length) {
 		btrfs_crit(root->fs_info, "mapping failed logical %llu bio len %llu len %llu",
-			(unsigned long long)logical,
-			(unsigned long long)length,
-			(unsigned long long)map_length);
+			logical, length, map_length);
 		BUG();
 	}
 
@@ -5314,23 +5591,72 @@ static struct btrfs_device *add_missing_dev(struct btrfs_root *root,
 	struct btrfs_device *device;
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 
-	device = kzalloc(sizeof(*device), GFP_NOFS);
-	if (!device)
+	device = btrfs_alloc_device(NULL, &devid, dev_uuid);
+	if (IS_ERR(device))
 		return NULL;
-	list_add(&device->dev_list,
-		 &fs_devices->devices);
-	device->devid = devid;
-	device->work.func = pending_bios_fn;
+
+	list_add(&device->dev_list, &fs_devices->devices);
 	device->fs_devices = fs_devices;
-	device->missing = 1;
 	fs_devices->num_devices++;
+
+	device->missing = 1;
 	fs_devices->missing_devices++;
-	spin_lock_init(&device->io_lock);
-	INIT_LIST_HEAD(&device->dev_alloc_list);
-	memcpy(device->uuid, dev_uuid, BTRFS_UUID_SIZE);
+
 	return device;
 }
 
+/**
+ * btrfs_alloc_device - allocate struct btrfs_device
+ * @fs_info:	used only for generating a new devid, can be NULL if
+ *		devid is provided (i.e. @devid != NULL).
+ * @devid:	a pointer to devid for this device.  If NULL a new devid
+ *		is generated.
+ * @uuid:	a pointer to UUID for this device.  If NULL a new UUID
+ *		is generated.
+ *
+ * Return: a pointer to a new &struct btrfs_device on success; ERR_PTR()
+ * on error.  Returned struct is not linked onto any lists and can be
+ * destroyed with kfree() right away.
+ */
+struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
+					const u64 *devid,
+					const u8 *uuid)
+{
+	struct btrfs_device *dev;
+	u64 tmp;
+
+	if (!devid && !fs_info) {
+		WARN_ON(1);
+		return ERR_PTR(-EINVAL);
+	}
+
+	dev = __alloc_device();
+	if (IS_ERR(dev))
+		return dev;
+
+	if (devid)
+		tmp = *devid;
+	else {
+		int ret;
+
+		ret = find_next_devid(fs_info, &tmp);
+		if (ret) {
+			kfree(dev);
+			return ERR_PTR(ret);
+		}
+	}
+	dev->devid = tmp;
+
+	if (uuid)
+		memcpy(dev->uuid, uuid, BTRFS_UUID_SIZE);
+	else
+		generate_random_uuid(dev->uuid);
+
+	dev->work.func = pending_bios_fn;
+
+	return dev;
+}
+
 static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 			  struct extent_buffer *leaf,
 			  struct btrfs_chunk *chunk)
@@ -5437,7 +5763,7 @@ static void fill_device_from_item(struct extent_buffer *leaf,
 	WARN_ON(device->devid == BTRFS_DEV_REPLACE_DEVID);
 	device->is_tgtdev_for_dev_replace = 0;
 
-	ptr = (unsigned long)btrfs_device_uuid(dev_item);
+	ptr = btrfs_device_uuid(dev_item);
 	read_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
 }
 
@@ -5500,11 +5826,9 @@ static int read_one_dev(struct btrfs_root *root,
 	u8 dev_uuid[BTRFS_UUID_SIZE];
 
 	devid = btrfs_device_id(leaf, dev_item);
-	read_extent_buffer(leaf, dev_uuid,
-			   (unsigned long)btrfs_device_uuid(dev_item),
+	read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
 			   BTRFS_UUID_SIZE);
-	read_extent_buffer(leaf, fs_uuid,
-			   (unsigned long)btrfs_device_fsid(dev_item),
+	read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
 			   BTRFS_UUID_SIZE);
 
 	if (memcmp(fs_uuid, root->fs_info->fsid, BTRFS_UUID_SIZE)) {
@@ -5519,8 +5843,7 @@ static int read_one_dev(struct btrfs_root *root,
 			return -EIO;
 
 		if (!device) {
-			btrfs_warn(root->fs_info, "devid %llu missing",
-				(unsigned long long)devid);
+			btrfs_warn(root->fs_info, "devid %llu missing", devid);
 			device = add_missing_dev(root, devid, dev_uuid);
 			if (!device)
 				return -ENOMEM;
@@ -5644,14 +5967,15 @@ int btrfs_read_chunk_tree(struct btrfs_root *root)
 	mutex_lock(&uuid_mutex);
 	lock_chunks(root);
 
