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

Pull btrfs updates from Chris Mason:
 "The largest set of changes here come from Miao Xie.  He's cleaning up
  and improving read recovery/repair for raid, and has a number of
  related fixes.

  I've merged another set of fsync fixes from Filipe, and he's also
  improved the way we handle metadata write errors to make sure we force
  the FS readonly if things go wrong.

  Otherwise we have a collection of fixes and cleanups.  Dave Sterba
  gets a cookie for removing the most lines (thanks Dave)"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (139 commits)
  btrfs: Fix compile error when CONFIG_SECURITY is not set.
  Btrfs: fix compiles when CONFIG_BTRFS_FS_RUN_SANITY_TESTS is off
  btrfs: Make btrfs handle security mount options internally to avoid losing security label.
  Btrfs: send, don't delay dir move if there's a new parent inode
  btrfs: add more superblock checks
  Btrfs: fix race in WAIT_SYNC ioctl
  Btrfs: be aware of btree inode write errors to avoid fs corruption
  Btrfs: remove redundant btrfs_verify_qgroup_counts declaration.
  btrfs: fix shadow warning on cmp
  Btrfs: fix compilation errors under DEBUG
  Btrfs: fix crash of btrfs_release_extent_buffer_page
  Btrfs: add missing end_page_writeback on submit_extent_page failure
  btrfs: Fix the wrong condition judgment about subset extent map
  Btrfs: fix build_backref_tree issue with multiple shared blocks
  Btrfs: cleanup error handling in build_backref_tree
  btrfs: move checks for DUMMY_ROOT into a helper
  btrfs: new define for the inline extent data start
  btrfs: kill extent_buffer_page helper
  btrfs: drop constant param from btrfs_release_extent_buffer_page
  btrfs: hide typecast to definition of BTRFS_SEND_TRANS_STUB
  ...

fs/btrfs/async-thread.c

@@ -74,6 +74,7 @@ BTRFS_WORK_HELPER(endio_helper);
 BTRFS_WORK_HELPER(endio_meta_helper);
 BTRFS_WORK_HELPER(endio_meta_write_helper);
 BTRFS_WORK_HELPER(endio_raid56_helper);
+BTRFS_WORK_HELPER(endio_repair_helper);
 BTRFS_WORK_HELPER(rmw_helper);
 BTRFS_WORK_HELPER(endio_write_helper);
 BTRFS_WORK_HELPER(freespace_write_helper);
@@ -91,7 +92,7 @@ __btrfs_alloc_workqueue(const char *name, int flags, int max_active,
 {
 	struct __btrfs_workqueue *ret = kzalloc(sizeof(*ret), GFP_NOFS);
 
-	if (unlikely(!ret))
+	if (!ret)
 		return NULL;
 
 	ret->max_active = max_active;
@@ -115,7 +116,7 @@ __btrfs_alloc_workqueue(const char *name, int flags, int max_active,
 		ret->normal_wq = alloc_workqueue("%s-%s", flags,
 						 ret->max_active, "btrfs",
 						 name);
-	if (unlikely(!ret->normal_wq)) {
+	if (!ret->normal_wq) {
 		kfree(ret);
 		return NULL;
 	}
@@ -137,12 +138,12 @@ struct btrfs_workqueue *btrfs_alloc_workqueue(const char *name,
 {
 	struct btrfs_workqueue *ret = kzalloc(sizeof(*ret), GFP_NOFS);
 
-	if (unlikely(!ret))
+	if (!ret)
 		return NULL;
 
 	ret->normal = __btrfs_alloc_workqueue(name, flags & ~WQ_HIGHPRI,
 					      max_active, thresh);
-	if (unlikely(!ret->normal)) {
+	if (!ret->normal) {
 		kfree(ret);
 		return NULL;
 	}
@@ -150,7 +151,7 @@ struct btrfs_workqueue *btrfs_alloc_workqueue(const char *name,
 	if (flags & WQ_HIGHPRI) {
 		ret->high = __btrfs_alloc_workqueue(name, flags, max_active,
 						    thresh);
-		if (unlikely(!ret->high)) {
+		if (!ret->high) {
 			__btrfs_destroy_workqueue(ret->normal);
 			kfree(ret);
 			return NULL;

fs/btrfs/async-thread.h

@@ -53,6 +53,7 @@ BTRFS_WORK_HELPER_PROTO(endio_helper);
 BTRFS_WORK_HELPER_PROTO(endio_meta_helper);
 BTRFS_WORK_HELPER_PROTO(endio_meta_write_helper);
 BTRFS_WORK_HELPER_PROTO(endio_raid56_helper);
+BTRFS_WORK_HELPER_PROTO(endio_repair_helper);
 BTRFS_WORK_HELPER_PROTO(rmw_helper);
 BTRFS_WORK_HELPER_PROTO(endio_write_helper);
 BTRFS_WORK_HELPER_PROTO(freespace_write_helper);

fs/btrfs/backref.c

@@ -25,6 +25,9 @@
 #include "delayed-ref.h"
 #include "locking.h"
 
+/* Just an arbitrary number so we can be sure this happened */
+#define BACKREF_FOUND_SHARED 6
+
 struct extent_inode_elem {
 	u64 inum;
 	u64 offset;
@@ -377,7 +380,8 @@ out:
 static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 				   struct btrfs_path *path, u64 time_seq,
 				   struct list_head *head,
-				   const u64 *extent_item_pos, u64 total_refs)
+				   const u64 *extent_item_pos, u64 total_refs,
+				   u64 root_objectid)
 {
 	int err;
 	int ret = 0;
@@ -402,6 +406,10 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 			continue;
 		if (ref->count == 0)
 			continue;
+		if (root_objectid && ref->root_id != root_objectid) {
+			ret = BACKREF_FOUND_SHARED;
+			goto out;
+		}
 		err = __resolve_indirect_ref(fs_info, path, time_seq, ref,
 					     parents, extent_item_pos,
 					     total_refs);
@@ -482,7 +490,7 @@ static int __add_missing_keys(struct btrfs_fs_info *fs_info,
 			continue;
 		BUG_ON(!ref->wanted_disk_byte);
 		eb = read_tree_block(fs_info->tree_root, ref->wanted_disk_byte,
-				     fs_info->tree_root->leafsize, 0);
+				     0);
 		if (!eb || !extent_buffer_uptodate(eb)) {
 			free_extent_buffer(eb);
 			return -EIO;
@@ -561,7 +569,8 @@ static void __merge_refs(struct list_head *head, int mode)
  * smaller or equal that seq to the list
  */
 static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
-			      struct list_head *prefs, u64 *total_refs)
+			      struct list_head *prefs, u64 *total_refs,
+			      u64 inum)
 {
 	struct btrfs_delayed_extent_op *extent_op = head->extent_op;
 	struct rb_node *n = &head->node.rb_node;
@@ -625,6 +634,16 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			key.objectid = ref->objectid;
 			key.type = BTRFS_EXTENT_DATA_KEY;
 			key.offset = ref->offset;
+
+			/*
+			 * Found a inum that doesn't match our known inum, we
+			 * know it's shared.
+			 */
+			if (inum && ref->objectid != inum) {
+				ret = BACKREF_FOUND_SHARED;
+				break;
+			}
+
 			ret = __add_prelim_ref(prefs, ref->root, &key, 0, 0,
 					       node->bytenr,
 					       node->ref_mod * sgn, GFP_ATOMIC);
@@ -659,7 +678,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 			     struct btrfs_path *path, u64 bytenr,
 			     int *info_level, struct list_head *prefs,
-			     u64 *total_refs)
+			     u64 *total_refs, u64 inum)
 {
 	int ret = 0;
 	int slot;
@@ -744,6 +763,12 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 								      dref);
 			key.type = BTRFS_EXTENT_DATA_KEY;
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
+
+			if (inum && key.objectid != inum) {
+				ret = BACKREF_FOUND_SHARED;
+				break;
+			}
+
 			root = btrfs_extent_data_ref_root(leaf, dref);
 			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
 					       bytenr, count, GFP_NOFS);
@@ -765,7 +790,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
  */
 static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 			    struct btrfs_path *path, u64 bytenr,
-			    int info_level, struct list_head *prefs)
+			    int info_level, struct list_head *prefs, u64 inum)
 {
 	struct btrfs_root *extent_root = fs_info->extent_root;
 	int ret;
@@ -827,6 +852,12 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 								      dref);
 			key.type = BTRFS_EXTENT_DATA_KEY;
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
+
+			if (inum && key.objectid != inum) {
+				ret = BACKREF_FOUND_SHARED;
+				break;
+			}
+
 			root = btrfs_extent_data_ref_root(leaf, dref);
 			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
 					       bytenr, count, GFP_NOFS);
@@ -854,7 +885,8 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 static int find_parent_nodes(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info, u64 bytenr,
 			     u64 time_seq, struct ulist *refs,
-			     struct ulist *roots, const u64 *extent_item_pos)
+			     struct ulist *roots, const u64 *extent_item_pos,
+			     u64 root_objectid, u64 inum)
 {
 	struct btrfs_key key;
 	struct btrfs_path *path;
@@ -929,7 +961,8 @@ again:
 			}
 			spin_unlock(&delayed_refs->lock);
 			ret = __add_delayed_refs(head, time_seq,
-						 &prefs_delayed, &total_refs);
+						 &prefs_delayed, &total_refs,
+						 inum);
 			mutex_unlock(&head->mutex);
 			if (ret)
 				goto out;
@@ -951,11 +984,11 @@ again:
 		     key.type == BTRFS_METADATA_ITEM_KEY)) {
 			ret = __add_inline_refs(fs_info, path, bytenr,
 						&info_level, &prefs,
-						&total_refs);
+						&total_refs, inum);
 			if (ret)
 				goto out;
 			ret = __add_keyed_refs(fs_info, path, bytenr,
-					       info_level, &prefs);
+					       info_level, &prefs, inum);
 			if (ret)
 				goto out;
 		}
@@ -971,7 +1004,8 @@ again:
 	__merge_refs(&prefs, 1);
 
 	ret = __resolve_indirect_refs(fs_info, path, time_seq, &prefs,
-				      extent_item_pos, total_refs);
+				      extent_item_pos, total_refs,
+				      root_objectid);
 	if (ret)
 		goto out;
 
@@ -981,6 +1015,11 @@ again:
 		ref = list_first_entry(&prefs, struct __prelim_ref, list);
 		WARN_ON(ref->count < 0);
 		if (roots && ref->count && ref->root_id && ref->parent == 0) {
+			if (root_objectid && ref->root_id != root_objectid) {
+				ret = BACKREF_FOUND_SHARED;
+				goto out;
+			}
+
 			/* no parent == root of tree */
 			ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS);
 			if (ret < 0)
@@ -989,12 +1028,10 @@ again:
 		if (ref->count && ref->parent) {
 			if (extent_item_pos && !ref->inode_list &&
 			    ref->level == 0) {
-				u32 bsz;
 				struct extent_buffer *eb;
-				bsz = btrfs_level_size(fs_info->extent_root,
-							ref->level);
+
 				eb = read_tree_block(fs_info->extent_root,
-							   ref->parent, bsz, 0);
+							   ref->parent, 0);
 				if (!eb || !extent_buffer_uptodate(eb)) {
 					free_extent_buffer(eb);
 					ret = -EIO;
@@ -1087,7 +1124,7 @@ static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 
 	ret = find_parent_nodes(trans, fs_info, bytenr,
-				time_seq, *leafs, NULL, extent_item_pos);
+				time_seq, *leafs, NULL, extent_item_pos, 0, 0);
 	if (ret < 0 && ret != -ENOENT) {
 		free_leaf_list(*leafs);
 		return ret;
@@ -1130,7 +1167,7 @@ static int __btrfs_find_all_roots(struct btrfs_trans_handle *trans,
 	ULIST_ITER_INIT(&uiter);
 	while (1) {
 		ret = find_parent_nodes(trans, fs_info, bytenr,
-					time_seq, tmp, *roots, NULL);
+					time_seq, tmp, *roots, NULL, 0, 0);
 		if (ret < 0 && ret != -ENOENT) {
 			ulist_free(tmp);
 			ulist_free(*roots);
@@ -1161,6 +1198,54 @@ int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+int btrfs_check_shared(struct btrfs_trans_handle *trans,
+		       struct btrfs_fs_info *fs_info, u64 root_objectid,
+		       u64 inum, u64 bytenr)
+{
+	struct ulist *tmp = NULL;
+	struct ulist *roots = NULL;
+	struct ulist_iterator uiter;
+	struct ulist_node *node;
+	struct seq_list elem = {};
+	int ret = 0;
+
+	tmp = ulist_alloc(GFP_NOFS);
+	roots = ulist_alloc(GFP_NOFS);
+	if (!tmp || !roots) {
+		ulist_free(tmp);
+		ulist_free(roots);
+		return -ENOMEM;
+	}
+
+	if (trans)
+		btrfs_get_tree_mod_seq(fs_info, &elem);
+	else
+		down_read(&fs_info->commit_root_sem);
+	ULIST_ITER_INIT(&uiter);
+	while (1) {
+		ret = find_parent_nodes(trans, fs_info, bytenr, elem.seq, tmp,
+					roots, NULL, root_objectid, inum);
+		if (ret == BACKREF_FOUND_SHARED) {
+			ret = 1;
+			break;
+		}
+		if (ret < 0 && ret != -ENOENT)
+			break;
+		node = ulist_next(tmp, &uiter);
+		if (!node)
+			break;
+		bytenr = node->val;
+		cond_resched();
+	}
+	if (trans)
+		btrfs_put_tree_mod_seq(fs_info, &elem);
+	else
+		up_read(&fs_info->commit_root_sem);
+	ulist_free(tmp);
+	ulist_free(roots);
+	return ret;
+}
+
 /*
  * this makes the path point to (inum INODE_ITEM ioff)
  */
@@ -1193,7 +1278,7 @@ int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
 	unsigned long ptr;
 
 	key.objectid = inode_objectid;
-	btrfs_set_key_type(&key, BTRFS_INODE_EXTREF_KEY);
+	key.type = BTRFS_INODE_EXTREF_KEY;
 	key.offset = start_off;
 
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
@@ -1233,7 +1318,7 @@ int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
 		ret = -ENOENT;
 		if (found_key.objectid != inode_objectid)
 			break;
-		if (btrfs_key_type(&found_key) != BTRFS_INODE_EXTREF_KEY)
+		if (found_key.type != BTRFS_INODE_EXTREF_KEY)
 			break;
 
 		ret = 0;
@@ -1366,7 +1451,7 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 	}
 	btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]);
 	if (found_key->type == BTRFS_METADATA_ITEM_KEY)
-		size = fs_info->extent_root->leafsize;
+		size = fs_info->extent_root->nodesize;
 	else if (found_key->type == BTRFS_EXTENT_ITEM_KEY)
 		size = found_key->offset;
 

fs/btrfs/backref.h

@@ -71,6 +71,9 @@ int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
 			  u64 start_off, struct btrfs_path *path,
 			  struct btrfs_inode_extref **ret_extref,
 			  u64 *found_off);
+int btrfs_check_shared(struct btrfs_trans_handle *trans,
+		       struct btrfs_fs_info *fs_info, u64 root_objectid,
+		       u64 inum, u64 bytenr);
 
 int __init btrfs_prelim_ref_init(void);
 void btrfs_prelim_ref_exit(void);

fs/btrfs/btrfs_inode.h

@@ -44,6 +44,17 @@
 #define BTRFS_INODE_IN_DELALLOC_LIST		9
 #define BTRFS_INODE_READDIO_NEED_LOCK		10
 #define BTRFS_INODE_HAS_PROPS		        11
+/*
+ * The following 3 bits are meant only for the btree inode.
+ * When any of them is set, it means an error happened while writing an
+ * extent buffer belonging to:
+ * 1) a non-log btree
+ * 2) a log btree and first log sub-transaction
+ * 3) a log btree and second log sub-transaction
+ */
+#define BTRFS_INODE_BTREE_ERR		        12
+#define BTRFS_INODE_BTREE_LOG1_ERR		13
+#define BTRFS_INODE_BTREE_LOG2_ERR		14
 
 /* in memory btrfs inode */
 struct btrfs_inode {
@@ -120,6 +131,12 @@ struct btrfs_inode {
 	 */
 	u64 delalloc_bytes;
 
+	/*
+	 * total number of bytes pending defrag, used by stat to check whether
+	 * it needs COW.
+	 */
+	u64 defrag_bytes;
+
 	/*
 	 * the size of the file stored in the metadata on disk.  data=ordered
 	 * means the in-memory i_size might be larger than the size on disk
@@ -248,8 +265,11 @@ static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
 	return 0;
 }
 
+#define BTRFS_DIO_ORIG_BIO_SUBMITTED	0x1
+
 struct btrfs_dio_private {
 	struct inode *inode;
+	unsigned long flags;
 	u64 logical_offset;
 	u64 disk_bytenr;
 	u64 bytes;
@@ -266,7 +286,12 @@ struct btrfs_dio_private {
 
 	/* dio_bio came from fs/direct-io.c */
 	struct bio *dio_bio;
-	u8 csum[0];
+
+	/*
+	 * The original bio may be splited to several sub-bios, this is
+	 * done during endio of sub-bios
+	 */
+	int (*subio_endio)(struct inode *, struct btrfs_io_bio *, int);
 };
 
 /*

fs/btrfs/check-integrity.c

@@ -807,7 +807,7 @@ static int btrfsic_process_superblock_dev_mirror(
 
 	/* super block bytenr is always the unmapped device bytenr */
 	dev_bytenr = btrfs_sb_offset(superblock_mirror_num);
-	if (dev_bytenr + BTRFS_SUPER_INFO_SIZE > device->total_bytes)
+	if (dev_bytenr + BTRFS_SUPER_INFO_SIZE > device->commit_total_bytes)
 		return -1;
 	bh = __bread(superblock_bdev, dev_bytenr / 4096,
 		     BTRFS_SUPER_INFO_SIZE);
@@ -820,7 +820,6 @@ static int btrfsic_process_superblock_dev_mirror(
 	    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 ||
 	    btrfs_super_sectorsize(super_tmp) != state->datablock_size) {
 		brelse(bh);
 		return 0;
@@ -1252,8 +1251,7 @@ static void btrfsic_read_from_block_data(
 
 	while (len > 0) {
 		cur = min(len, ((size_t)PAGE_CACHE_SIZE - offset_in_page));
-		BUG_ON(i >= (block_ctx->len + PAGE_CACHE_SIZE - 1) >>
-			    PAGE_CACHE_SHIFT);
+		BUG_ON(i >= DIV_ROUND_UP(block_ctx->len, PAGE_CACHE_SIZE));
 		kaddr = block_ctx->datav[i];
 		memcpy(dst, kaddr + offset_in_page, cur);
 
@@ -3120,24 +3118,12 @@ int btrfsic_mount(struct btrfs_root *root,
 	struct list_head *dev_head = &fs_devices->devices;
 	struct btrfs_device *device;
 
-	if (root->nodesize != root->leafsize) {
-		printk(KERN_INFO
-		       "btrfsic: cannot handle nodesize %d != leafsize %d!\n",
-		       root->nodesize, root->leafsize);
-		return -1;
-	}
 	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, 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, 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",

fs/btrfs/compression.c

@@ -91,8 +91,7 @@ static inline int compressed_bio_size(struct btrfs_root *root,
 	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
 
 	return sizeof(struct compressed_bio) +
-		((disk_size + root->sectorsize - 1) / root->sectorsize) *
-		csum_size;
+		(DIV_ROUND_UP(disk_size, root->sectorsize)) * csum_size;
 }
 
 static struct bio *compressed_bio_alloc(struct block_device *bdev,
@@ -389,7 +388,8 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
 			 * freed before we're done setting it up
 			 */
 			atomic_inc(&cb->pending_bios);
-			ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
+			ret = btrfs_bio_wq_end_io(root->fs_info, bio,
+					BTRFS_WQ_ENDIO_DATA);
 			BUG_ON(ret); /* -ENOMEM */
 
 			if (!skip_sum) {
@@ -420,7 +420,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
 	}
 	bio_get(bio);
 
-	ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
+	ret = btrfs_bio_wq_end_io(root->fs_info, bio, BTRFS_WQ_ENDIO_DATA);
 	BUG_ON(ret); /* -ENOMEM */
 
 	if (!skip_sum) {
@@ -615,8 +615,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	cb->compress_type = extent_compress_type(bio_flags);
 	cb->orig_bio = bio;
 
-	nr_pages = (compressed_len + PAGE_CACHE_SIZE - 1) /
-				 PAGE_CACHE_SIZE;
+	nr_pages = DIV_ROUND_UP(compressed_len, PAGE_CACHE_SIZE);
 	cb->compressed_pages = kzalloc(sizeof(struct page *) * nr_pages,
 				       GFP_NOFS);
 	if (!cb->compressed_pages)
@@ -670,7 +669,8 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 		    PAGE_CACHE_SIZE) {
 			bio_get(comp_bio);
 
-			ret = btrfs_bio_wq_end_io(root->fs_info, comp_bio, 0);
+			ret = btrfs_bio_wq_end_io(root->fs_info, comp_bio,
+					BTRFS_WQ_ENDIO_DATA);
 			BUG_ON(ret); /* -ENOMEM */
 
 			/*
@@ -686,8 +686,8 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 							comp_bio, sums);
 				BUG_ON(ret); /* -ENOMEM */
 			}
-			sums += (comp_bio->bi_iter.bi_size +
-				 root->sectorsize - 1) / root->sectorsize;
+			sums += DIV_ROUND_UP(comp_bio->bi_iter.bi_size,
+					     root->sectorsize);
 
 			ret = btrfs_map_bio(root, READ, comp_bio,
 					    mirror_num, 0);
@@ -708,7 +708,8 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	}
 	bio_get(comp_bio);
 
-	ret = btrfs_bio_wq_end_io(root->fs_info, comp_bio, 0);
+	ret = btrfs_bio_wq_end_io(root->fs_info, comp_bio,
+			BTRFS_WQ_ENDIO_DATA);
 	BUG_ON(ret); /* -ENOMEM */
 
 	if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)) {

fs/btrfs/ctree.c

@@ -258,9 +258,8 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
 	else
 		btrfs_node_key(buf, &disk_key, 0);
 
-	cow = btrfs_alloc_free_block(trans, root, buf->len, 0,
-				     new_root_objectid, &disk_key, level,
-				     buf->start, 0);
+	cow = btrfs_alloc_tree_block(trans, root, 0, new_root_objectid,
+			&disk_key, level, buf->start, 0);
 	if (IS_ERR(cow))
 		return PTR_ERR(cow);
 
@@ -1133,9 +1132,9 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 	} else
 		parent_start = 0;
 
-	cow = btrfs_alloc_free_block(trans, root, buf->len, parent_start,
-				     root->root_key.objectid, &disk_key,
-				     level, search_start, empty_size);
+	cow = btrfs_alloc_tree_block(trans, root, parent_start,
+			root->root_key.objectid, &disk_key, level,
+			search_start, empty_size);
 	if (IS_ERR(cow))
 		return PTR_ERR(cow);
 
@@ -1425,7 +1424,6 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 	struct tree_mod_root *old_root = NULL;
 	u64 old_generation = 0;
 	u64 logical;
-	u32 blocksize;
 
 	eb_root = btrfs_read_lock_root_node(root);
 	tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
@@ -1444,8 +1442,7 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 	if (old_root && tm && tm->op != MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
 		btrfs_tree_read_unlock(eb_root);
 		free_extent_buffer(eb_root);
-		blocksize = btrfs_level_size(root, old_root->level);
-		old = read_tree_block(root, logical, blocksize, 0);
+		old = read_tree_block(root, logical, 0);
 		if (WARN_ON(!old || !extent_buffer_uptodate(old))) {
 			free_extent_buffer(old);
 			btrfs_warn(root->fs_info,
@@ -1506,10 +1503,9 @@ static inline int should_cow_block(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   struct extent_buffer *buf)
 {
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state)))
+	if (btrfs_test_is_dummy_root(root))
 		return 0;
-#endif
+
 	/* ensure we can see the force_cow */
 	smp_rmb();
 
@@ -1651,7 +1647,7 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
 	WARN_ON(trans->transid != root->fs_info->generation);
 
 	parent_nritems = btrfs_header_nritems(parent);
-	blocksize = btrfs_level_size(root, parent_level - 1);
+	blocksize = root->nodesize;
 	end_slot = parent_nritems;
 
 	if (parent_nritems == 1)
@@ -1685,15 +1681,14 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
 			continue;
 		}
 
-		cur = btrfs_find_tree_block(root, blocknr, blocksize);
+		cur = btrfs_find_tree_block(root, blocknr);
 		if (cur)
 			uptodate = btrfs_buffer_uptodate(cur, gen, 0);
 		else
 			uptodate = 0;
 		if (!cur || !uptodate) {
 			if (!cur) {
-				cur = read_tree_block(root, blocknr,
-							 blocksize, gen);
+				cur = read_tree_block(root, blocknr, gen);
 				if (!cur || !extent_buffer_uptodate(cur)) {
 					free_extent_buffer(cur);
 					return -EIO;
@@ -1872,7 +1867,6 @@ static noinline struct extent_buffer *read_node_slot(struct btrfs_root *root,
 	BUG_ON(level == 0);
 
 	eb = read_tree_block(root, btrfs_node_blockptr(parent, slot),
-			     btrfs_level_size(root, level - 1),
 			     btrfs_node_ptr_generation(parent, slot));
 	if (eb && !extent_buffer_uptodate(eb)) {
 		free_extent_buffer(eb);
@@ -2267,8 +2261,8 @@ static void reada_for_search(struct btrfs_root *root,
 	node = path->nodes[level];
 
 	search = btrfs_node_blockptr(node, slot);
-	blocksize = btrfs_level_size(root, level - 1);
-	eb = btrfs_find_tree_block(root, search, blocksize);
+	blocksize = root->nodesize;
+	eb = btrfs_find_tree_block(root, search);
 	if (eb) {
 		free_extent_buffer(eb);
 		return;
@@ -2298,7 +2292,7 @@ static void reada_for_search(struct btrfs_root *root,
 		if ((search <= target && target - search <= 65536) ||
 		    (search > target && search - target <= 65536)) {
 			gen = btrfs_node_ptr_generation(node, nr);
-			readahead_tree_block(root, search, blocksize, gen);
+			readahead_tree_block(root, search, blocksize);
 			nread += blocksize;
 		}
 		nscan++;
@@ -2325,12 +2319,12 @@ static noinline void reada_for_balance(struct btrfs_root *root,
 
 	nritems = btrfs_header_nritems(parent);
 	slot = path->slots[level + 1];
-	blocksize = btrfs_level_size(root, level);
+	blocksize = root->nodesize;
 
 	if (slot > 0) {
 		block1 = btrfs_node_blockptr(parent, slot - 1);
 		gen = btrfs_node_ptr_generation(parent, slot - 1);
-		eb = btrfs_find_tree_block(root, block1, blocksize);
+		eb = btrfs_find_tree_block(root, block1);
 		/*
 		 * if we get -eagain from btrfs_buffer_uptodate, we
 		 * don't want to return eagain here.  That will loop
@@ -2343,16 +2337,16 @@ static noinline void reada_for_balance(struct btrfs_root *root,
 	if (slot + 1 < nritems) {
 		block2 = btrfs_node_blockptr(parent, slot + 1);
 		gen = btrfs_node_ptr_generation(parent, slot + 1);
-		eb = btrfs_find_tree_block(root, block2, blocksize);
+		eb = btrfs_find_tree_block(root, block2);
 		if (eb && btrfs_buffer_uptodate(eb, gen, 1) != 0)
 			block2 = 0;
 		free_extent_buffer(eb);
 	}
 
 	if (block1)
-		readahead_tree_block(root, block1, blocksize, 0);
+		readahead_tree_block(root, block1, blocksize);
 	if (block2)
-		readahead_tree_block(root, block2, blocksize, 0);
+		readahead_tree_block(root, block2, blocksize);
 }
 
 
@@ -2454,16 +2448,14 @@ read_block_for_search(struct btrfs_trans_handle *trans,
 {
 	u64 blocknr;
 	u64 gen;
-	u32 blocksize;
 	struct extent_buffer *b = *eb_ret;
 	struct extent_buffer *tmp;
 	int ret;
 
 	blocknr = btrfs_node_blockptr(b, slot);
 	gen = btrfs_node_ptr_generation(b, slot);
-	blocksize = btrfs_level_size(root, level - 1);
 
-	tmp = btrfs_find_tree_block(root, blocknr, blocksize);
+	tmp = btrfs_find_tree_block(root, blocknr);
 	if (tmp) {
 		/* first we do an atomic uptodate check */
 		if (btrfs_buffer_uptodate(tmp, gen, 1) > 0) {
@@ -2507,7 +2499,7 @@ read_block_for_search(struct btrfs_trans_handle *trans,
 	btrfs_release_path(p);
 
 	ret = -EAGAIN;
-	tmp = read_tree_block(root, blocknr, blocksize, 0);
+	tmp = read_tree_block(root, blocknr, 0);
 	if (tmp) {
 		/*
 		 * If the read above didn't mark this buffer up to date,
@@ -2792,8 +2784,6 @@ again:
 			if (!should_cow_block(trans, root, b))
 				goto cow_done;
 
-			btrfs_set_path_blocking(p);
-
 			/*
 			 * must have write locks on this node and the
 			 * parent
@@ -2807,6 +2797,7 @@ again:
 				goto again;
 			}
 
+			btrfs_set_path_blocking(p);
 			err = btrfs_cow_block(trans, root, b,
 					      p->nodes[level + 1],
 					      p->slots[level + 1], &b);
@@ -3362,9 +3353,8 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
 	else
 		btrfs_node_key(lower, &lower_key, 0);
 
-	c = btrfs_alloc_free_block(trans, root, root->nodesize, 0,
-				   root->root_key.objectid, &lower_key,
-				   level, root->node->start, 0);
+	c = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
+				   &lower_key, level, root->node->start, 0);
 	if (IS_ERR(c))
 		return PTR_ERR(c);
 
@@ -3502,9 +3492,8 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 	mid = (c_nritems + 1) / 2;
 	btrfs_node_key(c, &disk_key, mid);
 
-	split = btrfs_alloc_free_block(trans, root, root->nodesize, 0,
-					root->root_key.objectid,
-					&disk_key, level, c->start, 0);
+	split = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
+			&disk_key, level, c->start, 0);
 	if (IS_ERR(split))
 		return PTR_ERR(split);
 
@@ -4282,13 +4271,12 @@ again:
 	else
 		btrfs_item_key(l, &disk_key, mid);
 
-	right = btrfs_alloc_free_block(trans, root, root->leafsize, 0,
-					root->root_key.objectid,
-					&disk_key, 0, l->start, 0);
+	right = btrfs_alloc_tree_block(trans, root, 0, root->root_key.objectid,
+			&disk_key, 0, l->start, 0);
 	if (IS_ERR(right))
 		return PTR_ERR(right);
 
-	root_add_used(root, root->leafsize);
+	root_add_used(root, root->nodesize);
 
 	memset_extent_buffer(right, 0, 0, sizeof(struct btrfs_header));
 	btrfs_set_header_bytenr(right, right->start);
@@ -4626,8 +4614,7 @@ void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
 				ptr = btrfs_item_ptr_offset(leaf, slot);
 				memmove_extent_buffer(leaf, ptr,
 				      (unsigned long)fi,
-				      offsetof(struct btrfs_file_extent_item,
-						 disk_bytenr));
+				      BTRFS_FILE_EXTENT_INLINE_DATA_START);
 			}
 		}
 
@@ -4738,6 +4725,12 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 	int slot;
 	struct btrfs_map_token token;
 
+	if (path->slots[0] == 0) {
+		btrfs_cpu_key_to_disk(&disk_key, cpu_key);
+		fixup_low_keys(root, path, &disk_key, 1);
+	}
+	btrfs_unlock_up_safe(path, 1);
+
 	btrfs_init_map_token(&token);
 
 	leaf = path->nodes[0];
@@ -4798,12 +4791,6 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 	}
 
 	btrfs_set_header_nritems(leaf, nritems + nr);
-
-	if (slot == 0) {
-		btrfs_cpu_key_to_disk(&disk_key, cpu_key);
-		fixup_low_keys(root, path, &disk_key, 1);
-	}
-	btrfs_unlock_up_safe(path, 1);
 	btrfs_mark_buffer_dirty(leaf);
 
 	if (btrfs_leaf_free_space(root, leaf) < 0) {
@@ -5145,8 +5132,9 @@ int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key,
 	u32 nritems;
 	int level;
 	int ret = 1;
+	int keep_locks = path->keep_locks;
 
-	WARN_ON(!path->keep_locks);
+	path->keep_locks = 1;
 again:
 	cur = btrfs_read_lock_root_node(root);
 	level = btrfs_header_level(cur);
@@ -5210,7 +5198,6 @@ find_next_key:
 		path->slots[level] = slot;
 		if (level == path->lowest_level) {
 			ret = 0;
-			unlock_up(path, level, 1, 0, NULL);
 			goto out;
 		}
 		btrfs_set_path_blocking(path);
@@ -5225,9 +5212,12 @@ find_next_key:
 		btrfs_clear_path_blocking(path, NULL, 0);
 	}
 out:
-	if (ret == 0)
+	path->keep_locks = keep_locks;
+	if (ret == 0) {
+		btrfs_unlock_up_safe(path, path->lowest_level + 1);
+		btrfs_set_path_blocking(path);
 		memcpy(min_key, &found_key, sizeof(found_key));
-	btrfs_set_path_blocking(path);
+	}
 	return ret;
 }
 
@@ -5375,7 +5365,7 @@ int btrfs_compare_trees(struct btrfs_root *left_root,
 		goto out;
 	}
 
-	tmp_buf = kmalloc(left_root->leafsize, GFP_NOFS);
+	tmp_buf = kmalloc(left_root->nodesize, GFP_NOFS);
 	if (!tmp_buf) {
 		ret = -ENOMEM;
 		goto out;
@@ -5520,18 +5510,18 @@ int btrfs_compare_trees(struct btrfs_root *left_root,
 					goto out;
 				advance_right = ADVANCE;
 			} else {
-				enum btrfs_compare_tree_result cmp;
+				enum btrfs_compare_tree_result result;
 
 				WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
 				ret = tree_compare_item(left_root, left_path,
 						right_path, tmp_buf);
 				if (ret)
-					cmp = BTRFS_COMPARE_TREE_CHANGED;
+					result = BTRFS_COMPARE_TREE_CHANGED;
 				else
-					cmp = BTRFS_COMPARE_TREE_SAME;
+					result = BTRFS_COMPARE_TREE_SAME;
 				ret = changed_cb(left_root, right_root,
 						 left_path, right_path,
-						 &left_key, cmp, ctx);
+						 &left_key, result, ctx);
 				if (ret < 0)
 					goto out;
 				advance_left = ADVANCE;

fs/btrfs/ctree.h

@@ -34,6 +34,7 @@
 #include <linux/pagemap.h>
 #include <linux/btrfs.h>
 #include <linux/workqueue.h>
+#include <linux/security.h>
 #include "extent_io.h"
 #include "extent_map.h"
 #include "async-thread.h"
@@ -62,13 +63,6 @@ struct btrfs_ordered_sum;
 
 #define BTRFS_COMPAT_EXTENT_TREE_V0
 
-/*
- * files bigger than this get some pre-flushing when they are added
- * to the ordered operations list.  That way we limit the total
- * work done by the commit
- */
-#define BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT (8 * 1024 * 1024)
-
 /* holds pointers to all of the tree roots */
 #define BTRFS_ROOT_TREE_OBJECTID 1ULL
 
@@ -391,10 +385,12 @@ struct btrfs_header {
 				      sizeof(struct btrfs_header)) / \
 				     sizeof(struct btrfs_key_ptr))
 #define __BTRFS_LEAF_DATA_SIZE(bs) ((bs) - sizeof(struct btrfs_header))
-#define BTRFS_LEAF_DATA_SIZE(r) (__BTRFS_LEAF_DATA_SIZE(r->leafsize))
+#define BTRFS_LEAF_DATA_SIZE(r) (__BTRFS_LEAF_DATA_SIZE(r->nodesize))
+#define BTRFS_FILE_EXTENT_INLINE_DATA_START		\
+		(offsetof(struct btrfs_file_extent_item, disk_bytenr))
 #define BTRFS_MAX_INLINE_DATA_SIZE(r) (BTRFS_LEAF_DATA_SIZE(r) - \
 					sizeof(struct btrfs_item) - \
-					sizeof(struct btrfs_file_extent_item))
+					BTRFS_FILE_EXTENT_INLINE_DATA_START)
 #define BTRFS_MAX_XATTR_SIZE(r)	(BTRFS_LEAF_DATA_SIZE(r) - \
 				 sizeof(struct btrfs_item) -\
 				 sizeof(struct btrfs_dir_item))
@@ -474,7 +470,7 @@ struct btrfs_super_block {
 	__le64 num_devices;
 	__le32 sectorsize;
 	__le32 nodesize;
-	__le32 leafsize;
+	__le32 __unused_leafsize;
 	__le32 stripesize;
 	__le32 sys_chunk_array_size;
 	__le64 chunk_root_generation;
@@ -903,6 +899,8 @@ struct btrfs_file_extent_item {
 	/*
 	 * disk space consumed by the extent, checksum blocks are included
 	 * in these numbers
+	 *
+	 * At this offset in the structure, the inline extent data start.
 	 */
 	__le64 disk_bytenr;
 	__le64 disk_num_bytes;
@@ -1305,8 +1303,8 @@ struct btrfs_block_group_cache {
 	 */
 	struct list_head cluster_list;
 
-	/* For delayed block group creation */
-	struct list_head new_bg_list;
+	/* For delayed block group creation or deletion of empty block groups */
+	struct list_head bg_list;
 };
 
 /* delayed seq elem */
@@ -1545,6 +1543,7 @@ struct btrfs_fs_info {
 	struct btrfs_workqueue *endio_workers;
 	struct btrfs_workqueue *endio_meta_workers;
 	struct btrfs_workqueue *endio_raid56_workers;
+	struct btrfs_workqueue *endio_repair_workers;
 	struct btrfs_workqueue *rmw_workers;
 	struct btrfs_workqueue *endio_meta_write_workers;
 	struct btrfs_workqueue *endio_write_workers;
@@ -1574,6 +1573,7 @@ struct btrfs_fs_info {
 	int do_barriers;
 	int closing;
 	int log_root_recovering;
+	int open;
 
 	u64 total_pinned;
 
@@ -1723,6 +1723,12 @@ struct btrfs_fs_info {
 
 	/* Used to reclaim the metadata space in the background. */
 	struct work_struct async_reclaim_work;
+
+	spinlock_t unused_bgs_lock;
+	struct list_head unused_bgs;
+
+	/* For btrfs to record security options */
+	struct security_mnt_opts security_opts;
 };
 
 struct btrfs_subvolume_writers {
@@ -1776,12 +1782,12 @@ struct btrfs_root {
 
 	/* free ino cache stuff */
 	struct btrfs_free_space_ctl *free_ino_ctl;
-	enum btrfs_caching_type cached;
-	spinlock_t cache_lock;
-	wait_queue_head_t cache_wait;
+	enum btrfs_caching_type ino_cache_state;
+	spinlock_t ino_cache_lock;
+	wait_queue_head_t ino_cache_wait;
 	struct btrfs_free_space_ctl *free_ino_pinned;
-	u64 cache_progress;
-	struct inode *cache_inode;
+	u64 ino_cache_progress;
+	struct inode *ino_cache_inode;
 
 	struct mutex log_mutex;
 	wait_queue_head_t log_writer_wait;
@@ -1806,18 +1812,14 @@ struct btrfs_root {
 	/* node allocations are done in nodesize units */
 	u32 nodesize;
 
-	/* leaf allocations are done in leafsize units */
-	u32 leafsize;
-
 	u32 stripesize;
 
 	u32 type;
 
 	u64 highest_objectid;
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+	/* only used with CONFIG_BTRFS_FS_RUN_SANITY_TESTS is enabled */
 	u64 alloc_bytenr;
-#endif
 
 	u64 defrag_trans_start;
 	struct btrfs_key defrag_progress;
@@ -2094,6 +2096,7 @@ struct btrfs_ioctl_defrag_range_args {
 #define	BTRFS_MOUNT_CHANGE_INODE_CACHE	(1 << 24)
 
 #define BTRFS_DEFAULT_COMMIT_INTERVAL	(30)
+#define BTRFS_DEFAULT_MAX_INLINE	(8192)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2995,8 +2998,6 @@ BTRFS_SETGET_STACK_FUNCS(super_sectorsize, struct btrfs_super_block,
 			 sectorsize, 32);
 BTRFS_SETGET_STACK_FUNCS(super_nodesize, struct btrfs_super_block,
 			 nodesize, 32);
-BTRFS_SETGET_STACK_FUNCS(super_leafsize, struct btrfs_super_block,
-			 leafsize, 32);
 BTRFS_SETGET_STACK_FUNCS(super_stripesize, struct btrfs_super_block,
 			 stripesize, 32);
 BTRFS_SETGET_STACK_FUNCS(super_root_dir, struct btrfs_super_block,
@@ -3049,14 +3050,12 @@ BTRFS_SETGET_STACK_FUNCS(stack_file_extent_compression,
 static inline unsigned long
 btrfs_file_extent_inline_start(struct btrfs_file_extent_item *e)
 {
-	unsigned long offset = (unsigned long)e;
-	offset += offsetof(struct btrfs_file_extent_item, disk_bytenr);
-	return offset;
+	return (unsigned long)e + BTRFS_FILE_EXTENT_INLINE_DATA_START;
 }
 
 static inline u32 btrfs_file_extent_calc_inline_size(u32 datasize)
 {
-	return offsetof(struct btrfs_file_extent_item, disk_bytenr) + datasize;
+	return BTRFS_FILE_EXTENT_INLINE_DATA_START + datasize;
 }
 
 BTRFS_SETGET_FUNCS(file_extent_disk_bytenr, struct btrfs_file_extent_item,
@@ -3086,9 +3085,7 @@ BTRFS_SETGET_FUNCS(file_extent_other_encoding, struct btrfs_file_extent_item,
 static inline u32 btrfs_file_extent_inline_item_len(struct extent_buffer *eb,
 						    struct btrfs_item *e)
 {
-	unsigned long offset;
-	offset = offsetof(struct btrfs_file_extent_item, disk_bytenr);
-	return btrfs_item_size(eb, e) - offset;
+	return btrfs_item_size(eb, e) - BTRFS_FILE_EXTENT_INLINE_DATA_START;
 }
 
 /* this returns the number of file bytes represented by the inline item.
@@ -3232,13 +3229,6 @@ static inline struct btrfs_fs_info *btrfs_sb(struct super_block *sb)
 	return sb->s_fs_info;
 }
 
-static inline u32 btrfs_level_size(struct btrfs_root *root, int level)
-{
-	if (level == 0)
-		return root->leafsize;
-	return root->nodesize;
-}
-
 /* helper function to cast into the data area of the leaf. */
 #define btrfs_item_ptr(leaf, slot, type) \
 	((type *)(btrfs_leaf_data(leaf) + \
@@ -3263,7 +3253,7 @@ static inline gfp_t btrfs_alloc_write_mask(struct address_space *mapping)
 static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root,
 						 unsigned num_items)
 {
-	return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) *
+	return (root->nodesize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) *
 		2 * num_items;
 }
 
@@ -3274,8 +3264,7 @@ static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root,
 static inline u64 btrfs_calc_trunc_metadata_size(struct btrfs_root *root,
 						 unsigned num_items)
 {
-	return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) *
-		num_items;
+	return root->nodesize * BTRFS_MAX_LEVEL * num_items;
 }
 
 int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
@@ -3305,9 +3294,9 @@ struct btrfs_block_group_cache *btrfs_lookup_block_group(
 						 u64 bytenr);
 void btrfs_put_block_group(struct btrfs_block_group_cache *cache);
 int get_block_group_index(struct btrfs_block_group_cache *cache);
-struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
-					struct btrfs_root *root, u32 blocksize,
-					u64 parent, u64 root_objectid,
+struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
+					struct btrfs_root *root, u64 parent,
+					u64 root_objectid,
 					struct btrfs_disk_key *key, int level,
 					u64 hint, u64 empty_size);
 void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
@@ -3363,6 +3352,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 			   u64 size);
 int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root, u64 group_start);
+void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info);
 void btrfs_create_pending_block_groups(struct btrfs_trans_handle *trans,
 				       struct btrfs_root *root);
 u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data);
@@ -3604,6 +3594,7 @@ static inline void free_fs_info(struct btrfs_fs_info *fs_info)
 	kfree(fs_info->uuid_root);
 	kfree(fs_info->super_copy);
 	kfree(fs_info->super_for_commit);
+	security_free_mnt_opts(&fs_info->security_opts);
 	kfree(fs_info);
 }
 
@@ -3739,8 +3730,7 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
 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 btrfs_dio_private *dip, struct bio *bio,
-			      u64 logical_offset);
+			      struct bio *bio, u64 logical_offset);
 int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root,
 			     u64 objectid, u64 pos,
@@ -4141,8 +4131,15 @@ static inline int btrfs_defrag_cancelled(struct btrfs_fs_info *fs_info)
 /* Sanity test specific functions */
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 void btrfs_test_destroy_inode(struct inode *inode);
-int btrfs_verify_qgroup_counts(struct btrfs_fs_info *fs_info, u64 qgroupid,
-			       u64 rfer, u64 excl);
 #endif
 
+static inline int btrfs_test_is_dummy_root(struct btrfs_root *root)
+{
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state)))
+		return 1;
+#endif
+	return 0;
+}
+
 #endif

fs/btrfs/delayed-inode.c

@@ -1042,7 +1042,7 @@ static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
 	int ret;
 
 	key.objectid = node->inode_id;
-	btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
+	key.type = BTRFS_INODE_ITEM_KEY;
 	key.offset = 0;
 
 	if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &node->flags))
@@ -1099,7 +1099,7 @@ err_out:
 search:
 	btrfs_release_path(path);
 
-	btrfs_set_key_type(&key, BTRFS_INODE_EXTREF_KEY);
+	key.type = BTRFS_INODE_EXTREF_KEY;
 	key.offset = -1;
 	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 	if (ret < 0)
@@ -1473,7 +1473,7 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
 	}
 
 	delayed_item->key.objectid = btrfs_ino(dir);
-	btrfs_set_key_type(&delayed_item->key, BTRFS_DIR_INDEX_KEY);
+	delayed_item->key.type = BTRFS_DIR_INDEX_KEY;
 	delayed_item->key.offset = index;
 
 	dir_item = (struct btrfs_dir_item *)delayed_item->data;
@@ -1542,7 +1542,7 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
 		return PTR_ERR(node);
 
 	item_key.objectid = btrfs_ino(dir);
-	btrfs_set_key_type(&item_key, BTRFS_DIR_INDEX_KEY);
+	item_key.type = BTRFS_DIR_INDEX_KEY;
 	item_key.offset = index;
 
 	ret = btrfs_delete_delayed_insertion_item(root, node, &item_key);

fs/btrfs/dev-replace.c

@@ -168,8 +168,12 @@ no_valid_dev_replace_entry_found:
 					dev_replace->srcdev->total_bytes;
 				dev_replace->tgtdev->disk_total_bytes =
 					dev_replace->srcdev->disk_total_bytes;
+				dev_replace->tgtdev->commit_total_bytes =
+					dev_replace->srcdev->commit_total_bytes;
 				dev_replace->tgtdev->bytes_used =
 					dev_replace->srcdev->bytes_used;
+				dev_replace->tgtdev->commit_bytes_used =
+					dev_replace->srcdev->commit_bytes_used;
 			}
 			dev_replace->tgtdev->is_tgtdev_for_dev_replace = 1;
 			btrfs_init_dev_replace_tgtdev_for_resume(fs_info,
@@ -329,30 +333,34 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	    args->start.tgtdev_name[0] == '\0')
 		return -EINVAL;
 
-	mutex_lock(&fs_info->volume_mutex);
-	ret = btrfs_init_dev_replace_tgtdev(root, args->start.tgtdev_name,
-					    &tgt_device);
-	if (ret) {
-		btrfs_err(fs_info, "target device %s is invalid!",
-		       args->start.tgtdev_name);
-		mutex_unlock(&fs_info->volume_mutex);
-		return -EINVAL;
+	/*
+	 * Here we commit the transaction to make sure commit_total_bytes
+	 * of all the devices are updated.
+	 */
+	trans = btrfs_attach_transaction(root);
+	if (!IS_ERR(trans)) {
+		ret = btrfs_commit_transaction(trans, root);
+		if (ret)
+			return ret;
+	} else if (PTR_ERR(trans) != -ENOENT) {
+		return PTR_ERR(trans);
 	}
 
+	/* the disk copy procedure reuses the scrub code */
+	mutex_lock(&fs_info->volume_mutex);
 	ret = btrfs_dev_replace_find_srcdev(root, args->start.srcdevid,
 					    args->start.srcdev_name,
 					    &src_device);
-	mutex_unlock(&fs_info->volume_mutex);
 	if (ret) {
-		ret = -EINVAL;
-		goto leave_no_lock;
+		mutex_unlock(&fs_info->volume_mutex);
+		return ret;
 	}
 
-	if (tgt_device->total_bytes < src_device->total_bytes) {
-		btrfs_err(fs_info, "target device is smaller than source device!");
-		ret = -EINVAL;
-		goto leave_no_lock;
-	}
+	ret = btrfs_init_dev_replace_tgtdev(root, args->start.tgtdev_name,
+					    src_device, &tgt_device);
+	mutex_unlock(&fs_info->volume_mutex);
+	if (ret)
+		return ret;
 
 	btrfs_dev_replace_lock(dev_replace);
 	switch (dev_replace->replace_state) {
@@ -380,10 +388,6 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 		      src_device->devid,
 		      rcu_str_deref(tgt_device->name));
 
-	tgt_device->total_bytes = src_device->total_bytes;
-	tgt_device->disk_total_bytes = src_device->disk_total_bytes;
-	tgt_device->bytes_used = src_device->bytes_used;
-
 	/*
 	 * from now on, the writes to the srcdev are all duplicated to
 	 * go to the tgtdev as well (refer to btrfs_map_block()).
@@ -414,7 +418,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 
 	/* the disk copy procedure reuses the scrub code */
 	ret = btrfs_scrub_dev(fs_info, src_device->devid, 0,
-			      src_device->total_bytes,
+			      btrfs_device_get_total_bytes(src_device),
 			      &dev_replace->scrub_progress, 0, 1);
 
 	ret = btrfs_dev_replace_finishing(root->fs_info, ret);
@@ -426,9 +430,7 @@ leave:
 	dev_replace->srcdev = NULL;
 	dev_replace->tgtdev = NULL;
 	btrfs_dev_replace_unlock(dev_replace);
-leave_no_lock:
-	if (tgt_device)
-		btrfs_destroy_dev_replace_tgtdev(fs_info, tgt_device);
+	btrfs_destroy_dev_replace_tgtdev(fs_info, tgt_device);
 	return ret;
 }
 
@@ -507,9 +509,10 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	ret = btrfs_commit_transaction(trans, root);
 	WARN_ON(ret);
 
+	mutex_lock(&uuid_mutex);
 	/* keep away write_all_supers() during the finishing procedure */
-	mutex_lock(&root->fs_info->chunk_mutex);
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+	mutex_lock(&root->fs_info->chunk_mutex);
 	btrfs_dev_replace_lock(dev_replace);
 	dev_replace->replace_state =
 		scrub_ret ? BTRFS_IOCTL_DEV_REPLACE_STATE_CANCELED
@@ -532,8 +535,9 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 			      src_device->devid,
 			      rcu_str_deref(tgt_device->name), scrub_ret);
 		btrfs_dev_replace_unlock(dev_replace);
-		mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 		mutex_unlock(&root->fs_info->chunk_mutex);
+		mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+		mutex_unlock(&uuid_mutex);
 		if (tgt_device)
 			btrfs_destroy_dev_replace_tgtdev(fs_info, tgt_device);
 		mutex_unlock(&dev_replace->lock_finishing_cancel_unmount);
@@ -542,7 +546,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	}
 
 	printk_in_rcu(KERN_INFO
-		      "BTRFS: dev_replace from %s (devid %llu) to %s) finished\n",
+		      "BTRFS: dev_replace from %s (devid %llu) to %s finished\n",
 		      src_device->missing ? "<missing disk>" :
 		        rcu_str_deref(src_device->name),
 		      src_device->devid,
@@ -550,23 +554,29 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	tgt_device->is_tgtdev_for_dev_replace = 0;
 	tgt_device->devid = src_device->devid;
 	src_device->devid = BTRFS_DEV_REPLACE_DEVID;
-	tgt_device->bytes_used = src_device->bytes_used;
 	memcpy(uuid_tmp, tgt_device->uuid, sizeof(uuid_tmp));
 	memcpy(tgt_device->uuid, src_device->uuid, sizeof(tgt_device->uuid));
 	memcpy(src_device->uuid, uuid_tmp, sizeof(src_device->uuid));
-	tgt_device->total_bytes = src_device->total_bytes;
-	tgt_device->disk_total_bytes = src_device->disk_total_bytes;
-	tgt_device->bytes_used = src_device->bytes_used;
+	btrfs_device_set_total_bytes(tgt_device, src_device->total_bytes);
+	btrfs_device_set_disk_total_bytes(tgt_device,
+					  src_device->disk_total_bytes);
+	btrfs_device_set_bytes_used(tgt_device, src_device->bytes_used);
+	ASSERT(list_empty(&src_device->resized_list));
+	tgt_device->commit_total_bytes = src_device->commit_total_bytes;
+	tgt_device->commit_bytes_used = src_device->bytes_used;
 	if (fs_info->sb->s_bdev == src_device->bdev)
 		fs_info->sb->s_bdev = tgt_device->bdev;
 	if (fs_info->fs_devices->latest_bdev == src_device->bdev)
 		fs_info->fs_devices->latest_bdev = tgt_device->bdev;
 	list_add(&tgt_device->dev_alloc_list, &fs_info->fs_devices->alloc_list);
+	fs_info->fs_devices->rw_devices++;
 
 	/* replace the sysfs entry */
 	btrfs_kobj_rm_device(fs_info, src_device);
 	btrfs_kobj_add_device(fs_info, tgt_device);
 
+	btrfs_dev_replace_unlock(dev_replace);
+
 	btrfs_rm_dev_replace_blocked(fs_info);
 
 	btrfs_rm_dev_replace_srcdev(fs_info, src_device);
@@ -580,9 +590,9 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	 * superblock is scratched out so that it is no longer marked to
 	 * belong to this filesystem.
 	 */
-	btrfs_dev_replace_unlock(dev_replace);
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 	mutex_unlock(&root->fs_info->chunk_mutex);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+	mutex_unlock(&uuid_mutex);
 
 	/* write back the superblocks */
 	trans = btrfs_start_transaction(root, 0);
@@ -643,6 +653,7 @@ void btrfs_dev_replace_status(struct btrfs_fs_info *fs_info,
 			      struct btrfs_ioctl_dev_replace_args *args)
 {
 	struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
+	struct btrfs_device *srcdev;
 
 	btrfs_dev_replace_lock(dev_replace);
 	/* even if !dev_replace_is_valid, the values are good enough for
@@ -665,8 +676,9 @@ void btrfs_dev_replace_status(struct btrfs_fs_info *fs_info,
 		break;
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED:
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED:
+		srcdev = dev_replace->srcdev;
 		args->status.progress_1000 = div64_u64(dev_replace->cursor_left,
-			div64_u64(dev_replace->srcdev->total_bytes, 1000));
+			div64_u64(btrfs_device_get_total_bytes(srcdev), 1000));
 		break;
 	}
 	btrfs_dev_replace_unlock(dev_replace);
@@ -825,7 +837,7 @@ static int btrfs_dev_replace_continue_on_mount(struct btrfs_fs_info *fs_info)
 
 	ret = btrfs_scrub_dev(fs_info, dev_replace->srcdev->devid,
 			      dev_replace->committed_cursor_left,
-			      dev_replace->srcdev->total_bytes,
+			      btrfs_device_get_total_bytes(dev_replace->srcdev),
 			      &dev_replace->scrub_progress, 0, 1);
 	ret = btrfs_dev_replace_finishing(fs_info, ret);
 	WARN_ON(ret);

fs/btrfs/dir-item.c

@@ -86,7 +86,7 @@ int btrfs_insert_xattr_item(struct btrfs_trans_handle *trans,
 	BUG_ON(name_len + data_len > BTRFS_MAX_XATTR_SIZE(root));
 
 	key.objectid = objectid;
-	btrfs_set_key_type(&key, BTRFS_XATTR_ITEM_KEY);
+	key.type = BTRFS_XATTR_ITEM_KEY;
 	key.offset = btrfs_name_hash(name, name_len);
 
 	data_size = sizeof(*dir_item) + name_len + data_len;
@@ -137,7 +137,7 @@ int btrfs_insert_dir_item(struct btrfs_trans_handle *trans, struct btrfs_root
 	u32 data_size;
 
 	key.objectid = btrfs_ino(dir);
-	btrfs_set_key_type(&key, BTRFS_DIR_ITEM_KEY);
+	key.type = BTRFS_DIR_ITEM_KEY;
 	key.offset = btrfs_name_hash(name, name_len);
 
 	path = btrfs_alloc_path();
@@ -204,7 +204,7 @@ struct btrfs_dir_item *btrfs_lookup_dir_item(struct btrfs_trans_handle *trans,
 	int cow = mod != 0;
 
 	key.objectid = dir;
-	btrfs_set_key_type(&key, BTRFS_DIR_ITEM_KEY);
+	key.type = BTRFS_DIR_ITEM_KEY;
 
 	key.offset = btrfs_name_hash(name, name_len);
 
@@ -234,7 +234,7 @@ int btrfs_check_dir_item_collision(struct btrfs_root *root, u64 dir,
 		return -ENOMEM;
 
 	key.objectid = dir;
-	btrfs_set_key_type(&key, BTRFS_DIR_ITEM_KEY);
+	key.type = BTRFS_DIR_ITEM_KEY;
 	key.offset = btrfs_name_hash(name, name_len);
 
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
@@ -297,7 +297,7 @@ btrfs_lookup_dir_index_item(struct btrfs_trans_handle *trans,
 	int cow = mod != 0;
 
 	key.objectid = dir;
-	btrfs_set_key_type(&key, BTRFS_DIR_INDEX_KEY);
+	key.type = BTRFS_DIR_INDEX_KEY;
 	key.offset = objectid;
 
 	ret = btrfs_search_slot(trans, root, &key, path, ins_len, cow);
@@ -367,7 +367,7 @@ struct btrfs_dir_item *btrfs_lookup_xattr(struct btrfs_trans_handle *trans,
 	int cow = mod != 0;
 
 	key.objectid = dir;
-	btrfs_set_key_type(&key, BTRFS_XATTR_ITEM_KEY);
+	key.type = BTRFS_XATTR_ITEM_KEY;
 	key.offset = btrfs_name_hash(name, name_len);
 	ret = btrfs_search_slot(trans, root, &key, path, ins_len, cow);
 	if (ret < 0)

fs/btrfs/disk-io.c

@@ -72,21 +72,41 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root);
 static void btrfs_error_commit_super(struct btrfs_root *root);
 
 /*
- * end_io_wq structs are used to do processing in task context when an IO is
- * complete.  This is used during reads to verify checksums, and it is used
+ * btrfs_end_io_wq structs are used to do processing in task context when an IO
+ * is complete.  This is used during reads to verify checksums, and it is used
  * by writes to insert metadata for new file extents after IO is complete.
  */
-struct end_io_wq {
+struct btrfs_end_io_wq {
 	struct bio *bio;
 	bio_end_io_t *end_io;
 	void *private;
 	struct btrfs_fs_info *info;
 	int error;
-	int metadata;
+	enum btrfs_wq_endio_type metadata;
 	struct list_head list;
 	struct btrfs_work work;
 };
 
+static struct kmem_cache *btrfs_end_io_wq_cache;
+
+int __init btrfs_end_io_wq_init(void)
+{
+	btrfs_end_io_wq_cache = kmem_cache_create("btrfs_end_io_wq",
+					sizeof(struct btrfs_end_io_wq),
+					0,
+					SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD,
+					NULL);
+	if (!btrfs_end_io_wq_cache)
+		return -ENOMEM;
+	return 0;
+}
+
+void btrfs_end_io_wq_exit(void)
+{
+	if (btrfs_end_io_wq_cache)
+		kmem_cache_destroy(btrfs_end_io_wq_cache);
+}
+
 /*
  * async submit bios are used to offload expensive checksumming
  * onto the worker threads.  They checksum file and metadata bios
@@ -327,8 +347,7 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 {
 	struct extent_state *cached_state = NULL;
 	int ret;
-	bool need_lock = (current->journal_info ==
-			  (void *)BTRFS_SEND_TRANS_STUB);
+	bool need_lock = (current->journal_info == BTRFS_SEND_TRANS_STUB);
 
 	if (!parent_transid || btrfs_header_generation(eb) == parent_transid)
 		return 0;
@@ -348,9 +367,9 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 		ret = 0;
 		goto out;
 	}
-	printk_ratelimited("parent transid verify failed on %llu wanted %llu "
-		       "found %llu\n",
-		       eb->start, parent_transid, btrfs_header_generation(eb));
+	printk_ratelimited(KERN_INFO "BTRFS (device %s): parent transid verify failed on %llu wanted %llu found %llu\n",
+			eb->fs_info->sb->s_id, eb->start,
+			parent_transid, btrfs_header_generation(eb));
 	ret = 1;
 
 	/*
@@ -607,22 +626,22 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 		goto err;
 
 	eb->read_mirror = mirror;
-	if (test_bit(EXTENT_BUFFER_IOERR, &eb->bflags)) {
+	if (test_bit(EXTENT_BUFFER_READ_ERR, &eb->bflags)) {
 		ret = -EIO;
 		goto err;
 	}
 
 	found_start = btrfs_header_bytenr(eb);
 	if (found_start != eb->start) {
-		printk_ratelimited(KERN_INFO "BTRFS: bad tree block start "
+		printk_ratelimited(KERN_INFO "BTRFS (device %s): bad tree block start "
 			       "%llu %llu\n",
-			       found_start, eb->start);
+			       eb->fs_info->sb->s_id, 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",
-			       eb->start);
+		printk_ratelimited(KERN_INFO "BTRFS (device %s): bad fsid on block %llu\n",
+			       eb->fs_info->sb->s_id, eb->start);
 		ret = -EIO;
 		goto err;
 	}
@@ -680,7 +699,7 @@ static int btree_io_failed_hook(struct page *page, int failed_mirror)
 	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
 
 	eb = (struct extent_buffer *)page->private;
-	set_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
+	set_bit(EXTENT_BUFFER_READ_ERR, &eb->bflags);
 	eb->read_mirror = failed_mirror;
 	atomic_dec(&eb->io_pages);
 	if (test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags))
@@ -690,7 +709,7 @@ static int btree_io_failed_hook(struct page *page, int failed_mirror)
 
 static void end_workqueue_bio(struct bio *bio, int err)
 {
-	struct end_io_wq *end_io_wq = bio->bi_private;
+	struct btrfs_end_io_wq *end_io_wq = bio->bi_private;
 	struct btrfs_fs_info *fs_info;
 	struct btrfs_workqueue *wq;
 	btrfs_work_func_t func;
@@ -713,7 +732,11 @@ static void end_workqueue_bio(struct bio *bio, int err)
 			func = btrfs_endio_write_helper;
 		}
 	} else {
-		if (end_io_wq->metadata == BTRFS_WQ_ENDIO_RAID56) {
+		if (unlikely(end_io_wq->metadata ==
+			     BTRFS_WQ_ENDIO_DIO_REPAIR)) {
+			wq = fs_info->endio_repair_workers;
+			func = btrfs_endio_repair_helper;
+		} else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_RAID56) {
 			wq = fs_info->endio_raid56_workers;
 			func = btrfs_endio_raid56_helper;
 		} else if (end_io_wq->metadata) {
@@ -729,19 +752,12 @@ static void end_workqueue_bio(struct bio *bio, int err)
 	btrfs_queue_work(wq, &end_io_wq->work);
 }
 
-/*
- * For the metadata arg you want
- *
- * 0 - if data
- * 1 - if normal metadta
- * 2 - if writing to the free space cache area
- * 3 - raid parity work
- */
 int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
-			int metadata)
+			enum btrfs_wq_endio_type metadata)
 {
-	struct end_io_wq *end_io_wq;
-	end_io_wq = kmalloc(sizeof(*end_io_wq), GFP_NOFS);
+	struct btrfs_end_io_wq *end_io_wq;
+
+	end_io_wq = kmem_cache_alloc(btrfs_end_io_wq_cache, GFP_NOFS);
 	if (!end_io_wq)
 		return -ENOMEM;
 
@@ -925,7 +941,7 @@ static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 		 * can happen in the async kernel threads
 		 */
 		ret = btrfs_bio_wq_end_io(BTRFS_I(inode)->root->fs_info,
-					  bio, 1);
+					  bio, BTRFS_WQ_ENDIO_METADATA);
 		if (ret)
 			goto out_w_error;
 		ret = btrfs_map_bio(BTRFS_I(inode)->root, rw, bio,
@@ -1057,20 +1073,17 @@ static const struct address_space_operations btree_aops = {
 	.set_page_dirty = btree_set_page_dirty,
 };
 
-int readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize,
-			 u64 parent_transid)
+void readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize)
 {
 	struct extent_buffer *buf = NULL;
 	struct inode *btree_inode = root->fs_info->btree_inode;
-	int ret = 0;
 
 	buf = btrfs_find_create_tree_block(root, bytenr, blocksize);
 	if (!buf)
-		return 0;
+		return;
 	read_extent_buffer_pages(&BTRFS_I(btree_inode)->io_tree,
 				 buf, 0, WAIT_NONE, btree_get_extent, 0);
 	free_extent_buffer(buf);
-	return ret;
 }
 
 int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize,
@@ -1106,7 +1119,7 @@ int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize,
 }
 
 struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
-					    u64 bytenr, u32 blocksize)
+					    u64 bytenr)
 {
 	return find_extent_buffer(root->fs_info, bytenr);
 }
@@ -1114,11 +1127,9 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
 struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
 						 u64 bytenr, u32 blocksize)
 {
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state)))
+	if (btrfs_test_is_dummy_root(root))
 		return alloc_test_extent_buffer(root->fs_info, bytenr,
 						blocksize);
-#endif
 	return alloc_extent_buffer(root->fs_info, bytenr, blocksize);
 }
 
@@ -1136,12 +1147,12 @@ int btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
 }
 
 struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
-				      u32 blocksize, u64 parent_transid)
+				      u64 parent_transid)
 {
 	struct extent_buffer *buf = NULL;
 	int ret;
 
-	buf = btrfs_find_create_tree_block(root, bytenr, blocksize);
+	buf = btrfs_find_create_tree_block(root, bytenr, root->nodesize);
 	if (!buf)
 		return NULL;
 
@@ -1200,16 +1211,14 @@ btrfs_free_subvolume_writers(struct btrfs_subvolume_writers *writers)
 	kfree(writers);
 }
 
-static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
-			 u32 stripesize, struct btrfs_root *root,
-			 struct btrfs_fs_info *fs_info,
+static void __setup_root(u32 nodesize, u32 sectorsize, u32 stripesize,
+			 struct btrfs_root *root, struct btrfs_fs_info *fs_info,
 			 u64 objectid)
 {
 	root->node = NULL;
 	root->commit_root = NULL;
 	root->sectorsize = sectorsize;
 	root->nodesize = nodesize;
-	root->leafsize = leafsize;
 	root->stripesize = stripesize;
 	root->state = 0;
 	root->orphan_cleanup_state = 0;
@@ -1295,7 +1304,7 @@ struct btrfs_root *btrfs_alloc_dummy_root(void)
 	root = btrfs_alloc_root(NULL);
 	if (!root)
 		return ERR_PTR(-ENOMEM);
-	__setup_root(4096, 4096, 4096, 4096, root, NULL, 1);
+	__setup_root(4096, 4096, 4096, root, NULL, 1);
 	set_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state);
 	root->alloc_bytenr = 0;
 
@@ -1318,15 +1327,13 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	if (!root)
 		return ERR_PTR(-ENOMEM);
 
-	__setup_root(tree_root->nodesize, tree_root->leafsize,
-		     tree_root->sectorsize, tree_root->stripesize,
-		     root, fs_info, objectid);
+	__setup_root(tree_root->nodesize, tree_root->sectorsize,
+		tree_root->stripesize, root, fs_info, objectid);
 	root->root_key.objectid = objectid;
 	root->root_key.type = BTRFS_ROOT_ITEM_KEY;
 	root->root_key.offset = 0;
 
-	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
-				      0, objectid, NULL, 0, 0, 0);
+	leaf = btrfs_alloc_tree_block(trans, root, 0, objectid, NULL, 0, 0, 0);
 	if (IS_ERR(leaf)) {
 		ret = PTR_ERR(leaf);
 		leaf = NULL;
@@ -1396,9 +1403,9 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 	if (!root)
 		return ERR_PTR(-ENOMEM);
 
-	__setup_root(tree_root->nodesize, tree_root->leafsize,
-		     tree_root->sectorsize, tree_root->stripesize,
-		     root, fs_info, BTRFS_TREE_LOG_OBJECTID);
+	__setup_root(tree_root->nodesize, tree_root->sectorsize,
+		     tree_root->stripesize, root, fs_info,
+		     BTRFS_TREE_LOG_OBJECTID);
 
 	root->root_key.objectid = BTRFS_TREE_LOG_OBJECTID;
 	root->root_key.type = BTRFS_ROOT_ITEM_KEY;
@@ -1413,9 +1420,8 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 	 * updated (along with back refs to the log tree).
 	 */
 
-	leaf = btrfs_alloc_free_block(trans, root, root->leafsize, 0,
-				      BTRFS_TREE_LOG_OBJECTID, NULL,
-				      0, 0, 0);
+	leaf = btrfs_alloc_tree_block(trans, root, 0, BTRFS_TREE_LOG_OBJECTID,
+			NULL, 0, 0, 0);
 	if (IS_ERR(leaf)) {
 		kfree(root);
 		return ERR_CAST(leaf);
@@ -1465,7 +1471,7 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	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_nbytes(inode_item, root->nodesize);
 	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
 
 	btrfs_set_root_node(&log_root->root_item, log_root->node);
@@ -1485,7 +1491,6 @@ static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
 	struct btrfs_fs_info *fs_info = tree_root->fs_info;
 	struct btrfs_path *path;
 	u64 generation;
-	u32 blocksize;
 	int ret;
 
 	path = btrfs_alloc_path();
@@ -1498,9 +1503,8 @@ static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
 		goto alloc_fail;
 	}
 
-	__setup_root(tree_root->nodesize, tree_root->leafsize,
-		     tree_root->sectorsize, tree_root->stripesize,
-		     root, fs_info, key->objectid);
+	__setup_root(tree_root->nodesize, tree_root->sectorsize,
+		tree_root->stripesize, root, fs_info, key->objectid);
 
 	ret = btrfs_find_root(tree_root, key, path,
 			      &root->root_item, &root->root_key);
@@ -1511,9 +1515,8 @@ static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
 	}
 
 	generation = btrfs_root_generation(&root->root_item);
-	blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
 	root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
-				     blocksize, generation);
+				     generation);
 	if (!root->node) {
 		ret = -ENOMEM;
 		goto find_fail;
@@ -1573,8 +1576,8 @@ int btrfs_init_fs_root(struct btrfs_root *root)
 	root->subv_writers = writers;
 
 	btrfs_init_free_ino_ctl(root);
-	spin_lock_init(&root->cache_lock);
-	init_waitqueue_head(&root->cache_wait);
+	spin_lock_init(&root->ino_cache_lock);
+	init_waitqueue_head(&root->ino_cache_wait);
 
 	ret = get_anon_bdev(&root->anon_dev);
 	if (ret)
@@ -1708,10 +1711,6 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 	return ret;
 }
 
-/*
- * If this fails, caller must call bdi_destroy() to get rid of the
- * bdi again.
- */
 static int setup_bdi(struct btrfs_fs_info *info, struct backing_dev_info *bdi)
 {
 	int err;
@@ -1734,16 +1733,16 @@ static int setup_bdi(struct btrfs_fs_info *info, struct backing_dev_info *bdi)
 static void end_workqueue_fn(struct btrfs_work *work)
 {
 	struct bio *bio;
-	struct end_io_wq *end_io_wq;
+	struct btrfs_end_io_wq *end_io_wq;
 	int error;
 
-	end_io_wq = container_of(work, struct end_io_wq, work);
+	end_io_wq = container_of(work, struct btrfs_end_io_wq, work);
 	bio = end_io_wq->bio;
 
 	error = end_io_wq->error;
 	bio->bi_private = end_io_wq->private;
 	bio->bi_end_io = end_io_wq->end_io;
-	kfree(end_io_wq);
+	kmem_cache_free(btrfs_end_io_wq_cache, end_io_wq);
 	bio_endio_nodec(bio, error);
 }
 
@@ -1772,6 +1771,7 @@ static int cleaner_kthread(void *arg)
 		}
 
 		btrfs_run_delayed_iputs(root);
+		btrfs_delete_unused_bgs(root->fs_info);
 		again = btrfs_clean_one_deleted_snapshot(root);
 		mutex_unlock(&root->fs_info->cleaner_mutex);
 
@@ -2063,6 +2063,7 @@ static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info)
 	btrfs_destroy_workqueue(fs_info->endio_workers);
 	btrfs_destroy_workqueue(fs_info->endio_meta_workers);
 	btrfs_destroy_workqueue(fs_info->endio_raid56_workers);
+	btrfs_destroy_workqueue(fs_info->endio_repair_workers);
 	btrfs_destroy_workqueue(fs_info->rmw_workers);
 	btrfs_destroy_workqueue(fs_info->endio_meta_write_workers);
 	btrfs_destroy_workqueue(fs_info->endio_write_workers);
@@ -2143,8 +2144,6 @@ int open_ctree(struct super_block *sb,
 {
 	u32 sectorsize;
 	u32 nodesize;
-	u32 leafsize;
-	u32 blocksize;
 	u32 stripesize;
 	u64 generation;
 	u64 features;
@@ -2233,6 +2232,7 @@ int open_ctree(struct super_block *sb,
 	spin_lock_init(&fs_info->super_lock);
 	spin_lock_init(&fs_info->qgroup_op_lock);
 	spin_lock_init(&fs_info->buffer_lock);
+	spin_lock_init(&fs_info->unused_bgs_lock);
 	rwlock_init(&fs_info->tree_mod_log_lock);
 	mutex_init(&fs_info->reloc_mutex);
 	mutex_init(&fs_info->delalloc_root_mutex);
@@ -2242,6 +2242,7 @@ int open_ctree(struct super_block *sb,
 	INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
 	INIT_LIST_HEAD(&fs_info->space_info);
 	INIT_LIST_HEAD(&fs_info->tree_mod_seq_list);
+	INIT_LIST_HEAD(&fs_info->unused_bgs);
 	btrfs_mapping_init(&fs_info->mapping_tree);
 	btrfs_init_block_rsv(&fs_info->global_block_rsv,
 			     BTRFS_BLOCK_RSV_GLOBAL);
@@ -2260,7 +2261,7 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->qgroup_op_seq, 0);
 	atomic64_set(&fs_info->tree_mod_seq, 0);
 	fs_info->sb = sb;
-	fs_info->max_inline = 8192 * 1024;
+	fs_info->max_inline = BTRFS_DEFAULT_MAX_INLINE;
 	fs_info->metadata_ratio = 0;
 	fs_info->defrag_inodes = RB_ROOT;
 	fs_info->free_chunk_space = 0;
@@ -2389,7 +2390,7 @@ int open_ctree(struct super_block *sb,
 		goto fail_alloc;
 	}
 
-	__setup_root(4096, 4096, 4096, 4096, tree_root,
+	__setup_root(4096, 4096, 4096, tree_root,
 		     fs_info, BTRFS_ROOT_TREE_OBJECTID);
 
 	invalidate_bdev(fs_devices->latest_bdev);
@@ -2469,19 +2470,22 @@ int open_ctree(struct super_block *sb,
 		goto fail_alloc;
 	}
 
-	if (btrfs_super_leafsize(disk_super) !=
+	/*
+	 * Leafsize and nodesize were always equal, this is only a sanity check.
+	 */
+	if (le32_to_cpu(disk_super->__unused_leafsize) !=
 	    btrfs_super_nodesize(disk_super)) {
 		printk(KERN_ERR "BTRFS: couldn't mount because metadata "
 		       "blocksizes don't match.  node %d leaf %d\n",
 		       btrfs_super_nodesize(disk_super),
-		       btrfs_super_leafsize(disk_super));
+		       le32_to_cpu(disk_super->__unused_leafsize));
 		err = -EINVAL;
 		goto fail_alloc;
 	}
-	if (btrfs_super_leafsize(disk_super) > BTRFS_MAX_METADATA_BLOCKSIZE) {
+	if (btrfs_super_nodesize(disk_super) > BTRFS_MAX_METADATA_BLOCKSIZE) {
 		printk(KERN_ERR "BTRFS: couldn't mount because metadata "
 		       "blocksize (%d) was too large\n",
-		       btrfs_super_leafsize(disk_super));
+		       btrfs_super_nodesize(disk_super));
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2498,17 +2502,16 @@ int open_ctree(struct super_block *sb,
 	 * flag our filesystem as having big metadata blocks if
 	 * they are bigger than the page size
 	 */
-	if (btrfs_super_leafsize(disk_super) > PAGE_CACHE_SIZE) {
+	if (btrfs_super_nodesize(disk_super) > PAGE_CACHE_SIZE) {
 		if (!(features & BTRFS_FEATURE_INCOMPAT_BIG_METADATA))
 			printk(KERN_INFO "BTRFS: flagging fs with big metadata feature\n");
 		features |= BTRFS_FEATURE_INCOMPAT_BIG_METADATA;
 	}
 
 	nodesize = btrfs_super_nodesize(disk_super);
-	leafsize = btrfs_super_leafsize(disk_super);
 	sectorsize = btrfs_super_sectorsize(disk_super);
 	stripesize = btrfs_super_stripesize(disk_super);
-	fs_info->dirty_metadata_batch = leafsize * (1 + ilog2(nr_cpu_ids));
+	fs_info->dirty_metadata_batch = nodesize * (1 + ilog2(nr_cpu_ids));
 	fs_info->delalloc_batch = sectorsize * 512 * (1 + ilog2(nr_cpu_ids));
 
 	/*
@@ -2516,7 +2519,7 @@ int open_ctree(struct super_block *sb,
 	 * extent buffers for the same range.  It leads to corruptions
 	 */
 	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
-	    (sectorsize != leafsize)) {
+	    (sectorsize != nodesize)) {
 		printk(KERN_WARNING "BTRFS: unequal leaf/node/sector sizes "
 				"are not allowed for mixed block groups on %s\n",
 				sb->s_id);
@@ -2579,6 +2582,8 @@ int open_ctree(struct super_block *sb,
 		btrfs_alloc_workqueue("endio-meta-write", flags, max_active, 2);
 	fs_info->endio_raid56_workers =
 		btrfs_alloc_workqueue("endio-raid56", flags, max_active, 4);
+	fs_info->endio_repair_workers =
+		btrfs_alloc_workqueue("endio-repair", flags, 1, 0);
 	fs_info->rmw_workers =
 		btrfs_alloc_workqueue("rmw", flags, max_active, 2);
 	fs_info->endio_write_workers =
@@ -2600,11 +2605,12 @@ int open_ctree(struct super_block *sb,
 	      fs_info->submit_workers && fs_info->flush_workers &&
 	      fs_info->endio_workers && fs_info->endio_meta_workers &&
 	      fs_info->endio_meta_write_workers &&
+	      fs_info->endio_repair_workers &&
 	      fs_info->endio_write_workers && fs_info->endio_raid56_workers &&
 	      fs_info->endio_freespace_worker && fs_info->rmw_workers &&
 	      fs_info->caching_workers && fs_info->readahead_workers &&
 	      fs_info->fixup_workers && fs_info->delayed_workers &&
-	      fs_info->fixup_workers && fs_info->extent_workers &&
+	      fs_info->extent_workers &&
 	      fs_info->qgroup_rescan_workers)) {
 		err = -ENOMEM;
 		goto fail_sb_buffer;
@@ -2615,7 +2621,6 @@ int open_ctree(struct super_block *sb,
 				    4 * 1024 * 1024 / PAGE_CACHE_SIZE);
 
 	tree_root->nodesize = nodesize;
-	tree_root->leafsize = leafsize;
 	tree_root->sectorsize = sectorsize;
 	tree_root->stripesize = stripesize;
 
@@ -2642,16 +2647,14 @@ int open_ctree(struct super_block *sb,
 		goto fail_sb_buffer;
 	}
 
-	blocksize = btrfs_level_size(tree_root,
-				     btrfs_super_chunk_root_level(disk_super));
 	generation = btrfs_super_chunk_root_generation(disk_super);
 
-	__setup_root(nodesize, leafsize, sectorsize, stripesize,
-		     chunk_root, fs_info, BTRFS_CHUNK_TREE_OBJECTID);
+	__setup_root(nodesize, sectorsize, stripesize, chunk_root,
+		     fs_info, BTRFS_CHUNK_TREE_OBJECTID);
 
 	chunk_root->node = read_tree_block(chunk_root,
 					   btrfs_super_chunk_root(disk_super),
-					   blocksize, generation);
+					   generation);
 	if (!chunk_root->node ||
 	    !test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) {
 		printk(KERN_WARNING "BTRFS: failed to read chunk root on %s\n",
@@ -2684,13 +2687,11 @@ int open_ctree(struct super_block *sb,
 	}
 
 retry_root_backup:
-	blocksize = btrfs_level_size(tree_root,
-				     btrfs_super_root_level(disk_super));
 	generation = btrfs_super_generation(disk_super);
 
 	tree_root->node = read_tree_block(tree_root,
 					  btrfs_super_root(disk_super),
-					  blocksize, generation);
+					  generation);
 	if (!tree_root->node ||
 	    !test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) {
 		printk(KERN_WARNING "BTRFS: failed to read tree root on %s\n",
@@ -2859,9 +2860,6 @@ retry_root_backup:
 			err = -EIO;
 			goto fail_qgroup;
 		}
-		blocksize =
-		     btrfs_level_size(tree_root,
-				      btrfs_super_log_root_level(disk_super));
 
 		log_tree_root = btrfs_alloc_root(fs_info);
 		if (!log_tree_root) {
@@ -2869,11 +2867,10 @@ retry_root_backup:
 			goto fail_qgroup;
 		}
 
-		__setup_root(nodesize, leafsize, sectorsize, stripesize,
+		__setup_root(nodesize, sectorsize, stripesize,
 			     log_tree_root, fs_info, BTRFS_TREE_LOG_OBJECTID);
 
 		log_tree_root->node = read_tree_block(tree_root, bytenr,
-						      blocksize,
 						      generation + 1);
 		if (!log_tree_root->node ||
 		    !extent_buffer_uptodate(log_tree_root->node)) {
@@ -2980,6 +2977,8 @@ retry_root_backup:
 		fs_info->update_uuid_tree_gen = 1;
 	}
 
+	fs_info->open = 1;
+
 	return 0;
 
 fail_qgroup:
@@ -3139,7 +3138,8 @@ static int write_dev_supers(struct btrfs_device *device,
 
 	for (i = 0; i < max_mirrors; i++) {
 		bytenr = btrfs_sb_offset(i);
-		if (bytenr + BTRFS_SUPER_INFO_SIZE >= device->total_bytes)
+		if (bytenr + BTRFS_SUPER_INFO_SIZE >=
+		    device->commit_total_bytes)
 			break;
 
 		if (wait) {
@@ -3456,8 +3456,9 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 		btrfs_set_stack_device_type(dev_item, dev->type);
 		btrfs_set_stack_device_id(dev_item, dev->devid);
 		btrfs_set_stack_device_total_bytes(dev_item,
-						   dev->disk_total_bytes);
-		btrfs_set_stack_device_bytes_used(dev_item, dev->bytes_used);
+						   dev->commit_total_bytes);
+		btrfs_set_stack_device_bytes_used(dev_item,
+						  dev->commit_bytes_used);
 		btrfs_set_stack_device_io_align(dev_item, dev->io_align);
 		btrfs_set_stack_device_io_width(dev_item, dev->io_width);
 		btrfs_set_stack_device_sector_size(dev_item, dev->sector_size);
@@ -3532,7 +3533,7 @@ void btrfs_drop_and_free_fs_root(struct btrfs_fs_info *fs_info,
 
 static void free_fs_root(struct btrfs_root *root)
 {
-	iput(root->cache_inode);
+	iput(root->ino_cache_inode);
 	WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
 	btrfs_free_block_rsv(root, root->orphan_block_rsv);
 	root->orphan_block_rsv = NULL;
@@ -3623,7 +3624,7 @@ int btrfs_commit_super(struct btrfs_root *root)
 	return btrfs_commit_transaction(trans, root);
 }
 
-int close_ctree(struct btrfs_root *root)
+void close_ctree(struct btrfs_root *root)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	int ret;
@@ -3689,6 +3690,7 @@ int close_ctree(struct btrfs_root *root)
 	invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
 	btrfs_stop_all_workers(fs_info);
 
+	fs_info->open = 0;
 	free_root_pointers(fs_info, 1);
 
 	iput(fs_info->btree_inode);
@@ -3711,8 +3713,6 @@ int close_ctree(struct btrfs_root *root)
 
 	btrfs_free_block_rsv(root, root->orphan_block_rsv);
 	root->orphan_block_rsv = NULL;
-
-	return 0;
 }
 
 int btrfs_buffer_uptodate(struct extent_buffer *buf, u64 parent_transid,
@@ -3814,10 +3814,73 @@ int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid)
 static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
 			      int read_only)
 {
+	struct btrfs_super_block *sb = fs_info->super_copy;
+	int ret = 0;
+
+	if (sb->root_level > BTRFS_MAX_LEVEL) {
+		printk(KERN_ERR "BTRFS: tree_root level too big: %d > %d\n",
+				sb->root_level, BTRFS_MAX_LEVEL);
+		ret = -EINVAL;
+	}
+	if (sb->chunk_root_level > BTRFS_MAX_LEVEL) {
+		printk(KERN_ERR "BTRFS: chunk_root level too big: %d > %d\n",
+				sb->chunk_root_level, BTRFS_MAX_LEVEL);
+		ret = -EINVAL;
+	}
+	if (sb->log_root_level > BTRFS_MAX_LEVEL) {
+		printk(KERN_ERR "BTRFS: log_root level too big: %d > %d\n",
+				sb->log_root_level, BTRFS_MAX_LEVEL);
+		ret = -EINVAL;
+	}
+
 	/*
-	 * Placeholder for checks
+	 * The common minimum, we don't know if we can trust the nodesize/sectorsize
+	 * items yet, they'll be verified later. Issue just a warning.
 	 */
-	return 0;
+	if (!IS_ALIGNED(sb->root, 4096))
+		printk(KERN_WARNING "BTRFS: tree_root block unaligned: %llu\n",
+				sb->root);
+	if (!IS_ALIGNED(sb->chunk_root, 4096))
+		printk(KERN_WARNING "BTRFS: tree_root block unaligned: %llu\n",
+				sb->chunk_root);
+	if (!IS_ALIGNED(sb->log_root, 4096))
+		printk(KERN_WARNING "BTRFS: tree_root block unaligned: %llu\n",
+				sb->log_root);
+
+	if (memcmp(fs_info->fsid, sb->dev_item.fsid, BTRFS_UUID_SIZE) != 0) {
+		printk(KERN_ERR "BTRFS: dev_item UUID does not match fsid: %pU != %pU\n",
+				fs_info->fsid, sb->dev_item.fsid);
+		ret = -EINVAL;
+	}
+
+	/*
+	 * Hint to catch really bogus numbers, bitflips or so, more exact checks are
+	 * done later
+	 */
+	if (sb->num_devices > (1UL << 31))
+		printk(KERN_WARNING "BTRFS: suspicious number of devices: %llu\n",
+				sb->num_devices);
+
+	if (sb->bytenr != BTRFS_SUPER_INFO_OFFSET) {
+		printk(KERN_ERR "BTRFS: super offset mismatch %llu != %u\n",
+				sb->bytenr, BTRFS_SUPER_INFO_OFFSET);
+		ret = -EINVAL;
+	}
+
+	/*
+	 * The generation is a global counter, we'll trust it more than the others
+	 * but it's still possible that it's the one that's wrong.
+	 */
+	if (sb->generation < sb->chunk_root_generation)
+		printk(KERN_WARNING
+			"BTRFS: suspicious: generation < chunk_root_generation: %llu < %llu\n",
+			sb->generation, sb->chunk_root_generation);
+	if (sb->generation < sb->cache_generation && sb->cache_generation != (u64)-1)
+		printk(KERN_WARNING
+			"BTRFS: suspicious: generation < cache_generation: %llu < %llu\n",
+			sb->generation, sb->cache_generation);
+
+	return ret;
 }
 
 static void btrfs_error_commit_super(struct btrfs_root *root)
@@ -4009,9 +4072,8 @@ static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 
 		clear_extent_bits(dirty_pages, start, end, mark, GFP_NOFS);
 		while (start <= end) {
-			eb = btrfs_find_tree_block(root, start,
-						   root->leafsize);
-			start += root->leafsize;
+			eb = btrfs_find_tree_block(root, start);
+			start += root->nodesize;
 			if (!eb)
 				continue;
 			wait_on_extent_buffer_writeback(eb);

fs/btrfs/disk-io.h

@@ -25,11 +25,12 @@
 #define BTRFS_SUPER_MIRROR_MAX	 3
 #define BTRFS_SUPER_MIRROR_SHIFT 12
 
-enum {
+enum btrfs_wq_endio_type {
 	BTRFS_WQ_ENDIO_DATA = 0,
 	BTRFS_WQ_ENDIO_METADATA = 1,
 	BTRFS_WQ_ENDIO_FREE_SPACE = 2,
 	BTRFS_WQ_ENDIO_RAID56 = 3,
+	BTRFS_WQ_ENDIO_DIO_REPAIR = 4,
 };
 
 static inline u64 btrfs_sb_offset(int mirror)
@@ -44,9 +45,8 @@ struct btrfs_device;
 struct btrfs_fs_devices;
 
 struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
-				      u32 blocksize, u64 parent_transid);
-int readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize,
-			 u64 parent_transid);
+				      u64 parent_transid);
+void readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize);
 int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize,
 			 int mirror_num, struct extent_buffer **eb);
 struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
@@ -56,13 +56,13 @@ void clean_tree_block(struct btrfs_trans_handle *trans,
 int open_ctree(struct super_block *sb,
 	       struct btrfs_fs_devices *fs_devices,
 	       char *options);
-int close_ctree(struct btrfs_root *root);
+void close_ctree(struct btrfs_root *root);
 int write_ctree_super(struct btrfs_trans_handle *trans,
 		      struct btrfs_root *root, int max_mirrors);
 struct buffer_head *btrfs_read_dev_super(struct block_device *bdev);
 int btrfs_commit_super(struct btrfs_root *root);
 struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
-					    u64 bytenr, u32 blocksize);
+					    u64 bytenr);
 struct btrfs_root *btrfs_read_fs_root(struct btrfs_root *tree_root,
 				      struct btrfs_key *location);
 int btrfs_init_fs_root(struct btrfs_root *root);
@@ -119,7 +119,7 @@ int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid);
 u32 btrfs_csum_data(char *data, u32 seed, size_t len);
 void btrfs_csum_final(u32 crc, char *result);
 int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
-			int metadata);
+			enum btrfs_wq_endio_type metadata);
 int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
 			int rw, struct bio *bio, int mirror_num,
 			unsigned long bio_flags, u64 bio_offset,
@@ -141,6 +141,8 @@ int btree_lock_page_hook(struct page *page, void *data,
 				void (*flush_fn)(void *));
 int btrfs_calc_num_tolerated_disk_barrier_failures(
 	struct btrfs_fs_info *fs_info);
+int __init btrfs_end_io_wq_init(void);
+void btrfs_end_io_wq_exit(void);
 
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
 void btrfs_init_lockdep(void);

fs/btrfs/export.c

@@ -70,7 +70,7 @@ static struct dentry *btrfs_get_dentry(struct super_block *sb, u64 objectid,
 		return ERR_PTR(-ESTALE);
 
 	key.objectid = root_objectid;
-	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
+	key.type = BTRFS_ROOT_ITEM_KEY;
 	key.offset = (u64)-1;
 
 	index = srcu_read_lock(&fs_info->subvol_srcu);
@@ -82,7 +82,7 @@ static struct dentry *btrfs_get_dentry(struct super_block *sb, u64 objectid,
 	}
 
 	key.objectid = objectid;
-	btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
+	key.type = BTRFS_INODE_ITEM_KEY;
 	key.offset = 0;
 
 	inode = btrfs_iget(sb, &key, root, NULL);

fs/btrfs/extent-tree.c

@@ -491,7 +491,7 @@ next:
 							  key.objectid);
 			if (key.type == BTRFS_METADATA_ITEM_KEY)
 				last = key.objectid +
-					fs_info->tree_root->leafsize;
+					fs_info->tree_root->nodesize;
 			else
 				last = key.objectid + key.offset;
 
@@ -765,7 +765,7 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 	 * different
 	 */
 	if (metadata && !btrfs_fs_incompat(root->fs_info, SKINNY_METADATA)) {
-		offset = root->leafsize;
+		offset = root->nodesize;
 		metadata = 0;
 	}
 
@@ -799,13 +799,13 @@ again:
 					      path->slots[0]);
 			if (key.objectid == bytenr &&
 			    key.type == BTRFS_EXTENT_ITEM_KEY &&
-			    key.offset == root->leafsize)
+			    key.offset == root->nodesize)
 				ret = 0;
 		}
 		if (ret) {
 			key.objectid = bytenr;
 			key.type = BTRFS_EXTENT_ITEM_KEY;
-			key.offset = root->leafsize;
+			key.offset = root->nodesize;
 			btrfs_release_path(path);
 			goto again;
 		}
@@ -2651,7 +2651,7 @@ int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
 	num_bytes = btrfs_calc_trans_metadata_size(root, 1);
 	num_heads = heads_to_leaves(root, num_heads);
 	if (num_heads > 1)
-		num_bytes += (num_heads - 1) * root->leafsize;
+		num_bytes += (num_heads - 1) * root->nodesize;
 	num_bytes <<= 1;
 	global_rsv = &root->fs_info->global_block_rsv;
 