-	/* first we search for all of the device items, and then we
-	 * read in all of the chunk items.  This way we can create chunk
-	 * mappings that reference all of the devices that are afound
+	/*
+	 * Read all device items, and then all the chunk items. All
+	 * device items are found before any chunk item (their object id
+	 * is smaller than the lowest possible object id for a chunk
+	 * item - BTRFS_FIRST_CHUNK_TREE_OBJECTID).
 	 */
 	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
 	key.offset = 0;
 	key.type = 0;
-again:
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 	if (ret < 0)
 		goto error;
@@ -5667,17 +5991,13 @@ again:
 			break;
 		}
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
-		if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
-			if (found_key.objectid != BTRFS_DEV_ITEMS_OBJECTID)
-				break;
-			if (found_key.type == BTRFS_DEV_ITEM_KEY) {
-				struct btrfs_dev_item *dev_item;
-				dev_item = btrfs_item_ptr(leaf, slot,
+		if (found_key.type == BTRFS_DEV_ITEM_KEY) {
+			struct btrfs_dev_item *dev_item;
+			dev_item = btrfs_item_ptr(leaf, slot,
 						  struct btrfs_dev_item);
-				ret = read_one_dev(root, leaf, dev_item);
-				if (ret)
-					goto error;
-			}
+			ret = read_one_dev(root, leaf, dev_item);
+			if (ret)
+				goto error;
 		} else if (found_key.type == BTRFS_CHUNK_ITEM_KEY) {
 			struct btrfs_chunk *chunk;
 			chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
@@ -5687,11 +6007,6 @@ again:
 		}
 		path->slots[0]++;
 	}
-	if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
-		key.objectid = 0;
-		btrfs_release_path(path);
-		goto again;
-	}
 	ret = 0;
 error:
 	unlock_chunks(root);

fs/btrfs/volumes.h

@@ -152,6 +152,8 @@ struct btrfs_fs_devices {
 	int rotating;
 };
 
+#define BTRFS_BIO_INLINE_CSUM_SIZE	64
+
 /*
  * we need the mirror number and stripe index to be passed around
  * the call chain while we are processing end_io (especially errors).
@@ -161,9 +163,14 @@ struct btrfs_fs_devices {
  * we allocate are actually btrfs_io_bios.  We'll cram as much of
  * struct btrfs_bio as we can into this over time.
  */
+typedef void (btrfs_io_bio_end_io_t) (struct btrfs_io_bio *bio, int err);
 struct btrfs_io_bio {
 	unsigned long mirror_num;
 	unsigned long stripe_index;
+	u8 *csum;
+	u8 csum_inline[BTRFS_BIO_INLINE_CSUM_SIZE];
+	u8 *csum_allocated;
+	btrfs_io_bio_end_io_t *end_io;
 	struct bio bio;
 };
 
@@ -298,6 +305,9 @@ void btrfs_close_extra_devices(struct btrfs_fs_info *fs_info,
 int btrfs_find_device_missing_or_by_path(struct btrfs_root *root,
 					 char *device_path,
 					 struct btrfs_device **device);
+struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
+					const u64 *devid,
+					const u8 *uuid);
 int btrfs_rm_device(struct btrfs_root *root, char *device_path);
 void btrfs_cleanup_fs_uuids(void);
 int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len);
@@ -315,6 +325,8 @@ int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info);
 int btrfs_recover_balance(struct btrfs_fs_info *fs_info);
 int btrfs_pause_balance(struct btrfs_fs_info *fs_info);
 int btrfs_cancel_balance(struct btrfs_fs_info *fs_info);
+int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info);
+int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info);
 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset);
 int find_free_dev_extent(struct btrfs_trans_handle *trans,
 			 struct btrfs_device *device, u64 num_bytes,

include/trace/events/btrfs.h

@@ -439,7 +439,7 @@ TRACE_EVENT(btrfs_sync_fs,
 		{ BTRFS_UPDATE_DELAYED_HEAD, "UPDATE_DELAYED_HEAD" })
 			
 
-TRACE_EVENT(btrfs_delayed_tree_ref,
+DECLARE_EVENT_CLASS(btrfs_delayed_tree_ref,
 
 	TP_PROTO(struct btrfs_delayed_ref_node *ref,
 		 struct btrfs_delayed_tree_ref *full_ref,
@@ -481,7 +481,25 @@ TRACE_EVENT(btrfs_delayed_tree_ref,
 		  (unsigned long long)__entry->seq)
 );
 