@@ -3073,10 +3073,10 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
 	int (*process_func)(struct btrfs_trans_handle *, struct btrfs_root *,
 			    u64, u64, u64, u64, u64, u64, int);
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state)))
+
+	if (btrfs_test_is_dummy_root(root))
 		return 0;
-#endif
+
 	ref_root = btrfs_header_owner(buf);
 	nritems = btrfs_header_nritems(buf);
 	level = btrfs_header_level(buf);
@@ -3097,7 +3097,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
 	for (i = 0; i < nritems; i++) {
 		if (level == 0) {
 			btrfs_item_key_to_cpu(buf, &key, i);
-			if (btrfs_key_type(&key) != BTRFS_EXTENT_DATA_KEY)
+			if (key.type != BTRFS_EXTENT_DATA_KEY)
 				continue;
 			fi = btrfs_item_ptr(buf, i,
 					    struct btrfs_file_extent_item);
@@ -3117,7 +3117,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
 				goto fail;
 		} else {
 			bytenr = btrfs_node_blockptr(buf, i);
-			num_bytes = btrfs_level_size(root, level - 1);
+			num_bytes = root->nodesize;
 			ret = process_func(trans, root, bytenr, num_bytes,
 					   parent, ref_root, level - 1, 0,
 					   1);
@@ -4343,11 +4343,21 @@ static inline int need_do_async_reclaim(struct btrfs_space_info *space_info,
 }
 
 static int btrfs_need_do_async_reclaim(struct btrfs_space_info *space_info,
-				       struct btrfs_fs_info *fs_info)
+				       struct btrfs_fs_info *fs_info,
+				       int flush_state)
 {
 	u64 used;
 
 	spin_lock(&space_info->lock);
+	/*
+	 * We run out of space and have not got any free space via flush_space,
+	 * so don't bother doing async reclaim.
+	 */
+	if (flush_state > COMMIT_TRANS && space_info->full) {
+		spin_unlock(&space_info->lock);
+		return 0;
+	}
+
 	used = space_info->bytes_used + space_info->bytes_reserved +
 	       space_info->bytes_pinned + space_info->bytes_readonly +
 	       space_info->bytes_may_use;
@@ -4380,11 +4390,12 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work)
 		flush_space(fs_info->fs_root, space_info, to_reclaim,
 			    to_reclaim, flush_state);
 		flush_state++;
-		if (!btrfs_need_do_async_reclaim(space_info, fs_info))
+		if (!btrfs_need_do_async_reclaim(space_info, fs_info,
+						 flush_state))
 			return;
 	} while (flush_state <= COMMIT_TRANS);
 
-	if (btrfs_need_do_async_reclaim(space_info, fs_info))
+	if (btrfs_need_do_async_reclaim(space_info, fs_info, flush_state))
 		queue_work(system_unbound_wq, work);
 }
 
@@ -4502,7 +4513,13 @@ again:
 		space_info->flush = 1;
 	} else if (!ret && space_info->flags & BTRFS_BLOCK_GROUP_METADATA) {
 		used += orig_bytes;
-		if (need_do_async_reclaim(space_info, root->fs_info, used) &&
+		/*
+		 * We will do the space reservation dance during log replay,
+		 * which means we won't have fs_info->fs_root set, so don't do
+		 * the async reclaim as we will panic.
+		 */
+		if (!root->fs_info->log_root_recovering &&
+		    need_do_async_reclaim(space_info, root->fs_info, used) &&
 		    !work_busy(&root->fs_info->async_reclaim_work))
 			queue_work(system_unbound_wq,
 				   &root->fs_info->async_reclaim_work);
@@ -4839,7 +4856,7 @@ static u64 calc_global_metadata_size(struct btrfs_fs_info *fs_info)
 	if (num_bytes * 3 > meta_used)
 		num_bytes = div64_u64(meta_used, 3);
 
-	return ALIGN(num_bytes, fs_info->extent_root->leafsize << 10);
+	return ALIGN(num_bytes, fs_info->extent_root->nodesize << 10);
 }
 
 static void update_global_block_rsv(struct btrfs_fs_info *fs_info)
@@ -4988,7 +5005,7 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 
 	if (root->fs_info->quota_enabled) {
 		/* One for parent inode, two for dir entries */
-		num_bytes = 3 * root->leafsize;
+		num_bytes = 3 * root->nodesize;
 		ret = btrfs_qgroup_reserve(root, num_bytes);
 		if (ret)
 			return ret;
@@ -5176,7 +5193,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 
 	if (root->fs_info->quota_enabled) {
 		ret = btrfs_qgroup_reserve(root, num_bytes +
-					   nr_extents * root->leafsize);
+					   nr_extents * root->nodesize);
 		if (ret)
 			goto out_fail;
 	}
@@ -5185,7 +5202,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	if (unlikely(ret)) {
 		if (root->fs_info->quota_enabled)
 			btrfs_qgroup_free(root, num_bytes +
-						nr_extents * root->leafsize);
+						nr_extents * root->nodesize);
 		goto out_fail;
 	}
 
@@ -5301,7 +5318,7 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes)
 				      btrfs_ino(inode), to_free, 0);
 	if (root->fs_info->quota_enabled) {
 		btrfs_qgroup_free(root, num_bytes +
-					dropped * root->leafsize);
+					dropped * root->nodesize);
 	}
 
 	btrfs_block_rsv_release(root, &root->fs_info->delalloc_block_rsv,
@@ -5422,6 +5439,20 @@ static int update_block_group(struct btrfs_root *root,
 			spin_unlock(&cache->space_info->lock);
 		} else {
 			old_val -= num_bytes;
+
+			/*
+			 * No longer have used bytes in this block group, queue
+			 * it for deletion.
+			 */
+			if (old_val == 0) {
+				spin_lock(&info->unused_bgs_lock);
+				if (list_empty(&cache->bg_list)) {
+					btrfs_get_block_group(cache);
+					list_add_tail(&cache->bg_list,
+						      &info->unused_bgs);
+				}
+				spin_unlock(&info->unused_bgs_lock);
+			}
 			btrfs_set_block_group_used(&cache->item, old_val);
 			cache->pinned += num_bytes;
 			cache->space_info->bytes_pinned += num_bytes;
@@ -6233,10 +6264,9 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	int ret;
 	struct btrfs_fs_info *fs_info = root->fs_info;
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state)))
+	if (btrfs_test_is_dummy_root(root))
 		return 0;
-#endif
+
 	add_pinned_bytes(root->fs_info, num_bytes, owner, root_objectid);
 
 	/*
@@ -6263,14 +6293,6 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	return ret;
 }
 
-static u64 stripe_align(struct btrfs_root *root,
-			struct btrfs_block_group_cache *cache,
-			u64 val, u64 num_bytes)
-{
-	u64 ret = ALIGN(val, root->stripesize);
-	return ret;
-}
-
 /*
  * when we wait for progress in the block group caching, its because
  * our allocation attempt failed at least once.  So, we must sleep
@@ -6464,7 +6486,7 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	bool have_caching_bg = false;
 
 	WARN_ON(num_bytes < root->sectorsize);
-	btrfs_set_key_type(ins, BTRFS_EXTENT_ITEM_KEY);
+	ins->type = BTRFS_EXTENT_ITEM_KEY;
 	ins->objectid = 0;
 	ins->offset = 0;
 
@@ -6751,8 +6773,7 @@ unclustered_alloc:
 			goto loop;
 		}
 checks:
-		search_start = stripe_align(root, block_group,
-					    offset, num_bytes);
+		search_start = ALIGN(offset, root->stripesize);
 
 		/* move on to the next group */
 		if (search_start + num_bytes >
@@ -7077,7 +7098,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 	path = btrfs_alloc_path();
 	if (!path) {
 		btrfs_free_and_pin_reserved_extent(root, ins->objectid,
-						   root->leafsize);
+						   root->nodesize);
 		return -ENOMEM;
 	}
 
@@ -7086,7 +7107,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 				      ins, size);
 	if (ret) {
 		btrfs_free_and_pin_reserved_extent(root, ins->objectid,
-						   root->leafsize);
+						   root->nodesize);
 		btrfs_free_path(path);
 		return ret;
 	}
@@ -7101,7 +7122,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 
 	if (skinny_metadata) {
 		iref = (struct btrfs_extent_inline_ref *)(extent_item + 1);
-		num_bytes = root->leafsize;
+		num_bytes = root->nodesize;
 	} else {
 		block_info = (struct btrfs_tree_block_info *)(extent_item + 1);
 		btrfs_set_tree_block_key(leaf, block_info, key);
@@ -7131,14 +7152,14 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 			return ret;
 	}
 
-	ret = update_block_group(root, ins->objectid, root->leafsize, 1);
+	ret = update_block_group(root, ins->objectid, root->nodesize, 1);
 	if (ret) { /* -ENOENT, logic error */
 		btrfs_err(fs_info, "update block group failed for %llu %llu",
 			ins->objectid, ins->offset);
 		BUG();
 	}
 
-	trace_btrfs_reserved_extent_alloc(root, ins->objectid, root->leafsize);
+	trace_btrfs_reserved_extent_alloc(root, ins->objectid, root->nodesize);
 	return ret;
 }
 
@@ -7213,17 +7234,19 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	btrfs_set_buffer_uptodate(buf);
 
 	if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
+		buf->log_index = root->log_transid % 2;
 		/*
 		 * we allow two log transactions at a time, use different
 		 * EXENT bit to differentiate dirty pages.
 		 */
-		if (root->log_transid % 2 == 0)
+		if (buf->log_index == 0)
 			set_extent_dirty(&root->dirty_log_pages, buf->start,
 					buf->start + buf->len - 1, GFP_NOFS);
 		else
 			set_extent_new(&root->dirty_log_pages, buf->start,
 					buf->start + buf->len - 1, GFP_NOFS);
 	} else {
+		buf->log_index = -1;
 		set_extent_dirty(&trans->transaction->dirty_pages, buf->start,
 			 buf->start + buf->len - 1, GFP_NOFS);
 	}
@@ -7300,8 +7323,8 @@ static void unuse_block_rsv(struct btrfs_fs_info *fs_info,
  *
  * returns the tree buffer or NULL.
  */
-struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
-					struct btrfs_root *root, u32 blocksize,
+struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
+					struct btrfs_root *root,
 					u64 parent, u64 root_objectid,
 					struct btrfs_disk_key *key, int level,
 					u64 hint, u64 empty_size)
@@ -7311,18 +7334,18 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
 	struct extent_buffer *buf;
 	u64 flags = 0;
 	int ret;
+	u32 blocksize = root->nodesize;
 	bool skinny_metadata = btrfs_fs_incompat(root->fs_info,
 						 SKINNY_METADATA);
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state))) {
+	if (btrfs_test_is_dummy_root(root)) {
 		buf = btrfs_init_new_buffer(trans, root, root->alloc_bytenr,
 					    blocksize, level);
 		if (!IS_ERR(buf))
 			root->alloc_bytenr += blocksize;
 		return buf;
 	}
-#endif
+
 	block_rsv = use_block_rsv(trans, root, blocksize);
 	if (IS_ERR(block_rsv))
 		return ERR_CAST(block_rsv);
@@ -7417,7 +7440,7 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
 
 	eb = path->nodes[wc->level];
 	nritems = btrfs_header_nritems(eb);
-	blocksize = btrfs_level_size(root, wc->level - 1);
+	blocksize = root->nodesize;
 
 	for (slot = path->slots[wc->level]; slot < nritems; slot++) {
 		if (nread >= wc->reada_count)
@@ -7464,10 +7487,7 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
 				continue;
 		}
 reada:
-		ret = readahead_tree_block(root, bytenr, blocksize,
-					   generation);
-		if (ret)
-			break;
+		readahead_tree_block(root, bytenr, blocksize);
 		nread++;
 	}
 	wc->reada_slot = slot;
@@ -7626,7 +7646,6 @@ walk_down:
 	level = root_level;
 	while (level >= 0) {
 		if (path->nodes[level] == NULL) {
-			int child_bsize = root->nodesize;
 			int parent_slot;
 			u64 child_gen;
 			u64 child_bytenr;
@@ -7638,8 +7657,7 @@ walk_down:
 			child_bytenr = btrfs_node_blockptr(eb, parent_slot);
 			child_gen = btrfs_node_ptr_generation(eb, parent_slot);
 
-			eb = read_tree_block(root, child_bytenr, child_bsize,
-					     child_gen);
+			eb = read_tree_block(root, child_bytenr, child_gen);
 			if (!eb || !extent_buffer_uptodate(eb)) {
 				ret = -EIO;
 				goto out;
@@ -7655,7 +7673,7 @@ walk_down:
 			ret = btrfs_qgroup_record_ref(trans, root->fs_info,
 						root->objectid,
 						child_bytenr,
-						child_bsize,
+						root->nodesize,
 						BTRFS_QGROUP_OPER_SUB_SUBTREE,
 						0);
 			if (ret)
@@ -7806,9 +7824,9 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	}
 
 	bytenr = btrfs_node_blockptr(path->nodes[level], path->slots[level]);
-	blocksize = btrfs_level_size(root, level - 1);
+	blocksize = root->nodesize;
 
-	next = btrfs_find_tree_block(root, bytenr, blocksize);
+	next = btrfs_find_tree_block(root, bytenr);
 	if (!next) {
 		next = btrfs_find_create_tree_block(root, bytenr, blocksize);
 		if (!next)
@@ -7870,7 +7888,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	if (!next) {
 		if (reada && level == 1)
 			reada_walk_down(trans, root, wc, path);
-		next = read_tree_block(root, bytenr, blocksize, generation);
+		next = read_tree_block(root, bytenr, generation);
 		if (!next || !extent_buffer_uptodate(next)) {
 			free_extent_buffer(next);
 			return -EIO;
@@ -8853,6 +8871,16 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 	}
 	up_write(&info->commit_root_sem);
 
+	spin_lock(&info->unused_bgs_lock);
+	while (!list_empty(&info->unused_bgs)) {
+		block_group = list_first_entry(&info->unused_bgs,
+					       struct btrfs_block_group_cache,
+					       bg_list);
+		list_del_init(&block_group->bg_list);
+		btrfs_put_block_group(block_group);
+	}
+	spin_unlock(&info->unused_bgs_lock);
+
 	spin_lock(&info->block_group_cache_lock);
 	while ((n = rb_last(&info->block_group_cache_tree)) != NULL) {
 		block_group = rb_entry(n, struct btrfs_block_group_cache,
@@ -8987,7 +9015,7 @@ btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size)
 	init_rwsem(&cache->data_rwsem);
 	INIT_LIST_HEAD(&cache->list);
 	INIT_LIST_HEAD(&cache->cluster_list);
-	INIT_LIST_HEAD(&cache->new_bg_list);
+	INIT_LIST_HEAD(&cache->bg_list);
 	btrfs_init_free_space_ctl(cache);
 
 	return cache;
@@ -9009,7 +9037,7 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 	root = info->extent_root;
 	key.objectid = 0;
 	key.offset = 0;
-	btrfs_set_key_type(&key, BTRFS_BLOCK_GROUP_ITEM_KEY);
+	key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -9128,8 +9156,18 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		__link_block_group(space_info, cache);
 
 		set_avail_alloc_bits(root->fs_info, cache->flags);
-		if (btrfs_chunk_readonly(root, cache->key.objectid))
+		if (btrfs_chunk_readonly(root, cache->key.objectid)) {
 			set_block_group_ro(cache, 1);
+		} else if (btrfs_block_group_used(&cache->item) == 0) {
+			spin_lock(&info->unused_bgs_lock);
+			/* Should always be true but just in case. */
+			if (list_empty(&cache->bg_list)) {
+				btrfs_get_block_group(cache);
+				list_add_tail(&cache->bg_list,
+					      &info->unused_bgs);
+			}
+			spin_unlock(&info->unused_bgs_lock);
+		}
 	}
 
 	list_for_each_entry_rcu(space_info, &root->fs_info->space_info, list) {
@@ -9170,10 +9208,8 @@ void btrfs_create_pending_block_groups(struct btrfs_trans_handle *trans,
 	struct btrfs_key key;
 	int ret = 0;
 
-	list_for_each_entry_safe(block_group, tmp, &trans->new_bgs,
-				 new_bg_list) {
-		list_del_init(&block_group->new_bg_list);
-
+	list_for_each_entry_safe(block_group, tmp, &trans->new_bgs, bg_list) {
+		list_del_init(&block_group->bg_list);
 		if (ret)
 			continue;
 
@@ -9259,7 +9295,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 
 	__link_block_group(cache->space_info, cache);
 
-	list_add_tail(&cache->new_bg_list, &trans->new_bgs);
+	list_add_tail(&cache->bg_list, &trans->new_bgs);
 
 	set_avail_alloc_bits(extent_root->fs_info, type);
 
@@ -9413,8 +9449,6 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 
 	memcpy(&key, &block_group->key, sizeof(key));
 
-	btrfs_clear_space_info_full(root->fs_info);
-
 	btrfs_put_block_group(block_group);
 	btrfs_put_block_group(block_group);
 
@@ -9430,6 +9464,101 @@ out:
 	return ret;
 }
 
+/*
+ * Process the unused_bgs list and remove any that don't have any allocated
+ * space inside of them.
+ */
+void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_block_group_cache *block_group;
+	struct btrfs_space_info *space_info;
+	struct btrfs_root *root = fs_info->extent_root;
+	struct btrfs_trans_handle *trans;
+	int ret = 0;
+
+	if (!fs_info->open)
+		return;
+
+	spin_lock(&fs_info->unused_bgs_lock);
+	while (!list_empty(&fs_info->unused_bgs)) {
+		u64 start, end;
+
+		block_group = list_first_entry(&fs_info->unused_bgs,
+					       struct btrfs_block_group_cache,
+					       bg_list);
+		space_info = block_group->space_info;
+		list_del_init(&block_group->bg_list);
+		if (ret || btrfs_mixed_space_info(space_info)) {
+			btrfs_put_block_group(block_group);
+			continue;
+		}
+		spin_unlock(&fs_info->unused_bgs_lock);
+
+		/* Don't want to race with allocators so take the groups_sem */
+		down_write(&space_info->groups_sem);
+		spin_lock(&block_group->lock);
+		if (block_group->reserved ||
+		    btrfs_block_group_used(&block_group->item) ||
+		    block_group->ro) {
+			/*
+			 * We want to bail if we made new allocations or have
+			 * outstanding allocations in this block group.  We do
+			 * the ro check in case balance is currently acting on
+			 * this block group.
+			 */
+			spin_unlock(&block_group->lock);
+			up_write(&space_info->groups_sem);
+			goto next;
+		}
+		spin_unlock(&block_group->lock);
+
+		/* We don't want to force the issue, only flip if it's ok. */
+		ret = set_block_group_ro(block_group, 0);
+		up_write(&space_info->groups_sem);
+		if (ret < 0) {
+			ret = 0;
+			goto next;
+		}
+
+		/*
+		 * Want to do this before we do anything else so we can recover
+		 * properly if we fail to join the transaction.
+		 */
+		trans = btrfs_join_transaction(root);
+		if (IS_ERR(trans)) {
+			btrfs_set_block_group_rw(root, block_group);
+			ret = PTR_ERR(trans);
+			goto next;
+		}
+
+		/*
+		 * We could have pending pinned extents for this block group,
+		 * just delete them, we don't care about them anymore.
+		 */
+		start = block_group->key.objectid;
+		end = start + block_group->key.offset - 1;
+		clear_extent_bits(&fs_info->freed_extents[0], start, end,
+				  EXTENT_DIRTY, GFP_NOFS);
+		clear_extent_bits(&fs_info->freed_extents[1], start, end,
+				  EXTENT_DIRTY, GFP_NOFS);
+
+		/* Reset pinned so btrfs_put_block_group doesn't complain */
+		block_group->pinned = 0;
+
+		/*
+		 * Btrfs_remove_chunk will abort the transaction if things go
+		 * horribly wrong.
+		 */
+		ret = btrfs_remove_chunk(trans, root,
+					 block_group->key.objectid);
+		btrfs_end_transaction(trans, root);
+next:
+		btrfs_put_block_group(block_group);
+		spin_lock(&fs_info->unused_bgs_lock);
+	}
+	spin_unlock(&fs_info->unused_bgs_lock);
+}
+
 int btrfs_init_space_info(struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_space_info *space_info;
@@ -9561,7 +9690,7 @@ void btrfs_end_nocow_write(struct btrfs_root *root)
 
 int btrfs_start_nocow_write(struct btrfs_root *root)
 {
-	if (unlikely(atomic_read(&root->will_be_snapshoted)))
+	if (atomic_read(&root->will_be_snapshoted))
 		return 0;
 
 	percpu_counter_inc(&root->subv_writers->counter);
@@ -9569,7 +9698,7 @@ int btrfs_start_nocow_write(struct btrfs_root *root)
 	 * Make sure counter is updated before we check for snapshot creation.
 	 */
 	smp_mb();
-	if (unlikely(atomic_read(&root->will_be_snapshoted))) {
+	if (atomic_read(&root->will_be_snapshoted)) {
 		btrfs_end_nocow_write(root);
 		return 0;
 	}

fs/btrfs/extent_io.c

@@ -25,6 +25,11 @@ static struct kmem_cache *extent_state_cache;
 static struct kmem_cache *extent_buffer_cache;
 static struct bio_set *btrfs_bioset;
 
+static inline bool extent_state_in_tree(const struct extent_state *state)
+{
+	return !RB_EMPTY_NODE(&state->rb_node);
+}
+
 #ifdef CONFIG_BTRFS_DEBUG
 static LIST_HEAD(buffers);
 static LIST_HEAD(states);
@@ -59,9 +64,9 @@ void btrfs_leak_debug_check(void)
 
 	while (!list_empty(&states)) {
 		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",
-		       state->start, state->end, state->state, state->tree,
+		pr_err("BTRFS: state leak: start %llu end %llu state %lu in tree %d refs %d\n",
+		       state->start, state->end, state->state,
+		       extent_state_in_tree(state),
 		       atomic_read(&state->refs));
 		list_del(&state->leak_list);
 		kmem_cache_free(extent_state_cache, state);
@@ -209,7 +214,7 @@ static struct extent_state *alloc_extent_state(gfp_t mask)
 		return state;
 	state->state = 0;
 	state->private = 0;
-	state->tree = NULL;
+	RB_CLEAR_NODE(&state->rb_node);
 	btrfs_leak_debug_add(&state->leak_list, &states);
 	atomic_set(&state->refs, 1);
 	init_waitqueue_head(&state->wq);
@@ -222,7 +227,7 @@ void free_extent_state(struct extent_state *state)
 	if (!state)
 		return;
 	if (atomic_dec_and_test(&state->refs)) {
-		WARN_ON(state->tree);
+		WARN_ON(extent_state_in_tree(state));
 		btrfs_leak_debug_del(&state->leak_list);
 		trace_free_extent_state(state, _RET_IP_);
 		kmem_cache_free(extent_state_cache, state);
@@ -371,8 +376,8 @@ static void merge_state(struct extent_io_tree *tree,
 		    other->state == state->state) {
 			merge_cb(tree, state, other);
 			state->start = other->start;
-			other->tree = NULL;
 			rb_erase(&other->rb_node, &tree->state);
+			RB_CLEAR_NODE(&other->rb_node);
 			free_extent_state(other);
 		}
 	}
@@ -383,8 +388,8 @@ static void merge_state(struct extent_io_tree *tree,
 		    other->state == state->state) {
 			merge_cb(tree, state, other);
 			state->end = other->end;
-			other->tree = NULL;
 			rb_erase(&other->rb_node, &tree->state);
+			RB_CLEAR_NODE(&other->rb_node);
 			free_extent_state(other);
 		}
 	}
@@ -442,7 +447,6 @@ static int insert_state(struct extent_io_tree *tree,
 		       found->start, found->end, start, end);
 		return -EEXIST;
 	}
-	state->tree = tree;
 	merge_state(tree, state);
 	return 0;
 }
@@ -486,7 +490,6 @@ static int split_state(struct extent_io_tree *tree, struct extent_state *orig,
 		free_extent_state(prealloc);
 		return -EEXIST;
 	}
-	prealloc->tree = tree;
 	return 0;
 }
 
@@ -524,9 +527,9 @@ static struct extent_state *clear_state_bit(struct extent_io_tree *tree,
 		wake_up(&state->wq);
 	if (state->state == 0) {
 		next = next_state(state);
-		if (state->tree) {
+		if (extent_state_in_tree(state)) {
 			rb_erase(&state->rb_node, &tree->state);
-			state->tree = NULL;
+			RB_CLEAR_NODE(&state->rb_node);
 			free_extent_state(state);
 		} else {
 			WARN_ON(1);
@@ -606,8 +609,8 @@ again:
 			cached_state = NULL;
 		}
 
-		if (cached && cached->tree && cached->start <= start &&
-		    cached->end > start) {
+		if (cached && extent_state_in_tree(cached) &&
+		    cached->start <= start && cached->end > start) {
 			if (clear)
 				atomic_dec(&cached->refs);
 			state = cached;
@@ -843,7 +846,7 @@ again:
 	if (cached_state && *cached_state) {
 		state = *cached_state;
 		if (state->start <= start && state->end > start &&
-		    state->tree) {
+		    extent_state_in_tree(state)) {
 			node = &state->rb_node;
 			goto hit_next;
 		}
@@ -1069,7 +1072,7 @@ again:
 	if (cached_state && *cached_state) {
 		state = *cached_state;
 		if (state->start <= start && state->end > start &&
-		    state->tree) {
+		    extent_state_in_tree(state)) {
 			node = &state->rb_node;
 			goto hit_next;
 		}
@@ -1459,7 +1462,7 @@ int find_first_extent_bit(struct extent_io_tree *tree, u64 start,
 	spin_lock(&tree->lock);
 	if (cached_state && *cached_state) {
 		state = *cached_state;
-		if (state->end == start - 1 && state->tree) {
+		if (state->end == start - 1 && extent_state_in_tree(state)) {
 			n = rb_next(&state->rb_node);
 			while (n) {
 				state = rb_entry(n, struct extent_state,
@@ -1905,7 +1908,7 @@ int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,
 	int bitset = 0;
 
 	spin_lock(&tree->lock);
-	if (cached && cached->tree && cached->start <= start &&
+	if (cached && extent_state_in_tree(cached) && cached->start <= start &&
 	    cached->end > start)
 		node = &cached->rb_node;
 	else
@@ -1959,27 +1962,7 @@ static void check_page_uptodate(struct extent_io_tree *tree, struct page *page)
 		SetPageUptodate(page);
 }
 
-/*
- * When IO fails, either with EIO or csum verification fails, we
- * try other mirrors that might have a good copy of the data.  This
- * io_failure_record is used to record state as we go through all the
- * mirrors.  If another mirror has good data, the page is set up to date
- * and things continue.  If a good mirror can't be found, the original
- * bio end_io callback is called to indicate things have failed.
- */
-struct io_failure_record {
-	struct page *page;
-	u64 start;
-	u64 len;
-	u64 logical;
-	unsigned long bio_flags;
-	int this_mirror;
-	int failed_mirror;
-	int in_validation;
-};
-
-static int free_io_failure(struct inode *inode, struct io_failure_record *rec,
-				int did_repair)
+int free_io_failure(struct inode *inode, struct io_failure_record *rec)
 {
 	int ret;
 	int err = 0;
@@ -2012,10 +1995,10 @@ static int free_io_failure(struct inode *inode, struct io_failure_record *rec,
  * currently, there can be no more than two copies of every data bit. thus,
  * exactly one rewrite is required.
  */
-int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
-			u64 length, u64 logical, struct page *page,
-			int mirror_num)
+int repair_io_failure(struct inode *inode, u64 start, u64 length, u64 logical,
+		      struct page *page, unsigned int pg_offset, int mirror_num)
 {
+	struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
 	struct bio *bio;
 	struct btrfs_device *dev;
 	u64 map_length = 0;
@@ -2053,7 +2036,7 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
 		return -EIO;
 	}
 	bio->bi_bdev = dev->bdev;
-	bio_add_page(bio, page, length, start - page_offset(page));
+	bio_add_page(bio, page, length, pg_offset);
 
 	if (btrfsic_submit_bio_wait(WRITE_SYNC, bio)) {
 		/* try to remap that extent elsewhere? */
@@ -2063,10 +2046,9 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
 	}
 
 	printk_ratelimited_in_rcu(KERN_INFO
-			"BTRFS: read error corrected: ino %lu off %llu "
-		    "(dev %s sector %llu)\n", page->mapping->host->i_ino,
-		    start, rcu_str_deref(dev->name), sector);
-
+				  "BTRFS: read error corrected: ino %llu off %llu (dev %s sector %llu)\n",
+				  btrfs_ino(inode), start,
+				  rcu_str_deref(dev->name), sector);
 	bio_put(bio);
 	return 0;
 }
@@ -2082,9 +2064,11 @@ int repair_eb_io_failure(struct btrfs_root *root, struct extent_buffer *eb,
 		return -EROFS;
 
 	for (i = 0; i < num_pages; i++) {
-		struct page *p = extent_buffer_page(eb, i);
-		ret = repair_io_failure(root->fs_info, start, PAGE_CACHE_SIZE,
-					start, p, mirror_num);
+		struct page *p = eb->pages[i];
+
+		ret = repair_io_failure(root->fs_info->btree_inode, start,
+					PAGE_CACHE_SIZE, start, p,
+					start - page_offset(p), mirror_num);
 		if (ret)
 			break;
 		start += PAGE_CACHE_SIZE;
@@ -2097,16 +2081,15 @@ int repair_eb_io_failure(struct btrfs_root *root, struct extent_buffer *eb,
  * each time an IO finishes, we do a fast check in the IO failure tree
  * to see if we need to process or clean up an io_failure_record
  */
-static int clean_io_failure(u64 start, struct page *page)
+int clean_io_failure(struct inode *inode, u64 start, struct page *page,
+		     unsigned int pg_offset)
 {
 	u64 private;
 	u64 private_failure;
 	struct io_failure_record *failrec;
-	struct inode *inode = page->mapping->host;
 	struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
 	struct extent_state *state;
 	int num_copies;
-	int did_repair = 0;
 	int ret;
 
 	private = 0;
@@ -2127,7 +2110,6 @@ static int clean_io_failure(u64 start, struct page *page)
 		/* there was no real error, just free the record */
 		pr_debug("clean_io_failure: freeing dummy error at %llu\n",
 			 failrec->start);
-		did_repair = 1;
 		goto out;
 	}
 	if (fs_info->sb->s_flags & MS_RDONLY)
@@ -2144,55 +2126,70 @@ static int clean_io_failure(u64 start, struct page *page)
 		num_copies = btrfs_num_copies(fs_info, failrec->logical,
 					      failrec->len);
 		if (num_copies > 1)  {
-			ret = repair_io_failure(fs_info, start, failrec->len,
-						failrec->logical, page,
-						failrec->failed_mirror);
-			did_repair = !ret;
+			repair_io_failure(inode, start, failrec->len,
+					  failrec->logical, page,
+					  pg_offset, failrec->failed_mirror);
 		}
-		ret = 0;
 	}
 
 out:
-	if (!ret)
-		ret = free_io_failure(inode, failrec, did_repair);
+	free_io_failure(inode, failrec);
 
-	return ret;
+	return 0;
 }
 
 /*
- * this is a generic handler for readpage errors (default
- * readpage_io_failed_hook). if other copies exist, read those and write back
- * good data to the failed position. does not investigate in remapping the
- * failed extent elsewhere, hoping the device will be smart enough to do this as
- * needed
+ * Can be called when
+ * - hold extent lock
+ * - under ordered extent
+ * - the inode is freeing
  */
+void btrfs_free_io_failure_record(struct inode *inode, u64 start, u64 end)
+{
+	struct extent_io_tree *failure_tree = &BTRFS_I(inode)->io_failure_tree;
+	struct io_failure_record *failrec;
+	struct extent_state *state, *next;
 
-static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
-			      struct page *page, u64 start, u64 end,
-			      int failed_mirror)
+	if (RB_EMPTY_ROOT(&failure_tree->state))
+		return;
+
+	spin_lock(&failure_tree->lock);
+	state = find_first_extent_bit_state(failure_tree, start, EXTENT_DIRTY);
+	while (state) {
+		if (state->start > end)
+			break;
+
+		ASSERT(state->end <= end);
+
+		next = next_state(state);
+
+		failrec = (struct io_failure_record *)state->private;
+		free_extent_state(state);
+		kfree(failrec);
+
+		state = next;
+	}
+	spin_unlock(&failure_tree->lock);
+}
+
+int btrfs_get_io_failure_record(struct inode *inode, u64 start, u64 end,
+				struct io_failure_record **failrec_ret)
 {
-	struct io_failure_record *failrec = NULL;
+	struct io_failure_record *failrec;
 	u64 private;
 	struct extent_map *em;
-	struct inode *inode = page->mapping->host;
 	struct extent_io_tree *failure_tree = &BTRFS_I(inode)->io_failure_tree;
 	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;
 	u64 logical;
 
-	BUG_ON(failed_bio->bi_rw & REQ_WRITE);
-
 	ret = get_state_private(failure_tree, start, &private);
 	if (ret) {
 		failrec = kzalloc(sizeof(*failrec), GFP_NOFS);
 		if (!failrec)
 			return -ENOMEM;
+
 		failrec->start = start;
 		failrec->len = end - start + 1;
 		failrec->this_mirror = 0;
@@ -2212,11 +2209,11 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 			em = NULL;
 		}
 		read_unlock(&em_tree->lock);
-
 		if (!em) {
 			kfree(failrec);
 			return -EIO;
 		}
+
 		logical = start - em->start;
 		logical = em->block_start + logical;
 		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) {
@@ -2225,8 +2222,10 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 			extent_set_compress_type(&failrec->bio_flags,
 						 em->compress_type);
 		}
-		pr_debug("bio_readpage_error: (new) logical=%llu, start=%llu, "
-			 "len=%llu\n", logical, start, failrec->len);
+
+		pr_debug("Get IO Failure Record: (new) logical=%llu, start=%llu, len=%llu\n",
+			 logical, start, failrec->len);
+
 		failrec->logical = logical;
 		free_extent_map(em);
 
@@ -2246,8 +2245,7 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 		}
 	} else {
 		failrec = (struct io_failure_record *)(unsigned long)private;
-		pr_debug("bio_readpage_error: (found) logical=%llu, "
-			 "start=%llu, len=%llu, validation=%d\n",
+		pr_debug("Get IO Failure Record: (found) logical=%llu, start=%llu, len=%llu, validation=%d\n",
 			 failrec->logical, failrec->start, failrec->len,
 			 failrec->in_validation);
 		/*
@@ -2256,6 +2254,17 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 		 * clean_io_failure() clean all those errors at once.
 		 */
 	}
+
+	*failrec_ret = failrec;
+
+	return 0;
+}
+
+int btrfs_check_repairable(struct inode *inode, struct bio *failed_bio,
+			   struct io_failure_record *failrec, int failed_mirror)
+{
+	int num_copies;
+
 	num_copies = btrfs_num_copies(BTRFS_I(inode)->root->fs_info,
 				      failrec->logical, failrec->len);
 	if (num_copies == 1) {
@@ -2264,10 +2273,9 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 		 * 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=%d, next_mirror %d, failed_mirror %d\n",
+		pr_debug("Check Repairable: 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;
+		return 0;
 	}
 
 	/*
@@ -2287,7 +2295,6 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 		BUG_ON(failrec->in_validation);
 		failrec->in_validation = 1;
 		failrec->this_mirror = failed_mirror;
-		read_mode = READ_SYNC | REQ_FAILFAST_DEV;
 	} else {
 		/*
 		 * we're ready to fulfill a) and b) alongside. get a good copy
@@ -2303,25 +2310,36 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 		failrec->this_mirror++;
 		if (failrec->this_mirror == failed_mirror)
 			failrec->this_mirror++;
-		read_mode = READ_SYNC;
 	}
 
 	if (failrec->this_mirror > num_copies) {
-		pr_debug("bio_readpage_error: (fail) num_copies=%d, next_mirror %d, failed_mirror %d\n",
+		pr_debug("Check Repairable: (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;
+		return 0;
 	}
 
+	return 1;
+}
+
+
+struct bio *btrfs_create_repair_bio(struct inode *inode, struct bio *failed_bio,
+				    struct io_failure_record *failrec,
+				    struct page *page, int pg_offset, int icsum,
+				    bio_end_io_t *endio_func, void *data)
+{
+	struct bio *bio;
+	struct btrfs_io_bio *btrfs_failed_bio;
+	struct btrfs_io_bio *btrfs_bio;
+
 	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
-	if (!bio) {
-		free_io_failure(inode, failrec, 0);
-		return -EIO;
-	}
-	bio->bi_end_io = failed_bio->bi_end_io;
+	if (!bio)
+		return NULL;
+
+	bio->bi_end_io = endio_func;
 	bio->bi_iter.bi_sector = failrec->logical >> 9;
 	bio->bi_bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
 	bio->bi_iter.bi_size = 0;
+	bio->bi_private = data;
 
 	btrfs_failed_bio = btrfs_io_bio(failed_bio);
 	if (btrfs_failed_bio->csum) {
@@ -2330,21 +2348,73 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 
 		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,
+		icsum *= csum_size;
+		memcpy(btrfs_bio->csum, btrfs_failed_bio->csum + icsum,
 		       csum_size);
 	}
 
-	bio_add_page(bio, page, failrec->len, start - page_offset(page));
+	bio_add_page(bio, page, failrec->len, pg_offset);
+
+	return bio;
+}
+
+/*
+ * this is a generic handler for readpage errors (default
+ * readpage_io_failed_hook). if other copies exist, read those and write back
+ * good data to the failed position. does not investigate in remapping the
+ * failed extent elsewhere, hoping the device will be smart enough to do this as
+ * needed
+ */
+
+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;
+	struct inode *inode = page->mapping->host;
+	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;
+	struct bio *bio;
+	int read_mode;
+	int ret;
+
+	BUG_ON(failed_bio->bi_rw & REQ_WRITE);
+
+	ret = btrfs_get_io_failure_record(inode, start, end, &failrec);
+	if (ret)
+		return ret;
+
+	ret = btrfs_check_repairable(inode, failed_bio, failrec, failed_mirror);
+	if (!ret) {
+		free_io_failure(inode, failrec);
+		return -EIO;
+	}
+
+	if (failed_bio->bi_vcnt > 1)
+		read_mode = READ_SYNC | REQ_FAILFAST_DEV;
+	else
+		read_mode = READ_SYNC;
+
+	phy_offset >>= inode->i_sb->s_blocksize_bits;
+	bio = btrfs_create_repair_bio(inode, failed_bio, failrec, page,
+				      start - page_offset(page),
+				      (int)phy_offset, failed_bio->bi_end_io,
+				      NULL);
+	if (!bio) {
+		free_io_failure(inode, failrec);
+		return -EIO;
+	}
 
-	pr_debug("bio_readpage_error: submitting new read[%#x] to "
-		 "this_mirror=%d, num_copies=%d, in_validation=%d\n", read_mode,
-		 failrec->this_mirror, num_copies, failrec->in_validation);
+	pr_debug("Repair Read Error: submitting new read[%#x] to this_mirror=%d, in_validation=%d\n",
+		 read_mode, failrec->this_mirror, failrec->in_validation);
 
 	ret = tree->ops->submit_bio_hook(inode, read_mode, bio,
 					 failrec->this_mirror,
 					 failrec->bio_flags, 0);
+	if (ret) {
+		free_io_failure(inode, failrec);
+		bio_put(bio);
+	}
+
 	return ret;
 }
 
@@ -2469,7 +2539,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 		struct inode *inode = page->mapping->host;
 
 		pr_debug("end_bio_extent_readpage: bi_sector=%llu, err=%d, "
-			 "mirror=%lu\n", (u64)bio->bi_iter.bi_sector, err,
+			 "mirror=%u\n", (u64)bio->bi_iter.bi_sector, err,
 			 io_bio->mirror_num);
 		tree = &BTRFS_I(inode)->io_tree;
 
@@ -2503,7 +2573,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 			if (ret)
 				uptodate = 0;
 			else
-				clean_io_failure(start, page);
+				clean_io_failure(inode, start, page, 0);
 		}
 
 		if (likely(uptodate))
@@ -2540,12 +2610,12 @@ readpage_ok:
 		if (likely(uptodate)) {
 			loff_t i_size = i_size_read(inode);
 			pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
-			unsigned offset;
+			unsigned off;
 
 			/* Zero out the end if this page straddles i_size */
-			offset = i_size & (PAGE_CACHE_SIZE-1);
-			if (page->index == end_index && offset)
-				zero_user_segment(page, offset, PAGE_CACHE_SIZE);
+			off = i_size & (PAGE_CACHE_SIZE-1);
+			if (page->index == end_index && off)
+				zero_user_segment(page, off, PAGE_CACHE_SIZE);
 			SetPageUptodate(page);
 		} else {
 			ClearPageUptodate(page);
@@ -2618,9 +2688,18 @@ btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
 
 struct bio *btrfs_bio_clone(struct bio *bio, gfp_t gfp_mask)
 {
-	return bio_clone_bioset(bio, gfp_mask, btrfs_bioset);
-}
+	struct btrfs_io_bio *btrfs_bio;
+	struct bio *new;
 
+	new = bio_clone_bioset(bio, gfp_mask, btrfs_bioset);
+	if (new) {
+		btrfs_bio = btrfs_io_bio(new);
+		btrfs_bio->csum = NULL;
+		btrfs_bio->csum_allocated = NULL;
+		btrfs_bio->end_io = NULL;
+	}
+	return new;
+}
 
 /* this also allocates from the btrfs_bioset */
 struct bio *btrfs_io_bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs)
@@ -3501,7 +3580,7 @@ lock_extent_buffer_for_io(struct extent_buffer *eb,
 
 	num_pages = num_extent_pages(eb->start, eb->len);
 	for (i = 0; i < num_pages; i++) {
-		struct page *p = extent_buffer_page(eb, i);
+		struct page *p = eb->pages[i];
 
 		if (!trylock_page(p)) {
 			if (!flush) {
@@ -3522,6 +3601,68 @@ static void end_extent_buffer_writeback(struct extent_buffer *eb)
 	wake_up_bit(&eb->bflags, EXTENT_BUFFER_WRITEBACK);
 }
 
+static void set_btree_ioerr(struct page *page)
+{
+	struct extent_buffer *eb = (struct extent_buffer *)page->private;
+	struct btrfs_inode *btree_ino = BTRFS_I(eb->fs_info->btree_inode);
+
+	SetPageError(page);
+	if (test_and_set_bit(EXTENT_BUFFER_WRITE_ERR, &eb->bflags))
+		return;
+
+	/*
+	 * If writeback for a btree extent that doesn't belong to a log tree
+	 * failed, increment the counter transaction->eb_write_errors.
+	 * We do this because while the transaction is running and before it's
+	 * committing (when we call filemap_fdata[write|wait]_range against
+	 * the btree inode), we might have
+	 * btree_inode->i_mapping->a_ops->writepages() called by the VM - if it
+	 * returns an error or an error happens during writeback, when we're
+	 * committing the transaction we wouldn't know about it, since the pages
+	 * can be no longer dirty nor marked anymore for writeback (if a
+	 * subsequent modification to the extent buffer didn't happen before the
+	 * transaction commit), which makes filemap_fdata[write|wait]_range not
+	 * able to find the pages tagged with SetPageError at transaction
+	 * commit time. So if this happens we must abort the transaction,
+	 * otherwise we commit a super block with btree roots that point to
+	 * btree nodes/leafs whose content on disk is invalid - either garbage
+	 * or the content of some node/leaf from a past generation that got
+	 * cowed or deleted and is no longer valid.
+	 *
+	 * Note: setting AS_EIO/AS_ENOSPC in the btree inode's i_mapping would
+	 * not be enough - we need to distinguish between log tree extents vs
+	 * non-log tree extents, and the next filemap_fdatawait_range() call
+	 * will catch and clear such errors in the mapping - and that call might
+	 * be from a log sync and not from a transaction commit. Also, checking
+	 * for the eb flag EXTENT_BUFFER_WRITE_ERR at transaction commit time is
+	 * not done and would not be reliable - the eb might have been released
+	 * from memory and reading it back again means that flag would not be
+	 * set (since it's a runtime flag, not persisted on disk).
+	 *
+	 * Using the flags below in the btree inode also makes us achieve the
+	 * goal of AS_EIO/AS_ENOSPC when writepages() returns success, started
+	 * writeback for all dirty pages and before filemap_fdatawait_range()
+	 * is called, the writeback for all dirty pages had already finished
+	 * with errors - because we were not using AS_EIO/AS_ENOSPC,
+	 * filemap_fdatawait_range() would return success, as it could not know
+	 * that writeback errors happened (the pages were no longer tagged for
+	 * writeback).
+	 */
+	switch (eb->log_index) {
+	case -1:
+		set_bit(BTRFS_INODE_BTREE_ERR, &btree_ino->runtime_flags);
+		break;
+	case 0:
+		set_bit(BTRFS_INODE_BTREE_LOG1_ERR, &btree_ino->runtime_flags);
+		break;
+	case 1:
+		set_bit(BTRFS_INODE_BTREE_LOG2_ERR, &btree_ino->runtime_flags);
+		break;
+	default:
+		BUG(); /* unexpected, logic error */
+	}
+}
+
 static void end_bio_extent_buffer_writepage(struct bio *bio, int err)
 {
 	struct bio_vec *bvec;
@@ -3535,10 +3676,9 @@ static void end_bio_extent_buffer_writepage(struct bio *bio, int err)
 		BUG_ON(!eb);
 		done = atomic_dec_and_test(&eb->io_pages);
 
-		if (err || test_bit(EXTENT_BUFFER_IOERR, &eb->bflags)) {
-			set_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
+		if (err || test_bit(EXTENT_BUFFER_WRITE_ERR, &eb->bflags)) {
 			ClearPageUptodate(page);
-			SetPageError(page);
+			set_btree_ioerr(page);
 		}
 
 		end_page_writeback(page);
@@ -3565,14 +3705,14 @@ static noinline_for_stack int write_one_eb(struct extent_buffer *eb,
 	int rw = (epd->sync_io ? WRITE_SYNC : WRITE) | REQ_META;
 	int ret = 0;
 
-	clear_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
+	clear_bit(EXTENT_BUFFER_WRITE_ERR, &eb->bflags);
 	num_pages = num_extent_pages(eb->start, eb->len);
 	atomic_set(&eb->io_pages, num_pages);
 	if (btrfs_header_owner(eb) == BTRFS_TREE_LOG_OBJECTID)
 		bio_flags = EXTENT_BIO_TREE_LOG;
 
 	for (i = 0; i < num_pages; i++) {
-		struct page *p = extent_buffer_page(eb, i);
+		struct page *p = eb->pages[i];
 
 		clear_page_dirty_for_io(p);
 		set_page_writeback(p);
@@ -3582,8 +3722,8 @@ static noinline_for_stack int write_one_eb(struct extent_buffer *eb,
 					 0, epd->bio_flags, bio_flags);
 		epd->bio_flags = bio_flags;
 		if (ret) {
-			set_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
-			SetPageError(p);
+			set_btree_ioerr(p);
+			end_page_writeback(p);
 			if (atomic_sub_and_test(num_pages - i, &eb->io_pages))
 				end_extent_buffer_writeback(eb);
 			ret = -EIO;
@@ -3596,7 +3736,8 @@ static noinline_for_stack int write_one_eb(struct extent_buffer *eb,
 
 	if (unlikely(ret)) {
 		for (; i < num_pages; i++) {
-			struct page *p = extent_buffer_page(eb, i);
+			struct page *p = eb->pages[i];
+			clear_page_dirty_for_io(p);
 			unlock_page(p);
 		}
 	}
@@ -4166,19 +4307,6 @@ static struct extent_map *get_extent_skip_holes(struct inode *inode,
 	return NULL;
 }
 
-static noinline int count_ext_ref(u64 inum, u64 offset, u64 root_id, void *ctx)
-{
-	unsigned long cnt = *((unsigned long *)ctx);
-
-	cnt++;
-	*((unsigned long *)ctx) = cnt;
-
-	/* Now we're sure that the extent is shared. */
-	if (cnt > 1)
-		return 1;
-	return 0;
-}
-
 int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len, get_extent_t *get_extent)
 {
@@ -4195,6 +4323,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	struct extent_map *em = NULL;
 	struct extent_state *cached_state = NULL;
 	struct btrfs_path *path;
+	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int end = 0;
 	u64 em_start = 0;
 	u64 em_len = 0;
@@ -4215,8 +4344,8 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	 * lookup the last file extent.  We're not using i_size here
 	 * because there might be preallocation past i_size
 	 */
-	ret = btrfs_lookup_file_extent(NULL, BTRFS_I(inode)->root,
-				       path, btrfs_ino(inode), -1, 0);
+	ret = btrfs_lookup_file_extent(NULL, root, path, btrfs_ino(inode), -1,
+				       0);
 	if (ret < 0) {
 		btrfs_free_path(path);
 		return ret;
@@ -4224,7 +4353,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	WARN_ON(!ret);
 	path->slots[0]--;
 	btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
-	found_type = btrfs_key_type(&found_key);
+	found_type = found_key.type;
 
 	/* No extents, but there might be delalloc bits */
 	if (found_key.objectid != btrfs_ino(inode) ||
@@ -4309,25 +4438,27 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		} else if (em->block_start == EXTENT_MAP_DELALLOC) {
 			flags |= (FIEMAP_EXTENT_DELALLOC |
 				  FIEMAP_EXTENT_UNKNOWN);
-		} else {
-			unsigned long ref_cnt = 0;
+		} else if (fieinfo->fi_extents_max) {
+			u64 bytenr = em->block_start -
+				(em->start - em->orig_start);
 
 			disko = em->block_start + offset_in_extent;
 
 			/*
 			 * As btrfs supports shared space, this information
 			 * can be exported to userspace tools via
-			 * flag FIEMAP_EXTENT_SHARED.
+			 * flag FIEMAP_EXTENT_SHARED.  If fi_extents_max == 0
+			 * then we're just getting a count and we can skip the
+			 * lookup stuff.
 			 */
-			ret = iterate_inodes_from_logical(
-					em->block_start,
-					BTRFS_I(inode)->root->fs_info,
-					path, count_ext_ref, &ref_cnt);
-			if (ret < 0 && ret != -ENOENT)
+			ret = btrfs_check_shared(NULL, root->fs_info,
+						 root->objectid,
+						 btrfs_ino(inode), bytenr);
+			if (ret < 0)
 				goto out_free;
-
-			if (ref_cnt > 1)
+			if (ret)
 				flags |= FIEMAP_EXTENT_SHARED;
+			ret = 0;
 		}
 		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags))
 			flags |= FIEMAP_EXTENT_ENCODED;
@@ -4381,24 +4512,21 @@ int extent_buffer_under_io(struct extent_buffer *eb)
 /*
  * Helper for releasing extent buffer page.
  */
-static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
-						unsigned long start_idx)
+static void btrfs_release_extent_buffer_page(struct extent_buffer *eb)
 {
 	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)
+	index = num_extent_pages(eb->start, eb->len);
+	if (index == 0)
 		return;
 
 	do {
 		index--;
-		page = extent_buffer_page(eb, index);
+		page = eb->pages[index];
 		if (page && mapped) {
 			spin_lock(&page->mapping->private_lock);
 			/*
@@ -4429,7 +4557,7 @@ static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
 			/* One for when we alloced the page */
 			page_cache_release(page);
 		}
-	} while (index != start_idx);
+	} while (index != 0);
 }
 
 /*
@@ -4437,7 +4565,7 @@ static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
  */
 static inline void btrfs_release_extent_buffer(struct extent_buffer *eb)
 {
-	btrfs_release_extent_buffer_page(eb, 0);
+	btrfs_release_extent_buffer_page(eb);
 	__free_extent_buffer(eb);
 }
 
@@ -4580,7 +4708,8 @@ static void mark_extent_buffer_accessed(struct extent_buffer *eb,
 
 	num_pages = num_extent_pages(eb->start, eb->len);
 	for (i = 0; i < num_pages; i++) {
-		struct page *p = extent_buffer_page(eb, i);
+		struct page *p = eb->pages[i];
+
 		if (p != accessed)
 			mark_page_accessed(p);
 	}
@@ -4749,7 +4878,7 @@ again:
 	 */
 	SetPageChecked(eb->pages[0]);
 	for (i = 1; i < num_pages; i++) {
-		p = extent_buffer_page(eb, i);
+		p = eb->pages[i];
 		ClearPageChecked(p);
 		unlock_page(p);
 	}
@@ -4794,7 +4923,7 @@ static int release_extent_buffer(struct extent_buffer *eb)
 		}
 
 		/* Should be safe to release our pages at this point */
-		btrfs_release_extent_buffer_page(eb, 0);
+		btrfs_release_extent_buffer_page(eb);
 		call_rcu(&eb->rcu_head, btrfs_release_extent_buffer_rcu);
 		return 1;
 	}
@@ -4860,7 +4989,7 @@ void clear_extent_buffer_dirty(struct extent_buffer *eb)
 	num_pages = num_extent_pages(eb->start, eb->len);
 
 	for (i = 0; i < num_pages; i++) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		if (!PageDirty(page))
 			continue;
 
@@ -4896,7 +5025,7 @@ int set_extent_buffer_dirty(struct extent_buffer *eb)
 	WARN_ON(!test_bit(EXTENT_BUFFER_TREE_REF, &eb->bflags));
 
 	for (i = 0; i < num_pages; i++)
-		set_page_dirty(extent_buffer_page(eb, i));
+		set_page_dirty(eb->pages[i]);
 	return was_dirty;
 }
 
@@ -4909,7 +5038,7 @@ int clear_extent_buffer_uptodate(struct extent_buffer *eb)
 	clear_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags);
 	num_pages = num_extent_pages(eb->start, eb->len);
 	for (i = 0; i < num_pages; i++) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		if (page)
 			ClearPageUptodate(page);
 	}
@@ -4925,7 +5054,7 @@ int set_extent_buffer_uptodate(struct extent_buffer *eb)
 	set_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags);
 	num_pages = num_extent_pages(eb->start, eb->len);
 	for (i = 0; i < num_pages; i++) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		SetPageUptodate(page);
 	}
 	return 0;
@@ -4965,7 +5094,7 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 
 	num_pages = num_extent_pages(eb->start, eb->len);
 	for (i = start_i; i < num_pages; i++) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		if (wait == WAIT_NONE) {
 			if (!trylock_page(page))
 				goto unlock_exit;
@@ -4984,11 +5113,11 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 		goto unlock_exit;
 	}
 
-	clear_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
+	clear_bit(EXTENT_BUFFER_READ_ERR, &eb->bflags);
 	eb->read_mirror = 0;
 	atomic_set(&eb->io_pages, num_reads);
 	for (i = start_i; i < num_pages; i++) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		if (!PageUptodate(page)) {
 			ClearPageError(page);
 			err = __extent_read_full_page(tree, page,
@@ -5013,7 +5142,7 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 		return ret;
 
 	for (i = start_i; i < num_pages; i++) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		wait_on_page_locked(page);
 		if (!PageUptodate(page))
 			ret = -EIO;
@@ -5024,7 +5153,7 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 unlock_exit:
 	i = start_i;
 	while (locked_pages > 0) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		i++;
 		unlock_page(page);
 		locked_pages--;
@@ -5050,7 +5179,7 @@ void read_extent_buffer(struct extent_buffer *eb, void *dstv,
 	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 
 		cur = min(len, (PAGE_CACHE_SIZE - offset));
 		kaddr = page_address(page);
@@ -5082,7 +5211,7 @@ int read_extent_buffer_to_user(struct extent_buffer *eb, void __user *dstv,
 	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 
 		cur = min(len, (PAGE_CACHE_SIZE - offset));
 		kaddr = page_address(page);
@@ -5131,7 +5260,7 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long start,
 		return -EINVAL;
 	}
 
-	p = extent_buffer_page(eb, i);
+	p = eb->pages[i];
 	kaddr = page_address(p);
 	*map = kaddr + offset;
 	*map_len = PAGE_CACHE_SIZE - offset;
@@ -5157,7 +5286,7 @@ int memcmp_extent_buffer(struct extent_buffer *eb, const void *ptrv,
 	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 
 		cur = min(len, (PAGE_CACHE_SIZE - offset));
 
@@ -5191,7 +5320,7 @@ void write_extent_buffer(struct extent_buffer *eb, const void *srcv,
 	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		WARN_ON(!PageUptodate(page));
 
 		cur = min(len, PAGE_CACHE_SIZE - offset);
@@ -5221,7 +5350,7 @@ void memset_extent_buffer(struct extent_buffer *eb, char c,
 	offset = (start_offset + start) & (PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
-		page = extent_buffer_page(eb, i);
+		page = eb->pages[i];
 		WARN_ON(!PageUptodate(page));
 
 		cur = min(len, PAGE_CACHE_SIZE - offset);
@@ -5252,7 +5381,7 @@ void copy_extent_buffer(struct extent_buffer *dst, struct extent_buffer *src,
 		(PAGE_CACHE_SIZE - 1);
 
 	while (len > 0) {
-		page = extent_buffer_page(dst, i);
+		page = dst->pages[i];
 		WARN_ON(!PageUptodate(page));
 
 		cur = min(len, (unsigned long)(PAGE_CACHE_SIZE - offset));
@@ -5330,8 +5459,7 @@ void memcpy_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 		cur = min_t(unsigned long, cur,
 			(unsigned long)(PAGE_CACHE_SIZE - dst_off_in_page));
 
-		copy_pages(extent_buffer_page(dst, dst_i),
-			   extent_buffer_page(dst, src_i),
+		copy_pages(dst->pages[dst_i], dst->pages[src_i],
 			   dst_off_in_page, src_off_in_page, cur);
 
 		src_offset += cur;
@@ -5377,8 +5505,7 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 
 		cur = min_t(unsigned long, len, src_off_in_page + 1);
 		cur = min(cur, dst_off_in_page + 1);
-		copy_pages(extent_buffer_page(dst, dst_i),
-			   extent_buffer_page(dst, src_i),
+		copy_pages(dst->pages[dst_i], dst->pages[src_i],
 			   dst_off_in_page - cur + 1,
 			   src_off_in_page - cur + 1, cur);
 

fs/btrfs/extent_io.h

@@ -11,8 +11,6 @@
 #define EXTENT_NEW (1 << 4)
 #define EXTENT_DELALLOC (1 << 5)
 #define EXTENT_DEFRAG (1 << 6)
-#define EXTENT_DEFRAG_DONE (1 << 7)
-#define EXTENT_BUFFER_FILLED (1 << 8)
 #define EXTENT_BOUNDARY (1 << 9)
 #define EXTENT_NODATASUM (1 << 10)
 #define EXTENT_DO_ACCOUNTING (1 << 11)
@@ -34,16 +32,16 @@
 
 /* these are bit numbers for test/set bit */
 #define EXTENT_BUFFER_UPTODATE 0
-#define EXTENT_BUFFER_BLOCKING 1
 #define EXTENT_BUFFER_DIRTY 2
 #define EXTENT_BUFFER_CORRUPT 3
 #define EXTENT_BUFFER_READAHEAD 4	/* this got triggered by readahead */
 #define EXTENT_BUFFER_TREE_REF 5
 #define EXTENT_BUFFER_STALE 6
 #define EXTENT_BUFFER_WRITEBACK 7
-#define EXTENT_BUFFER_IOERR 8
+#define EXTENT_BUFFER_READ_ERR 8        /* read IO error */
 #define EXTENT_BUFFER_DUMMY 9
 #define EXTENT_BUFFER_IN_TREE 10
+#define EXTENT_BUFFER_WRITE_ERR 11    /* write IO error */
 
 /* these are flags for extent_clear_unlock_delalloc */
 #define PAGE_UNLOCK		(1 << 0)
@@ -57,7 +55,6 @@
  * map has page->private set to one.
  */
 #define EXTENT_PAGE_PRIVATE 1
-#define EXTENT_PAGE_PRIVATE_FIRST_PAGE 3
 
 struct extent_state;
 struct btrfs_root;
@@ -108,7 +105,6 @@ struct extent_state {
 	struct rb_node rb_node;
 
 	/* ADD NEW ELEMENTS AFTER THIS */
-	struct extent_io_tree *tree;
 	wait_queue_head_t wq;
 	atomic_t refs;
 	unsigned long state;
@@ -126,8 +122,6 @@ struct extent_state {
 struct extent_buffer {
 	u64 start;
 	unsigned long len;
-	unsigned long map_start;
-	unsigned long map_len;
 	unsigned long bflags;
 	struct btrfs_fs_info *fs_info;
 	spinlock_t refs_lock;
@@ -144,7 +138,9 @@ struct extent_buffer {
 	atomic_t blocking_readers;
 	atomic_t spinning_readers;
 	atomic_t spinning_writers;
-	int lock_nested;
+	short lock_nested;
+	/* >= 0 if eb belongs to a log tree, -1 otherwise */
+	short log_index;
 
 	/* protects write locks */
 	rwlock_t lock;
@@ -286,12 +282,6 @@ static inline unsigned long num_extent_pages(u64 start, u64 len)
 		(start >> PAGE_CACHE_SHIFT);
 }
 
-static inline struct page *extent_buffer_page(struct extent_buffer *eb,
-					      unsigned long i)
-{
-	return eb->pages[i];
-}
-
 static inline void extent_buffer_get(struct extent_buffer *eb)
 {
 	atomic_inc(&eb->refs);
@@ -341,18 +331,50 @@ struct bio *btrfs_bio_clone(struct bio *bio, gfp_t gfp_mask);
 
 struct btrfs_fs_info;
 
-int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
-			u64 length, u64 logical, struct page *page,
-			int mirror_num);
+int repair_io_failure(struct inode *inode, u64 start, u64 length, u64 logical,
+		      struct page *page, unsigned int pg_offset,
+		      int mirror_num);
+int clean_io_failure(struct inode *inode, u64 start, struct page *page,
+		     unsigned int pg_offset);
 int end_extent_writepage(struct page *page, int err, u64 start, u64 end);
 int repair_eb_io_failure(struct btrfs_root *root, struct extent_buffer *eb,
 			 int mirror_num);
+
+/*
+ * When IO fails, either with EIO or csum verification fails, we
+ * try other mirrors that might have a good copy of the data.  This
+ * io_failure_record is used to record state as we go through all the
+ * mirrors.  If another mirror has good data, the page is set up to date
+ * and things continue.  If a good mirror can't be found, the original
+ * bio end_io callback is called to indicate things have failed.
+ */
+struct io_failure_record {
+	struct page *page;
+	u64 start;
+	u64 len;
+	u64 logical;
+	unsigned long bio_flags;
+	int this_mirror;
+	int failed_mirror;
+	int in_validation;
+};
+
+void btrfs_free_io_failure_record(struct inode *inode, u64 start, u64 end);
+int btrfs_get_io_failure_record(struct inode *inode, u64 start, u64 end,
+				struct io_failure_record **failrec_ret);
+int btrfs_check_repairable(struct inode *inode, struct bio *failed_bio,
+			   struct io_failure_record *failrec, int fail_mirror);
+struct bio *btrfs_create_repair_bio(struct inode *inode, struct bio *failed_bio,
+				    struct io_failure_record *failrec,
+				    struct page *page, int pg_offset, int icsum,
+				    bio_end_io_t *endio_func, void *data);
+int free_io_failure(struct inode *inode, struct io_failure_record *rec);
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 noinline u64 find_lock_delalloc_range(struct inode *inode,
 				      struct extent_io_tree *tree,
 				      struct page *locked_page, u64 *start,
 				      u64 *end, u64 max_bytes);
+#endif
 struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info,
 					       u64 start, unsigned long len);
 #endif
-#endif

fs/btrfs/file-item.c

@@ -55,7 +55,7 @@ int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	file_key.objectid = objectid;
 	file_key.offset = pos;
-	btrfs_set_key_type(&file_key, BTRFS_EXTENT_DATA_KEY);
+	file_key.type = BTRFS_EXTENT_DATA_KEY;
 
 	path->leave_spinning = 1;
 	ret = btrfs_insert_empty_item(trans, root, path, &file_key,
@@ -100,7 +100,7 @@ btrfs_lookup_csum(struct btrfs_trans_handle *trans,
 
 	file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
 	file_key.offset = bytenr;
-	btrfs_set_key_type(&file_key, BTRFS_EXTENT_CSUM_KEY);
+	file_key.type = BTRFS_EXTENT_CSUM_KEY;
 	ret = btrfs_search_slot(trans, root, &file_key, path, 0, cow);
 	if (ret < 0)
 		goto fail;
@@ -111,7 +111,7 @@ btrfs_lookup_csum(struct btrfs_trans_handle *trans,
 			goto fail;
 		path->slots[0]--;
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
-		if (btrfs_key_type(&found_key) != BTRFS_EXTENT_CSUM_KEY)
+		if (found_key.type != BTRFS_EXTENT_CSUM_KEY)
 			goto fail;
 
 		csum_offset = (bytenr - found_key.offset) >>
@@ -148,7 +148,7 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
 
 	file_key.objectid = objectid;
 	file_key.offset = offset;
-	btrfs_set_key_type(&file_key, BTRFS_EXTENT_DATA_KEY);
+	file_key.type = BTRFS_EXTENT_DATA_KEY;
 	ret = btrfs_search_slot(trans, root, &file_key, path, ins_len, cow);
 	return ret;
 }
@@ -299,19 +299,9 @@ 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 btrfs_dio_private *dip, struct bio *bio,
-			      u64 offset)
+			      struct bio *bio, u64 offset)
 {
-	int len = (bio->bi_iter.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;
+	return __btrfs_lookup_bio_sums(root, inode, bio, offset, NULL, 1);
 }
 
 int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
@@ -329,8 +319,8 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 	u64 csum_end;
 	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
 
-	ASSERT(start == ALIGN(start, root->sectorsize) &&
-	       (end + 1) == ALIGN(end + 1, root->sectorsize));
+	ASSERT(IS_ALIGNED(start, root->sectorsize) &&
+	       IS_ALIGNED(end + 1, root->sectorsize));
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -720,7 +710,7 @@ again:
 	bytenr = sums->bytenr + total_bytes;
 	file_key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
 	file_key.offset = bytenr;
-	btrfs_set_key_type(&file_key, BTRFS_EXTENT_CSUM_KEY);
+	file_key.type = BTRFS_EXTENT_CSUM_KEY;
 
 	item = btrfs_lookup_csum(trans, root, path, bytenr, 1);
 	if (!IS_ERR(item)) {
@@ -790,7 +780,7 @@ again:
 	csum_offset = (bytenr - found_key.offset) >>
 			root->fs_info->sb->s_blocksize_bits;
 
-	if (btrfs_key_type(&found_key) != BTRFS_EXTENT_CSUM_KEY ||
+	if (found_key.type != BTRFS_EXTENT_CSUM_KEY ||
 	    found_key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
 	    csum_offset >= MAX_CSUM_ITEMS(root, csum_size)) {
 		goto insert;

fs/btrfs/file.c

@@ -299,7 +299,7 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info,
 
 	/* get the inode */
 	key.objectid = defrag->root;
-	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
+	key.type = BTRFS_ROOT_ITEM_KEY;
 	key.offset = (u64)-1;
 
 	index = srcu_read_lock(&fs_info->subvol_srcu);
@@ -311,7 +311,7 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info,
 	}
 
 	key.objectid = defrag->ino;
-	btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
+	key.type = BTRFS_INODE_ITEM_KEY;
 	key.offset = 0;
 	inode = btrfs_iget(fs_info->sb, &key, inode_root, NULL);
 	if (IS_ERR(inode)) {
@@ -452,7 +452,7 @@ static noinline int btrfs_copy_from_user(loff_t pos, int num_pages,
 		if (unlikely(copied == 0))
 			break;
 
-		if (unlikely(copied < PAGE_CACHE_SIZE - offset)) {
+		if (copied < PAGE_CACHE_SIZE - offset) {
 			offset += copied;
 		} else {
 			pg++;
@@ -1481,9 +1481,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 	bool force_page_uptodate = false;
 	bool need_unlock;
 
-	nrptrs = min((iov_iter_count(i) + PAGE_CACHE_SIZE - 1) /
-		     PAGE_CACHE_SIZE, PAGE_CACHE_SIZE /
-		     (sizeof(struct page *)));
+	nrptrs = min(DIV_ROUND_UP(iov_iter_count(i), PAGE_CACHE_SIZE),
+			PAGE_CACHE_SIZE / (sizeof(struct page *)));
 	nrptrs = min(nrptrs, current->nr_dirtied_pause - current->nr_dirtied);
 	nrptrs = max(nrptrs, 8);
 	pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL);
@@ -1497,8 +1496,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 		size_t write_bytes = min(iov_iter_count(i),
 					 nrptrs * (size_t)PAGE_CACHE_SIZE -
 					 offset);
-		size_t num_pages = (write_bytes + offset +
-				    PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+		size_t num_pages = DIV_ROUND_UP(write_bytes + offset,
+						PAGE_CACHE_SIZE);
 		size_t reserve_bytes;
 		size_t dirty_pages;
 		size_t copied;
@@ -1526,9 +1525,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 				 * our prealloc extent may be smaller than
 				 * write_bytes, so scale down.
 				 */
-				num_pages = (write_bytes + offset +
-					     PAGE_CACHE_SIZE - 1) >>
-					PAGE_CACHE_SHIFT;
+				num_pages = DIV_ROUND_UP(write_bytes + offset,
+							 PAGE_CACHE_SIZE);
 				reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
 				ret = 0;
 			} else {
@@ -1590,9 +1588,8 @@ again:
 			dirty_pages = 0;
 		} else {
 			force_page_uptodate = false;
-			dirty_pages = (copied + offset +
-				       PAGE_CACHE_SIZE - 1) >>
-				       PAGE_CACHE_SHIFT;
+			dirty_pages = DIV_ROUND_UP(copied + offset,
+						   PAGE_CACHE_SIZE);
 		}
 
 		/*
@@ -1653,7 +1650,7 @@ again:
 		cond_resched();
 
 		balance_dirty_pages_ratelimited(inode->i_mapping);
-		if (dirty_pages < (root->leafsize >> PAGE_CACHE_SHIFT) + 1)
+		if (dirty_pages < (root->nodesize >> PAGE_CACHE_SHIFT) + 1)
 			btrfs_btree_balance_dirty(root);
 
 		pos += copied;
@@ -1795,7 +1792,7 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb,
 	if (sync)
 		atomic_inc(&BTRFS_I(inode)->sync_writers);
 
-	if (unlikely(file->f_flags & O_DIRECT)) {
+	if (file->f_flags & O_DIRECT) {
 		num_written = __btrfs_direct_write(iocb, from, pos);
 	} else {
 		num_written = __btrfs_buffered_write(file, from, pos);
@@ -1852,6 +1849,20 @@ int btrfs_release_file(struct inode *inode, struct file *filp)
 	return 0;
 }
 
+static int start_ordered_ops(struct inode *inode, loff_t start, loff_t end)
+{
+	int ret;
+
+	atomic_inc(&BTRFS_I(inode)->sync_writers);
+	ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
+	if (!ret && test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
+			     &BTRFS_I(inode)->runtime_flags))
+		ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
+	atomic_dec(&BTRFS_I(inode)->sync_writers);
+
+	return ret;
+}
+
 /*
  * fsync call for both files and directories.  This logs the inode into
  * the tree log instead of forcing full commits whenever possible.
@@ -1881,30 +1892,64 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	 * multi-task, and make the performance up.  See
 	 * btrfs_wait_ordered_range for an explanation of the ASYNC check.
 	 */
-	atomic_inc(&BTRFS_I(inode)->sync_writers);
-	ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
-	if (!ret && test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
-			     &BTRFS_I(inode)->runtime_flags))
-		ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
-	atomic_dec(&BTRFS_I(inode)->sync_writers);
+	ret = start_ordered_ops(inode, start, end);
 	if (ret)
 		return ret;
 
 	mutex_lock(&inode->i_mutex);
-
-	/*
-	 * We flush the dirty pages again to avoid some dirty pages in the
-	 * range being left.
-	 */
 	atomic_inc(&root->log_batch);
 	full_sync = test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
 			     &BTRFS_I(inode)->runtime_flags);
+	/*
+	 * We might have have had more pages made dirty after calling
+	 * start_ordered_ops and before acquiring the inode's i_mutex.
+	 */
 	if (full_sync) {
+		/*
+		 * For a full sync, we need to make sure any ordered operations
+		 * start and finish before we start logging the inode, so that
+		 * all extents are persisted and the respective file extent
+		 * items are in the fs/subvol btree.
+		 */
 		ret = btrfs_wait_ordered_range(inode, start, end - start + 1);
-		if (ret) {
-			mutex_unlock(&inode->i_mutex);
-			goto out;
-		}
+	} else {
+		/*
+		 * Start any new ordered operations before starting to log the
+		 * inode. We will wait for them to finish in btrfs_sync_log().
+		 *
+		 * Right before acquiring the inode's mutex, we might have new
+		 * writes dirtying pages, which won't immediately start the
+		 * respective ordered operations - that is done through the
+		 * fill_delalloc callbacks invoked from the writepage and
+		 * writepages address space operations. So make sure we start
+		 * all ordered operations before starting to log our inode. Not
+		 * doing this means that while logging the inode, writeback
+		 * could start and invoke writepage/writepages, which would call
+		 * the fill_delalloc callbacks (cow_file_range,
+		 * submit_compressed_extents). These callbacks add first an
+		 * extent map to the modified list of extents and then create
+		 * the respective ordered operation, which means in
+		 * tree-log.c:btrfs_log_inode() we might capture all existing
+		 * ordered operations (with btrfs_get_logged_extents()) before
+		 * the fill_delalloc callback adds its ordered operation, and by
+		 * the time we visit the modified list of extent maps (with
+		 * btrfs_log_changed_extents()), we see and process the extent
+		 * map they created. We then use the extent map to construct a
+		 * file extent item for logging without waiting for the
+		 * respective ordered operation to finish - this file extent
+		 * item points to a disk location that might not have yet been
+		 * written to, containing random data - so after a crash a log
+		 * replay will make our inode have file extent items that point
+		 * to disk locations containing invalid data, as we returned
+		 * success to userspace without waiting for the respective
+		 * ordered operation to finish, because it wasn't captured by
+		 * btrfs_get_logged_extents().
+		 */
+		ret = start_ordered_ops(inode, start, end);
+	}
+	if (ret) {
+		mutex_unlock(&inode->i_mutex);
+		goto out;
 	}
 	atomic_inc(&root->log_batch);
 
@@ -1984,6 +2029,25 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	 */
 	mutex_unlock(&inode->i_mutex);
 
+	/*
+	 * If any of the ordered extents had an error, just return it to user
+	 * space, so that the application knows some writes didn't succeed and
+	 * can take proper action (retry for e.g.). Blindly committing the
+	 * transaction in this case, would fool userspace that everything was
+	 * successful. And we also want to make sure our log doesn't contain
+	 * file extent items pointing to extents that weren't fully written to -
+	 * just like in the non fast fsync path, where we check for the ordered
+	 * operation's error flag before writing to the log tree and return -EIO
+	 * if any of them had this flag set (btrfs_wait_ordered_range) -
+	 * therefore we need to check for errors in the ordered operations,
+	 * which are indicated by ctx.io_err.
+	 */
+	if (ctx.io_err) {
+		btrfs_end_transaction(trans, root);
+		ret = ctx.io_err;
+		goto out;
+	}
+
 	if (ret != BTRFS_NO_LOG_SYNC) {
 		if (!ret) {
 			ret = btrfs_sync_log(trans, root, &ctx);
@@ -2621,23 +2685,28 @@ static int find_desired_extent(struct inode *inode, loff_t *offset, int whence)
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct extent_map *em = NULL;
 	struct extent_state *cached_state = NULL;
-	u64 lockstart = *offset;
-	u64 lockend = i_size_read(inode);
-	u64 start = *offset;
-	u64 len = i_size_read(inode);
+	u64 lockstart;
+	u64 lockend;
+	u64 start;
+	u64 len;
 	int ret = 0;
 
-	lockend = max_t(u64, root->sectorsize, lockend);
+	if (inode->i_size == 0)
+		return -ENXIO;
+
+	/*
+	 * *offset can be negative, in this case we start finding DATA/HOLE from
+	 * the very start of the file.
+	 */
+	start = max_t(loff_t, 0, *offset);
+
+	lockstart = round_down(start, root->sectorsize);
+	lockend = round_up(i_size_read(inode), root->sectorsize);
 	if (lockend <= lockstart)
 		lockend = lockstart + root->sectorsize;
-
 	lockend--;
 	len = lockend - lockstart + 1;
 
-	len = max_t(u64, len, root->sectorsize);
-	if (inode->i_size == 0)
-		return -ENXIO;
-
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend, 0,
 			 &cached_state);
 

fs/btrfs/free-space-cache.c

@@ -279,8 +279,7 @@ static int io_ctl_init(struct io_ctl *io_ctl, struct inode *inode,
 	int num_pages;
 	int check_crcs = 0;
 
-	num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >>
-		    PAGE_CACHE_SHIFT;
+	num_pages = DIV_ROUND_UP(i_size_read(inode), PAGE_CACHE_SIZE);
 
 	if (btrfs_ino(inode) != BTRFS_FREE_INO_OBJECTID)
 		check_crcs = 1;
@@ -1998,6 +1997,128 @@ static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl,
 	return merged;
 }
 
+static bool steal_from_bitmap_to_end(struct btrfs_free_space_ctl *ctl,
+				     struct btrfs_free_space *info,
+				     bool update_stat)
+{
+	struct btrfs_free_space *bitmap;
+	unsigned long i;
+	unsigned long j;
+	const u64 end = info->offset + info->bytes;
+	const u64 bitmap_offset = offset_to_bitmap(ctl, end);
+	u64 bytes;
+
+	bitmap = tree_search_offset(ctl, bitmap_offset, 1, 0);
+	if (!bitmap)
+		return false;
+
+	i = offset_to_bit(bitmap->offset, ctl->unit, end);
+	j = find_next_zero_bit(bitmap->bitmap, BITS_PER_BITMAP, i);
+	if (j == i)
+		return false;
+	bytes = (j - i) * ctl->unit;
+	info->bytes += bytes;
+
+	if (update_stat)
+		bitmap_clear_bits(ctl, bitmap, end, bytes);
+	else
+		__bitmap_clear_bits(ctl, bitmap, end, bytes);
+
+	if (!bitmap->bytes)
+		free_bitmap(ctl, bitmap);
+
+	return true;
+}
+
+static bool steal_from_bitmap_to_front(struct btrfs_free_space_ctl *ctl,
+				       struct btrfs_free_space *info,
+				       bool update_stat)
+{
+	struct btrfs_free_space *bitmap;
+	u64 bitmap_offset;
+	unsigned long i;
+	unsigned long j;
+	unsigned long prev_j;
+	u64 bytes;
+
+	bitmap_offset = offset_to_bitmap(ctl, info->offset);
+	/* If we're on a boundary, try the previous logical bitmap. */
+	if (bitmap_offset == info->offset) {
+		if (info->offset == 0)
+			return false;
+		bitmap_offset = offset_to_bitmap(ctl, info->offset - 1);
+	}
+
+	bitmap = tree_search_offset(ctl, bitmap_offset, 1, 0);
+	if (!bitmap)
+		return false;
+
+	i = offset_to_bit(bitmap->offset, ctl->unit, info->offset) - 1;
+	j = 0;
+	prev_j = (unsigned long)-1;
+	for_each_clear_bit_from(j, bitmap->bitmap, BITS_PER_BITMAP) {
+		if (j > i)
+			break;
+		prev_j = j;
+	}
+	if (prev_j == i)
+		return false;
+
+	if (prev_j == (unsigned long)-1)
+		bytes = (i + 1) * ctl->unit;
+	else
+		bytes = (i - prev_j) * ctl->unit;
+
+	info->offset -= bytes;
+	info->bytes += bytes;
+
+	if (update_stat)
+		bitmap_clear_bits(ctl, bitmap, info->offset, bytes);
+	else
+		__bitmap_clear_bits(ctl, bitmap, info->offset, bytes);
+
+	if (!bitmap->bytes)
+		free_bitmap(ctl, bitmap);
+
+	return true;
+}
+
+/*
+ * We prefer always to allocate from extent entries, both for clustered and
+ * non-clustered allocation requests. So when attempting to add a new extent
+ * entry, try to see if there's adjacent free space in bitmap entries, and if
+ * there is, migrate that space from the bitmaps to the extent.
+ * Like this we get better chances of satisfying space allocation requests
+ * because we attempt to satisfy them based on a single cache entry, and never
+ * on 2 or more entries - even if the entries represent a contiguous free space
+ * region (e.g. 1 extent entry + 1 bitmap entry starting where the extent entry
+ * ends).
+ */
+static void steal_from_bitmap(struct btrfs_free_space_ctl *ctl,
+			      struct btrfs_free_space *info,
+			      bool update_stat)
+{
+	/*
+	 * Only work with disconnected entries, as we can change their offset,
+	 * and must be extent entries.
+	 */
+	ASSERT(!info->bitmap);
+	ASSERT(RB_EMPTY_NODE(&info->offset_index));
+
+	if (ctl->total_bitmaps > 0) {
+		bool stole_end;
+		bool stole_front = false;
+
+		stole_end = steal_from_bitmap_to_end(ctl, info, update_stat);
+		if (ctl->total_bitmaps > 0)
+			stole_front = steal_from_bitmap_to_front(ctl, info,
+								 update_stat);
+
+		if (stole_end || stole_front)
+			try_merge_free_space(ctl, info, update_stat);
+	}
+}
+
 int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl,
 			   u64 offset, u64 bytes)
 {
@@ -2010,6 +2131,7 @@ int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl,
 
 	info->offset = offset;
 	info->bytes = bytes;
+	RB_CLEAR_NODE(&info->offset_index);
 
 	spin_lock(&ctl->tree_lock);
 
@@ -2029,6 +2151,14 @@ int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl,
 		goto out;
 	}
 link:
+	/*
+	 * Only steal free space from adjacent bitmaps if we're sure we're not
+	 * going to add the new free space to existing bitmap entries - because
+	 * that would mean unnecessary work that would be reverted. Therefore
+	 * attempt to steal space from bitmaps if we're adding an extent entry.
+	 */
+	steal_from_bitmap(ctl, info, true);
+
 	ret = link_free_space(ctl, info);
 	if (ret)
 		kmem_cache_free(btrfs_free_space_cachep, info);
@@ -2205,10 +2335,13 @@ __btrfs_return_cluster_to_free_space(
 		entry = rb_entry(node, struct btrfs_free_space, offset_index);
 		node = rb_next(&entry->offset_index);
 		rb_erase(&entry->offset_index, &cluster->root);
+		RB_CLEAR_NODE(&entry->offset_index);
 
 		bitmap = (entry->bitmap != NULL);
-		if (!bitmap)
+		if (!bitmap) {
 			try_merge_free_space(ctl, entry, false);
+			steal_from_bitmap(ctl, entry, false);
+		}
 		tree_insert_offset(&ctl->free_space_offset,
 				   entry->offset, &entry->offset_index, bitmap);
 	}
@@ -3033,10 +3166,10 @@ struct inode *lookup_free_ino_inode(struct btrfs_root *root,
 {
 	struct inode *inode = NULL;
 
-	spin_lock(&root->cache_lock);
-	if (root->cache_inode)
-		inode = igrab(root->cache_inode);
-	spin_unlock(&root->cache_lock);
+	spin_lock(&root->ino_cache_lock);
+	if (root->ino_cache_inode)
+		inode = igrab(root->ino_cache_inode);
+	spin_unlock(&root->ino_cache_lock);
 	if (inode)
 		return inode;
 
@@ -3044,10 +3177,10 @@ struct inode *lookup_free_ino_inode(struct btrfs_root *root,
 	if (IS_ERR(inode))
 		return inode;
 
-	spin_lock(&root->cache_lock);
+	spin_lock(&root->ino_cache_lock);
 	if (!btrfs_fs_closing(root->fs_info))
-		root->cache_inode = igrab(inode);
-	spin_unlock(&root->cache_lock);
+		root->ino_cache_inode = igrab(inode);
+	spin_unlock(&root->ino_cache_lock);
 
 	return inode;
 }
@@ -3176,6 +3309,7 @@ again:
 		map = NULL;
 		add_new_bitmap(ctl, info, offset);
 		bitmap_info = info;
+		info = NULL;
 	}
 
 	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
@@ -3186,6 +3320,8 @@ again:
 	if (bytes)
 		goto again;
 
+	if (info)
+		kmem_cache_free(btrfs_free_space_cachep, info);
 	if (map)
 		kfree(map);
 	return 0;
@@ -3260,6 +3396,7 @@ have_info:
 			goto have_info;
 		}
 
+		ret = 0;
 		goto out;
 	}
 

fs/btrfs/hash.c

@@ -20,10 +20,8 @@ static struct crypto_shash *tfm;
 int __init btrfs_hash_init(void)
 {
 	tfm = crypto_alloc_shash("crc32c", 0, 0);
-	if (IS_ERR(tfm))
-		return PTR_ERR(tfm);
 
-	return 0;
+	return PTR_ERR_OR_ZERO(tfm);
 }
 
 void btrfs_hash_exit(void)

fs/btrfs/inode-item.c

@@ -135,7 +135,7 @@ static int btrfs_del_inode_extref(struct btrfs_trans_handle *trans,
 	u32 item_size;
 
 	key.objectid = inode_objectid;
-	btrfs_set_key_type(&key, BTRFS_INODE_EXTREF_KEY);
+	key.type = BTRFS_INODE_EXTREF_KEY;
 	key.offset = btrfs_extref_hash(ref_objectid, name, name_len);
 
 	path = btrfs_alloc_path();
@@ -209,7 +209,7 @@ int btrfs_del_inode_ref(struct btrfs_trans_handle *trans,
 
 	key.objectid = inode_objectid;
 	key.offset = ref_objectid;
-	btrfs_set_key_type(&key, BTRFS_INODE_REF_KEY);
+	key.type = BTRFS_INODE_REF_KEY;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -337,7 +337,7 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans,
 
 	key.objectid = inode_objectid;
 	key.offset = ref_objectid;
-	btrfs_set_key_type(&key, BTRFS_INODE_REF_KEY);
+	key.type = BTRFS_INODE_REF_KEY;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -400,7 +400,7 @@ int btrfs_insert_empty_inode(struct btrfs_trans_handle *trans,
 	struct btrfs_key key;
 	int ret;
 	key.objectid = objectid;
-	btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
+	key.type = BTRFS_INODE_ITEM_KEY;
 	key.offset = 0;
 
 	ret = btrfs_insert_empty_item(trans, root, path, &key,
@@ -420,13 +420,13 @@ int btrfs_lookup_inode(struct btrfs_trans_handle *trans, struct btrfs_root
 	struct btrfs_key found_key;
 
 	ret = btrfs_search_slot(trans, root, location, path, ins_len, cow);
-	if (ret > 0 && btrfs_key_type(location) == BTRFS_ROOT_ITEM_KEY &&
+	if (ret > 0 && location->type == BTRFS_ROOT_ITEM_KEY &&
 	    location->offset == (u64)-1 && path->slots[0] != 0) {
 		slot = path->slots[0] - 1;
 		leaf = path->nodes[0];
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
 		if (found_key.objectid == location->objectid &&
-		    btrfs_key_type(&found_key) == btrfs_key_type(location)) {
+		    found_key.type == location->type) {
 			path->slots[0]--;
 			return 0;
 		}

fs/btrfs/inode-map.c

@@ -87,7 +87,7 @@ again:
 				 */
 				btrfs_item_key_to_cpu(leaf, &key, 0);
 				btrfs_release_path(path);
-				root->cache_progress = last;
+				root->ino_cache_progress = last;
 				up_read(&fs_info->commit_root_sem);
 				schedule_timeout(1);
 				goto again;
@@ -106,7 +106,7 @@ again:
 		if (last != (u64)-1 && last + 1 != key.objectid) {
 			__btrfs_add_free_space(ctl, last + 1,
 					       key.objectid - last - 1);
-			wake_up(&root->cache_wait);
+			wake_up(&root->ino_cache_wait);
 		}
 
 		last = key.objectid;
@@ -119,14 +119,14 @@ next:
 				       root->highest_objectid - last - 1);
 	}
 
-	spin_lock(&root->cache_lock);
-	root->cached = BTRFS_CACHE_FINISHED;
-	spin_unlock(&root->cache_lock);
+	spin_lock(&root->ino_cache_lock);
+	root->ino_cache_state = BTRFS_CACHE_FINISHED;
+	spin_unlock(&root->ino_cache_lock);
 
-	root->cache_progress = (u64)-1;
+	root->ino_cache_progress = (u64)-1;
 	btrfs_unpin_free_ino(root);
 out:
-	wake_up(&root->cache_wait);
+	wake_up(&root->ino_cache_wait);
 	up_read(&fs_info->commit_root_sem);
 
 	btrfs_free_path(path);
@@ -144,20 +144,20 @@ static void start_caching(struct btrfs_root *root)
 	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
 		return;
 
-	spin_lock(&root->cache_lock);
-	if (root->cached != BTRFS_CACHE_NO) {
-		spin_unlock(&root->cache_lock);
+	spin_lock(&root->ino_cache_lock);
+	if (root->ino_cache_state != BTRFS_CACHE_NO) {
+		spin_unlock(&root->ino_cache_lock);
 		return;
 	}
 
-	root->cached = BTRFS_CACHE_STARTED;
-	spin_unlock(&root->cache_lock);
+	root->ino_cache_state = BTRFS_CACHE_STARTED;
+	spin_unlock(&root->ino_cache_lock);
 
 	ret = load_free_ino_cache(root->fs_info, root);
 	if (ret == 1) {
-		spin_lock(&root->cache_lock);
-		root->cached = BTRFS_CACHE_FINISHED;
-		spin_unlock(&root->cache_lock);
+		spin_lock(&root->ino_cache_lock);
+		root->ino_cache_state = BTRFS_CACHE_FINISHED;
+		spin_unlock(&root->ino_cache_lock);
 		return;
 	}
 
@@ -196,11 +196,11 @@ again:
 
 	start_caching(root);
 
-	wait_event(root->cache_wait,
-		   root->cached == BTRFS_CACHE_FINISHED ||
+	wait_event(root->ino_cache_wait,
+		   root->ino_cache_state == BTRFS_CACHE_FINISHED ||
 		   root->free_ino_ctl->free_space > 0);
 
-	if (root->cached == BTRFS_CACHE_FINISHED &&
+	if (root->ino_cache_state == BTRFS_CACHE_FINISHED &&
 	    root->free_ino_ctl->free_space == 0)
 		return -ENOSPC;
 	else
@@ -214,17 +214,17 @@ void btrfs_return_ino(struct btrfs_root *root, u64 objectid)
 	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
 		return;
 again:
-	if (root->cached == BTRFS_CACHE_FINISHED) {
+	if (root->ino_cache_state == BTRFS_CACHE_FINISHED) {
 		__btrfs_add_free_space(pinned, objectid, 1);
 	} else {
 		down_write(&root->fs_info->commit_root_sem);
-		spin_lock(&root->cache_lock);
-		if (root->cached == BTRFS_CACHE_FINISHED) {
-			spin_unlock(&root->cache_lock);
+		spin_lock(&root->ino_cache_lock);
+		if (root->ino_cache_state == BTRFS_CACHE_FINISHED) {
+			spin_unlock(&root->ino_cache_lock);
 			up_write(&root->fs_info->commit_root_sem);
 			goto again;
 		}
-		spin_unlock(&root->cache_lock);
+		spin_unlock(&root->ino_cache_lock);
 
 		start_caching(root);
 
@@ -235,10 +235,10 @@ again:
 }
 
 /*
- * When a transaction is committed, we'll move those inode numbers which
- * are smaller than root->cache_progress from pinned tree to free_ino tree,
- * and others will just be dropped, because the commit root we were
- * searching has changed.
+ * When a transaction is committed, we'll move those inode numbers which are
+ * smaller than root->ino_cache_progress from pinned tree to free_ino tree, and
+ * others will just be dropped, because the commit root we were searching has
+ * changed.
  *
  * Must be called with root->fs_info->commit_root_sem held
  */
@@ -261,10 +261,10 @@ void btrfs_unpin_free_ino(struct btrfs_root *root)
 		info = rb_entry(n, struct btrfs_free_space, offset_index);
 		BUG_ON(info->bitmap); /* Logic error */
 
-		if (info->offset > root->cache_progress)
+		if (info->offset > root->ino_cache_progress)
 			goto free;
-		else if (info->offset + info->bytes > root->cache_progress)
-			count = root->cache_progress - info->offset + 1;
+		else if (info->offset + info->bytes > root->ino_cache_progress)
+			count = root->ino_cache_progress - info->offset + 1;
 		else
 			count = info->bytes;
 
@@ -462,13 +462,13 @@ again:
 		}
 	}
 
-	spin_lock(&root->cache_lock);
-	if (root->cached != BTRFS_CACHE_FINISHED) {
+	spin_lock(&root->ino_cache_lock);
+	if (root->ino_cache_state != BTRFS_CACHE_FINISHED) {
 		ret = -1;
-		spin_unlock(&root->cache_lock);
+		spin_unlock(&root->ino_cache_lock);
 		goto out_put;
 	}
-	spin_unlock(&root->cache_lock);
+	spin_unlock(&root->ino_cache_lock);
 
 	spin_lock(&ctl->tree_lock);
 	prealloc = sizeof(struct btrfs_free_space) * ctl->free_extents;

fs/btrfs/inode.c

@@ -153,7 +153,7 @@ static int insert_inline_extent(struct btrfs_trans_handle *trans,
 
 		key.objectid = btrfs_ino(inode);
 		key.offset = start;
-		btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
+		key.type = BTRFS_EXTENT_DATA_KEY;
 
 		datasize = btrfs_file_extent_calc_inline_size(cur_size);
 		path->leave_spinning = 1;
@@ -249,8 +249,8 @@ static noinline int cow_file_range_inline(struct btrfs_root *root,
 		data_len = compressed_size;
 
 	if (start > 0 ||
-	    actual_end >= PAGE_CACHE_SIZE ||
-	    data_len >= BTRFS_MAX_INLINE_DATA_SIZE(root) ||
+	    actual_end > PAGE_CACHE_SIZE ||
+	    data_len > BTRFS_MAX_INLINE_DATA_SIZE(root) ||
 	    (!compressed_size &&
 	    (actual_end & (root->sectorsize - 1)) == 0) ||
 	    end + 1 < isize ||
@@ -348,6 +348,23 @@ static noinline int add_async_extent(struct async_cow *cow,
 	return 0;
 }
 