-TRACE_EVENT(btrfs_delayed_data_ref,
+DEFINE_EVENT(btrfs_delayed_tree_ref,  add_delayed_tree_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_tree_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DEFINE_EVENT(btrfs_delayed_tree_ref,  run_delayed_tree_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_tree_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DECLARE_EVENT_CLASS(btrfs_delayed_data_ref,
 
 	TP_PROTO(struct btrfs_delayed_ref_node *ref,
 		 struct btrfs_delayed_data_ref *full_ref,
@@ -527,7 +545,25 @@ TRACE_EVENT(btrfs_delayed_data_ref,
 		  (unsigned long long)__entry->seq)
 );
 
-TRACE_EVENT(btrfs_delayed_ref_head,
+DEFINE_EVENT(btrfs_delayed_data_ref,  add_delayed_data_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_data_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DEFINE_EVENT(btrfs_delayed_data_ref,  run_delayed_data_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_data_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DECLARE_EVENT_CLASS(btrfs_delayed_ref_head,
 
 	TP_PROTO(struct btrfs_delayed_ref_node *ref,
 		 struct btrfs_delayed_ref_head *head_ref,
@@ -556,6 +592,24 @@ TRACE_EVENT(btrfs_delayed_ref_head,
 		  __entry->is_data)
 );
 
+DEFINE_EVENT(btrfs_delayed_ref_head,  add_delayed_ref_head,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_ref_head *head_ref,
+		 int action),
+
+	TP_ARGS(ref, head_ref, action)
+);
+
+DEFINE_EVENT(btrfs_delayed_ref_head,  run_delayed_ref_head,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_ref_head *head_ref,
+		 int action),
+
+	TP_ARGS(ref, head_ref, action)
+);
+
 #define show_chunk_type(type)					\
 	__print_flags(type, "|",				\
 		{ BTRFS_BLOCK_GROUP_DATA, 	"DATA"	},	\

include/uapi/linux/btrfs.h

@@ -305,6 +305,31 @@ struct btrfs_ioctl_clone_range_args {
 #define BTRFS_DEFRAG_RANGE_COMPRESS 1
 #define BTRFS_DEFRAG_RANGE_START_IO 2
 
+#define BTRFS_SAME_DATA_DIFFERS	1
+/* For extent-same ioctl */
+struct btrfs_ioctl_same_extent_info {
+	__s64 fd;		/* in - destination file */
+	__u64 logical_offset;	/* in - start of extent in destination */
+	__u64 bytes_deduped;	/* out - total # of bytes we were able
+				 * to dedupe from this file */
+	/* status of this dedupe operation:
+	 * 0 if dedup succeeds
+	 * < 0 for error
+	 * == BTRFS_SAME_DATA_DIFFERS if data differs
+	 */
+	__s32 status;		/* out - see above description */
+	__u32 reserved;
+};
+
+struct btrfs_ioctl_same_args {
+	__u64 logical_offset;	/* in - start of extent in source */
+	__u64 length;		/* in - length of extent */
+	__u16 dest_count;	/* in - total elements in info array */
+	__u16 reserved1;
+	__u32 reserved2;
+	struct btrfs_ioctl_same_extent_info info[0];
+};
+
 struct btrfs_ioctl_space_info {
 	__u64 flags;
 	__u64 total_bytes;
@@ -524,7 +549,7 @@ static inline char *btrfs_err_str(enum btrfs_err_code err_code)
 				   struct btrfs_ioctl_search_args)
 #define BTRFS_IOC_INO_LOOKUP _IOWR(BTRFS_IOCTL_MAGIC, 18, \
 				   struct btrfs_ioctl_ino_lookup_args)
-#define BTRFS_IOC_DEFAULT_SUBVOL _IOW(BTRFS_IOCTL_MAGIC, 19, u64)
+#define BTRFS_IOC_DEFAULT_SUBVOL _IOW(BTRFS_IOCTL_MAGIC, 19, __u64)
 #define BTRFS_IOC_SPACE_INFO _IOWR(BTRFS_IOCTL_MAGIC, 20, \
 				    struct btrfs_ioctl_space_args)
 #define BTRFS_IOC_START_SYNC _IOR(BTRFS_IOCTL_MAGIC, 24, __u64)
@@ -579,4 +604,7 @@ static inline char *btrfs_err_str(enum btrfs_err_code err_code)
 				      struct btrfs_ioctl_get_dev_stats)
 #define BTRFS_IOC_DEV_REPLACE _IOWR(BTRFS_IOCTL_MAGIC, 53, \
 				    struct btrfs_ioctl_dev_replace_args)
+#define BTRFS_IOC_FILE_EXTENT_SAME _IOWR(BTRFS_IOCTL_MAGIC, 54, \
+					 struct btrfs_ioctl_same_args)
+
 #endif /* _UAPI_LINUX_BTRFS_H */