+static inline int inode_need_compress(struct inode *inode)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+
+	/* force compress */
+	if (btrfs_test_opt(root, FORCE_COMPRESS))
+		return 1;
+	/* bad compression ratios */
+	if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS)
+		return 0;
+	if (btrfs_test_opt(root, COMPRESS) ||
+	    BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS ||
+	    BTRFS_I(inode)->force_compress)
+		return 1;
+	return 0;
+}
+
 /*
  * we create compressed extents in two phases.  The first
  * phase compresses a range of pages that have already been
@@ -444,10 +461,7 @@ again:
 	 * inode has not been flagged as nocompress.  This flag can
 	 * change at any time if we discover bad compression ratios.
 	 */
-	if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS) &&
-	    (btrfs_test_opt(root, COMPRESS) ||
-	     (BTRFS_I(inode)->force_compress) ||
-	     (BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS))) {
+	if (inode_need_compress(inode)) {
 		WARN_ON(pages);
 		pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS);
 		if (!pages) {
@@ -1094,7 +1108,8 @@ static int cow_file_range_async(struct inode *inode, struct page *locked_page,
 		async_cow->locked_page = locked_page;
 		async_cow->start = start;
 
-		if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS)
+		if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS &&
+		    !btrfs_test_opt(root, FORCE_COMPRESS))
 			cur_end = end;
 		else
 			cur_end = min(end, start + 512 * 1024 - 1);
@@ -1445,6 +1460,26 @@ error:
 	return ret;
 }
 
+static inline int need_force_cow(struct inode *inode, u64 start, u64 end)
+{
+
+	if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW) &&
+	    !(BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC))
+		return 0;
+
+	/*
+	 * @defrag_bytes is a hint value, no spinlock held here,
+	 * if is not zero, it means the file is defragging.
+	 * Force cow if given extent needs to be defragged.
+	 */
+	if (BTRFS_I(inode)->defrag_bytes &&
+	    test_range_bit(&BTRFS_I(inode)->io_tree, start, end,
+			   EXTENT_DEFRAG, 0, NULL))
+		return 1;
+
+	return 0;
+}
+
 /*
  * extent_io.c call back to do delayed allocation processing
  */
@@ -1453,17 +1488,15 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page,
 			      unsigned long *nr_written)
 {
 	int ret;
-	struct btrfs_root *root = BTRFS_I(inode)->root;
+	int force_cow = need_force_cow(inode, start, end);
 
-	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW) {
+	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW && !force_cow) {
 		ret = run_delalloc_nocow(inode, locked_page, start, end,
 					 page_started, 1, nr_written);
-	} else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC) {
+	} else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC && !force_cow) {
 		ret = run_delalloc_nocow(inode, locked_page, start, end,
 					 page_started, 0, nr_written);
-	} else if (!btrfs_test_opt(root, COMPRESS) &&
-		   !(BTRFS_I(inode)->force_compress) &&
-		   !(BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS)) {
+	} else if (!inode_need_compress(inode)) {
 		ret = cow_file_range(inode, locked_page, start, end,
 				      page_started, nr_written, 1);
 	} else {
@@ -1555,6 +1588,8 @@ static void btrfs_set_bit_hook(struct inode *inode,
 			       struct extent_state *state, unsigned long *bits)
 {
 
+	if ((*bits & EXTENT_DEFRAG) && !(*bits & EXTENT_DELALLOC))
+		WARN_ON(1);
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
 	 * but in this case, we are only testing for the DELALLOC
@@ -1577,6 +1612,8 @@ static void btrfs_set_bit_hook(struct inode *inode,
 				     root->fs_info->delalloc_batch);
 		spin_lock(&BTRFS_I(inode)->lock);
 		BTRFS_I(inode)->delalloc_bytes += len;
+		if (*bits & EXTENT_DEFRAG)
+			BTRFS_I(inode)->defrag_bytes += len;
 		if (do_list && !test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
 					 &BTRFS_I(inode)->runtime_flags))
 			btrfs_add_delalloc_inodes(root, inode);
@@ -1591,6 +1628,13 @@ static void btrfs_clear_bit_hook(struct inode *inode,
 				 struct extent_state *state,
 				 unsigned long *bits)
 {
+	u64 len = state->end + 1 - state->start;
+
+	spin_lock(&BTRFS_I(inode)->lock);
+	if ((state->state & EXTENT_DEFRAG) && (*bits & EXTENT_DEFRAG))
+		BTRFS_I(inode)->defrag_bytes -= len;
+	spin_unlock(&BTRFS_I(inode)->lock);
+
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
 	 * but in this case, we are only testing for the DELALLOC
@@ -1598,7 +1642,6 @@ static void btrfs_clear_bit_hook(struct inode *inode,
 	 */
 	if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
 		struct btrfs_root *root = BTRFS_I(inode)->root;
-		u64 len = state->end + 1 - state->start;
 		bool do_list = !btrfs_is_free_space_inode(inode);
 
 		if (*bits & EXTENT_FIRST_DELALLOC) {
@@ -2660,6 +2703,10 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 		goto out;
 	}
 
+	btrfs_free_io_failure_record(inode, ordered_extent->file_offset,
+				     ordered_extent->file_offset +
+				     ordered_extent->len - 1);
+
 	if (test_bit(BTRFS_ORDERED_TRUNCATED, &ordered_extent->flags)) {
 		truncated = true;
 		logical_len = ordered_extent->truncated_len;
@@ -2856,6 +2903,40 @@ static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
 	return 0;
 }
 
+static int __readpage_endio_check(struct inode *inode,
+				  struct btrfs_io_bio *io_bio,
+				  int icsum, struct page *page,
+				  int pgoff, u64 start, size_t len)
+{
+	char *kaddr;
+	u32 csum_expected;
+	u32 csum = ~(u32)0;
+	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
+				      DEFAULT_RATELIMIT_BURST);
+
+	csum_expected = *(((u32 *)io_bio->csum) + icsum);
+
+	kaddr = kmap_atomic(page);
+	csum = btrfs_csum_data(kaddr + pgoff, csum,  len);
+	btrfs_csum_final(csum, (char *)&csum);
+	if (csum != csum_expected)
+		goto zeroit;
+
+	kunmap_atomic(kaddr);
+	return 0;
+zeroit:
+	if (__ratelimit(&_rs))
+		btrfs_info(BTRFS_I(inode)->root->fs_info,
+			   "csum failed ino %llu off %llu csum %u expected csum %u",
+			   btrfs_ino(inode), start, csum, csum_expected);
+	memset(kaddr + pgoff, 1, len);
+	flush_dcache_page(page);
+	kunmap_atomic(kaddr);
+	if (csum_expected == 0)
+		return 0;
+	return -EIO;
+}
+
 /*
  * when reads are done, we need to check csums to verify the data is correct
  * if there's a match, we allow the bio to finish.  If not, the code in
@@ -2868,20 +2949,15 @@ static int btrfs_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 	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;
 	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);
 
 	if (PageChecked(page)) {
 		ClearPageChecked(page);
-		goto good;
+		return 0;
 	}
 
 	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)
-		goto good;
+		return 0;
 
 	if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID &&
 	    test_range_bit(io_tree, start, end, EXTENT_NODATASUM, 1, NULL)) {
@@ -2891,28 +2967,8 @@ static int btrfs_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 	}
 
 	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 != csum_expected)
-		goto zeroit;
-
-	kunmap_atomic(kaddr);
-good:
-	return 0;
-
-zeroit:
-	if (__ratelimit(&_rs))
-		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 (csum_expected == 0)
-		return 0;
-	return -EIO;
+	return __readpage_endio_check(inode, io_bio, phy_offset, page, offset,
+				      start, (size_t)(end - start + 1));
 }
 
 struct delayed_iput {
@@ -3159,7 +3215,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 	path->reada = -1;
 
 	key.objectid = BTRFS_ORPHAN_OBJECTID;
-	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
+	key.type = BTRFS_ORPHAN_ITEM_KEY;
 	key.offset = (u64)-1;
 
 	while (1) {
@@ -3186,7 +3242,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 		/* make sure the item matches what we want */
 		if (found_key.objectid != BTRFS_ORPHAN_OBJECTID)
 			break;
-		if (btrfs_key_type(&found_key) != BTRFS_ORPHAN_ITEM_KEY)
+		if (found_key.type != BTRFS_ORPHAN_ITEM_KEY)
 			break;
 
 		/* release the path since we're done with it */
@@ -3662,7 +3718,8 @@ noinline int btrfs_update_inode(struct btrfs_trans_handle *trans,
 	 * without delay
 	 */
 	if (!btrfs_is_free_space_inode(inode)
-	    && root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID) {
+	    && root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID
+	    && !root->fs_info->log_root_recovering) {
 		btrfs_update_root_times(trans, root);
 
 		ret = btrfs_delayed_update_inode(trans, root, inode);
@@ -4085,7 +4142,7 @@ search_again:
 		fi = NULL;
 		leaf = path->nodes[0];
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
-		found_type = btrfs_key_type(&found_key);
+		found_type = found_key.type;
 
 		if (found_key.objectid != ino)
 			break;
@@ -4747,6 +4804,8 @@ void btrfs_evict_inode(struct inode *inode)
 	/* do we really want it for ->i_nlink > 0 and zero btrfs_root_refs? */
 	btrfs_wait_ordered_range(inode, 0, (u64)-1);
 
+	btrfs_free_io_failure_record(inode, 0, (u64)-1);
+
 	if (root->fs_info->log_root_recovering) {
 		BUG_ON(test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
 				 &BTRFS_I(inode)->runtime_flags));
@@ -5331,7 +5390,7 @@ static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
 		btrfs_get_delayed_items(inode, &ins_list, &del_list);
 	}
 
-	btrfs_set_key_type(&key, key_type);
+	key.type = key_type;
 	key.offset = ctx->pos;
 	key.objectid = btrfs_ino(inode);
 
@@ -5356,7 +5415,7 @@ static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
 
 		if (found_key.objectid != key.objectid)
 			break;
-		if (btrfs_key_type(&found_key) != key_type)
+		if (found_key.type != key_type)
 			break;
 		if (found_key.offset < ctx->pos)
 			goto next;
@@ -5568,7 +5627,7 @@ static int btrfs_set_inode_index_count(struct inode *inode)
 	int ret;
 
 	key.objectid = btrfs_ino(inode);
-	btrfs_set_key_type(&key, BTRFS_DIR_INDEX_KEY);
+	key.type = BTRFS_DIR_INDEX_KEY;
 	key.offset = (u64)-1;
 
 	path = btrfs_alloc_path();
@@ -5600,7 +5659,7 @@ static int btrfs_set_inode_index_count(struct inode *inode)
 	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 
 	if (found_key.objectid != btrfs_ino(inode) ||
-	    btrfs_key_type(&found_key) != BTRFS_DIR_INDEX_KEY) {
+	    found_key.type != BTRFS_DIR_INDEX_KEY) {
 		BTRFS_I(inode)->index_cnt = 2;
 		goto out;
 	}
@@ -5718,7 +5777,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &BTRFS_I(inode)->runtime_flags);
 
 	key[0].objectid = objectid;
-	btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY);
+	key[0].type = BTRFS_INODE_ITEM_KEY;
 	key[0].offset = 0;
 
 	sizes[0] = sizeof(struct btrfs_inode_item);
@@ -5731,7 +5790,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 		 * add more hard links than can fit in the ref item.
 		 */
 		key[1].objectid = objectid;
-		btrfs_set_key_type(&key[1], BTRFS_INODE_REF_KEY);
+		key[1].type = BTRFS_INODE_REF_KEY;
 		key[1].offset = ref_objectid;
 
 		sizes[1] = name_len + sizeof(*ref);
@@ -5740,7 +5799,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	location = &BTRFS_I(inode)->location;
 	location->objectid = objectid;
 	location->offset = 0;
-	btrfs_set_key_type(location, BTRFS_INODE_ITEM_KEY);
+	location->type = BTRFS_INODE_ITEM_KEY;
 
 	ret = btrfs_insert_inode_locked(inode);
 	if (ret < 0)
@@ -5832,7 +5891,7 @@ int btrfs_add_link(struct btrfs_trans_handle *trans,
 		memcpy(&key, &BTRFS_I(inode)->root->root_key, sizeof(key));
 	} else {
 		key.objectid = ino;
-		btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
+		key.type = BTRFS_INODE_ITEM_KEY;
 		key.offset = 0;
 	}
 
@@ -6191,21 +6250,60 @@ out_fail_inode:
 	goto out_fail;
 }
 
+/* Find next extent map of a given extent map, caller needs to ensure locks */
+static struct extent_map *next_extent_map(struct extent_map *em)
+{
+	struct rb_node *next;
+
+	next = rb_next(&em->rb_node);
+	if (!next)
+		return NULL;
+	return container_of(next, struct extent_map, rb_node);
+}
+
+static struct extent_map *prev_extent_map(struct extent_map *em)
+{
+	struct rb_node *prev;
+
+	prev = rb_prev(&em->rb_node);
+	if (!prev)
+		return NULL;
+	return container_of(prev, struct extent_map, rb_node);
+}
+
 /* helper for btfs_get_extent.  Given an existing extent in the tree,
+ * the existing extent is the nearest extent to map_start,
  * and an extent that you want to insert, deal with overlap and insert
- * the new extent into the tree.
+ * the best fitted new extent into the tree.
  */
 static int merge_extent_mapping(struct extent_map_tree *em_tree,
 				struct extent_map *existing,
 				struct extent_map *em,
 				u64 map_start)
 {
+	struct extent_map *prev;
+	struct extent_map *next;
+	u64 start;
+	u64 end;
 	u64 start_diff;
 
 	BUG_ON(map_start < em->start || map_start >= extent_map_end(em));
-	start_diff = map_start - em->start;
-	em->start = map_start;
-	em->len = existing->start - em->start;
+
+	if (existing->start > map_start) {
+		next = existing;
+		prev = prev_extent_map(next);
+	} else {
+		prev = existing;
+		next = next_extent_map(prev);
+	}
+
+	start = prev ? extent_map_end(prev) : em->start;
+	start = max_t(u64, start, em->start);
+	end = next ? next->start : extent_map_end(em);
+	end = min_t(u64, end, extent_map_end(em));
+	start_diff = start - em->start;
+	em->start = start;
+	em->len = end - start;
 	if (em->block_start < EXTENT_MAP_LAST_BYTE &&
 	    !test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) {
 		em->block_start += start_diff;
@@ -6333,7 +6431,7 @@ again:
 			      struct btrfs_file_extent_item);
 	/* are we inside the extent that was found? */
 	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
-	found_type = btrfs_key_type(&found_key);
+	found_type = found_key.type;
 	if (found_key.objectid != objectid ||
 	    found_type != BTRFS_EXTENT_DATA_KEY) {
 		/*
@@ -6482,25 +6580,21 @@ insert:
 
 		ret = 0;
 
-		existing = lookup_extent_mapping(em_tree, start, len);
-		if (existing && (existing->start > start ||
-		    existing->start + existing->len <= start)) {
+		existing = search_extent_mapping(em_tree, start, len);
+		/*
+		 * existing will always be non-NULL, since there must be
+		 * extent causing the -EEXIST.
+		 */
+		if (start >= extent_map_end(existing) ||
+		    start <= existing->start) {
+			/*
+			 * The existing extent map is the one nearest to
+			 * the [start, start + len) range which overlaps
+			 */
+			err = merge_extent_mapping(em_tree, existing,
+						   em, start);
 			free_extent_map(existing);
-			existing = NULL;
-		}
-		if (!existing) {
-			existing = lookup_extent_mapping(em_tree, em->start,
-							 em->len);
-			if (existing) {
-				err = merge_extent_mapping(em_tree, existing,
-							   em, start);
-				free_extent_map(existing);
-				if (err) {
-					free_extent_map(em);
-					em = NULL;
-				}
-			} else {
-				err = -EIO;
+			if (err) {
 				free_extent_map(em);
 				em = NULL;
 			}
@@ -7112,8 +7206,10 @@ 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))
+				if (IS_ERR(em)) {
+					ret = PTR_ERR(em);
 					goto unlock_err;
+				}
 			}
 
 			ret = btrfs_add_ordered_extent_dio(inode, start,
@@ -7188,45 +7284,277 @@ unlock_err:
 	return ret;
 }
 
-static void btrfs_endio_direct_read(struct bio *bio, int err)
+static inline int submit_dio_repair_bio(struct inode *inode, struct bio *bio,
+					int rw, int mirror_num)
 {
-	struct btrfs_dio_private *dip = bio->bi_private;
-	struct bio_vec *bvec;
-	struct inode *inode = dip->inode;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct bio *dio_bio;
-	u32 *csums = (u32 *)dip->csum;
+	int ret;
+
+	BUG_ON(rw & REQ_WRITE);
+
+	bio_get(bio);
+
+	ret = btrfs_bio_wq_end_io(root->fs_info, bio,
+				  BTRFS_WQ_ENDIO_DIO_REPAIR);
+	if (ret)
+		goto err;
+
+	ret = btrfs_map_bio(root, rw, bio, mirror_num, 0);
+err:
+	bio_put(bio);
+	return ret;
+}
+
+static int btrfs_check_dio_repairable(struct inode *inode,
+				      struct bio *failed_bio,
+				      struct io_failure_record *failrec,
+				      int failed_mirror)
+{
+	int num_copies;
+
+	num_copies = btrfs_num_copies(BTRFS_I(inode)->root->fs_info,
+				      failrec->logical, failrec->len);
+	if (num_copies == 1) {
+		/*
+		 * we only have a single copy of the data, so don't bother with
+		 * all the retry and error correction code that follows. no
+		 * matter what the error is, it is very likely to persist.
+		 */
+		pr_debug("Check DIO Repairable: cannot repair, num_copies=%d, next_mirror %d, failed_mirror %d\n",
+			 num_copies, failrec->this_mirror, failed_mirror);
+		return 0;
+	}
+
+	failrec->failed_mirror = failed_mirror;
+	failrec->this_mirror++;
+	if (failrec->this_mirror == failed_mirror)
+		failrec->this_mirror++;
+
+	if (failrec->this_mirror > num_copies) {
+		pr_debug("Check DIO Repairable: (fail) num_copies=%d, next_mirror %d, failed_mirror %d\n",
+			 num_copies, failrec->this_mirror, failed_mirror);
+		return 0;
+	}
+
+	return 1;
+}
+
+static int dio_read_error(struct inode *inode, struct bio *failed_bio,
+			  struct page *page, u64 start, u64 end,
+			  int failed_mirror, bio_end_io_t *repair_endio,
+			  void *repair_arg)
+{
+	struct io_failure_record *failrec;
+	struct bio *bio;
+	int isector;
+	int read_mode;
+	int ret;
+
+	BUG_ON(failed_bio->bi_rw & REQ_WRITE);
+
+	ret = btrfs_get_io_failure_record(inode, start, end, &failrec);
+	if (ret)
+		return ret;
+
+	ret = btrfs_check_dio_repairable(inode, failed_bio, failrec,
+					 failed_mirror);
+	if (!ret) {
+		free_io_failure(inode, failrec);
+		return -EIO;
+	}
+
+	if (failed_bio->bi_vcnt > 1)
+		read_mode = READ_SYNC | REQ_FAILFAST_DEV;
+	else
+		read_mode = READ_SYNC;
+
+	isector = start - btrfs_io_bio(failed_bio)->logical;
+	isector >>= inode->i_sb->s_blocksize_bits;
+	bio = btrfs_create_repair_bio(inode, failed_bio, failrec, page,
+				      0, isector, repair_endio, repair_arg);
+	if (!bio) {
+		free_io_failure(inode, failrec);
+		return -EIO;
+	}
+
+	btrfs_debug(BTRFS_I(inode)->root->fs_info,
+		    "Repair DIO Read Error: submitting new dio read[%#x] to this_mirror=%d, in_validation=%d\n",
+		    read_mode, failrec->this_mirror, failrec->in_validation);
+
+	ret = submit_dio_repair_bio(inode, bio, read_mode,
+				    failrec->this_mirror);
+	if (ret) {
+		free_io_failure(inode, failrec);
+		bio_put(bio);
+	}
+
+	return ret;
+}
+
+struct btrfs_retry_complete {
+	struct completion done;
+	struct inode *inode;
+	u64 start;
+	int uptodate;
+};
+
+static void btrfs_retry_endio_nocsum(struct bio *bio, int err)
+{
+	struct btrfs_retry_complete *done = bio->bi_private;
+	struct bio_vec *bvec;
+	int i;
+
+	if (err)
+		goto end;
+
+	done->uptodate = 1;
+	bio_for_each_segment_all(bvec, bio, i)
+		clean_io_failure(done->inode, done->start, bvec->bv_page, 0);
+end:
+	complete(&done->done);
+	bio_put(bio);
+}
+
+static int __btrfs_correct_data_nocsum(struct inode *inode,
+				       struct btrfs_io_bio *io_bio)
+{
+	struct bio_vec *bvec;
+	struct btrfs_retry_complete done;
 	u64 start;
 	int i;
+	int ret;
+
+	start = io_bio->logical;
+	done.inode = inode;
+
+	bio_for_each_segment_all(bvec, &io_bio->bio, i) {
+try_again:
+		done.uptodate = 0;
+		done.start = start;
+		init_completion(&done.done);
+
+		ret = dio_read_error(inode, &io_bio->bio, bvec->bv_page, start,
+				     start + bvec->bv_len - 1,
+				     io_bio->mirror_num,
+				     btrfs_retry_endio_nocsum, &done);
+		if (ret)
+			return ret;
+
+		wait_for_completion(&done.done);
+
+		if (!done.uptodate) {
+			/* We might have another mirror, so try again */
+			goto try_again;
+		}
+
+		start += bvec->bv_len;
+	}
+
+	return 0;
+}
+
+static void btrfs_retry_endio(struct bio *bio, int err)
+{
+	struct btrfs_retry_complete *done = bio->bi_private;
+	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
+	struct bio_vec *bvec;
+	int uptodate;
+	int ret;
+	int i;
+
+	if (err)
+		goto end;
 
-	start = dip->logical_offset;
+	uptodate = 1;
 	bio_for_each_segment_all(bvec, bio, i) {
-		if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)) {
-			struct page *page = bvec->bv_page;
-			char *kaddr;
-			u32 csum = ~(u32)0;
-			unsigned long flags;
-
-			local_irq_save(flags);
-			kaddr = kmap_atomic(page);
-			csum = btrfs_csum_data(kaddr + bvec->bv_offset,
-					       csum, bvec->bv_len);
-			btrfs_csum_final(csum, (char *)&csum);
-			kunmap_atomic(kaddr);
-			local_irq_restore(flags);
-
-			flush_dcache_page(bvec->bv_page);
-			if (csum != csums[i]) {
-				btrfs_err(root->fs_info, "csum failed ino %llu off %llu csum %u expected csum %u",
-					  btrfs_ino(inode), start, csum,
-					  csums[i]);
-				err = -EIO;
-			}
+		ret = __readpage_endio_check(done->inode, io_bio, i,
+					     bvec->bv_page, 0,
+					     done->start, bvec->bv_len);
+		if (!ret)
+			clean_io_failure(done->inode, done->start,
+					 bvec->bv_page, 0);
+		else
+			uptodate = 0;
+	}
+
+	done->uptodate = uptodate;
+end:
+	complete(&done->done);
+	bio_put(bio);
+}
+
+static int __btrfs_subio_endio_read(struct inode *inode,
+				    struct btrfs_io_bio *io_bio, int err)
+{
+	struct bio_vec *bvec;
+	struct btrfs_retry_complete done;
+	u64 start;
+	u64 offset = 0;
+	int i;
+	int ret;
+
+	err = 0;
+	start = io_bio->logical;
+	done.inode = inode;
+
+	bio_for_each_segment_all(bvec, &io_bio->bio, i) {
+		ret = __readpage_endio_check(inode, io_bio, i, bvec->bv_page,
+					     0, start, bvec->bv_len);
+		if (likely(!ret))
+			goto next;
+try_again:
+		done.uptodate = 0;
+		done.start = start;
+		init_completion(&done.done);
+
+		ret = dio_read_error(inode, &io_bio->bio, bvec->bv_page, start,
+				     start + bvec->bv_len - 1,
+				     io_bio->mirror_num,
+				     btrfs_retry_endio, &done);
+		if (ret) {
+			err = ret;
+			goto next;
 		}
 
+		wait_for_completion(&done.done);
+
+		if (!done.uptodate) {
+			/* We might have another mirror, so try again */
+			goto try_again;
+		}
+next:
+		offset += bvec->bv_len;
 		start += bvec->bv_len;
 	}
 
+	return err;
+}
+
+static int btrfs_subio_endio_read(struct inode *inode,
+				  struct btrfs_io_bio *io_bio, int err)
+{
+	bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
+
+	if (skip_csum) {
+		if (unlikely(err))
+			return __btrfs_correct_data_nocsum(inode, io_bio);
+		else
+			return 0;
+	} else {
+		return __btrfs_subio_endio_read(inode, io_bio, err);
+	}
+}
+
+static void btrfs_endio_direct_read(struct bio *bio, int err)
+{
+	struct btrfs_dio_private *dip = bio->bi_private;
+	struct inode *inode = dip->inode;
+	struct bio *dio_bio;
+	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
+
+	if (dip->flags & BTRFS_DIO_ORIG_BIO_SUBMITTED)
+		err = btrfs_subio_endio_read(inode, io_bio, err);
+
 	unlock_extent(&BTRFS_I(inode)->io_tree, dip->logical_offset,
 		      dip->logical_offset + dip->bytes - 1);
 	dio_bio = dip->dio_bio;
@@ -7237,6 +7565,9 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 	if (err)
 		clear_bit(BIO_UPTODATE, &dio_bio->bi_flags);
 	dio_end_io(dio_bio, err);
+
+	if (io_bio->end_io)
+		io_bio->end_io(io_bio, err);
 	bio_put(bio);
 }
 
@@ -7302,12 +7633,17 @@ static void btrfs_end_dio_bio(struct bio *bio, int err)
 {
 	struct btrfs_dio_private *dip = bio->bi_private;
 
+	if (err)
+		btrfs_warn(BTRFS_I(dip->inode)->root->fs_info,
+			   "direct IO failed ino %llu rw %lu sector %#Lx len %u err no %d",
+			   btrfs_ino(dip->inode), bio->bi_rw,
+			   (unsigned long long)bio->bi_iter.bi_sector,
+			   bio->bi_iter.bi_size, err);
+
+	if (dip->subio_endio)
+		err = dip->subio_endio(dip->inode, btrfs_io_bio(bio), err);
+
 	if (err) {
-		btrfs_err(BTRFS_I(dip->inode)->root->fs_info,
-			  "direct IO failed ino %llu rw %lu sector %#Lx len %u err no %d",
-		      btrfs_ino(dip->inode), bio->bi_rw,
-		      (unsigned long long)bio->bi_iter.bi_sector,
-		      bio->bi_iter.bi_size, err);
 		dip->errors = 1;
 
 		/*
@@ -7338,6 +7674,38 @@ static struct bio *btrfs_dio_bio_alloc(struct block_device *bdev,
 	return btrfs_bio_alloc(bdev, first_sector, nr_vecs, gfp_flags);
 }
 
+static inline int btrfs_lookup_and_bind_dio_csum(struct btrfs_root *root,
+						 struct inode *inode,
+						 struct btrfs_dio_private *dip,
+						 struct bio *bio,
+						 u64 file_offset)
+{
+	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
+	struct btrfs_io_bio *orig_io_bio = btrfs_io_bio(dip->orig_bio);
+	int ret;
+
+	/*
+	 * We load all the csum data we need when we submit
+	 * the first bio to reduce the csum tree search and
+	 * contention.
+	 */
+	if (dip->logical_offset == file_offset) {
+		ret = btrfs_lookup_bio_sums_dio(root, inode, dip->orig_bio,
+						file_offset);
+		if (ret)
+			return ret;
+	}
+
+	if (bio == dip->orig_bio)
+		return 0;
+
+	file_offset -= dip->logical_offset;
+	file_offset >>= inode->i_sb->s_blocksize_bits;
+	io_bio->csum = (u8 *)(((u32 *)orig_io_bio->csum) + file_offset);
+
+	return 0;
+}
+
 static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
 					 int rw, u64 file_offset, int skip_sum,
 					 int async_submit)
@@ -7353,7 +7721,8 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
 	bio_get(bio);
 
 	if (!write) {
-		ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
+		ret = btrfs_bio_wq_end_io(root->fs_info, bio,
+				BTRFS_WQ_ENDIO_DATA);
 		if (ret)
 			goto err;
 	}
@@ -7376,13 +7745,12 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
 		ret = btrfs_csum_one_bio(root, inode, bio, file_offset, 1);
 		if (ret)
 			goto err;
-	} else if (!skip_sum) {
-		ret = btrfs_lookup_bio_sums_dio(root, inode, dip, bio,
-						file_offset);
+	} else {
+		ret = btrfs_lookup_and_bind_dio_csum(root, inode, dip, bio,
+						     file_offset);
 		if (ret)
 			goto err;
 	}
-
 map:
 	ret = btrfs_map_bio(root, rw, bio, 0, async_submit);
 err:
@@ -7403,7 +7771,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 	u64 submit_len = 0;
 	u64 map_length;
 	int nr_pages = 0;
-	int ret = 0;
+	int ret;
 	int async_submit = 0;
 
 	map_length = orig_bio->bi_iter.bi_size;
@@ -7414,6 +7782,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 
 	if (map_length >= orig_bio->bi_iter.bi_size) {
 		bio = orig_bio;
+		dip->flags |= BTRFS_DIO_ORIG_BIO_SUBMITTED;
 		goto submit;
 	}
 
@@ -7430,12 +7799,13 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 
 	bio->bi_private = dip;
 	bio->bi_end_io = btrfs_end_dio_bio;
+	btrfs_io_bio(bio)->logical = file_offset;
 	atomic_inc(&dip->pending_bios);
 
 	while (bvec <= (orig_bio->bi_io_vec + orig_bio->bi_vcnt - 1)) {
-		if (unlikely(map_length < submit_len + bvec->bv_len ||
+		if (map_length < submit_len + bvec->bv_len ||
 		    bio_add_page(bio, bvec->bv_page, bvec->bv_len,
-				 bvec->bv_offset) < bvec->bv_len)) {
+				 bvec->bv_offset) < bvec->bv_len) {
 			/*
 			 * inc the count before we submit the bio so
 			 * we know the end IO handler won't happen before
@@ -7464,6 +7834,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip,
 				goto out_err;
 			bio->bi_private = dip;
 			bio->bi_end_io = btrfs_end_dio_bio;
+			btrfs_io_bio(bio)->logical = file_offset;
 
 			map_length = orig_bio->bi_iter.bi_size;
 			ret = btrfs_map_block(root->fs_info, rw,
@@ -7507,11 +7878,10 @@ static void btrfs_submit_direct(int rw, struct bio *dio_bio,
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_dio_private *dip;
 	struct bio *io_bio;
+	struct btrfs_io_bio *btrfs_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;
 
@@ -7521,16 +7891,7 @@ static void btrfs_submit_direct(int rw, struct bio *dio_bio,
 		goto free_ordered;
 	}
 
-	if (!skip_sum && !write) {
-		csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
-		sum_len = dio_bio->bi_iter.bi_size >>
-			inode->i_sb->s_blocksize_bits;
-		sum_len *= csum_size;
-	} else {
-		sum_len = 0;
-	}
-
-	dip = kmalloc(sizeof(*dip) + sum_len, GFP_NOFS);
+	dip = kzalloc(sizeof(*dip), GFP_NOFS);
 	if (!dip) {
 		ret = -ENOMEM;
 		goto free_io_bio;
@@ -7542,20 +7903,25 @@ static void btrfs_submit_direct(int rw, struct bio *dio_bio,
 	dip->bytes = dio_bio->bi_iter.bi_size;
 	dip->disk_bytenr = (u64)dio_bio->bi_iter.bi_sector << 9;
 	io_bio->bi_private = dip;
-	dip->errors = 0;
 	dip->orig_bio = io_bio;
 	dip->dio_bio = dio_bio;
 	atomic_set(&dip->pending_bios, 0);
+	btrfs_bio = btrfs_io_bio(io_bio);
+	btrfs_bio->logical = file_offset;
 
-	if (write)
+	if (write) {
 		io_bio->bi_end_io = btrfs_endio_direct_write;
-	else
+	} else {
 		io_bio->bi_end_io = btrfs_endio_direct_read;
+		dip->subio_endio = btrfs_subio_endio_read;
+	}
 
 	ret = btrfs_submit_direct_hook(rw, dip, skip_sum);
 	if (!ret)
 		return;
 
+	if (btrfs_bio->end_io)
+		btrfs_bio->end_io(btrfs_bio, ret);
 free_io_bio:
 	bio_put(io_bio);
 
@@ -7652,8 +8018,8 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 		ret = btrfs_delalloc_reserve_space(inode, count);
 		if (ret)
 			goto out;
-	} else if (unlikely(test_bit(BTRFS_INODE_READDIO_NEED_LOCK,
-				     &BTRFS_I(inode)->runtime_flags))) {
+	} else if (test_bit(BTRFS_INODE_READDIO_NEED_LOCK,
+				     &BTRFS_I(inode)->runtime_flags)) {
 		inode_dio_done(inode);
 		flags = DIO_LOCKING | DIO_SKIP_HOLES;
 		wakeup = false;
@@ -8173,6 +8539,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	ei->last_sub_trans = 0;
 	ei->logged_trans = 0;
 	ei->delalloc_bytes = 0;
+	ei->defrag_bytes = 0;
 	ei->disk_i_size = 0;
 	ei->flags = 0;
 	ei->csum_bytes = 0;
@@ -8231,6 +8598,7 @@ void btrfs_destroy_inode(struct inode *inode)
 	WARN_ON(BTRFS_I(inode)->reserved_extents);
 	WARN_ON(BTRFS_I(inode)->delalloc_bytes);
 	WARN_ON(BTRFS_I(inode)->csum_bytes);
+	WARN_ON(BTRFS_I(inode)->defrag_bytes);
 
 	/*
 	 * This can happen where we create an inode, but somebody else also
@@ -8646,7 +9014,7 @@ static int __start_delalloc_inodes(struct btrfs_root *root, int delay_iput,
 		spin_unlock(&root->delalloc_lock);
 
 		work = btrfs_alloc_delalloc_work(inode, 0, delay_iput);
-		if (unlikely(!work)) {
+		if (!work) {
 			if (delay_iput)
 				btrfs_add_delayed_iput(inode);
 			else
@@ -8832,7 +9200,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 	}
 	key.objectid = btrfs_ino(inode);
 	key.offset = 0;
-	btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
+	key.type = BTRFS_EXTENT_DATA_KEY;
 	datasize = btrfs_file_extent_calc_inline_size(name_len);
 	err = btrfs_insert_empty_item(trans, root, path, &key,
 				      datasize);

fs/btrfs/ioctl.c

@@ -332,6 +332,9 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 			goto out_drop;
 
 	} else {
+		ret = btrfs_set_prop(inode, "btrfs.compression", NULL, 0, 0);
+		if (ret && ret != -ENODATA)
+			goto out_drop;
 		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
 	}
 
@@ -477,8 +480,7 @@ static noinline int create_subvol(struct inode *dir,
 	if (ret)
 		goto fail;
 
-	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
-				      0, objectid, NULL, 0, 0, 0);
+	leaf = btrfs_alloc_tree_block(trans, root, 0, objectid, NULL, 0, 0, 0);
 	if (IS_ERR(leaf)) {
 		ret = PTR_ERR(leaf);
 		goto fail;
@@ -503,7 +505,7 @@ static noinline int create_subvol(struct inode *dir,
 	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_nbytes(inode_item, root->nodesize);
 	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
 
 	btrfs_set_root_flags(&root_item, 0);
@@ -535,7 +537,7 @@ static noinline int create_subvol(struct inode *dir,
 
 	key.objectid = objectid;
 	key.offset = 0;
-	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
+	key.type = BTRFS_ROOT_ITEM_KEY;
 	ret = btrfs_insert_root(trans, root->fs_info->tree_root, &key,
 				&root_item);
 	if (ret)
@@ -882,7 +884,7 @@ out_unlock:
  * file you want to defrag, we return 0 to let you know to skip this
  * part of the file
  */
-static int check_defrag_in_cache(struct inode *inode, u64 offset, int thresh)
+static int check_defrag_in_cache(struct inode *inode, u64 offset, u32 thresh)
 {
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	struct extent_map *em = NULL;
@@ -917,7 +919,7 @@ static int check_defrag_in_cache(struct inode *inode, u64 offset, int thresh)
  */
 static int find_new_extents(struct btrfs_root *root,
 			    struct inode *inode, u64 newer_than,
-			    u64 *off, int thresh)
+			    u64 *off, u32 thresh)
 {
 	struct btrfs_path *path;
 	struct btrfs_key min_key;
@@ -936,12 +938,9 @@ static int find_new_extents(struct btrfs_root *root,
 	min_key.offset = *off;
 
 	while (1) {
-		path->keep_locks = 1;
 		ret = btrfs_search_forward(root, &min_key, path, newer_than);
 		if (ret != 0)
 			goto none;
-		path->keep_locks = 0;
-		btrfs_unlock_up_safe(path, 1);
 process_slot:
 		if (min_key.objectid != ino)
 			goto none;
@@ -1029,7 +1028,7 @@ static bool defrag_check_next_extent(struct inode *inode, struct extent_map *em)
 	return ret;
 }
 
-static int should_defrag_range(struct inode *inode, u64 start, int thresh,
+static int should_defrag_range(struct inode *inode, u64 start, u32 thresh,
 			       u64 *last_len, u64 *skip, u64 *defrag_end,
 			       int compress)
 {
@@ -1259,7 +1258,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	int ret;
 	int defrag_count = 0;
 	int compress_type = BTRFS_COMPRESS_ZLIB;
-	int extent_thresh = range->extent_thresh;
+	u32 extent_thresh = range->extent_thresh;
 	unsigned long max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
 	unsigned long cluster = max_cluster;
 	u64 new_align = ~((u64)128 * 1024 - 1);
@@ -1335,8 +1334,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 		inode->i_mapping->writeback_index = i;
 
 	while (i <= last_index && defrag_count < max_to_defrag &&
-	       (i < (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >>
-		PAGE_CACHE_SHIFT)) {
+	       (i < DIV_ROUND_UP(i_size_read(inode), PAGE_CACHE_SIZE))) {
 		/*
 		 * make sure we stop running if someone unmounts
 		 * the FS
@@ -1359,7 +1357,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			 * the should_defrag function tells us how much to skip
 			 * bump our counter by the suggested amount
 			 */
-			next = (skip + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+			next = DIV_ROUND_UP(skip, PAGE_CACHE_SIZE);
 			i = max(i + 1, next);
 			continue;
 		}
@@ -1554,7 +1552,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 		goto out_free;
 	}
 
-	old_size = device->total_bytes;
+	old_size = btrfs_device_get_total_bytes(device);
 
 	if (mod < 0) {
 		if (new_size > old_size) {
@@ -2089,8 +2087,6 @@ static noinline int search_ioctl(struct inode *inode,
 	key.type = sk->min_type;
 	key.offset = sk->min_offset;
 
-	path->keep_locks = 1;
-
 	while (1) {
 		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
 		if (ret != 0) {
@@ -2526,9 +2522,9 @@ out_unlock:
 		ASSERT(dest->send_in_progress == 0);
 
 		/* the last ref */
-		if (dest->cache_inode) {
-			iput(dest->cache_inode);
-			dest->cache_inode = NULL;
+		if (dest->ino_cache_inode) {
+			iput(dest->ino_cache_inode);
+			dest->ino_cache_inode = NULL;
 		}
 	}
 out_dput:
@@ -2634,6 +2630,9 @@ static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
 	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
 	ret = btrfs_init_new_device(root, vol_args->name);
 
+	if (!ret)
+		btrfs_info(root->fs_info, "disk added %s",vol_args->name);
+
 	kfree(vol_args);
 out:
 	mutex_unlock(&root->fs_info->volume_mutex);
@@ -2673,6 +2672,9 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
 	mutex_unlock(&root->fs_info->volume_mutex);
 	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
 
+	if (!ret)
+		btrfs_info(root->fs_info, "disk deleted %s",vol_args->name);
+
 out:
 	kfree(vol_args);
 err_drop:
@@ -2737,8 +2739,8 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 	}
 
 	di_args->devid = dev->devid;
-	di_args->bytes_used = dev->bytes_used;
-	di_args->total_bytes = dev->total_bytes;
+	di_args->bytes_used = btrfs_device_get_bytes_used(dev);
+	di_args->total_bytes = btrfs_device_get_total_bytes(dev);
 	memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid));
 	if (dev->name) {
 		struct rcu_string *name;
@@ -3164,7 +3166,7 @@ static void clone_update_extent_map(struct inode *inode,
 					em->start + em->len - 1, 0);
 	}
 
-	if (unlikely(ret))
+	if (ret)
 		set_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
 			&BTRFS_I(inode)->runtime_flags);
 }
@@ -3199,7 +3201,7 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 	u64 last_dest_end = destoff;
 
 	ret = -ENOMEM;
-	buf = vmalloc(btrfs_level_size(root, 0));
+	buf = vmalloc(root->nodesize);
 	if (!buf)
 		return ret;
 
@@ -3252,11 +3254,11 @@ process_slot:
 		slot = path->slots[0];
 
 		btrfs_item_key_to_cpu(leaf, &key, slot);
-		if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
+		if (key.type > BTRFS_EXTENT_DATA_KEY ||
 		    key.objectid != btrfs_ino(src))
 			break;
 
-		if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
+		if (key.type == BTRFS_EXTENT_DATA_KEY) {
 			struct btrfs_file_extent_item *extent;
 			int type;
 			u32 size;
@@ -5283,6 +5285,12 @@ long btrfs_ioctl(struct file *file, unsigned int
 		if (ret)
 			return ret;
 		ret = btrfs_sync_fs(file->f_dentry->d_sb, 1);
+		/*
+		 * The transaction thread may want to do more work,
+		 * namely it pokes the cleaner ktread that will start
+		 * processing uncleaned subvols.
+		 */
+		wake_up_process(root->fs_info->transaction_kthread);
 		return ret;
 	}
 	case BTRFS_IOC_START_SYNC:

fs/btrfs/lzo.c

@@ -266,8 +266,7 @@ static int lzo_decompress_biovec(struct list_head *ws,
 	char *data_in;
 	unsigned long page_in_index = 0;
 	unsigned long page_out_index = 0;
-	unsigned long total_pages_in = (srclen + PAGE_CACHE_SIZE - 1) /
-					PAGE_CACHE_SIZE;
+	unsigned long total_pages_in = DIV_ROUND_UP(srclen, PAGE_CACHE_SIZE);
 	unsigned long buf_start;
 	unsigned long buf_offset = 0;
 	unsigned long bytes;

fs/btrfs/orphan.c

@@ -27,7 +27,7 @@ int btrfs_insert_orphan_item(struct btrfs_trans_handle *trans,
 	int ret = 0;
 
 	key.objectid = BTRFS_ORPHAN_OBJECTID;
-	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
+	key.type = BTRFS_ORPHAN_ITEM_KEY;
 	key.offset = offset;
 
 	path = btrfs_alloc_path();
@@ -48,7 +48,7 @@ int btrfs_del_orphan_item(struct btrfs_trans_handle *trans,
 	int ret = 0;
 
 	key.objectid = BTRFS_ORPHAN_OBJECTID;
-	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
+	key.type = BTRFS_ORPHAN_ITEM_KEY;
 	key.offset = offset;
 
 	path = btrfs_alloc_path();

fs/btrfs/print-tree.c

@@ -195,7 +195,7 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 	for (i = 0 ; i < nr ; i++) {
 		item = btrfs_item_nr(i);
 		btrfs_item_key_to_cpu(l, &key, i);
-		type = btrfs_key_type(&key);
+		type = key.type;
 		printk(KERN_INFO "\titem %d key (%llu %u %llu) itemoff %d "
 		       "itemsize %d\n",
 			i, key.objectid, type, key.offset,
@@ -336,7 +336,6 @@ void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c)
 	for (i = 0; i < nr; i++) {
 		struct extent_buffer *next = read_tree_block(root,
 					btrfs_node_blockptr(c, i),
-					btrfs_level_size(root, level - 1),
 					btrfs_node_ptr_generation(c, i));
 		if (btrfs_is_leaf(next) &&
 		   level != 1)

fs/btrfs/qgroup.c

@@ -539,10 +539,9 @@ static int add_qgroup_item(struct btrfs_trans_handle *trans,
 	struct extent_buffer *leaf;
 	struct btrfs_key key;
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &quota_root->state)))
+	if (btrfs_test_is_dummy_root(quota_root))
 		return 0;
-#endif
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -551,9 +550,15 @@ static int add_qgroup_item(struct btrfs_trans_handle *trans,
 	key.type = BTRFS_QGROUP_INFO_KEY;
 	key.offset = qgroupid;
 
+	/*
+	 * Avoid a transaction abort by catching -EEXIST here. In that
+	 * case, we proceed by re-initializing the existing structure
+	 * on disk.
+	 */
+
 	ret = btrfs_insert_empty_item(trans, quota_root, path, &key,
 				      sizeof(*qgroup_info));
-	if (ret)
+	if (ret && ret != -EEXIST)
 		goto out;
 
 	leaf = path->nodes[0];
@@ -572,7 +577,7 @@ static int add_qgroup_item(struct btrfs_trans_handle *trans,
 	key.type = BTRFS_QGROUP_LIMIT_KEY;
 	ret = btrfs_insert_empty_item(trans, quota_root, path, &key,
 				      sizeof(*qgroup_limit));
-	if (ret)
+	if (ret && ret != -EEXIST)
 		goto out;
 
 	leaf = path->nodes[0];
@@ -692,10 +697,9 @@ static int update_qgroup_info_item(struct btrfs_trans_handle *trans,
 	int ret;
 	int slot;
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-	if (unlikely(test_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state)))
+	if (btrfs_test_is_dummy_root(root))
 		return 0;
-#endif
+
 	key.objectid = 0;
 	key.type = BTRFS_QGROUP_INFO_KEY;
 	key.offset = qgroup->qgroupid;
@@ -1335,6 +1339,8 @@ int btrfs_qgroup_record_ref(struct btrfs_trans_handle *trans,
 	INIT_LIST_HEAD(&oper->elem.list);
 	oper->elem.seq = 0;
 
+	trace_btrfs_qgroup_record_ref(oper);
+
 	if (type == BTRFS_QGROUP_OPER_SUB_SUBTREE) {
 		/*
 		 * If any operation for this bytenr/ref_root combo
@@ -2077,6 +2083,8 @@ static int btrfs_qgroup_account(struct btrfs_trans_handle *trans,
 
 	ASSERT(is_fstree(oper->ref_root));
 
+	trace_btrfs_qgroup_account(oper);
+
 	switch (oper->type) {
 	case BTRFS_QGROUP_OPER_ADD_EXCL:
 	case BTRFS_QGROUP_OPER_SUB_EXCL:
@@ -2237,7 +2245,6 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans,
 	if (srcid) {
 		struct btrfs_root *srcroot;
 		struct btrfs_key srckey;
-		int srcroot_level;
 
 		srckey.objectid = srcid;
 		srckey.type = BTRFS_ROOT_ITEM_KEY;
@@ -2249,8 +2256,7 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans,
 		}
 
 		rcu_read_lock();
-		srcroot_level = btrfs_header_level(srcroot->node);
-		level_size = btrfs_level_size(srcroot, srcroot_level);
+		level_size = srcroot->nodesize;
 		rcu_read_unlock();
 	}
 
@@ -2566,7 +2572,7 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 		    found.type != BTRFS_METADATA_ITEM_KEY)
 			continue;
 		if (found.type == BTRFS_METADATA_ITEM_KEY)
-			num_bytes = fs_info->extent_root->leafsize;
+			num_bytes = fs_info->extent_root->nodesize;
 		else
 			num_bytes = found.offset;
 

fs/btrfs/raid56.c

@@ -912,7 +912,7 @@ static struct page *page_in_rbio(struct btrfs_raid_bio *rbio,
 static unsigned long rbio_nr_pages(unsigned long stripe_len, int nr_stripes)
 {
 	unsigned long nr = stripe_len * nr_stripes;
-	return (nr + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+	return DIV_ROUND_UP(nr, PAGE_CACHE_SIZE);
 }
 
 /*
@@ -1442,7 +1442,7 @@ static int raid56_rmw_stripe(struct btrfs_raid_bio *rbio)
 	struct btrfs_bio *bbio = rbio->bbio;
 	struct bio_list bio_list;
 	int ret;
-	int nr_pages = (rbio->stripe_len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+	int nr_pages = DIV_ROUND_UP(rbio->stripe_len, PAGE_CACHE_SIZE);
 	int pagenr;
 	int stripe;
 	struct bio *bio;
@@ -1725,7 +1725,7 @@ static void __raid_recover_end_io(struct btrfs_raid_bio *rbio)
 	int pagenr, stripe;
 	void **pointers;
 	int faila = -1, failb = -1;
-	int nr_pages = (rbio->stripe_len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+	int nr_pages = DIV_ROUND_UP(rbio->stripe_len, PAGE_CACHE_SIZE);
 	struct page *page;
 	int err;
 	int i;
@@ -1940,7 +1940,7 @@ static int __raid56_parity_recover(struct btrfs_raid_bio *rbio)
 	struct btrfs_bio *bbio = rbio->bbio;
 	struct bio_list bio_list;
 	int ret;
-	int nr_pages = (rbio->stripe_len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+	int nr_pages = DIV_ROUND_UP(rbio->stripe_len, PAGE_CACHE_SIZE);
 	int pagenr;
 	int stripe;
 	struct bio *bio;

fs/btrfs/reada.c

@@ -347,7 +347,7 @@ static struct reada_extent *reada_find_extent(struct btrfs_root *root,
 	if (!re)
 		return NULL;
 
-	blocksize = btrfs_level_size(root, level);
+	blocksize = root->nodesize;
 	re->logical = logical;
 	re->blocksize = blocksize;
 	re->top = *top;

fs/btrfs/relocation.c

@@ -736,7 +736,8 @@ again:
 		err = ret;
 		goto out;
 	}
-	BUG_ON(!ret || !path1->slots[0]);
+	ASSERT(ret);
+	ASSERT(path1->slots[0]);
 
 	path1->slots[0]--;
 
@@ -746,10 +747,10 @@ again:
 		 * the backref was added previously when processing
 		 * backref of type BTRFS_TREE_BLOCK_REF_KEY
 		 */
-		BUG_ON(!list_is_singular(&cur->upper));
+		ASSERT(list_is_singular(&cur->upper));
 		edge = list_entry(cur->upper.next, struct backref_edge,
 				  list[LOWER]);
-		BUG_ON(!list_empty(&edge->list[UPPER]));
+		ASSERT(list_empty(&edge->list[UPPER]));
 		exist = edge->node[UPPER];
 		/*
 		 * add the upper level block to pending list if we need
@@ -831,7 +832,7 @@ again:
 					cur->cowonly = 1;
 			}
 #else
-		BUG_ON(key.type == BTRFS_EXTENT_REF_V0_KEY);
+		ASSERT(key.type != BTRFS_EXTENT_REF_V0_KEY);
 		if (key.type == BTRFS_SHARED_BLOCK_REF_KEY) {
 #endif
 			if (key.objectid == key.offset) {
@@ -840,7 +841,7 @@ again:
 				 * backref of this type.
 				 */
 				root = find_reloc_root(rc, cur->bytenr);
-				BUG_ON(!root);
+				ASSERT(root);
 				cur->root = root;
 				break;
 			}
@@ -868,7 +869,7 @@ again:
 			} else {
 				upper = rb_entry(rb_node, struct backref_node,
 						 rb_node);
-				BUG_ON(!upper->checked);
+				ASSERT(upper->checked);
 				INIT_LIST_HEAD(&edge->list[UPPER]);
 			}
 			list_add_tail(&edge->list[LOWER], &cur->upper);
@@ -892,7 +893,7 @@ again:
 
 		if (btrfs_root_level(&root->root_item) == cur->level) {
 			/* tree root */
-			BUG_ON(btrfs_root_bytenr(&root->root_item) !=
+			ASSERT(btrfs_root_bytenr(&root->root_item) ==
 			       cur->bytenr);
 			if (should_ignore_root(root))
 				list_add(&cur->list, &useless);
@@ -927,7 +928,7 @@ again:
 		need_check = true;
 		for (; level < BTRFS_MAX_LEVEL; level++) {
 			if (!path2->nodes[level]) {
-				BUG_ON(btrfs_root_bytenr(&root->root_item) !=
+				ASSERT(btrfs_root_bytenr(&root->root_item) ==
 				       lower->bytenr);
 				if (should_ignore_root(root))
 					list_add(&lower->list, &useless);
@@ -977,12 +978,15 @@ again:
 					need_check = false;
 					list_add_tail(&edge->list[UPPER],
 						      &list);
-				} else
+				} else {
+					if (upper->checked)
+						need_check = true;
 					INIT_LIST_HEAD(&edge->list[UPPER]);
+				}
 			} else {
 				upper = rb_entry(rb_node, struct backref_node,
 						 rb_node);
-				BUG_ON(!upper->checked);
+				ASSERT(upper->checked);
 				INIT_LIST_HEAD(&edge->list[UPPER]);
 				if (!upper->owner)
 					upper->owner = btrfs_header_owner(eb);
@@ -1026,7 +1030,7 @@ next:
 	 * everything goes well, connect backref nodes and insert backref nodes
 	 * into the cache.
 	 */
-	BUG_ON(!node->checked);
+	ASSERT(node->checked);
 	cowonly = node->cowonly;
 	if (!cowonly) {
 		rb_node = tree_insert(&cache->rb_root, node->bytenr,
@@ -1062,8 +1066,21 @@ next:
 			continue;
 		}
 
-		BUG_ON(!upper->checked);
-		BUG_ON(cowonly != upper->cowonly);
+		if (!upper->checked) {
+			/*
+			 * Still want to blow up for developers since this is a
+			 * logic bug.
+			 */
+			ASSERT(0);
+			err = -EINVAL;
+			goto out;
+		}
+		if (cowonly != upper->cowonly) {
+			ASSERT(0);
+			err = -EINVAL;
+			goto out;
+		}
+
 		if (!cowonly) {
 			rb_node = tree_insert(&cache->rb_root, upper->bytenr,
 					      &upper->rb_node);
@@ -1086,7 +1103,7 @@ next:
 	while (!list_empty(&useless)) {
 		upper = list_entry(useless.next, struct backref_node, list);
 		list_del_init(&upper->list);
-		BUG_ON(!list_empty(&upper->upper));
+		ASSERT(list_empty(&upper->upper));
 		if (upper == node)
 			node = NULL;
 		if (upper->lowest) {
@@ -1119,29 +1136,45 @@ out:
 	if (err) {
 		while (!list_empty(&useless)) {
 			lower = list_entry(useless.next,
-					   struct backref_node, upper);
-			list_del_init(&lower->upper);
+					   struct backref_node, list);
+			list_del_init(&lower->list);
 		}
-		upper = node;
-		INIT_LIST_HEAD(&list);
-		while (upper) {
-			if (RB_EMPTY_NODE(&upper->rb_node)) {
-				list_splice_tail(&upper->upper, &list);
-				free_backref_node(cache, upper);
-			}
-
-			if (list_empty(&list))
-				break;
-
-			edge = list_entry(list.next, struct backref_edge,
-					  list[LOWER]);
+		while (!list_empty(&list)) {
+			edge = list_first_entry(&list, struct backref_edge,
+						list[UPPER]);
+			list_del(&edge->list[UPPER]);
 			list_del(&edge->list[LOWER]);
+			lower = edge->node[LOWER];
 			upper = edge->node[UPPER];
 			free_backref_edge(cache, edge);
+
+			/*
+			 * Lower is no longer linked to any upper backref nodes
+			 * and isn't in the cache, we can free it ourselves.
+			 */
+			if (list_empty(&lower->upper) &&
+			    RB_EMPTY_NODE(&lower->rb_node))
+				list_add(&lower->list, &useless);
+
+			if (!RB_EMPTY_NODE(&upper->rb_node))
+				continue;
+
+			/* Add this guy's upper edges to the list to proces */
+			list_for_each_entry(edge, &upper->upper, list[LOWER])
+				list_add_tail(&edge->list[UPPER], &list);
+			if (list_empty(&upper->upper))
+				list_add(&upper->list, &useless);
+		}
+
+		while (!list_empty(&useless)) {
+			lower = list_entry(useless.next,
+					   struct backref_node, list);
+			list_del_init(&lower->list);
+			free_backref_node(cache, lower);
 		}
 		return ERR_PTR(err);
 	}
-	BUG_ON(node && node->detached);
+	ASSERT(!node || !node->detached);
 	return node;
 }
 
@@ -1787,7 +1820,7 @@ again:
 			btrfs_node_key_to_cpu(parent, next_key, slot + 1);
 
 		old_bytenr = btrfs_node_blockptr(parent, slot);
-		blocksize = btrfs_level_size(dest, level - 1);
+		blocksize = dest->nodesize;
 		old_ptr_gen = btrfs_node_ptr_generation(parent, slot);
 
 		if (level <= max_level) {
@@ -1813,8 +1846,7 @@ again:
 				break;
 			}
 
-			eb = read_tree_block(dest, old_bytenr, blocksize,
-					     old_ptr_gen);
+			eb = read_tree_block(dest, old_bytenr, old_ptr_gen);
 			if (!eb || !extent_buffer_uptodate(eb)) {
 				ret = (!eb) ? -ENOMEM : -EIO;
 				free_extent_buffer(eb);
@@ -1944,7 +1976,6 @@ int walk_down_reloc_tree(struct btrfs_root *root, struct btrfs_path *path,
 	u64 bytenr;
 	u64 ptr_gen = 0;
 	u64 last_snapshot;
-	u32 blocksize;
 	u32 nritems;
 
 	last_snapshot = btrfs_root_last_snapshot(&root->root_item);
@@ -1970,8 +2001,7 @@ int walk_down_reloc_tree(struct btrfs_root *root, struct btrfs_path *path,
 		}
 
 		bytenr = btrfs_node_blockptr(eb, path->slots[i]);
-		blocksize = btrfs_level_size(root, i - 1);
-		eb = read_tree_block(root, bytenr, blocksize, ptr_gen);
+		eb = read_tree_block(root, bytenr, ptr_gen);
 		if (!eb || !extent_buffer_uptodate(eb)) {
 			free_extent_buffer(eb);
 			return -EIO;
@@ -2316,7 +2346,7 @@ void free_reloc_roots(struct list_head *list)
 }
 
 static noinline_for_stack
-int merge_reloc_roots(struct reloc_control *rc)
+void merge_reloc_roots(struct reloc_control *rc)
 {
 	struct btrfs_root *root;
 	struct btrfs_root *reloc_root;
@@ -2397,7 +2427,6 @@ out:
 	}
 
 	BUG_ON(!RB_EMPTY_ROOT(&rc->reloc_root_tree.rb_root));
-	return ret;
 }
 
 static void free_block_list(struct rb_root *blocks)
@@ -2544,8 +2573,7 @@ u64 calcu_metadata_size(struct reloc_control *rc,
 			if (next->processed && (reserve || next != node))
 				break;
 
-			num_bytes += btrfs_level_size(rc->extent_root,
-						      next->level);
+			num_bytes += rc->extent_root->nodesize;
 
 			if (list_empty(&next->upper))
 				break;
@@ -2679,9 +2707,9 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 				goto next;
 		}
 
-		blocksize = btrfs_level_size(root, node->level);
+		blocksize = root->nodesize;
 		generation = btrfs_node_ptr_generation(upper->eb, slot);
-		eb = read_tree_block(root, bytenr, blocksize, generation);
+		eb = read_tree_block(root, bytenr, generation);
 		if (!eb || !extent_buffer_uptodate(eb)) {
 			free_extent_buffer(eb);
 			err = -EIO;
@@ -2789,7 +2817,7 @@ static void __mark_block_processed(struct reloc_control *rc,
 	u32 blocksize;
 	if (node->level == 0 ||
 	    in_block_group(node->bytenr, rc->block_group)) {
-		blocksize = btrfs_level_size(rc->extent_root, node->level);
+		blocksize = rc->extent_root->nodesize;
 		mark_block_processed(rc, node->bytenr, blocksize);
 	}
 	node->processed = 1;
@@ -2843,7 +2871,7 @@ static int get_tree_block_key(struct reloc_control *rc,
 
 	BUG_ON(block->key_ready);
 	eb = read_tree_block(rc->extent_root, block->bytenr,
-			     block->key.objectid, block->key.offset);
+			     block->key.offset);
 	if (!eb || !extent_buffer_uptodate(eb)) {
 		free_extent_buffer(eb);
 		return -EIO;
@@ -2858,20 +2886,6 @@ static int get_tree_block_key(struct reloc_control *rc,
 	return 0;
 }
 
-static int reada_tree_block(struct reloc_control *rc,
-			    struct tree_block *block)
-{
-	BUG_ON(block->key_ready);
-	if (block->key.type == BTRFS_METADATA_ITEM_KEY)
-		readahead_tree_block(rc->extent_root, block->bytenr,
-				     block->key.objectid,
-				     rc->extent_root->leafsize);
-	else
-		readahead_tree_block(rc->extent_root, block->bytenr,
-				     block->key.objectid, block->key.offset);
-	return 0;
-}
-
 /*
  * helper function to relocate a tree block
  */
@@ -2951,7 +2965,8 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans,
 	while (rb_node) {
 		block = rb_entry(rb_node, struct tree_block, rb_node);
 		if (!block->key_ready)
-			reada_tree_block(rc, block);
+			readahead_tree_block(rc->extent_root, block->bytenr,
+					block->key.objectid);
 		rb_node = rb_next(rb_node);
 	}
 
@@ -3313,7 +3328,7 @@ static int add_tree_block(struct reloc_control *rc,
 		return -ENOMEM;
 
 	block->bytenr = extent_key->objectid;
-	block->key.objectid = rc->extent_root->leafsize;
+	block->key.objectid = rc->extent_root->nodesize;
 	block->key.offset = generation;
 	block->level = level;
 	block->key_ready = 0;
@@ -3640,7 +3655,7 @@ int add_data_references(struct reloc_control *rc,
 	struct btrfs_extent_inline_ref *iref;
 	unsigned long ptr;
 	unsigned long end;
-	u32 blocksize = btrfs_level_size(rc->extent_root, 0);
+	u32 blocksize = rc->extent_root->nodesize;
 	int ret = 0;
 	int err = 0;
 
@@ -3783,7 +3798,7 @@ next:
 		}
 
 		if (key.type == BTRFS_METADATA_ITEM_KEY &&
-		    key.objectid + rc->extent_root->leafsize <=
+		    key.objectid + rc->extent_root->nodesize <=
 		    rc->search_start) {
 			path->slots[0]++;
 			goto next;
@@ -3801,7 +3816,7 @@ next:
 				rc->search_start = key.objectid + key.offset;
 			else
 				rc->search_start = key.objectid +
-					rc->extent_root->leafsize;
+					rc->extent_root->nodesize;
 			memcpy(extent_key, &key, sizeof(key));
 			return 0;
 		}
@@ -4096,7 +4111,6 @@ static int __insert_orphan_inode(struct btrfs_trans_handle *trans,
 	btrfs_set_inode_flags(leaf, item, BTRFS_INODE_NOCOMPRESS |
 					  BTRFS_INODE_PREALLOC);
 	btrfs_mark_buffer_dirty(leaf);
-	btrfs_release_path(path);
 out:
 	btrfs_free_path(path);
 	return ret;

fs/btrfs/scrub.c

@@ -137,7 +137,6 @@ struct scrub_ctx {
 	int			pages_per_rd_bio;
 	u32			sectorsize;
 	u32			nodesize;
-	u32			leafsize;
 
 	int			is_dev_replace;
 	struct scrub_wr_ctx	wr_ctx;
@@ -178,17 +177,12 @@ struct scrub_copy_nocow_ctx {
 struct scrub_warning {
 	struct btrfs_path	*path;
 	u64			extent_item_size;
-	char			*scratch_buf;
-	char			*msg_buf;
 	const char		*errstr;
 	sector_t		sector;
 	u64			logical;
 	struct btrfs_device	*dev;
-	int			msg_bufsize;
-	int			scratch_bufsize;
 };
 
-
 static void scrub_pending_bio_inc(struct scrub_ctx *sctx);
 static void scrub_pending_bio_dec(struct scrub_ctx *sctx);
 static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx);
@@ -438,7 +432,6 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace)
 	}
 	sctx->first_free = 0;
 	sctx->nodesize = dev->dev_root->nodesize;
-	sctx->leafsize = dev->dev_root->leafsize;
 	sctx->sectorsize = dev->dev_root->sectorsize;
 	atomic_set(&sctx->bios_in_flight, 0);
 	atomic_set(&sctx->workers_pending, 0);
@@ -553,7 +546,6 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock)
 	u64 ref_root;
 	u32 item_size;
 	u8 ref_level;
-	const int bufsize = 4096;
 	int ret;
 
 	WARN_ON(sblock->page_count < 1);
@@ -561,18 +553,13 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock)
 	fs_info = sblock->sctx->dev_root->fs_info;
 
 	path = btrfs_alloc_path();
+	if (!path)
+		return;
 
-	swarn.scratch_buf = kmalloc(bufsize, GFP_NOFS);
-	swarn.msg_buf = kmalloc(bufsize, GFP_NOFS);
 	swarn.sector = (sblock->pagev[0]->physical) >> 9;
 	swarn.logical = sblock->pagev[0]->logical;
 	swarn.errstr = errstr;
 	swarn.dev = NULL;
-	swarn.msg_bufsize = bufsize;
-	swarn.scratch_bufsize = bufsize;
-
-	if (!path || !swarn.scratch_buf || !swarn.msg_buf)
-		goto out;
 
 	ret = extent_from_logical(fs_info, swarn.logical, path, &found_key,
 				  &flags);
@@ -613,8 +600,6 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock)
 
 out:
 	btrfs_free_path(path);
-	kfree(swarn.scratch_buf);
-	kfree(swarn.msg_buf);
 }
 
 static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx)
@@ -681,9 +666,9 @@ static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx)
 			ret = -EIO;
 			goto out;
 		}
-		fs_info = BTRFS_I(inode)->root->fs_info;
-		ret = repair_io_failure(fs_info, offset, PAGE_SIZE,
+		ret = repair_io_failure(inode, offset, PAGE_SIZE,
 					fixup->logical, page,
+					offset - page_offset(page),
 					fixup->mirror_num);
 		unlock_page(page);
 		corrected = !ret;
@@ -1361,6 +1346,16 @@ static void scrub_recheck_block(struct btrfs_fs_info *fs_info,
 	return;
 }
 
+static inline int scrub_check_fsid(u8 fsid[],
+				   struct scrub_page *spage)
+{
+	struct btrfs_fs_devices *fs_devices = spage->dev->fs_devices;
+	int ret;
+
+	ret = memcmp(fsid, fs_devices->fsid, BTRFS_UUID_SIZE);
+	return !ret;
+}
+
 static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
 					 struct scrub_block *sblock,
 					 int is_metadata, int have_csum,
@@ -1380,7 +1375,7 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
 		h = (struct btrfs_header *)mapped_buffer;
 
 		if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h) ||
-		    memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE) ||
+		    !scrub_check_fsid(h->fsid, sblock->pagev[0]) ||
 		    memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
 			   BTRFS_UUID_SIZE)) {
 			sblock->header_error = 1;
@@ -1751,14 +1746,13 @@ static int scrub_checksum_tree_block(struct scrub_block *sblock)
 	if (sblock->pagev[0]->generation != btrfs_stack_header_generation(h))
 		++fail;
 
-	if (memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
+	if (!scrub_check_fsid(h->fsid, sblock->pagev[0]))
 		++fail;
 
 	if (memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
 		   BTRFS_UUID_SIZE))
 		++fail;
 
-	WARN_ON(sctx->nodesize != sctx->leafsize);
 	len = sctx->nodesize - BTRFS_CSUM_SIZE;
 	mapped_size = PAGE_SIZE - BTRFS_CSUM_SIZE;
 	p = ((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE;
@@ -1791,8 +1785,6 @@ static int scrub_checksum_super(struct scrub_block *sblock)
 {
 	struct btrfs_super_block *s;
 	struct scrub_ctx *sctx = sblock->sctx;
-	struct btrfs_root *root = sctx->dev_root;
-	struct btrfs_fs_info *fs_info = root->fs_info;
 	u8 calculated_csum[BTRFS_CSUM_SIZE];
 	u8 on_disk_csum[BTRFS_CSUM_SIZE];
 	struct page *page;
@@ -1817,7 +1809,7 @@ static int scrub_checksum_super(struct scrub_block *sblock)
 	if (sblock->pagev[0]->generation != btrfs_super_generation(s))
 		++fail_gen;
 
-	if (memcmp(s->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
+	if (!scrub_check_fsid(s->fsid, sblock->pagev[0]))
 		++fail_cor;
 
 	len = BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE;
@@ -2196,7 +2188,6 @@ static int scrub_extent(struct scrub_ctx *sctx, u64 logical, u64 len,
 		sctx->stat.data_bytes_scrubbed += len;
 		spin_unlock(&sctx->stat_lock);
 	} else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
-		WARN_ON(sctx->nodesize != sctx->leafsize);
 		blocksize = sctx->nodesize;
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.tree_extents_scrubbed++;
@@ -2487,7 +2478,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 			btrfs_item_key_to_cpu(l, &key, slot);
 
 			if (key.type == BTRFS_METADATA_ITEM_KEY)
-				bytes = root->leafsize;
+				bytes = root->nodesize;
 			else
 				bytes = key.offset;
 
@@ -2714,7 +2705,7 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		if (found_key.objectid != scrub_dev->devid)
 			break;
 
-		if (btrfs_key_type(&found_key) != BTRFS_DEV_EXTENT_KEY)
+		if (found_key.type != BTRFS_DEV_EXTENT_KEY)
 			break;
 
 		if (found_key.offset >= end)
@@ -2828,11 +2819,16 @@ static noinline_for_stack int scrub_supers(struct scrub_ctx *sctx,
 	if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))
 		return -EIO;
 
-	gen = root->fs_info->last_trans_committed;
+	/* Seed devices of a new filesystem has their own generation. */
+	if (scrub_dev->fs_devices != root->fs_info->fs_devices)
+		gen = scrub_dev->generation;
+	else
+		gen = root->fs_info->last_trans_committed;
 
 	for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
 		bytenr = btrfs_sb_offset(i);
-		if (bytenr + BTRFS_SUPER_INFO_SIZE > scrub_dev->total_bytes)
+		if (bytenr + BTRFS_SUPER_INFO_SIZE >
+		    scrub_dev->commit_total_bytes)
 			break;
 
 		ret = scrub_pages(sctx, bytenr, BTRFS_SUPER_INFO_SIZE, bytenr,
@@ -2910,17 +2906,6 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	if (btrfs_fs_closing(fs_info))
 		return -EINVAL;
 
-	/*
-	 * check some assumptions
-	 */
-	if (fs_info->chunk_root->nodesize != fs_info->chunk_root->leafsize) {
-		btrfs_err(fs_info,
-			   "scrub: size assumption nodesize == leafsize (%d == %d) fails",
-		       fs_info->chunk_root->nodesize,
-		       fs_info->chunk_root->leafsize);
-		return -EINVAL;
-	}
-
 	if (fs_info->chunk_root->nodesize > BTRFS_STRIPE_LEN) {
 		/*
 		 * in this case scrub is unable to calculate the checksum

fs/btrfs/send.c

@@ -515,7 +515,8 @@ static int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off)
 	set_fs(KERNEL_DS);
 
 	while (pos < len) {
-		ret = vfs_write(filp, (char *)buf + pos, len - pos, off);
+		ret = vfs_write(filp, (__force const char __user *)buf + pos,
+				len - pos, off);
 		/* TODO handle that correctly */
 		/*if (ret == -ERESTARTSYS) {
 			continue;
@@ -985,11 +986,13 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 	int num;
 	u8 type;
 
-	if (found_key->type == BTRFS_XATTR_ITEM_KEY)
-		buf_len = BTRFS_MAX_XATTR_SIZE(root);
-	else
-		buf_len = PATH_MAX;
-
+	/*
+	 * Start with a small buffer (1 page). If later we end up needing more
+	 * space, which can happen for xattrs on a fs with a leaf size greater
+	 * then the page size, attempt to increase the buffer. Typically xattr
+	 * values are small.
+	 */
+	buf_len = PATH_MAX;
 	buf = kmalloc(buf_len, GFP_NOFS);
 	if (!buf) {
 		ret = -ENOMEM;
@@ -1016,7 +1019,7 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 				ret = -ENAMETOOLONG;
 				goto out;
 			}
-			if (name_len + data_len > buf_len) {
+			if (name_len + data_len > BTRFS_MAX_XATTR_SIZE(root)) {
 				ret = -E2BIG;
 				goto out;
 			}
@@ -1024,12 +1027,34 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 			/*
 			 * Path too long
 			 */
-			if (name_len + data_len > buf_len) {
+			if (name_len + data_len > PATH_MAX) {
 				ret = -ENAMETOOLONG;
 				goto out;
 			}
 		}
 
+		if (name_len + data_len > buf_len) {
+			buf_len = name_len + data_len;
+			if (is_vmalloc_addr(buf)) {
+				vfree(buf);
+				buf = NULL;
+			} else {
+				char *tmp = krealloc(buf, buf_len,
+						     GFP_NOFS | __GFP_NOWARN);
+
+				if (!tmp)
+					kfree(buf);
+				buf = tmp;
+			}
+			if (!buf) {
+				buf = vmalloc(buf_len);
+				if (!buf) {
+					ret = -ENOMEM;
+					goto out;
+				}
+			}
+		}
+
 		read_extent_buffer(eb, buf, (unsigned long)(di + 1),
 				name_len + data_len);
 
@@ -1050,7 +1075,7 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 	}
 
 out:
-	kfree(buf);
+	kvfree(buf);
 	return ret;
 }
 
@@ -3302,7 +3327,7 @@ static int wait_for_parent_move(struct send_ctx *sctx,
 		if (ret < 0 && ret != -ENOENT) {
 			goto out;
 		} else if (ret == -ENOENT) {
-			ret = 1;
+			ret = 0;
 			break;
 		}
 
@@ -5703,7 +5728,7 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 			NULL);
 	sort_clone_roots = 1;
 
-	current->journal_info = (void *)BTRFS_SEND_TRANS_STUB;
+	current->journal_info = BTRFS_SEND_TRANS_STUB;
 	ret = send_subvol(sctx);
 	current->journal_info = NULL;
 	if (ret < 0)

fs/btrfs/super.c

@@ -60,6 +60,7 @@
 #include "backref.h"
 #include "tests/btrfs-tests.h"
 
+#include "qgroup.h"
 #define CREATE_TRACE_POINTS
 #include <trace/events/btrfs.h>
 
@@ -307,13 +308,7 @@ void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
 
 static void btrfs_put_super(struct super_block *sb)
 {
-	(void)close_ctree(btrfs_sb(sb)->tree_root);
-	/* FIXME: need to fix VFS to return error? */
-	/* AV: return it _where_?  ->put_super() can be triggered by any number
-	 * of async events, up to and including delivery of SIGKILL to the
-	 * last process that kept it busy.  Or segfault in the aforementioned
-	 * process...  Whom would you report that to?
-	 */
+	close_ctree(btrfs_sb(sb)->tree_root);
 }
 
 enum {
@@ -400,7 +395,6 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 	int ret = 0;
 	char *compress_type;
 	bool compress_force = false;
-	bool compress = false;
 
 	cache_gen = btrfs_super_cache_generation(root->fs_info->super_copy);
 	if (cache_gen)
@@ -478,7 +472,6 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			/* Fallthrough */
 		case Opt_compress:
 		case Opt_compress_type:
-			compress = true;
 			if (token == Opt_compress ||
 			    token == Opt_compress_force ||
 			    strcmp(args[0].from, "zlib") == 0) {
@@ -508,11 +501,18 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				btrfs_set_and_info(root, FORCE_COMPRESS,
 						   "force %s compression",
 						   compress_type);
-			} else if (compress) {
+			} else {
 				if (!btrfs_test_opt(root, COMPRESS))
 					btrfs_info(root->fs_info,
 						   "btrfs: use %s compression",
 						   compress_type);
+				/*
+				 * If we remount from compress-force=xxx to
+				 * compress=xxx, we need clear FORCE_COMPRESS
+				 * flag, otherwise, there is no way for users
+				 * to disable forcible compression separately.
+				 */
+				btrfs_clear_opt(info->mount_opt, FORCE_COMPRESS);
 			}
 			break;
 		case Opt_ssd:
@@ -1014,7 +1014,7 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",nodatacow");
 	if (btrfs_test_opt(root, NOBARRIER))
 		seq_puts(seq, ",nobarrier");
-	if (info->max_inline != 8192 * 1024)
+	if (info->max_inline != BTRFS_DEFAULT_MAX_INLINE)
 		seq_printf(seq, ",max_inline=%llu", info->max_inline);
 	if (info->alloc_start != 0)
 		seq_printf(seq, ",alloc_start=%llu", info->alloc_start);
@@ -1215,6 +1215,56 @@ static struct dentry *mount_subvol(const char *subvol_name, int flags,
 	return root;
 }
 
+static int parse_security_options(char *orig_opts,
+				  struct security_mnt_opts *sec_opts)
+{
+	char *secdata = NULL;
+	int ret = 0;
+
+	secdata = alloc_secdata();
+	if (!secdata)
+		return -ENOMEM;
+	ret = security_sb_copy_data(orig_opts, secdata);
+	if (ret) {
+		free_secdata(secdata);
+		return ret;
+	}
+	ret = security_sb_parse_opts_str(secdata, sec_opts);
+	free_secdata(secdata);
+	return ret;
+}
+
+static int setup_security_options(struct btrfs_fs_info *fs_info,
+				  struct super_block *sb,
+				  struct security_mnt_opts *sec_opts)
+{
+	int ret = 0;
+
+	/*
+	 * Call security_sb_set_mnt_opts() to check whether new sec_opts
+	 * is valid.
+	 */
+	ret = security_sb_set_mnt_opts(sb, sec_opts, 0, NULL);
+	if (ret)
+		return ret;
+
+#ifdef CONFIG_SECURITY
+	if (!fs_info->security_opts.num_mnt_opts) {
+		/* first time security setup, copy sec_opts to fs_info */
+		memcpy(&fs_info->security_opts, sec_opts, sizeof(*sec_opts));
+	} else {
+		/*
+		 * Since SELinux(the only one supports security_mnt_opts) does
+		 * NOT support changing context during remount/mount same sb,
+		 * This must be the same or part of the same security options,
+		 * just free it.
+		 */
+		security_free_mnt_opts(sec_opts);
+	}
+#endif
+	return ret;
+}
+
 /*
  * Find a superblock for the given device / mount point.
  *
@@ -1229,6 +1279,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 	struct dentry *root;
 	struct btrfs_fs_devices *fs_devices = NULL;
 	struct btrfs_fs_info *fs_info = NULL;
+	struct security_mnt_opts new_sec_opts;
 	fmode_t mode = FMODE_READ;
 	char *subvol_name = NULL;
 	u64 subvol_objectid = 0;
@@ -1251,9 +1302,16 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 		return root;
 	}
 
+	security_init_mnt_opts(&new_sec_opts);
+	if (data) {
+		error = parse_security_options(data, &new_sec_opts);
+		if (error)
+			return ERR_PTR(error);
+	}
+
 	error = btrfs_scan_one_device(device_name, mode, fs_type, &fs_devices);
 	if (error)
-		return ERR_PTR(error);
+		goto error_sec_opts;
 
 	/*
 	 * Setup a dummy root and fs_info for test/set super.  This is because
@@ -1262,13 +1320,16 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 	 * then open_ctree will properly initialize everything later.
 	 */
 	fs_info = kzalloc(sizeof(struct btrfs_fs_info), GFP_NOFS);
-	if (!fs_info)
-		return ERR_PTR(-ENOMEM);
+	if (!fs_info) {
+		error = -ENOMEM;
+		goto error_sec_opts;
+	}
 
 	fs_info->fs_devices = fs_devices;
 
 	fs_info->super_copy = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS);
 	fs_info->super_for_commit = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS);
+	security_init_mnt_opts(&fs_info->security_opts);
 	if (!fs_info->super_copy || !fs_info->super_for_commit) {
 		error = -ENOMEM;
 		goto error_fs_info;
@@ -1306,8 +1367,19 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 	}
 
 	root = !error ? get_default_root(s, subvol_objectid) : ERR_PTR(error);
-	if (IS_ERR(root))
+	if (IS_ERR(root)) {
+		deactivate_locked_super(s);
+		error = PTR_ERR(root);
+		goto error_sec_opts;
+	}
+
+	fs_info = btrfs_sb(s);
+	error = setup_security_options(fs_info, s, &new_sec_opts);
+	if (error) {
+		dput(root);
 		deactivate_locked_super(s);
+		goto error_sec_opts;
+	}
 
 	return root;
 
@@ -1315,6 +1387,8 @@ error_close_devices:
 	btrfs_close_devices(fs_devices);
 error_fs_info:
 	free_fs_info(fs_info);
+error_sec_opts:
+	security_free_mnt_opts(&new_sec_opts);
 	return ERR_PTR(error);
 }
 
@@ -1396,6 +1470,21 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 	sync_filesystem(sb);
 	btrfs_remount_prepare(fs_info);
 
+	if (data) {
+		struct security_mnt_opts new_sec_opts;
+
+		security_init_mnt_opts(&new_sec_opts);
+		ret = parse_security_options(data, &new_sec_opts);
+		if (ret)
+			goto restore;
+		ret = setup_security_options(fs_info, sb,
+					     &new_sec_opts);
+		if (ret) {
+			security_free_mnt_opts(&new_sec_opts);
+			goto restore;
+		}
+	}
+
 	ret = btrfs_parse_options(root, data);
 	if (ret) {
 		ret = -EINVAL;
@@ -1694,7 +1783,11 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
 	int ret;
 
-	/* holding chunk_muext to avoid allocating new chunks */
+	/*
+	 * holding chunk_muext to avoid allocating new chunks, holding
+	 * device_list_mutex to avoid the device being removed
+	 */
+	mutex_lock(&fs_info->fs_devices->device_list_mutex);
 	mutex_lock(&fs_info->chunk_mutex);
 	rcu_read_lock();
 	list_for_each_entry_rcu(found, head, list) {
@@ -1735,11 +1828,13 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	ret = btrfs_calc_avail_data_space(fs_info->tree_root, &total_free_data);
 	if (ret) {
 		mutex_unlock(&fs_info->chunk_mutex);
+		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 		return ret;
 	}
 	buf->f_bavail += div_u64(total_free_data, factor);
 	buf->f_bavail = buf->f_bavail >> bits;
 	mutex_unlock(&fs_info->chunk_mutex);
+	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 
 	buf->f_type = BTRFS_SUPER_MAGIC;
 	buf->f_bsize = dentry->d_sb->s_blocksize;
@@ -1769,7 +1864,7 @@ static struct file_system_type btrfs_fs_type = {
 	.name		= "btrfs",
 	.mount		= btrfs_mount,
 	.kill_sb	= btrfs_kill_super,
-	.fs_flags	= FS_REQUIRES_DEV,
+	.fs_flags	= FS_REQUIRES_DEV | FS_BINARY_MOUNTDATA,
 };
 MODULE_ALIAS_FS("btrfs");
 
@@ -1992,12 +2087,16 @@ static int __init init_btrfs_fs(void)
 		goto free_auto_defrag;
 
 	err = btrfs_prelim_ref_init();
+	if (err)
+		goto free_delayed_ref;
+
+	err = btrfs_end_io_wq_init();
 	if (err)
 		goto free_prelim_ref;
 
 	err = btrfs_interface_init();
 	if (err)
-		goto free_delayed_ref;
+		goto free_end_io_wq;
 
 	btrfs_init_lockdep();
 
@@ -2015,6 +2114,8 @@ static int __init init_btrfs_fs(void)
 
 unregister_ioctl:
 	btrfs_interface_exit();
+free_end_io_wq:
+	btrfs_end_io_wq_exit();
 free_prelim_ref:
 	btrfs_prelim_ref_exit();
 free_delayed_ref:

fs/btrfs/sysfs.c

@@ -242,7 +242,7 @@ static ssize_t global_rsv_size_show(struct kobject *kobj,
 	struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
 	return btrfs_show_u64(&block_rsv->size, &block_rsv->lock, buf);
 }
-BTRFS_ATTR(global_rsv_size, 0444, global_rsv_size_show);
+BTRFS_ATTR(global_rsv_size, global_rsv_size_show);
 
 static ssize_t global_rsv_reserved_show(struct kobject *kobj,
 					struct kobj_attribute *a, char *buf)
@@ -251,7 +251,7 @@ static ssize_t global_rsv_reserved_show(struct kobject *kobj,
 	struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
 	return btrfs_show_u64(&block_rsv->reserved, &block_rsv->lock, buf);
 }
-BTRFS_ATTR(global_rsv_reserved, 0444, global_rsv_reserved_show);
+BTRFS_ATTR(global_rsv_reserved, global_rsv_reserved_show);
 
 #define to_space_info(_kobj) container_of(_kobj, struct btrfs_space_info, kobj)
 #define to_raid_kobj(_kobj) container_of(_kobj, struct raid_kobject, kobj)
@@ -306,7 +306,7 @@ static ssize_t btrfs_space_info_show_##field(struct kobject *kobj,	\
 	struct btrfs_space_info *sinfo = to_space_info(kobj);		\
 	return btrfs_show_u64(&sinfo->field, &sinfo->lock, buf);	\
 }									\
-BTRFS_ATTR(field, 0444, btrfs_space_info_show_##field)
+BTRFS_ATTR(field, btrfs_space_info_show_##field)
 
 static ssize_t btrfs_space_info_show_total_bytes_pinned(struct kobject *kobj,
 						       struct kobj_attribute *a,
@@ -325,7 +325,7 @@ SPACE_INFO_ATTR(bytes_reserved);
 SPACE_INFO_ATTR(bytes_may_use);
 SPACE_INFO_ATTR(disk_used);
 SPACE_INFO_ATTR(disk_total);
-BTRFS_ATTR(total_bytes_pinned, 0444, btrfs_space_info_show_total_bytes_pinned);
+BTRFS_ATTR(total_bytes_pinned, btrfs_space_info_show_total_bytes_pinned);
 
 static struct attribute *space_info_attrs[] = {
 	BTRFS_ATTR_PTR(flags),
@@ -363,7 +363,8 @@ static ssize_t btrfs_label_show(struct kobject *kobj,
 				struct kobj_attribute *a, char *buf)
 {
 	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
-	return snprintf(buf, PAGE_SIZE, "%s\n", fs_info->super_copy->label);
+	char *label = fs_info->super_copy->label;
+	return snprintf(buf, PAGE_SIZE, label[0] ? "%s\n" : "%s", label);
 }
 
 static ssize_t btrfs_label_store(struct kobject *kobj,
@@ -374,8 +375,18 @@ static ssize_t btrfs_label_store(struct kobject *kobj,
 	struct btrfs_trans_handle *trans;
 	struct btrfs_root *root = fs_info->fs_root;
 	int ret;
+	size_t p_len;
 
-	if (len >= BTRFS_LABEL_SIZE)
+	if (fs_info->sb->s_flags & MS_RDONLY)
+		return -EROFS;
+
+	/*
+	 * p_len is the len until the first occurrence of either
+	 * '\n' or '\0'
+	 */
+	p_len = strcspn(buf, "\n");
+
+	if (p_len >= BTRFS_LABEL_SIZE)
 		return -EINVAL;
 
 	trans = btrfs_start_transaction(root, 0);
@@ -383,7 +394,8 @@ static ssize_t btrfs_label_store(struct kobject *kobj,
 		return PTR_ERR(trans);
 
 	spin_lock(&root->fs_info->super_lock);
-	strcpy(fs_info->super_copy->label, buf);
+	memset(fs_info->super_copy->label, 0, BTRFS_LABEL_SIZE);
+	memcpy(fs_info->super_copy->label, buf, p_len);
 	spin_unlock(&root->fs_info->super_lock);
 	ret = btrfs_commit_transaction(trans, root);
 
@@ -392,14 +404,7 @@ static ssize_t btrfs_label_store(struct kobject *kobj,
 
 	return ret;
 }
-BTRFS_ATTR_RW(label, 0644, btrfs_label_show, btrfs_label_store);
-
-static ssize_t btrfs_no_store(struct kobject *kobj,
-				 struct kobj_attribute *a,
-				 const char *buf, size_t len)
-{
-	return -EPERM;
-}
+BTRFS_ATTR_RW(label, btrfs_label_show, btrfs_label_store);
 
 static ssize_t btrfs_nodesize_show(struct kobject *kobj,
 				struct kobj_attribute *a, char *buf)
@@ -409,7 +414,7 @@ static ssize_t btrfs_nodesize_show(struct kobject *kobj,
 	return snprintf(buf, PAGE_SIZE, "%u\n", fs_info->super_copy->nodesize);
 }
 
-BTRFS_ATTR_RW(nodesize, 0444, btrfs_nodesize_show, btrfs_no_store);
+BTRFS_ATTR(nodesize, btrfs_nodesize_show);
 
 static ssize_t btrfs_sectorsize_show(struct kobject *kobj,
 				struct kobj_attribute *a, char *buf)
@@ -419,7 +424,7 @@ static ssize_t btrfs_sectorsize_show(struct kobject *kobj,
 	return snprintf(buf, PAGE_SIZE, "%u\n", fs_info->super_copy->sectorsize);
 }
 
-BTRFS_ATTR_RW(sectorsize, 0444, btrfs_sectorsize_show, btrfs_no_store);
+BTRFS_ATTR(sectorsize, btrfs_sectorsize_show);
 
 static ssize_t btrfs_clone_alignment_show(struct kobject *kobj,
 				struct kobj_attribute *a, char *buf)
@@ -429,7 +434,7 @@ static ssize_t btrfs_clone_alignment_show(struct kobject *kobj,
 	return snprintf(buf, PAGE_SIZE, "%u\n", fs_info->super_copy->sectorsize);
 }
 
-BTRFS_ATTR_RW(clone_alignment, 0444, btrfs_clone_alignment_show, btrfs_no_store);
+BTRFS_ATTR(clone_alignment, btrfs_clone_alignment_show);
 
 static struct attribute *btrfs_attrs[] = {
 	BTRFS_ATTR_PTR(label),

fs/btrfs/sysfs.h

@@ -20,16 +20,20 @@ enum btrfs_feature_set {
 	.store	= _store,						\
 }
 
-#define BTRFS_ATTR_RW(_name, _mode, _show, _store)			\
-static struct kobj_attribute btrfs_attr_##_name =			\
-			__INIT_KOBJ_ATTR(_name, _mode, _show, _store)
-#define BTRFS_ATTR(_name, _mode, _show)					\
-	BTRFS_ATTR_RW(_name, _mode, _show, NULL)
+#define BTRFS_ATTR_RW(_name, _show, _store)			\
+	static struct kobj_attribute btrfs_attr_##_name =		\
+			__INIT_KOBJ_ATTR(_name, 0644, _show, _store)
+
+#define BTRFS_ATTR(_name, _show)					\
+	static struct kobj_attribute btrfs_attr_##_name =		\
+			__INIT_KOBJ_ATTR(_name, 0444, _show, NULL)
+
 #define BTRFS_ATTR_PTR(_name)    (&btrfs_attr_##_name.attr)
 
 #define BTRFS_RAID_ATTR(_name, _show)					\
-static struct kobj_attribute btrfs_raid_attr_##_name =			\
+	static struct kobj_attribute btrfs_raid_attr_##_name =		\
 			__INIT_KOBJ_ATTR(_name, 0444, _show, NULL)
+
 #define BTRFS_RAID_ATTR_PTR(_name)    (&btrfs_raid_attr_##_name.attr)
 
 

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

@@ -40,11 +40,12 @@ static struct btrfs_block_group_cache *init_test_block_group(void)
 	cache->key.offset = 1024 * 1024 * 1024;
 	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
 	cache->sectorsize = 4096;
+	cache->full_stripe_len = 4096;
 
 	spin_lock_init(&cache->lock);
 	INIT_LIST_HEAD(&cache->list);
 	INIT_LIST_HEAD(&cache->cluster_list);
-	INIT_LIST_HEAD(&cache->new_bg_list);
+	INIT_LIST_HEAD(&cache->bg_list);
 
 	btrfs_init_free_space_ctl(cache);
 
@@ -364,6 +365,517 @@ static int test_bitmaps_and_extents(struct btrfs_block_group_cache *cache)
 	return 0;
 }
 
+/* Used by test_steal_space_from_bitmap_to_extent(). */
+static bool test_use_bitmap(struct btrfs_free_space_ctl *ctl,
+			    struct btrfs_free_space *info)
+{
+	return ctl->free_extents > 0;
+}
+
+/* Used by test_steal_space_from_bitmap_to_extent(). */
+static int
+check_num_extents_and_bitmaps(const struct btrfs_block_group_cache *cache,
+			      const int num_extents,
+			      const int num_bitmaps)
+{
+	if (cache->free_space_ctl->free_extents != num_extents) {
+		test_msg("Incorrect # of extent entries in the cache: %d, expected %d\n",
+			 cache->free_space_ctl->free_extents, num_extents);
+		return -EINVAL;
+	}
+	if (cache->free_space_ctl->total_bitmaps != num_bitmaps) {
+		test_msg("Incorrect # of extent entries in the cache: %d, expected %d\n",
+			 cache->free_space_ctl->total_bitmaps, num_bitmaps);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+/* Used by test_steal_space_from_bitmap_to_extent(). */
+static int check_cache_empty(struct btrfs_block_group_cache *cache)
+{
+	u64 offset;
+	u64 max_extent_size;
+
+	/*
+	 * Now lets confirm that there's absolutely no free space left to
+	 * allocate.
+	 */
+	if (cache->free_space_ctl->free_space != 0) {
+		test_msg("Cache free space is not 0\n");
+		return -EINVAL;
+	}
+
+	/* And any allocation request, no matter how small, should fail now. */
+	offset = btrfs_find_space_for_alloc(cache, 0, 4096, 0,
+					    &max_extent_size);
+	if (offset != 0) {
+		test_msg("Space allocation did not fail, returned offset: %llu",
+			 offset);
+		return -EINVAL;
+	}
+
+	/* And no extent nor bitmap entries in the cache anymore. */
+	return check_num_extents_and_bitmaps(cache, 0, 0);
+}
+
+/*
+ * Before we were able to steal free space from a bitmap entry to an extent
+ * entry, we could end up with 2 entries representing a contiguous free space.
+ * One would be an extent entry and the other a bitmap entry. Since in order
+ * to allocate space to a caller we use only 1 entry, we couldn't return that
+ * whole range to the caller if it was requested. This forced the caller to
+ * either assume ENOSPC or perform several smaller space allocations, which
+ * wasn't optimal as they could be spread all over the block group while under
+ * concurrency (extra overhead and fragmentation).
+ *
+ * This stealing approach is benefical, since we always prefer to allocate from
+ * extent entries, both for clustered and non-clustered allocation requests.
+ */
+static int
+test_steal_space_from_bitmap_to_extent(struct btrfs_block_group_cache *cache)
+{
+	int ret;
+	u64 offset;
+	u64 max_extent_size;
+
+	bool (*use_bitmap_op)(struct btrfs_free_space_ctl *,
+			      struct btrfs_free_space *);
+
+	test_msg("Running space stealing from bitmap to extent\n");
+
+	/*
+	 * For this test, we want to ensure we end up with an extent entry
+	 * immediately adjacent to a bitmap entry, where the bitmap starts
+	 * at an offset where the extent entry ends. We keep adding and
+	 * removing free space to reach into this state, but to get there
+	 * we need to reach a point where marking new free space doesn't
+	 * result in adding new extent entries or merging the new space
+	 * with existing extent entries - the space ends up being marked
+	 * in an existing bitmap that covers the new free space range.
+	 *
+	 * To get there, we need to reach the threshold defined set at
+	 * cache->free_space_ctl->extents_thresh, which currently is
+	 * 256 extents on a x86_64 system at least, and a few other
+	 * conditions (check free_space_cache.c). Instead of making the
+	 * test much longer and complicated, use a "use_bitmap" operation
+	 * that forces use of bitmaps as soon as we have at least 1
+	 * extent entry.
+	 */
+	use_bitmap_op = cache->free_space_ctl->op->use_bitmap;
+	cache->free_space_ctl->op->use_bitmap = test_use_bitmap;
+
+	/*
+	 * Extent entry covering free space range [128Mb - 256Kb, 128Mb - 128Kb[
+	 */
+	ret = test_add_free_space_entry(cache, 128 * 1024 * 1024 - 256 * 1024,
+					128 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	/* Bitmap entry covering free space range [128Mb + 512Kb, 256Mb[ */
+	ret = test_add_free_space_entry(cache, 128 * 1024 * 1024 + 512 * 1024,
+					128 * 1024 * 1024 - 512 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * Now make only the first 256Kb of the bitmap marked as free, so that
+	 * we end up with only the following ranges marked as free space:
+	 *
+	 * [128Mb - 256Kb, 128Mb - 128Kb[
+	 * [128Mb + 512Kb, 128Mb + 768Kb[
+	 */
+	ret = btrfs_remove_free_space(cache,
+				      128 * 1024 * 1024 + 768 * 1024,
+				      128 * 1024 * 1024 - 768 * 1024);
+	if (ret) {
+		test_msg("Failed to free part of bitmap space %d\n", ret);
+		return ret;
+	}
+
+	/* Confirm that only those 2 ranges are marked as free. */
+	if (!test_check_exists(cache, 128 * 1024 * 1024 - 256 * 1024,
+			       128 * 1024)) {
+		test_msg("Free space range missing\n");
+		return -ENOENT;
+	}
+	if (!test_check_exists(cache, 128 * 1024 * 1024 + 512 * 1024,
+			       256 * 1024)) {
+		test_msg("Free space range missing\n");
+		return -ENOENT;
+	}
+
+	/*
+	 * Confirm that the bitmap range [128Mb + 768Kb, 256Mb[ isn't marked
+	 * as free anymore.
+	 */
+	if (test_check_exists(cache, 128 * 1024 * 1024 + 768 * 1024,
+			      128 * 1024 * 1024 - 768 * 1024)) {
+		test_msg("Bitmap region not removed from space cache\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Confirm that the region [128Mb + 256Kb, 128Mb + 512Kb[, which is
+	 * covered by the bitmap, isn't marked as free.
+	 */
+	if (test_check_exists(cache, 128 * 1024 * 1024 + 256 * 1024,
+			      256 * 1024)) {
+		test_msg("Invalid bitmap region marked as free\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Confirm that the region [128Mb, 128Mb + 256Kb[, which is covered
+	 * by the bitmap too, isn't marked as free either.
+	 */
+	if (test_check_exists(cache, 128 * 1024 * 1024,
+			      256 * 1024)) {
+		test_msg("Invalid bitmap region marked as free\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Now lets mark the region [128Mb, 128Mb + 512Kb[ as free too. But,
+	 * lets make sure the free space cache marks it as free in the bitmap,
+	 * and doesn't insert a new extent entry to represent this region.
+	 */
+	ret = btrfs_add_free_space(cache, 128 * 1024 * 1024, 512 * 1024);
+	if (ret) {
+		test_msg("Error adding free space: %d\n", ret);
+		return ret;
+	}
+	/* Confirm the region is marked as free. */
+	if (!test_check_exists(cache, 128 * 1024 * 1024, 512 * 1024)) {
+		test_msg("Bitmap region not marked as free\n");
+		return -ENOENT;
+	}
+
+	/*
+	 * Confirm that no new extent entries or bitmap entries were added to
+	 * the cache after adding that free space region.
+	 */
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * Now lets add a small free space region to the right of the previous
+	 * one, which is not contiguous with it and is part of the bitmap too.
+	 * The goal is to test that the bitmap entry space stealing doesn't
+	 * steal this space region.
+	 */
+	ret = btrfs_add_free_space(cache, 128 * 1024 * 1024 + 16 * 1024 * 1024,
+				   4096);
+	if (ret) {
+		test_msg("Error adding free space: %d\n", ret);
+		return ret;
+	}
+
+	/*
+	 * Confirm that no new extent entries or bitmap entries were added to
+	 * the cache after adding that free space region.
+	 */
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * Now mark the region [128Mb - 128Kb, 128Mb[ as free too. This will
+	 * expand the range covered by the existing extent entry that represents
+	 * the free space [128Mb - 256Kb, 128Mb - 128Kb[.
+	 */
+	ret = btrfs_add_free_space(cache, 128 * 1024 * 1024 - 128 * 1024,
+				   128 * 1024);
+	if (ret) {
+		test_msg("Error adding free space: %d\n", ret);
+		return ret;
+	}
+	/* Confirm the region is marked as free. */
+	if (!test_check_exists(cache, 128 * 1024 * 1024 - 128 * 1024,
+			       128 * 1024)) {
+		test_msg("Extent region not marked as free\n");
+		return -ENOENT;
+	}
+
+	/*
+	 * Confirm that our extent entry didn't stole all free space from the
+	 * bitmap, because of the small 4Kb free space region.
+	 */
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * So now we have the range [128Mb - 256Kb, 128Mb + 768Kb[ as free
+	 * space. Without stealing bitmap free space into extent entry space,
+	 * we would have all this free space represented by 2 entries in the
+	 * cache:
+	 *
+	 * extent entry covering range: [128Mb - 256Kb, 128Mb[
+	 * bitmap entry covering range: [128Mb, 128Mb + 768Kb[
+	 *
+	 * Attempting to allocate the whole free space (1Mb) would fail, because
+	 * we can't allocate from multiple entries.
+	 * With the bitmap free space stealing, we get a single extent entry
+	 * that represents the 1Mb free space, and therefore we're able to
+	 * allocate the whole free space at once.
+	 */
+	if (!test_check_exists(cache, 128 * 1024 * 1024 - 256 * 1024,
+			       1 * 1024 * 1024)) {
+		test_msg("Expected region not marked as free\n");
+		return -ENOENT;
+	}
+
+	if (cache->free_space_ctl->free_space != (1 * 1024 * 1024 + 4096)) {
+		test_msg("Cache free space is not 1Mb + 4Kb\n");
+		return -EINVAL;
+	}
+
+	offset = btrfs_find_space_for_alloc(cache,
+					    0, 1 * 1024 * 1024, 0,
+					    &max_extent_size);
+	if (offset != (128 * 1024 * 1024 - 256 * 1024)) {
+		test_msg("Failed to allocate 1Mb from space cache, returned offset is: %llu\n",
+			 offset);
+		return -EINVAL;
+	}
+
+	/* All that remains is a 4Kb free space region in a bitmap. Confirm. */
+	ret = check_num_extents_and_bitmaps(cache, 1, 1);
+	if (ret)
+		return ret;
+
+	if (cache->free_space_ctl->free_space != 4096) {
+		test_msg("Cache free space is not 4Kb\n");
+		return -EINVAL;
+	}
+
+	offset = btrfs_find_space_for_alloc(cache,
+					    0, 4096, 0,
+					    &max_extent_size);
+	if (offset != (128 * 1024 * 1024 + 16 * 1024 * 1024)) {
+		test_msg("Failed to allocate 4Kb from space cache, returned offset is: %llu\n",
+			 offset);
+		return -EINVAL;
+	}
+
+	ret = check_cache_empty(cache);
+	if (ret)
+		return ret;
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	/*
+	 * Now test a similar scenario, but where our extent entry is located
+	 * to the right of the bitmap entry, so that we can check that stealing
+	 * space from a bitmap to the front of an extent entry works.
+	 */
+
+	/*
+	 * Extent entry covering free space range [128Mb + 128Kb, 128Mb + 256Kb[
+	 */
+	ret = test_add_free_space_entry(cache, 128 * 1024 * 1024 + 128 * 1024,
+					128 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	/* Bitmap entry covering free space range [0, 128Mb - 512Kb[ */
+	ret = test_add_free_space_entry(cache, 0,
+					128 * 1024 * 1024 - 512 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * Now make only the last 256Kb of the bitmap marked as free, so that
+	 * we end up with only the following ranges marked as free space:
+	 *
+	 * [128Mb + 128b, 128Mb + 256Kb[
+	 * [128Mb - 768Kb, 128Mb - 512Kb[
+	 */
+	ret = btrfs_remove_free_space(cache,
+				      0,
+				      128 * 1024 * 1024 - 768 * 1024);
+	if (ret) {
+		test_msg("Failed to free part of bitmap space %d\n", ret);
+		return ret;
+	}
+
+	/* Confirm that only those 2 ranges are marked as free. */
+	if (!test_check_exists(cache, 128 * 1024 * 1024 + 128 * 1024,
+			       128 * 1024)) {
+		test_msg("Free space range missing\n");
+		return -ENOENT;
+	}
+	if (!test_check_exists(cache, 128 * 1024 * 1024 - 768 * 1024,
+			       256 * 1024)) {
+		test_msg("Free space range missing\n");
+		return -ENOENT;
+	}
+
+	/*
+	 * Confirm that the bitmap range [0, 128Mb - 768Kb[ isn't marked
+	 * as free anymore.
+	 */
+	if (test_check_exists(cache, 0,
+			      128 * 1024 * 1024 - 768 * 1024)) {
+		test_msg("Bitmap region not removed from space cache\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Confirm that the region [128Mb - 512Kb, 128Mb[, which is
+	 * covered by the bitmap, isn't marked as free.
+	 */
+	if (test_check_exists(cache, 128 * 1024 * 1024 - 512 * 1024,
+			      512 * 1024)) {
+		test_msg("Invalid bitmap region marked as free\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Now lets mark the region [128Mb - 512Kb, 128Mb[ as free too. But,
+	 * lets make sure the free space cache marks it as free in the bitmap,
+	 * and doesn't insert a new extent entry to represent this region.
+	 */
+	ret = btrfs_add_free_space(cache, 128 * 1024 * 1024 - 512 * 1024,
+				   512 * 1024);
+	if (ret) {
+		test_msg("Error adding free space: %d\n", ret);
+		return ret;
+	}
+	/* Confirm the region is marked as free. */
+	if (!test_check_exists(cache, 128 * 1024 * 1024 - 512 * 1024,
+			       512 * 1024)) {
+		test_msg("Bitmap region not marked as free\n");
+		return -ENOENT;
+	}
+
+	/*
+	 * Confirm that no new extent entries or bitmap entries were added to
+	 * the cache after adding that free space region.
+	 */
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * Now lets add a small free space region to the left of the previous
+	 * one, which is not contiguous with it and is part of the bitmap too.
+	 * The goal is to test that the bitmap entry space stealing doesn't
+	 * steal this space region.
+	 */
+	ret = btrfs_add_free_space(cache, 32 * 1024 * 1024, 8192);
+	if (ret) {
+		test_msg("Error adding free space: %d\n", ret);
+		return ret;
+	}
+
+	/*
+	 * Now mark the region [128Mb, 128Mb + 128Kb[ as free too. This will
+	 * expand the range covered by the existing extent entry that represents
+	 * the free space [128Mb + 128Kb, 128Mb + 256Kb[.
+	 */
+	ret = btrfs_add_free_space(cache, 128 * 1024 * 1024, 128 * 1024);
+	if (ret) {
+		test_msg("Error adding free space: %d\n", ret);
+		return ret;
+	}
+	/* Confirm the region is marked as free. */
+	if (!test_check_exists(cache, 128 * 1024 * 1024, 128 * 1024)) {
+		test_msg("Extent region not marked as free\n");
+		return -ENOENT;
+	}
+
+	/*
+	 * Confirm that our extent entry didn't stole all free space from the
+	 * bitmap, because of the small 8Kb free space region.
+	 */
+	ret = check_num_extents_and_bitmaps(cache, 2, 1);
+	if (ret)
+		return ret;
+
+	/*
+	 * So now we have the range [128Mb - 768Kb, 128Mb + 256Kb[ as free
+	 * space. Without stealing bitmap free space into extent entry space,
+	 * we would have all this free space represented by 2 entries in the
+	 * cache:
+	 *
+	 * extent entry covering range: [128Mb, 128Mb + 256Kb[
+	 * bitmap entry covering range: [128Mb - 768Kb, 128Mb[
+	 *
+	 * Attempting to allocate the whole free space (1Mb) would fail, because
+	 * we can't allocate from multiple entries.
+	 * With the bitmap free space stealing, we get a single extent entry
+	 * that represents the 1Mb free space, and therefore we're able to
+	 * allocate the whole free space at once.
+	 */
+	if (!test_check_exists(cache, 128 * 1024 * 1024 - 768 * 1024,
+			       1 * 1024 * 1024)) {
+		test_msg("Expected region not marked as free\n");
+		return -ENOENT;
+	}
+
+	if (cache->free_space_ctl->free_space != (1 * 1024 * 1024 + 8192)) {
+		test_msg("Cache free space is not 1Mb + 8Kb\n");
+		return -EINVAL;
+	}
+
+	offset = btrfs_find_space_for_alloc(cache,
+					    0, 1 * 1024 * 1024, 0,
+					    &max_extent_size);
+	if (offset != (128 * 1024 * 1024 - 768 * 1024)) {
+		test_msg("Failed to allocate 1Mb from space cache, returned offset is: %llu\n",
+			 offset);
+		return -EINVAL;
+	}
+
+	/* All that remains is a 8Kb free space region in a bitmap. Confirm. */
+	ret = check_num_extents_and_bitmaps(cache, 1, 1);
+	if (ret)
+		return ret;
+
+	if (cache->free_space_ctl->free_space != 8192) {
+		test_msg("Cache free space is not 8Kb\n");
+		return -EINVAL;
+	}
+
+	offset = btrfs_find_space_for_alloc(cache,
+					    0, 8192, 0,
+					    &max_extent_size);
+	if (offset != (32 * 1024 * 1024)) {
+		test_msg("Failed to allocate 8Kb from space cache, returned offset is: %llu\n",
+			 offset);
+		return -EINVAL;
+	}
+
+	ret = check_cache_empty(cache);
+	if (ret)
+		return ret;
+
+	cache->free_space_ctl->op->use_bitmap = use_bitmap_op;
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	return 0;
+}
+
 int btrfs_test_free_space_cache(void)
 {
 	struct btrfs_block_group_cache *cache;
@@ -386,6 +898,8 @@ int btrfs_test_free_space_cache(void)
 	ret = test_bitmaps_and_extents(cache);
 	if (ret)
 		goto out;
+
+	ret = test_steal_space_from_bitmap_to_extent(cache);
 out:
 	__btrfs_remove_free_space_cache(cache->free_space_ctl);
 	kfree(cache->free_space_ctl);

fs/btrfs/transaction.c

@@ -386,7 +386,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 	int ret;
 
 	/* Send isn't supposed to start transactions. */
-	ASSERT(current->journal_info != (void *)BTRFS_SEND_TRANS_STUB);
+	ASSERT(current->journal_info != BTRFS_SEND_TRANS_STUB);
 
 	if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))
 		return ERR_PTR(-EROFS);
@@ -408,7 +408,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 	if (num_items > 0 && root != root->fs_info->chunk_root) {
 		if (root->fs_info->quota_enabled &&
 		    is_fstree(root->root_key.objectid)) {
-			qgroup_reserved = num_items * root->leafsize;
+			qgroup_reserved = num_items * root->nodesize;
 			ret = btrfs_qgroup_reserve(root, qgroup_reserved);
 			if (ret)
 				return ERR_PTR(ret);
@@ -418,7 +418,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 		/*
 		 * Do the reservation for the relocation root creation
 		 */
-		if (unlikely(need_reserve_reloc_root(root))) {
+		if (need_reserve_reloc_root(root)) {
 			num_bytes += root->nodesize;
 			reloc_reserved = true;
 		}
@@ -609,7 +609,6 @@ int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid)
 		if (transid <= root->fs_info->last_trans_committed)
 			goto out;
 
-		ret = -EINVAL;
 		/* find specified transaction */
 		spin_lock(&root->fs_info->trans_lock);
 		list_for_each_entry(t, &root->fs_info->trans_list, list) {
@@ -625,9 +624,16 @@ int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid)
 			}
 		}
 		spin_unlock(&root->fs_info->trans_lock);
-		/* The specified transaction doesn't exist */
-		if (!cur_trans)
+
+		/*
+		 * The specified transaction doesn't exist, or we
+		 * raced with btrfs_commit_transaction
+		 */
+		if (!cur_trans) {
+			if (transid > root->fs_info->last_trans_committed)
+				ret = -EINVAL;
 			goto out;
+		}
 	} else {
 		/* find newest transaction that is committing | committed */
 		spin_lock(&root->fs_info->trans_lock);
@@ -851,6 +857,8 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
 	struct extent_state *cached_state = NULL;
 	u64 start = 0;
 	u64 end;
+	struct btrfs_inode *btree_ino = BTRFS_I(root->fs_info->btree_inode);
+	bool errors = false;
 
 	while (!find_first_extent_bit(dirty_pages, start, &start, &end,
 				      EXTENT_NEED_WAIT, &cached_state)) {
@@ -864,6 +872,26 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
 	}
 	if (err)
 		werr = err;
+
+	if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
+		if ((mark & EXTENT_DIRTY) &&
+		    test_and_clear_bit(BTRFS_INODE_BTREE_LOG1_ERR,
+				       &btree_ino->runtime_flags))
+			errors = true;
+
+		if ((mark & EXTENT_NEW) &&
+		    test_and_clear_bit(BTRFS_INODE_BTREE_LOG2_ERR,
+				       &btree_ino->runtime_flags))
+			errors = true;
+	} else {
+		if (test_and_clear_bit(BTRFS_INODE_BTREE_ERR,
+				       &btree_ino->runtime_flags))
+			errors = true;
+	}
+
+	if (errors && !werr)
+		werr = -EIO;
+
 	return werr;
 }
 
@@ -1629,6 +1657,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_transaction *cur_trans = trans->transaction;
 	struct btrfs_transaction *prev_trans = NULL;
+	struct btrfs_inode *btree_ino = BTRFS_I(root->fs_info->btree_inode);
 	int ret;
 
 	/* Stop the commit early if ->aborted is set */
@@ -1868,6 +1897,12 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	memcpy(root->fs_info->super_for_commit, root->fs_info->super_copy,
 	       sizeof(*root->fs_info->super_copy));
 
+	btrfs_update_commit_device_size(root->fs_info);
+	btrfs_update_commit_device_bytes_used(root, cur_trans);
+
+	clear_bit(BTRFS_INODE_BTREE_LOG1_ERR, &btree_ino->runtime_flags);
+	clear_bit(BTRFS_INODE_BTREE_LOG2_ERR, &btree_ino->runtime_flags);
+
 	spin_lock(&root->fs_info->trans_lock);
 	cur_trans->state = TRANS_STATE_UNBLOCKED;
 	root->fs_info->running_transaction = NULL;
@@ -1981,9 +2016,6 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 		ret = btrfs_drop_snapshot(root, NULL, 0, 0);
 	else
 		ret = btrfs_drop_snapshot(root, NULL, 1, 0);
-	/*
-	 * If we encounter a transaction abort during snapshot cleaning, we
-	 * don't want to crash here
-	 */
+
 	return (ret < 0) ? 0 : 1;
 }

fs/btrfs/transaction.h

@@ -79,7 +79,7 @@ struct btrfs_transaction {
 #define TRANS_EXTWRITERS	(__TRANS_USERSPACE | __TRANS_START |	\
 				 __TRANS_ATTACH)
 
-#define BTRFS_SEND_TRANS_STUB	1
+#define BTRFS_SEND_TRANS_STUB	((void *)1)
 
 struct btrfs_trans_handle {
 	u64 transid;

fs/btrfs/tree-log.c

@@ -97,7 +97,8 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root, struct inode *inode,
 			   int inode_only,
 			   const loff_t start,
-			   const loff_t end);
+			   const loff_t end,
+			   struct btrfs_log_ctx *ctx);
 static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root,
 			     struct btrfs_path *path, u64 objectid);
@@ -1498,7 +1499,7 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
 		return -EIO;
 
 	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
-	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
+	key.type = BTRFS_ORPHAN_ITEM_KEY;
 	key.offset = objectid;
 
 	ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
@@ -1637,6 +1638,7 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 	    found_key.type == log_key.type &&
 	    found_key.offset == log_key.offset &&
 	    btrfs_dir_type(path->nodes[0], dst_di) == log_type) {
+		update_size = false;
 		goto out;
 	}
 
@@ -2157,7 +2159,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 
 		bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
 		ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
-		blocksize = btrfs_level_size(root, *level - 1);
+		blocksize = root->nodesize;
 
 		parent = path->nodes[*level];
 		root_owner = btrfs_header_owner(parent);
@@ -2983,8 +2985,6 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 	min_key.type = key_type;
 	min_key.offset = min_offset;
 
-	path->keep_locks = 1;
-
 	ret = btrfs_search_forward(root, &min_key, path, trans->transid);
 
 	/*
@@ -3364,7 +3364,7 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 		 * or deletes of this inode don't have to relog the inode
 		 * again
 		 */
-		if (btrfs_key_type(ins_keys + i) == BTRFS_EXTENT_DATA_KEY &&
+		if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY &&
 		    !skip_csum) {
 			int found_type;
 			extent = btrfs_item_ptr(src, start_slot + i,
@@ -3573,107 +3573,33 @@ static int extent_cmp(void *priv, struct list_head *a, struct list_head *b)
 	return 0;
 }
 
-static int log_one_extent(struct btrfs_trans_handle *trans,
-			  struct inode *inode, struct btrfs_root *root,
-			  struct extent_map *em, struct btrfs_path *path,
-			  struct list_head *logged_list)
+static int wait_ordered_extents(struct btrfs_trans_handle *trans,
+				struct inode *inode,
+				struct btrfs_root *root,
+				const struct extent_map *em,
+				const struct list_head *logged_list,
+				bool *ordered_io_error)
 {
-	struct btrfs_root *log = root->log_root;
-	struct btrfs_file_extent_item *fi;
-	struct extent_buffer *leaf;
 	struct btrfs_ordered_extent *ordered;
-	struct list_head ordered_sums;
-	struct btrfs_map_token token;
-	struct btrfs_key key;
+	struct btrfs_root *log = root->log_root;
 	u64 mod_start = em->mod_start;
 	u64 mod_len = em->mod_len;
+	const bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 	u64 csum_offset;
 	u64 csum_len;
-	u64 extent_offset = em->start - em->orig_start;
-	u64 block_len;
-	int ret;
-	bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
-	int extent_inserted = 0;
-
-	INIT_LIST_HEAD(&ordered_sums);
-	btrfs_init_map_token(&token);
-
-	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
-				   em->start + em->len, NULL, 0, 1,
-				   sizeof(*fi), &extent_inserted);
-	if (ret)
-		return ret;
-
-	if (!extent_inserted) {
-		key.objectid = btrfs_ino(inode);
-		key.type = BTRFS_EXTENT_DATA_KEY;
-		key.offset = em->start;
-
-		ret = btrfs_insert_empty_item(trans, log, path, &key,
-					      sizeof(*fi));
-		if (ret)
-			return ret;
-	}
-	leaf = path->nodes[0];
-	fi = btrfs_item_ptr(leaf, path->slots[0],
-			    struct btrfs_file_extent_item);
-
-	btrfs_set_token_file_extent_generation(leaf, fi, em->generation,
-					       &token);
-	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
-		skip_csum = true;
-		btrfs_set_token_file_extent_type(leaf, fi,
-						 BTRFS_FILE_EXTENT_PREALLOC,
-						 &token);
-	} else {
-		btrfs_set_token_file_extent_type(leaf, fi,
-						 BTRFS_FILE_EXTENT_REG,
-						 &token);
-		if (em->block_start == EXTENT_MAP_HOLE)
-			skip_csum = true;
-	}
-
-	block_len = max(em->block_len, em->orig_block_len);
-	if (em->compress_type != BTRFS_COMPRESS_NONE) {
-		btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
-							em->block_start,
-							&token);
-		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
-							   &token);
-	} else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
-		btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
-							em->block_start -
-							extent_offset, &token);
-		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
-							   &token);
-	} else {
-		btrfs_set_token_file_extent_disk_bytenr(leaf, fi, 0, &token);
-		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, 0,
-							   &token);
-	}
-
-	btrfs_set_token_file_extent_offset(leaf, fi,
-					   em->start - em->orig_start,
-					   &token);
-	btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
-	btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token);
-	btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type,
-						&token);
-	btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token);
-	btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token);
-	btrfs_mark_buffer_dirty(leaf);
+	LIST_HEAD(ordered_sums);
+	int ret = 0;
 
-	btrfs_release_path(path);
-	if (ret) {
-		return ret;
-	}
+	*ordered_io_error = false;
 
-	if (skip_csum)
+	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags) ||
+	    em->block_start == EXTENT_MAP_HOLE)
 		return 0;
 
 	/*
-	 * First check and see if our csums are on our outstanding ordered
-	 * extents.
+	 * Wait far any ordered extent that covers our extent map. If it
+	 * finishes without an error, first check and see if our csums are on
+	 * our outstanding ordered extents.
 	 */
 	list_for_each_entry(ordered, logged_list, log_list) {
 		struct btrfs_ordered_sum *sum;
@@ -3685,6 +3611,24 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 		    mod_start + mod_len <= ordered->file_offset)
 			continue;
 
+		if (!test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) &&
+		    !test_bit(BTRFS_ORDERED_IOERR, &ordered->flags) &&
+		    !test_bit(BTRFS_ORDERED_DIRECT, &ordered->flags)) {
+			const u64 start = ordered->file_offset;
+			const u64 end = ordered->file_offset + ordered->len - 1;
+
+			WARN_ON(ordered->inode != inode);
+			filemap_fdatawrite_range(inode->i_mapping, start, end);
+		}
+
+		wait_event(ordered->wait,
+			   (test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) ||
+			    test_bit(BTRFS_ORDERED_IOERR, &ordered->flags)));
+
+		if (test_bit(BTRFS_ORDERED_IOERR, &ordered->flags)) {
+			*ordered_io_error = true;
+			break;
+		}
 		/*
 		 * We are going to copy all the csums on this ordered extent, so
 		 * go ahead and adjust mod_start and mod_len in case this
@@ -3716,6 +3660,9 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 			}
 		}
 
+		if (skip_csum)
+			continue;
+
 		/*
 		 * To keep us from looping for the above case of an ordered
 		 * extent that falls inside of the logged extent.
@@ -3733,18 +3680,16 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 		list_for_each_entry(sum, &ordered->list, list) {
 			ret = btrfs_csum_file_blocks(trans, log, sum);
 			if (ret)
-				goto unlocked;
+				break;
 		}
-
 	}
-unlocked:
 
-	if (!mod_len || ret)
+	if (*ordered_io_error || !mod_len || ret || skip_csum)
 		return ret;
 
 	if (em->compress_type) {
 		csum_offset = 0;
-		csum_len = block_len;
+		csum_len = max(em->block_len, em->orig_block_len);
 	} else {
 		csum_offset = mod_start - em->start;
 		csum_len = mod_len;
@@ -3771,11 +3716,106 @@ unlocked:
 	return ret;
 }
 
+static int log_one_extent(struct btrfs_trans_handle *trans,
+			  struct inode *inode, struct btrfs_root *root,
+			  const struct extent_map *em,
+			  struct btrfs_path *path,
+			  const struct list_head *logged_list,
+			  struct btrfs_log_ctx *ctx)
+{
+	struct btrfs_root *log = root->log_root;
+	struct btrfs_file_extent_item *fi;
+	struct extent_buffer *leaf;
+	struct btrfs_map_token token;
+	struct btrfs_key key;
+	u64 extent_offset = em->start - em->orig_start;
+	u64 block_len;
+	int ret;
+	int extent_inserted = 0;
+	bool ordered_io_err = false;
+
+	ret = wait_ordered_extents(trans, inode, root, em, logged_list,
+				   &ordered_io_err);
+	if (ret)
+		return ret;
+
+	if (ordered_io_err) {
+		ctx->io_err = -EIO;
+		return 0;
+	}
+
+	btrfs_init_map_token(&token);
+
+	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
+				   em->start + em->len, NULL, 0, 1,
+				   sizeof(*fi), &extent_inserted);
+	if (ret)
+		return ret;
+
+	if (!extent_inserted) {
+		key.objectid = btrfs_ino(inode);
+		key.type = BTRFS_EXTENT_DATA_KEY;
+		key.offset = em->start;
+
+		ret = btrfs_insert_empty_item(trans, log, path, &key,
+					      sizeof(*fi));
+		if (ret)
+			return ret;
+	}
+	leaf = path->nodes[0];
+	fi = btrfs_item_ptr(leaf, path->slots[0],
+			    struct btrfs_file_extent_item);
+
+	btrfs_set_token_file_extent_generation(leaf, fi, em->generation,
+					       &token);
+	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags))
+		btrfs_set_token_file_extent_type(leaf, fi,
+						 BTRFS_FILE_EXTENT_PREALLOC,
+						 &token);
+	else
+		btrfs_set_token_file_extent_type(leaf, fi,
+						 BTRFS_FILE_EXTENT_REG,
+						 &token);
+
+	block_len = max(em->block_len, em->orig_block_len);
+	if (em->compress_type != BTRFS_COMPRESS_NONE) {
+		btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
+							em->block_start,
+							&token);
+		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
+							   &token);
+	} else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
+		btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
+							em->block_start -
+							extent_offset, &token);
+		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
+							   &token);
+	} else {
+		btrfs_set_token_file_extent_disk_bytenr(leaf, fi, 0, &token);
+		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, 0,
+							   &token);
+	}
+
+	btrfs_set_token_file_extent_offset(leaf, fi, extent_offset, &token);
+	btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
+	btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token);
+	btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type,
+						&token);
+	btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token);
+	btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token);
+	btrfs_mark_buffer_dirty(leaf);
+
+	btrfs_release_path(path);
+
+	return ret;
+}
+
 static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     struct inode *inode,
 				     struct btrfs_path *path,
-				     struct list_head *logged_list)
+				     struct list_head *logged_list,
+				     struct btrfs_log_ctx *ctx)
 {
 	struct extent_map *em, *n;
 	struct list_head extents;
@@ -3833,7 +3873,8 @@ process:
 
 		write_unlock(&tree->lock);
 
-		ret = log_one_extent(trans, inode, root, em, path, logged_list);
+		ret = log_one_extent(trans, inode, root, em, path, logged_list,
+				     ctx);
 		write_lock(&tree->lock);
 		clear_em_logging(tree, em);
 		free_extent_map(em);
@@ -3863,7 +3904,8 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root, struct inode *inode,
 			   int inode_only,
 			   const loff_t start,
-			   const loff_t end)
+			   const loff_t end,
+			   struct btrfs_log_ctx *ctx)
 {
 	struct btrfs_path *path;
 	struct btrfs_path *dst_path;
@@ -3964,7 +4006,6 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 		err = ret;
 		goto out_unlock;
 	}
-	path->keep_locks = 1;
 
 	while (1) {
 		ins_nr = 0;
@@ -4049,7 +4090,7 @@ log_extents:
 	btrfs_release_path(dst_path);
 	if (fast_search) {
 		ret = btrfs_log_changed_extents(trans, root, inode, dst_path,
-						&logged_list);
+						&logged_list, ctx);
 		if (ret) {
 			err = ret;
 			goto out_unlock;
@@ -4239,7 +4280,7 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 	if (ret)
 		goto end_no_trans;
 
-	ret = btrfs_log_inode(trans, root, inode, inode_only, start, end);
+	ret = btrfs_log_inode(trans, root, inode, inode_only, start, end, ctx);
 	if (ret)
 		goto end_trans;
 
@@ -4268,7 +4309,7 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 		if (BTRFS_I(inode)->generation >
 		    root->fs_info->last_trans_committed) {
 			ret = btrfs_log_inode(trans, root, inode, inode_only,
-					      0, LLONG_MAX);
+					      0, LLONG_MAX, ctx);
 			if (ret)
 				goto end_trans;
 		}
@@ -4360,7 +4401,7 @@ int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
 again:
 	key.objectid = BTRFS_TREE_LOG_OBJECTID;
 	key.offset = (u64)-1;
-	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
+	key.type = BTRFS_ROOT_ITEM_KEY;
 
 	while (1) {
 		ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);

fs/btrfs/tree-log.h

@@ -28,6 +28,7 @@
 struct btrfs_log_ctx {
 	int log_ret;
 	int log_transid;
+	int io_err;
 	struct list_head list;
 };
 
@@ -35,6 +36,7 @@ static inline void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx)
 {
 	ctx->log_ret = 0;
 	ctx->log_transid = 0;
+	ctx->io_err = 0;
 	INIT_LIST_HEAD(&ctx->list);
 }
 

fs/btrfs/uuid-tree.c

@@ -279,7 +279,6 @@ int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
 	key.offset = 0;
 
 again_search_slot:
-	path->keep_locks = 1;
 	ret = btrfs_search_forward(root, &key, path, 0);
 	if (ret) {
 		if (ret > 0)

fs/btrfs/volumes.c

@@ -50,7 +50,7 @@ static void __btrfs_reset_dev_stats(struct btrfs_device *dev);
 static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev);
 static void btrfs_dev_stat_print_on_load(struct btrfs_device *device);
 
-static DEFINE_MUTEX(uuid_mutex);
+DEFINE_MUTEX(uuid_mutex);
 static LIST_HEAD(fs_uuids);
 
 static void lock_chunks(struct btrfs_root *root)
@@ -74,6 +74,7 @@ static struct btrfs_fs_devices *__alloc_fs_devices(void)
 	mutex_init(&fs_devs->device_list_mutex);
 
 	INIT_LIST_HEAD(&fs_devs->devices);
+	INIT_LIST_HEAD(&fs_devs->resized_devices);
 	INIT_LIST_HEAD(&fs_devs->alloc_list);
 	INIT_LIST_HEAD(&fs_devs->list);
 
@@ -154,11 +155,13 @@ static struct btrfs_device *__alloc_device(void)
 
 	INIT_LIST_HEAD(&dev->dev_list);
 	INIT_LIST_HEAD(&dev->dev_alloc_list);
+	INIT_LIST_HEAD(&dev->resized_list);
 
 	spin_lock_init(&dev->io_lock);
 
 	spin_lock_init(&dev->reada_lock);
 	atomic_set(&dev->reada_in_flight, 0);
+	atomic_set(&dev->dev_stats_ccnt, 0);
 	INIT_RADIX_TREE(&dev->reada_zones, GFP_NOFS & ~__GFP_WAIT);
 	INIT_RADIX_TREE(&dev->reada_extents, GFP_NOFS & ~__GFP_WAIT);
 
@@ -474,14 +477,13 @@ static noinline int device_list_add(const char *path,
 			return PTR_ERR(fs_devices);
 
 		list_add(&fs_devices->list, &fs_uuids);
-		fs_devices->latest_devid = devid;
-		fs_devices->latest_trans = found_transid;
 
 		device = NULL;
 	} else {
 		device = __find_device(&fs_devices->devices, devid,
 				       disk_super->dev_item.uuid);
 	}
+
 	if (!device) {
 		if (fs_devices->opened)
 			return -EBUSY;
@@ -565,10 +567,6 @@ static noinline int device_list_add(const char *path,
 	if (!fs_devices->opened)
 		device->generation = found_transid;
 
-	if (found_transid > fs_devices->latest_trans) {
-		fs_devices->latest_devid = devid;
-		fs_devices->latest_trans = found_transid;
-	}
 	*fs_devices_ret = fs_devices;
 
 	return ret;
@@ -584,8 +582,7 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 	if (IS_ERR(fs_devices))
 		return fs_devices;
 
-	fs_devices->latest_devid = orig->latest_devid;
-	fs_devices->latest_trans = orig->latest_trans;
+	mutex_lock(&orig->device_list_mutex);
 	fs_devices->total_devices = orig->total_devices;
 
 	/* We have held the volume lock, it is safe to get the devices. */
@@ -614,8 +611,10 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 		device->fs_devices = fs_devices;
 		fs_devices->num_devices++;
 	}
+	mutex_unlock(&orig->device_list_mutex);
 	return fs_devices;
 error:
+	mutex_unlock(&orig->device_list_mutex);
 	free_fs_devices(fs_devices);
 	return ERR_PTR(-ENOMEM);
 }
@@ -624,10 +623,7 @@ void btrfs_close_extra_devices(struct btrfs_fs_info *fs_info,
 			       struct btrfs_fs_devices *fs_devices, int step)
 {
 	struct btrfs_device *device, *next;
-
-	struct block_device *latest_bdev = NULL;
-	u64 latest_devid = 0;
-	u64 latest_transid = 0;
+	struct btrfs_device *latest_dev = NULL;
 
 	mutex_lock(&uuid_mutex);
 again:
@@ -635,11 +631,9 @@ again:
 	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
 		if (device->in_fs_metadata) {
 			if (!device->is_tgtdev_for_dev_replace &&
-			    (!latest_transid ||
-			     device->generation > latest_transid)) {
-				latest_devid = device->devid;
-				latest_transid = device->generation;
-				latest_bdev = device->bdev;
+			    (!latest_dev ||
+			     device->generation > latest_dev->generation)) {
+				latest_dev = device;
 			}
 			continue;
 		}
@@ -681,9 +675,7 @@ again:
 		goto again;
 	}
 
-	fs_devices->latest_bdev = latest_bdev;
-	fs_devices->latest_devid = latest_devid;
-	fs_devices->latest_trans = latest_transid;
+	fs_devices->latest_bdev = latest_dev->bdev;
 
 	mutex_unlock(&uuid_mutex);
 }
@@ -732,8 +724,6 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 			fs_devices->rw_devices--;
 		}
 
-		if (device->can_discard)
-			fs_devices->num_can_discard--;
 		if (device->missing)
 			fs_devices->missing_devices--;
 
@@ -798,11 +788,9 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 	struct block_device *bdev;
 	struct list_head *head = &fs_devices->devices;
 	struct btrfs_device *device;
-	struct block_device *latest_bdev = NULL;
+	struct btrfs_device *latest_dev = NULL;
 	struct buffer_head *bh;
 	struct btrfs_super_block *disk_super;
-	u64 latest_devid = 0;
-	u64 latest_transid = 0;
 	u64 devid;
 	int seeding = 1;
 	int ret = 0;
@@ -830,11 +818,9 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 			goto error_brelse;
 
 		device->generation = btrfs_super_generation(disk_super);
-		if (!latest_transid || device->generation > latest_transid) {
-			latest_devid = devid;
-			latest_transid = device->generation;
-			latest_bdev = bdev;
-		}
+		if (!latest_dev ||
+		    device->generation > latest_dev->generation)
+			latest_dev = device;
 
 		if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_SEEDING) {
 			device->writeable = 0;
@@ -844,10 +830,8 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 		}
 
 		q = bdev_get_queue(bdev);
-		if (blk_queue_discard(q)) {
+		if (blk_queue_discard(q))
 			device->can_discard = 1;
-			fs_devices->num_can_discard++;
-		}
 
 		device->bdev = bdev;
 		device->in_fs_metadata = 0;
@@ -877,9 +861,7 @@ error_brelse:
 	}
 	fs_devices->seeding = seeding;
 	fs_devices->opened = 1;
-	fs_devices->latest_bdev = latest_bdev;
-	fs_devices->latest_devid = latest_devid;
-	fs_devices->latest_trans = latest_transid;
+	fs_devices->latest_bdev = latest_dev->bdev;
 	fs_devices->total_rw_bytes = 0;
 out:
 	return ret;
@@ -1053,7 +1035,7 @@ int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
 		if (key.objectid > device->devid)
 			break;
 
-		if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
+		if (key.type != BTRFS_DEV_EXTENT_KEY)
 			goto next;
 
 		dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
@@ -1205,7 +1187,7 @@ again:
 		if (key.objectid > device->devid)
 			break;
 
-		if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
+		if (key.type != BTRFS_DEV_EXTENT_KEY)
 			goto next;
 
 		if (key.offset > search_start) {
@@ -1284,7 +1266,7 @@ out:
 
 static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans,
 			  struct btrfs_device *device,
-			  u64 start)
+			  u64 start, u64 *dev_extent_len)
 {
 	int ret;
 	struct btrfs_path *path;
@@ -1326,13 +1308,8 @@ again:
 		goto out;
 	}
 
-	if (device->bytes_used > 0) {
-		u64 len = btrfs_dev_extent_length(leaf, extent);
-		device->bytes_used -= len;
-		spin_lock(&root->fs_info->free_chunk_lock);
-		root->fs_info->free_chunk_space += len;
-		spin_unlock(&root->fs_info->free_chunk_lock);
-	}
+	*dev_extent_len = btrfs_dev_extent_length(leaf, extent);
+
 	ret = btrfs_del_item(trans, root, path);
 	if (ret) {
 		btrfs_error(root->fs_info, ret,
@@ -1482,8 +1459,10 @@ static int btrfs_add_device(struct btrfs_trans_handle *trans,
 	btrfs_set_device_io_align(leaf, dev_item, device->io_align);
 	btrfs_set_device_io_width(leaf, dev_item, device->io_width);
 	btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
-	btrfs_set_device_total_bytes(leaf, dev_item, device->disk_total_bytes);
-	btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used);
+	btrfs_set_device_total_bytes(leaf, dev_item,
+				     btrfs_device_get_disk_total_bytes(device));
+	btrfs_set_device_bytes_used(leaf, dev_item,
+				    btrfs_device_get_bytes_used(device));
 	btrfs_set_device_group(leaf, dev_item, 0);
 	btrfs_set_device_seek_speed(leaf, dev_item, 0);
 	btrfs_set_device_bandwidth(leaf, dev_item, 0);
@@ -1539,7 +1518,6 @@ static int btrfs_rm_dev_item(struct btrfs_root *root,
 	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
 	key.type = BTRFS_DEV_ITEM_KEY;
 	key.offset = device->devid;
-	lock_chunks(root);
 
 	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 	if (ret < 0)
@@ -1555,7 +1533,6 @@ static int btrfs_rm_dev_item(struct btrfs_root *root,
 		goto out;
 out:
 	btrfs_free_path(path);
-	unlock_chunks(root);
 	btrfs_commit_transaction(trans, root);
 	return ret;
 }
@@ -1671,8 +1648,8 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	if (device->writeable) {
 		lock_chunks(root);
 		list_del_init(&device->dev_alloc_list);
+		device->fs_devices->rw_devices--;
 		unlock_chunks(root);
-		root->fs_info->fs_devices->rw_devices--;
 		clear_super = true;
 	}
 
@@ -1691,11 +1668,6 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	if (ret)
 		goto error_undo;
 
-	spin_lock(&root->fs_info->free_chunk_lock);
-	root->fs_info->free_chunk_space = device->total_bytes -
-		device->bytes_used;
-	spin_unlock(&root->fs_info->free_chunk_lock);
-
 	device->in_fs_metadata = 0;
 	btrfs_scrub_cancel_dev(root->fs_info, device);
 
@@ -1749,9 +1721,7 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 			fs_devices = fs_devices->seed;
 		}
 		cur_devices->seed = NULL;
-		lock_chunks(root);
 		__btrfs_close_devices(cur_devices);
-		unlock_chunks(root);
 		free_fs_devices(cur_devices);
 	}
 
@@ -1824,8 +1794,8 @@ error_undo:
 		lock_chunks(root);
 		list_add(&device->dev_alloc_list,
 			 &root->fs_info->fs_devices->alloc_list);
+		device->fs_devices->rw_devices++;
 		unlock_chunks(root);
-		root->fs_info->fs_devices->rw_devices++;
 	}
 	goto error_brelse;
 }
@@ -1833,29 +1803,57 @@ error_undo:
 void btrfs_rm_dev_replace_srcdev(struct btrfs_fs_info *fs_info,
 				 struct btrfs_device *srcdev)
 {
+	struct btrfs_fs_devices *fs_devices;
+
 	WARN_ON(!mutex_is_locked(&fs_info->fs_devices->device_list_mutex));
 
+	/*
+	 * in case of fs with no seed, srcdev->fs_devices will point
+	 * to fs_devices of fs_info. However when the dev being replaced is
+	 * a seed dev it will point to the seed's local fs_devices. In short
+	 * srcdev will have its correct fs_devices in both the cases.
+	 */
+	fs_devices = srcdev->fs_devices;
+
 	list_del_rcu(&srcdev->dev_list);
 	list_del_rcu(&srcdev->dev_alloc_list);
-	fs_info->fs_devices->num_devices--;
-	if (srcdev->missing) {
-		fs_info->fs_devices->missing_devices--;
-		fs_info->fs_devices->rw_devices++;
-	}
-	if (srcdev->can_discard)
-		fs_info->fs_devices->num_can_discard--;
-	if (srcdev->bdev) {
-		fs_info->fs_devices->open_devices--;
+	fs_devices->num_devices--;
+	if (srcdev->missing)
+		fs_devices->missing_devices--;
 
-		/*
-		 * zero out the old super if it is not writable
-		 * (e.g. seed device)
-		 */
-		if (srcdev->writeable)
-			btrfs_scratch_superblock(srcdev);
+	if (srcdev->writeable) {
+		fs_devices->rw_devices--;
+		/* zero out the old super if it is writable */
+		btrfs_scratch_superblock(srcdev);
 	}
 
+	if (srcdev->bdev)
+		fs_devices->open_devices--;
+
 	call_rcu(&srcdev->rcu, free_device);
+
+	/*
+	 * unless fs_devices is seed fs, num_devices shouldn't go
+	 * zero
+	 */
+	BUG_ON(!fs_devices->num_devices && !fs_devices->seeding);
+
+	/* if this is no devs we rather delete the fs_devices */
+	if (!fs_devices->num_devices) {
+		struct btrfs_fs_devices *tmp_fs_devices;
+
+		tmp_fs_devices = fs_info->fs_devices;
+		while (tmp_fs_devices) {
+			if (tmp_fs_devices->seed == fs_devices) {
+				tmp_fs_devices->seed = fs_devices->seed;
+				break;
+			}
+			tmp_fs_devices = tmp_fs_devices->seed;
+		}
+		fs_devices->seed = NULL;
+		__btrfs_close_devices(fs_devices);
+		free_fs_devices(fs_devices);
+	}
 }
 
 void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
@@ -1863,6 +1861,7 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 {
 	struct btrfs_device *next_device;
 
+	mutex_lock(&uuid_mutex);
 	WARN_ON(!tgtdev);
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
 	if (tgtdev->bdev) {
@@ -1870,8 +1869,6 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 		fs_info->fs_devices->open_devices--;
 	}
 	fs_info->fs_devices->num_devices--;
-	if (tgtdev->can_discard)
-		fs_info->fs_devices->num_can_discard++;
 
 	next_device = list_entry(fs_info->fs_devices->devices.next,
 				 struct btrfs_device, dev_list);
@@ -1884,6 +1881,7 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	call_rcu(&tgtdev->rcu, free_device);
 
 	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
+	mutex_unlock(&uuid_mutex);
 }
 
 static int btrfs_find_device_by_path(struct btrfs_root *root, char *device_path,
@@ -1982,17 +1980,17 @@ 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);
+	list_for_each_entry(device, &seed_devices->devices, dev_list)
+		device->fs_devices = seed_devices;
 
+	lock_chunks(root);
 	list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list);
-	list_for_each_entry(device, &seed_devices->devices, dev_list) {
-		device->fs_devices = seed_devices;
-	}
+	unlock_chunks(root);
 
 	fs_devices->seeding = 0;
 	fs_devices->num_devices = 0;
 	fs_devices->open_devices = 0;
 	fs_devices->missing_devices = 0;
-	fs_devices->num_can_discard = 0;
 	fs_devices->rotating = 0;
 	fs_devices->seed = seed_devices;
 
@@ -2092,7 +2090,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	struct list_head *devices;
 	struct super_block *sb = root->fs_info->sb;
 	struct rcu_string *name;
-	u64 total_bytes;
+	u64 tmp;
 	int seeding_dev = 0;
 	int ret = 0;
 
@@ -2148,8 +2146,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 		goto error;
 	}
 
-	lock_chunks(root);
-
 	q = bdev_get_queue(bdev);
 	if (blk_queue_discard(q))
 		device->can_discard = 1;
@@ -2160,6 +2156,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	device->sector_size = root->sectorsize;
 	device->total_bytes = i_size_read(bdev->bd_inode);
 	device->disk_total_bytes = device->total_bytes;
+	device->commit_total_bytes = device->total_bytes;
 	device->dev_root = root->fs_info->dev_root;
 	device->bdev = bdev;
 	device->in_fs_metadata = 1;
@@ -2177,6 +2174,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	device->fs_devices = root->fs_info->fs_devices;
 
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+	lock_chunks(root);
 	list_add_rcu(&device->dev_list, &root->fs_info->fs_devices->devices);
 	list_add(&device->dev_alloc_list,
 		 &root->fs_info->fs_devices->alloc_list);
@@ -2184,8 +2182,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	root->fs_info->fs_devices->open_devices++;
 	root->fs_info->fs_devices->rw_devices++;
 	root->fs_info->fs_devices->total_devices++;
-	if (device->can_discard)
-		root->fs_info->fs_devices->num_can_discard++;
 	root->fs_info->fs_devices->total_rw_bytes += device->total_bytes;
 
 	spin_lock(&root->fs_info->free_chunk_lock);
@@ -2195,26 +2191,45 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	if (!blk_queue_nonrot(bdev_get_queue(bdev)))
 		root->fs_info->fs_devices->rotating = 1;
 
-	total_bytes = btrfs_super_total_bytes(root->fs_info->super_copy);
+	tmp = btrfs_super_total_bytes(root->fs_info->super_copy);
 	btrfs_set_super_total_bytes(root->fs_info->super_copy,
-				    total_bytes + device->total_bytes);
+				    tmp + device->total_bytes);
 
-	total_bytes = btrfs_super_num_devices(root->fs_info->super_copy);
+	tmp = btrfs_super_num_devices(root->fs_info->super_copy);
 	btrfs_set_super_num_devices(root->fs_info->super_copy,
-				    total_bytes + 1);
+				    tmp + 1);
 
 	/* add sysfs device entry */
 	btrfs_kobj_add_device(root->fs_info, device);
 
+	/*
+	 * we've got more storage, clear any full flags on the space
+	 * infos
+	 */
+	btrfs_clear_space_info_full(root->fs_info);
+
+	unlock_chunks(root);
 	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	if (seeding_dev) {
-		char fsid_buf[BTRFS_UUID_UNPARSED_SIZE];
+		lock_chunks(root);
 		ret = init_first_rw_device(trans, root, device);
+		unlock_chunks(root);
 		if (ret) {
 			btrfs_abort_transaction(trans, root, ret);
 			goto error_trans;
 		}
+	}
+
+	ret = btrfs_add_device(trans, root, device);
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto error_trans;
+	}
+
+	if (seeding_dev) {
+		char fsid_buf[BTRFS_UUID_UNPARSED_SIZE];
+
 		ret = btrfs_finish_sprout(trans, root);
 		if (ret) {
 			btrfs_abort_transaction(trans, root, ret);
@@ -2228,21 +2243,8 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 						root->fs_info->fsid);
 		if (kobject_rename(&root->fs_info->super_kobj, fsid_buf))
 			goto error_trans;
-	} else {
-		ret = btrfs_add_device(trans, root, device);
-		if (ret) {
-			btrfs_abort_transaction(trans, root, ret);
-			goto error_trans;
-		}
 	}
 
-	/*
-	 * we've got more storage, clear any full flags on the space
-	 * infos
-	 */
-	btrfs_clear_space_info_full(root->fs_info);
-
-	unlock_chunks(root);
 	root->fs_info->num_tolerated_disk_barrier_failures =
 		btrfs_calc_num_tolerated_disk_barrier_failures(root->fs_info);
 	ret = btrfs_commit_transaction(trans, root);
@@ -2274,7 +2276,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	return ret;
 
 error_trans:
-	unlock_chunks(root);
 	btrfs_end_transaction(trans, root);
 	rcu_string_free(device->name);
 	btrfs_kobj_rm_device(root->fs_info, device);
@@ -2289,6 +2290,7 @@ error:
 }
 
 int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
+				  struct btrfs_device *srcdev,
 				  struct btrfs_device **device_out)
 {
 	struct request_queue *q;
@@ -2301,24 +2303,38 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 	int ret = 0;
 
 	*device_out = NULL;
-	if (fs_info->fs_devices->seeding)
+	if (fs_info->fs_devices->seeding) {
+		btrfs_err(fs_info, "the filesystem is a seed filesystem!");
 		return -EINVAL;
+	}
 
 	bdev = blkdev_get_by_path(device_path, FMODE_WRITE | FMODE_EXCL,
 				  fs_info->bdev_holder);
-	if (IS_ERR(bdev))
+	if (IS_ERR(bdev)) {
+		btrfs_err(fs_info, "target device %s is invalid!", device_path);
 		return PTR_ERR(bdev);
+	}
 
 	filemap_write_and_wait(bdev->bd_inode->i_mapping);
 
 	devices = &fs_info->fs_devices->devices;
 	list_for_each_entry(device, devices, dev_list) {
 		if (device->bdev == bdev) {
+			btrfs_err(fs_info, "target device is in the filesystem!");
 			ret = -EEXIST;
 			goto error;
 		}
 	}
 
+
+	if (i_size_read(bdev->bd_inode) <
+	    btrfs_device_get_total_bytes(srcdev)) {
+		btrfs_err(fs_info, "target device is smaller than source device!");
+		ret = -EINVAL;
+		goto error;
+	}
+
+
 	device = btrfs_alloc_device(NULL, &devid, NULL);
 	if (IS_ERR(device)) {
 		ret = PTR_ERR(device);
@@ -2342,8 +2358,12 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 	device->io_width = root->sectorsize;
 	device->io_align = root->sectorsize;
 	device->sector_size = root->sectorsize;
-	device->total_bytes = i_size_read(bdev->bd_inode);
-	device->disk_total_bytes = device->total_bytes;
+	device->total_bytes = btrfs_device_get_total_bytes(srcdev);
+	device->disk_total_bytes = btrfs_device_get_disk_total_bytes(srcdev);
+	device->bytes_used = btrfs_device_get_bytes_used(srcdev);
+	ASSERT(list_empty(&srcdev->resized_list));
+	device->commit_total_bytes = srcdev->commit_total_bytes;
+	device->commit_bytes_used = device->bytes_used;
 	device->dev_root = fs_info->dev_root;
 	device->bdev = bdev;
 	device->in_fs_metadata = 1;
@@ -2355,8 +2375,6 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 	list_add(&device->dev_list, &fs_info->fs_devices->devices);
 	fs_info->fs_devices->num_devices++;
 	fs_info->fs_devices->open_devices++;
-	if (device->can_discard)
-		fs_info->fs_devices->num_can_discard++;
 	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	*device_out = device;
@@ -2415,8 +2433,10 @@ static noinline int btrfs_update_device(struct btrfs_trans_handle *trans,
 	btrfs_set_device_io_align(leaf, dev_item, device->io_align);
 	btrfs_set_device_io_width(leaf, dev_item, device->io_width);
 	btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
-	btrfs_set_device_total_bytes(leaf, dev_item, device->disk_total_bytes);
-	btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used);
+	btrfs_set_device_total_bytes(leaf, dev_item,
+				     btrfs_device_get_disk_total_bytes(device));
+	btrfs_set_device_bytes_used(leaf, dev_item,
+				    btrfs_device_get_bytes_used(device));
 	btrfs_mark_buffer_dirty(leaf);
 
 out:
@@ -2424,40 +2444,44 @@ out:
 	return ret;
 }
 
-static int __btrfs_grow_device(struct btrfs_trans_handle *trans,
+int btrfs_grow_device(struct btrfs_trans_handle *trans,
 		      struct btrfs_device *device, u64 new_size)
 {
 	struct btrfs_super_block *super_copy =
 		device->dev_root->fs_info->super_copy;
-	u64 old_total = btrfs_super_total_bytes(super_copy);
-	u64 diff = new_size - device->total_bytes;
+	struct btrfs_fs_devices *fs_devices;
+	u64 old_total;
+	u64 diff;
 
 	if (!device->writeable)
 		return -EACCES;
+
+	lock_chunks(device->dev_root);
+	old_total = btrfs_super_total_bytes(super_copy);
+	diff = new_size - device->total_bytes;
+
 	if (new_size <= device->total_bytes ||
-	    device->is_tgtdev_for_dev_replace)
+	    device->is_tgtdev_for_dev_replace) {
+		unlock_chunks(device->dev_root);
 		return -EINVAL;
+	}
+
+	fs_devices = device->dev_root->fs_info->fs_devices;
 
 	btrfs_set_super_total_bytes(super_copy, old_total + diff);
 	device->fs_devices->total_rw_bytes += diff;
 
-	device->total_bytes = new_size;
-	device->disk_total_bytes = new_size;
+	btrfs_device_set_total_bytes(device, new_size);
+	btrfs_device_set_disk_total_bytes(device, new_size);
 	btrfs_clear_space_info_full(device->dev_root->fs_info);
+	if (list_empty(&device->resized_list))
+		list_add_tail(&device->resized_list,
+			      &fs_devices->resized_devices);
+	unlock_chunks(device->dev_root);
 
 	return btrfs_update_device(trans, device);
 }
 
-int btrfs_grow_device(struct btrfs_trans_handle *trans,
-		      struct btrfs_device *device, u64 new_size)
-{
-	int ret;
-	lock_chunks(device->dev_root);
-	ret = __btrfs_grow_device(trans, device, new_size);
-	unlock_chunks(device->dev_root);
-	return ret;
-}
-
 static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
 			    struct btrfs_root *root,
 			    u64 chunk_tree, u64 chunk_objectid,
@@ -2509,6 +2533,7 @@ static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64
 	u32 cur;
 	struct btrfs_key key;
 
+	lock_chunks(root);
 	array_size = btrfs_super_sys_array_size(super_copy);
 
 	ptr = super_copy->sys_chunk_array;
@@ -2538,79 +2563,95 @@ static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64
 			cur += len;
 		}
 	}
+	unlock_chunks(root);
 	return ret;
 }
 
-static int btrfs_relocate_chunk(struct btrfs_root *root,
-			 u64 chunk_tree, u64 chunk_objectid,
-			 u64 chunk_offset)
+int btrfs_remove_chunk(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root, u64 chunk_offset)
 {
 	struct extent_map_tree *em_tree;
-	struct btrfs_root *extent_root;
-	struct btrfs_trans_handle *trans;
 	struct extent_map *em;
+	struct btrfs_root *extent_root = root->fs_info->extent_root;
 	struct map_lookup *map;
-	int ret;
-	int i;
+	u64 dev_extent_len = 0;
+	u64 chunk_objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
+	u64 chunk_tree = root->fs_info->chunk_root->objectid;
+	int i, ret = 0;
 
+	/* Just in case */
 	root = root->fs_info->chunk_root;
-	extent_root = root->fs_info->extent_root;
 	em_tree = &root->fs_info->mapping_tree.map_tree;
 
-	ret = btrfs_can_relocate(extent_root, chunk_offset);
-	if (ret)
-		return -ENOSPC;
-
-	/* step one, relocate all the extents inside this chunk */
-	ret = btrfs_relocate_block_group(extent_root, chunk_offset);
-	if (ret)
-		return ret;
-
-	trans = btrfs_start_transaction(root, 0);
-	if (IS_ERR(trans)) {
-		ret = PTR_ERR(trans);
-		btrfs_std_error(root->fs_info, ret);
-		return ret;
-	}
-
-	lock_chunks(root);
-
-	/*
-	 * step two, delete the device extents and the
-	 * chunk tree entries
-	 */
 	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, chunk_offset, 1);
 	read_unlock(&em_tree->lock);
 
-	BUG_ON(!em || em->start > chunk_offset ||
-	       em->start + em->len < chunk_offset);
+	if (!em || em->start > chunk_offset ||
+	    em->start + em->len < chunk_offset) {
+		/*
+		 * This is a logic error, but we don't want to just rely on the
+		 * user having built with ASSERT enabled, so if ASSERT doens't
+		 * do anything we still error out.
+		 */
+		ASSERT(0);
+		if (em)
+			free_extent_map(em);
+		return -EINVAL;
+	}
 	map = (struct map_lookup *)em->bdev;
 
 	for (i = 0; i < map->num_stripes; i++) {
-		ret = btrfs_free_dev_extent(trans, map->stripes[i].dev,
-					    map->stripes[i].physical);
-		BUG_ON(ret);
+		struct btrfs_device *device = map->stripes[i].dev;
+		ret = btrfs_free_dev_extent(trans, device,
+					    map->stripes[i].physical,
+					    &dev_extent_len);
+		if (ret) {
+			btrfs_abort_transaction(trans, root, ret);
+			goto out;
+		}
+
+		if (device->bytes_used > 0) {
+			lock_chunks(root);
+			btrfs_device_set_bytes_used(device,
+					device->bytes_used - dev_extent_len);
+			spin_lock(&root->fs_info->free_chunk_lock);
+			root->fs_info->free_chunk_space += dev_extent_len;
+			spin_unlock(&root->fs_info->free_chunk_lock);
+			btrfs_clear_space_info_full(root->fs_info);
+			unlock_chunks(root);
+		}
 
 		if (map->stripes[i].dev) {
 			ret = btrfs_update_device(trans, map->stripes[i].dev);
-			BUG_ON(ret);
+			if (ret) {
+				btrfs_abort_transaction(trans, root, ret);
+				goto out;
+			}
 		}
 	}
 	ret = btrfs_free_chunk(trans, root, chunk_tree, chunk_objectid,
 			       chunk_offset);
-
-	BUG_ON(ret);
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto out;
+	}
 
 	trace_btrfs_chunk_free(root, map, chunk_offset, em->len);
 
 	if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
 		ret = btrfs_del_sys_chunk(root, chunk_objectid, chunk_offset);
-		BUG_ON(ret);
+		if (ret) {
+			btrfs_abort_transaction(trans, root, ret);
+			goto out;
+		}
 	}
 
 	ret = btrfs_remove_block_group(trans, extent_root, chunk_offset);
-	BUG_ON(ret);
+	if (ret) {
+		btrfs_abort_transaction(trans, extent_root, ret);
+		goto out;
+	}
 
 	write_lock(&em_tree->lock);
 	remove_extent_mapping(em_tree, em);
@@ -2618,12 +2659,46 @@ static int btrfs_relocate_chunk(struct btrfs_root *root,
 
 	/* once for the tree */
 	free_extent_map(em);
+out:
 	/* once for us */
 	free_extent_map(em);
+	return ret;
+}
 
-	unlock_chunks(root);
+static int btrfs_relocate_chunk(struct btrfs_root *root,
+			 u64 chunk_tree, u64 chunk_objectid,
+			 u64 chunk_offset)
+{
+	struct btrfs_root *extent_root;
+	struct btrfs_trans_handle *trans;
+	int ret;
+
+	root = root->fs_info->chunk_root;
+	extent_root = root->fs_info->extent_root;
+
+	ret = btrfs_can_relocate(extent_root, chunk_offset);
+	if (ret)
+		return -ENOSPC;
+
+	/* step one, relocate all the extents inside this chunk */
+	ret = btrfs_relocate_block_group(extent_root, chunk_offset);
+	if (ret)
+		return ret;
+
+	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		btrfs_std_error(root->fs_info, ret);
+		return ret;
+	}
+
+	/*
+	 * step two, delete the device extents and the
+	 * chunk tree entries
+	 */
+	ret = btrfs_remove_chunk(trans, root, chunk_offset);
 	btrfs_end_transaction(trans, root);
-	return 0;
+	return ret;
 }
 
 static int btrfs_relocate_sys_chunks(struct btrfs_root *root)
@@ -2676,8 +2751,8 @@ again:
 						   found_key.offset);
 			if (ret == -ENOSPC)
 				failed++;
-			else if (ret)
-				BUG();
+			else
+				BUG_ON(ret);
 		}
 
 		if (found_key.offset == 0)
@@ -3084,11 +3159,12 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info)
 	/* step one make some room on all the devices */
 	devices = &fs_info->fs_devices->devices;
 	list_for_each_entry(device, devices, dev_list) {
-		old_size = device->total_bytes;
+		old_size = btrfs_device_get_total_bytes(device);
 		size_to_free = div_factor(old_size, 1);
 		size_to_free = min(size_to_free, (u64)1 * 1024 * 1024);
 		if (!device->writeable ||
-		    device->total_bytes - device->bytes_used > size_to_free ||
+		    btrfs_device_get_total_bytes(device) -
+		    btrfs_device_get_bytes_used(device) > size_to_free ||
 		    device->is_tgtdev_for_dev_replace)
 			continue;
 
@@ -3643,8 +3719,6 @@ static int btrfs_uuid_scan_kthread(void *data)
 	max_key.type = BTRFS_ROOT_ITEM_KEY;
 	max_key.offset = (u64)-1;
 
-	path->keep_locks = 1;
-
 	while (1) {
 		ret = btrfs_search_forward(root, &key, path, 0);
 		if (ret) {
@@ -3896,8 +3970,8 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size)
 	struct btrfs_key key;
 	struct btrfs_super_block *super_copy = root->fs_info->super_copy;
 	u64 old_total = btrfs_super_total_bytes(super_copy);
-	u64 old_size = device->total_bytes;
-	u64 diff = device->total_bytes - new_size;
+	u64 old_size = btrfs_device_get_total_bytes(device);
+	u64 diff = old_size - new_size;
 
 	if (device->is_tgtdev_for_dev_replace)
 		return -EINVAL;
@@ -3910,7 +3984,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size)
 
 	lock_chunks(root);
 
-	device->total_bytes = new_size;
+	btrfs_device_set_total_bytes(device, new_size);
 	if (device->writeable) {
 		device->fs_devices->total_rw_bytes -= diff;
 		spin_lock(&root->fs_info->free_chunk_lock);
@@ -3976,7 +4050,7 @@ again:
 		ret = -ENOSPC;
 		lock_chunks(root);
 
-		device->total_bytes = old_size;
+		btrfs_device_set_total_bytes(device, old_size);
 		if (device->writeable)
 			device->fs_devices->total_rw_bytes += diff;
 		spin_lock(&root->fs_info->free_chunk_lock);
@@ -3994,18 +4068,17 @@ again:
 	}
 
 	lock_chunks(root);
+	btrfs_device_set_disk_total_bytes(device, new_size);
+	if (list_empty(&device->resized_list))
+		list_add_tail(&device->resized_list,
+			      &root->fs_info->fs_devices->resized_devices);
 
-	device->disk_total_bytes = new_size;
-	/* Now btrfs_update_device() will change the on-disk size. */
-	ret = btrfs_update_device(trans, device);
-	if (ret) {
-		unlock_chunks(root);
-		btrfs_end_transaction(trans, root);
-		goto done;
-	}
 	WARN_ON(diff > old_total);
 	btrfs_set_super_total_bytes(super_copy, old_total - diff);
 	unlock_chunks(root);
+
+	/* Now btrfs_update_device() will change the on-disk size. */
+	ret = btrfs_update_device(trans, device);
 	btrfs_end_transaction(trans, root);
 done:
 	btrfs_free_path(path);
@@ -4021,10 +4094,13 @@ static int btrfs_add_system_chunk(struct btrfs_root *root,
 	u32 array_size;
 	u8 *ptr;
 
+	lock_chunks(root);
 	array_size = btrfs_super_sys_array_size(super_copy);
 	if (array_size + item_size + sizeof(disk_key)
-			> BTRFS_SYSTEM_CHUNK_ARRAY_SIZE)
+			> BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) {
+		unlock_chunks(root);
 		return -EFBIG;
+	}
 
 	ptr = super_copy->sys_chunk_array + array_size;
 	btrfs_cpu_key_to_disk(&disk_key, key);
@@ -4033,6 +4109,8 @@ static int btrfs_add_system_chunk(struct btrfs_root *root,
 	memcpy(ptr, chunk, item_size);
 	item_size += sizeof(disk_key);
 	btrfs_set_super_sys_array_size(super_copy, array_size + item_size);
+	unlock_chunks(root);
+
 	return 0;
 }
 
@@ -4402,6 +4480,16 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 	if (ret)
 		goto error_del_extent;
 
+	for (i = 0; i < map->num_stripes; i++) {
+		num_bytes = map->stripes[i].dev->bytes_used + stripe_size;
+		btrfs_device_set_bytes_used(map->stripes[i].dev, num_bytes);
+	}
+
+	spin_lock(&extent_root->fs_info->free_chunk_lock);
+	extent_root->fs_info->free_chunk_space -= (stripe_size *
+						   map->num_stripes);
+	spin_unlock(&extent_root->fs_info->free_chunk_lock);
+
 	free_extent_map(em);
 	check_raid56_incompat_flag(extent_root->fs_info, type);
 
@@ -4473,7 +4561,6 @@ int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
 		device = map->stripes[i].dev;
 		dev_offset = map->stripes[i].physical;
 
-		device->bytes_used += stripe_size;
 		ret = btrfs_update_device(trans, device);
 		if (ret)
 			goto out;
@@ -4486,11 +4573,6 @@ int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
 			goto out;
 	}
 
-	spin_lock(&extent_root->fs_info->free_chunk_lock);
-	extent_root->fs_info->free_chunk_space -= (stripe_size *
-						   map->num_stripes);
-	spin_unlock(&extent_root->fs_info->free_chunk_lock);
-
 	stripe = &chunk->stripe;
 	for (i = 0; i < map->num_stripes; i++) {
 		device = map->stripes[i].dev;
@@ -4570,16 +4652,25 @@ static noinline int init_first_rw_device(struct btrfs_trans_handle *trans,
 	alloc_profile = btrfs_get_alloc_profile(fs_info->chunk_root, 0);
 	ret = __btrfs_alloc_chunk(trans, extent_root, sys_chunk_offset,
 				  alloc_profile);
-	if (ret) {
-		btrfs_abort_transaction(trans, root, ret);
-		goto out;
+	return ret;
+}
+
+static inline int btrfs_chunk_max_errors(struct map_lookup *map)
+{
+	int max_errors;
+
+	if (map->type & (BTRFS_BLOCK_GROUP_RAID1 |
+			 BTRFS_BLOCK_GROUP_RAID10 |
+			 BTRFS_BLOCK_GROUP_RAID5 |
+			 BTRFS_BLOCK_GROUP_DUP)) {
+		max_errors = 1;
+	} else if (map->type & BTRFS_BLOCK_GROUP_RAID6) {
+		max_errors = 2;
+	} else {
+		max_errors = 0;
 	}
 
-	ret = btrfs_add_device(trans, fs_info->chunk_root, device);
-	if (ret)
-		btrfs_abort_transaction(trans, root, ret);
-out:
-	return ret;
+	return max_errors;
 }
 
 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
@@ -4588,6 +4679,7 @@ int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
 	struct map_lookup *map;
 	struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
 	int readonly = 0;
+	int miss_ndevs = 0;
 	int i;
 
 	read_lock(&map_tree->map_tree.lock);
@@ -4596,18 +4688,27 @@ int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
 	if (!em)
 		return 1;
 
-	if (btrfs_test_opt(root, DEGRADED)) {
-		free_extent_map(em);
-		return 0;
-	}
-
 	map = (struct map_lookup *)em->bdev;
 	for (i = 0; i < map->num_stripes; i++) {
+		if (map->stripes[i].dev->missing) {
+			miss_ndevs++;
+			continue;
+		}
+
 		if (!map->stripes[i].dev->writeable) {
 			readonly = 1;
-			break;
+			goto end;
 		}
 	}
+
+	/*
+	 * If the number of missing devices is larger than max errors,
+	 * we can not write the data into that chunk successfully, so
+	 * set it readonly.
+	 */
+	if (miss_ndevs > btrfs_chunk_max_errors(map))
+		readonly = 1;
+end:
 	free_extent_map(em);
 	return readonly;
 }
@@ -5008,6 +5109,8 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 			num_stripes = min_t(u64, map->num_stripes,
 					    stripe_nr_end - stripe_nr_orig);
 		stripe_index = do_div(stripe_nr, map->num_stripes);
+		if (!(rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS)))
+			mirror_num = 1;
 	} else if (map->type & BTRFS_BLOCK_GROUP_RAID1) {
 		if (rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS))
 			num_stripes = map->num_stripes;
@@ -5111,6 +5214,9 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 			/* We distribute the parity blocks across stripes */
 			tmp = stripe_nr + stripe_index;
 			stripe_index = do_div(tmp, map->num_stripes);
+			if (!(rw & (REQ_WRITE | REQ_DISCARD |
+				    REQ_GET_READ_MIRRORS)) && mirror_num <= 1)
+				mirror_num = 1;
 		}
 	} else {
 		/*
@@ -5218,16 +5324,8 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 		}
 	}
 
-	if (rw & (REQ_WRITE | REQ_GET_READ_MIRRORS)) {
-		if (map->type & (BTRFS_BLOCK_GROUP_RAID1 |
-				 BTRFS_BLOCK_GROUP_RAID10 |
-				 BTRFS_BLOCK_GROUP_RAID5 |
-				 BTRFS_BLOCK_GROUP_DUP)) {
-			max_errors = 1;
-		} else if (map->type & BTRFS_BLOCK_GROUP_RAID6) {
-			max_errors = 2;
-		}
-	}
+	if (rw & (REQ_WRITE | REQ_GET_READ_MIRRORS))
+		max_errors = btrfs_chunk_max_errors(map);
 
 	if (dev_replace_is_ongoing && (rw & (REQ_WRITE | REQ_DISCARD)) &&
 	    dev_replace->tgtdev != NULL) {
@@ -5610,8 +5708,8 @@ static void submit_stripe_bio(struct btrfs_root *root, struct btrfs_bio *bbio,
 		name = rcu_dereference(dev->name);
 		pr_debug("btrfs_map_bio: rw %d, sector=%llu, dev=%lu "
 			 "(%s id %llu), size=%u\n", rw,
-			 (u64)bio->bi_sector, (u_long)dev->bdev->bd_dev,
-			 name->str, dev->devid, bio->bi_size);
+			 (u64)bio->bi_iter.bi_sector, (u_long)dev->bdev->bd_dev,
+			 name->str, dev->devid, bio->bi_iter.bi_size);
 		rcu_read_unlock();
 	}
 #endif
@@ -5789,10 +5887,10 @@ struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
 }
 
 static struct btrfs_device *add_missing_dev(struct btrfs_root *root,
+					    struct btrfs_fs_devices *fs_devices,
 					    u64 devid, u8 *dev_uuid)
 {
 	struct btrfs_device *device;
-	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 
 	device = btrfs_alloc_device(NULL, &devid, dev_uuid);
 	if (IS_ERR(device))
@@ -5929,7 +6027,8 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 		}
 		if (!map->stripes[i].dev) {
 			map->stripes[i].dev =
-				add_missing_dev(root, devid, uuid);
+				add_missing_dev(root, root->fs_info->fs_devices,
+						devid, uuid);
 			if (!map->stripes[i].dev) {
 				free_extent_map(em);
 				return -EIO;
@@ -5956,7 +6055,9 @@ static void fill_device_from_item(struct extent_buffer *leaf,
 	device->devid = btrfs_device_id(leaf, dev_item);
 	device->disk_total_bytes = btrfs_device_total_bytes(leaf, dev_item);
 	device->total_bytes = device->disk_total_bytes;
+	device->commit_total_bytes = device->disk_total_bytes;
 	device->bytes_used = btrfs_device_bytes_used(leaf, dev_item);
+	device->commit_bytes_used = device->bytes_used;
 	device->type = btrfs_device_type(leaf, dev_item);
 	device->io_align = btrfs_device_io_align(leaf, dev_item);
 	device->io_width = btrfs_device_io_width(leaf, dev_item);
@@ -5968,7 +6069,8 @@ static void fill_device_from_item(struct extent_buffer *leaf,
 	read_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
 }
 
-static int open_seed_devices(struct btrfs_root *root, u8 *fsid)
+static struct btrfs_fs_devices *open_seed_devices(struct btrfs_root *root,
+						  u8 *fsid)
 {
 	struct btrfs_fs_devices *fs_devices;
 	int ret;
@@ -5977,49 +6079,56 @@ static int open_seed_devices(struct btrfs_root *root, u8 *fsid)
 
 	fs_devices = root->fs_info->fs_devices->seed;
 	while (fs_devices) {
-		if (!memcmp(fs_devices->fsid, fsid, BTRFS_UUID_SIZE)) {
-			ret = 0;
-			goto out;
-		}
+		if (!memcmp(fs_devices->fsid, fsid, BTRFS_UUID_SIZE))
+			return fs_devices;
+
 		fs_devices = fs_devices->seed;
 	}
 
 	fs_devices = find_fsid(fsid);
 	if (!fs_devices) {
-		ret = -ENOENT;
-		goto out;
+		if (!btrfs_test_opt(root, DEGRADED))
+			return ERR_PTR(-ENOENT);
+
+		fs_devices = alloc_fs_devices(fsid);
+		if (IS_ERR(fs_devices))
+			return fs_devices;
+
+		fs_devices->seeding = 1;
+		fs_devices->opened = 1;
+		return fs_devices;
 	}
 
 	fs_devices = clone_fs_devices(fs_devices);
-	if (IS_ERR(fs_devices)) {
-		ret = PTR_ERR(fs_devices);
-		goto out;
-	}
+	if (IS_ERR(fs_devices))
+		return fs_devices;
 
 	ret = __btrfs_open_devices(fs_devices, FMODE_READ,
 				   root->fs_info->bdev_holder);
 	if (ret) {
 		free_fs_devices(fs_devices);
+		fs_devices = ERR_PTR(ret);
 		goto out;
 	}
 
 	if (!fs_devices->seeding) {
 		__btrfs_close_devices(fs_devices);
 		free_fs_devices(fs_devices);
-		ret = -EINVAL;
+		fs_devices = ERR_PTR(-EINVAL);
 		goto out;
 	}
 
 	fs_devices->seed = root->fs_info->fs_devices->seed;
 	root->fs_info->fs_devices->seed = fs_devices;
 out:
-	return ret;
+	return fs_devices;
 }
 
 static int read_one_dev(struct btrfs_root *root,
 			struct extent_buffer *leaf,
 			struct btrfs_dev_item *dev_item)
 {
+	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 	struct btrfs_device *device;
 	u64 devid;
 	int ret;
@@ -6033,31 +6142,48 @@ static int read_one_dev(struct btrfs_root *root,
 			   BTRFS_UUID_SIZE);
 
 	if (memcmp(fs_uuid, root->fs_info->fsid, BTRFS_UUID_SIZE)) {
-		ret = open_seed_devices(root, fs_uuid);
-		if (ret && !btrfs_test_opt(root, DEGRADED))
-			return ret;
+		fs_devices = open_seed_devices(root, fs_uuid);
+		if (IS_ERR(fs_devices))
+			return PTR_ERR(fs_devices);
 	}
 
 	device = btrfs_find_device(root->fs_info, devid, dev_uuid, fs_uuid);
-	if (!device || !device->bdev) {
+	if (!device) {
 		if (!btrfs_test_opt(root, DEGRADED))
 			return -EIO;
 
-		if (!device) {
-			btrfs_warn(root->fs_info, "devid %llu missing", devid);
-			device = add_missing_dev(root, devid, dev_uuid);
-			if (!device)
-				return -ENOMEM;
-		} else if (!device->missing) {
+		btrfs_warn(root->fs_info, "devid %llu missing", devid);
+		device = add_missing_dev(root, fs_devices, devid, dev_uuid);
+		if (!device)
+			return -ENOMEM;
+	} else {
+		if (!device->bdev && !btrfs_test_opt(root, DEGRADED))
+			return -EIO;
+
+		if(!device->bdev && !device->missing) {
 			/*
 			 * this happens when a device that was properly setup
 			 * in the device info lists suddenly goes bad.
 			 * device->bdev is NULL, and so we have to set
 			 * device->missing to one here
 			 */
-			root->fs_info->fs_devices->missing_devices++;
+			device->fs_devices->missing_devices++;
 			device->missing = 1;
 		}
+
+		/* Move the device to its own fs_devices */
+		if (device->fs_devices != fs_devices) {
+			ASSERT(device->missing);
+
+			list_move(&device->dev_list, &fs_devices->devices);
+			device->fs_devices->num_devices--;
+			fs_devices->num_devices++;
+
+			device->fs_devices->missing_devices--;
+			fs_devices->missing_devices++;
+
+			device->fs_devices = fs_devices;
+		}
 	}
 
 	if (device->fs_devices != root->fs_info->fs_devices) {
@@ -6373,16 +6499,18 @@ int btrfs_run_dev_stats(struct btrfs_trans_handle *trans,
 	struct btrfs_root *dev_root = fs_info->dev_root;
 	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
 	struct btrfs_device *device;
+	int stats_cnt;
 	int ret = 0;
 
 	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry(device, &fs_devices->devices, dev_list) {
-		if (!device->dev_stats_valid || !device->dev_stats_dirty)
+		if (!device->dev_stats_valid || !btrfs_dev_stats_dirty(device))
 			continue;
 
+		stats_cnt = atomic_read(&device->dev_stats_ccnt);
 		ret = update_dev_stat_item(trans, dev_root, device);
 		if (!ret)
-			device->dev_stats_dirty = 0;
+			atomic_sub(stats_cnt, &device->dev_stats_ccnt);
 	}
 	mutex_unlock(&fs_devices->device_list_mutex);
 
@@ -6481,3 +6609,51 @@ int btrfs_scratch_superblock(struct btrfs_device *device)
 
 	return 0;
 }
+
+/*
+ * Update the size of all devices, which is used for writing out the
+ * super blocks.
+ */
+void btrfs_update_commit_device_size(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
+	struct btrfs_device *curr, *next;
+
+	if (list_empty(&fs_devices->resized_devices))
+		return;
+
+	mutex_lock(&fs_devices->device_list_mutex);
+	lock_chunks(fs_info->dev_root);
+	list_for_each_entry_safe(curr, next, &fs_devices->resized_devices,
+				 resized_list) {
+		list_del_init(&curr->resized_list);
+		curr->commit_total_bytes = curr->disk_total_bytes;
+	}
+	unlock_chunks(fs_info->dev_root);
+	mutex_unlock(&fs_devices->device_list_mutex);
+}
+
+/* Must be invoked during the transaction commit */
+void btrfs_update_commit_device_bytes_used(struct btrfs_root *root,
+					struct btrfs_transaction *transaction)
+{
+	struct extent_map *em;
+	struct map_lookup *map;
+	struct btrfs_device *dev;
+	int i;
+
+	if (list_empty(&transaction->pending_chunks))
+		return;
+
+	/* In order to kick the device replace finish process */
+	lock_chunks(root);
+	list_for_each_entry(em, &transaction->pending_chunks, list) {
+		map = (struct map_lookup *)em->bdev;
+
+		for (i = 0; i < map->num_stripes; i++) {
+			dev = map->stripes[i].dev;
+			dev->commit_bytes_used = dev->bytes_used;
+		}
+	}
+	unlock_chunks(root);
+}

fs/btrfs/volumes.h

@@ -24,6 +24,8 @@
 #include <linux/btrfs.h>
 #include "async-thread.h"
 
+extern struct mutex uuid_mutex;
+
 #define BTRFS_STRIPE_LEN	(64 * 1024)
 
 struct buffer_head;
@@ -32,41 +34,59 @@ struct btrfs_pending_bios {
 	struct bio *tail;
 };
 
+/*
+ * Use sequence counter to get consistent device stat data on
+ * 32-bit processors.
+ */
+#if BITS_PER_LONG==32 && defined(CONFIG_SMP)
+#include <linux/seqlock.h>
+#define __BTRFS_NEED_DEVICE_DATA_ORDERED
+#define btrfs_device_data_ordered_init(device)	\
+	seqcount_init(&device->data_seqcount)
+#else
+#define btrfs_device_data_ordered_init(device) do { } while (0)
+#endif
+
 struct btrfs_device {
 	struct list_head dev_list;
 	struct list_head dev_alloc_list;
 	struct btrfs_fs_devices *fs_devices;
+
 	struct btrfs_root *dev_root;
 
+	struct rcu_string *name;
+
+	u64 generation;
+
+	spinlock_t io_lock ____cacheline_aligned;
+	int running_pending;
 	/* regular prio bios */
 	struct btrfs_pending_bios pending_bios;
 	/* WRITE_SYNC bios */
 	struct btrfs_pending_bios pending_sync_bios;
 
-	u64 generation;
-	int running_pending;
+	struct block_device *bdev;
+
+	/* the mode sent to blkdev_get */
+	fmode_t mode;
+
 	int writeable;
 	int in_fs_metadata;
 	int missing;
 	int can_discard;
 	int is_tgtdev_for_dev_replace;
 
-	spinlock_t io_lock;
-	/* the mode sent to blkdev_get */
-	fmode_t mode;
-
-	struct block_device *bdev;
-
-
-	struct rcu_string *name;
+#ifdef __BTRFS_NEED_DEVICE_DATA_ORDERED
+	seqcount_t data_seqcount;
+#endif
 
 	/* the internal btrfs device id */
 	u64 devid;
 
-	/* size of the device */
+	/* size of the device in memory */
 	u64 total_bytes;
 
-	/* size of the disk */
+	/* size of the device on disk */
 	u64 disk_total_bytes;
 
 	/* bytes used */
@@ -83,10 +103,26 @@ struct btrfs_device {
 	/* minimal io size for this device */
 	u32 sector_size;
 
-
 	/* physical drive uuid (or lvm uuid) */
 	u8 uuid[BTRFS_UUID_SIZE];
 
+	/*
+	 * size of the device on the current transaction
+	 *
+	 * This variant is update when committing the transaction,
+	 * and protected by device_list_mutex
+	 */
+	u64 commit_total_bytes;
+
+	/* bytes used on the current transaction */
+	u64 commit_bytes_used;
+	/*
+	 * used to manage the device which is resized
+	 *
+	 * It is protected by chunk_lock.
+	 */
+	struct list_head resized_list;
+
 	/* for sending down flush barriers */
 	int nobarriers;
 	struct bio *flush_bio;
@@ -107,26 +143,90 @@ struct btrfs_device {
 	struct radix_tree_root reada_zones;
 	struct radix_tree_root reada_extents;
 
-
 	/* disk I/O failure stats. For detailed description refer to
 	 * enum btrfs_dev_stat_values in ioctl.h */
 	int dev_stats_valid;
-	int dev_stats_dirty; /* counters need to be written to disk */
+
+	/* Counter to record the change of device stats */
+	atomic_t dev_stats_ccnt;
 	atomic_t dev_stat_values[BTRFS_DEV_STAT_VALUES_MAX];
 };
 
+/*
+ * If we read those variants at the context of their own lock, we needn't
+ * use the following helpers, reading them directly is safe.
+ */
+#if BITS_PER_LONG==32 && defined(CONFIG_SMP)
+#define BTRFS_DEVICE_GETSET_FUNCS(name)					\
+static inline u64							\
+btrfs_device_get_##name(const struct btrfs_device *dev)			\
+{									\
+	u64 size;							\
+	unsigned int seq;						\
+									\
+	do {								\
+		seq = read_seqcount_begin(&dev->data_seqcount);		\
+		size = dev->name;					\
+	} while (read_seqcount_retry(&dev->data_seqcount, seq));	\
+	return size;							\
+}									\
+									\
+static inline void							\
+btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
+{									\
+	preempt_disable();						\
+	write_seqcount_begin(&dev->data_seqcount);			\
+	dev->name = size;						\
+	write_seqcount_end(&dev->data_seqcount);			\
+	preempt_enable();						\
+}
+#elif BITS_PER_LONG==32 && defined(CONFIG_PREEMPT)
+#define BTRFS_DEVICE_GETSET_FUNCS(name)					\
+static inline u64							\
+btrfs_device_get_##name(const struct btrfs_device *dev)			\
+{									\
+	u64 size;							\
+									\
+	preempt_disable();						\
+	size = dev->name;						\
+	preempt_enable();						\
+	return size;							\
+}									\
+									\
+static inline void							\
+btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
+{									\
+	preempt_disable();						\
+	dev->name = size;						\
+	preempt_enable();						\
+}
+#else
+#define BTRFS_DEVICE_GETSET_FUNCS(name)					\
+static inline u64							\
+btrfs_device_get_##name(const struct btrfs_device *dev)			\
+{									\
+	return dev->name;						\
+}									\
+									\
+static inline void							\
+btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
+{									\
+	dev->name = size;						\
+}
+#endif
+
+BTRFS_DEVICE_GETSET_FUNCS(total_bytes);
+BTRFS_DEVICE_GETSET_FUNCS(disk_total_bytes);
+BTRFS_DEVICE_GETSET_FUNCS(bytes_used);
+
 struct btrfs_fs_devices {
 	u8 fsid[BTRFS_FSID_SIZE]; /* FS specific uuid */
 
-	/* the device with this id has the most recent copy of the super */
-	u64 latest_devid;
-	u64 latest_trans;
 	u64 num_devices;
 	u64 open_devices;
 	u64 rw_devices;
 	u64 missing_devices;
 	u64 total_rw_bytes;
-	u64 num_can_discard;
 	u64 total_devices;
 	struct block_device *latest_bdev;
 
@@ -139,6 +239,7 @@ struct btrfs_fs_devices {
 	struct mutex device_list_mutex;
 	struct list_head devices;
 
+	struct list_head resized_devices;
 	/* devices not currently being allocated */
 	struct list_head alloc_list;
 	struct list_head list;
@@ -167,8 +268,9 @@ struct btrfs_fs_devices {
  */
 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;
+	unsigned int mirror_num;
+	unsigned int stripe_index;
+	u64 logical;
 	u8 *csum;
 	u8 csum_inline[BTRFS_BIO_INLINE_CSUM_SIZE];
 	u8 *csum_allocated;
@@ -325,6 +427,7 @@ struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
 int btrfs_shrink_device(struct btrfs_device *device, u64 new_size);
 int btrfs_init_new_device(struct btrfs_root *root, char *path);
 int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
+				  struct btrfs_device *srcdev,
 				  struct btrfs_device **device_out);
 int btrfs_balance(struct btrfs_balance_control *bctl,
 		  struct btrfs_ioctl_balance_args *bargs);
@@ -360,11 +463,20 @@ unsigned long btrfs_full_stripe_len(struct btrfs_root *root,
 int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
 				struct btrfs_root *extent_root,
 				u64 chunk_offset, u64 chunk_size);
+int btrfs_remove_chunk(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root, u64 chunk_offset);
+
+static inline int btrfs_dev_stats_dirty(struct btrfs_device *dev)
+{
+	return atomic_read(&dev->dev_stats_ccnt);
+}
+
 static inline void btrfs_dev_stat_inc(struct btrfs_device *dev,
 				      int index)
 {
 	atomic_inc(dev->dev_stat_values + index);
-	dev->dev_stats_dirty = 1;
+	smp_mb__before_atomic();
+	atomic_inc(&dev->dev_stats_ccnt);
 }
 
 static inline int btrfs_dev_stat_read(struct btrfs_device *dev,
@@ -379,7 +491,8 @@ static inline int btrfs_dev_stat_read_and_reset(struct btrfs_device *dev,
 	int ret;
 
 	ret = atomic_xchg(dev->dev_stat_values + index, 0);
-	dev->dev_stats_dirty = 1;
+	smp_mb__before_atomic();
+	atomic_inc(&dev->dev_stats_ccnt);
 	return ret;
 }
 
@@ -387,7 +500,8 @@ static inline void btrfs_dev_stat_set(struct btrfs_device *dev,
 				      int index, unsigned long val)
 {
 	atomic_set(dev->dev_stat_values + index, val);
-	dev->dev_stats_dirty = 1;
+	smp_mb__before_atomic();
+	atomic_inc(&dev->dev_stats_ccnt);
 }
 
 static inline void btrfs_dev_stat_reset(struct btrfs_device *dev,
@@ -395,4 +509,8 @@ static inline void btrfs_dev_stat_reset(struct btrfs_device *dev,
 {
 	btrfs_dev_stat_set(dev, index, 0);
 }
+
+void btrfs_update_commit_device_size(struct btrfs_fs_info *fs_info);
+void btrfs_update_commit_device_bytes_used(struct btrfs_root *root,
+					struct btrfs_transaction *transaction);
 #endif

fs/btrfs/xattr.c

@@ -237,7 +237,7 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size)
 	 * first xattr that we find and walk forward
 	 */
 	key.objectid = btrfs_ino(inode);
-	btrfs_set_key_type(&key, BTRFS_XATTR_ITEM_KEY);
+	key.type = BTRFS_XATTR_ITEM_KEY;
 	key.offset = 0;
 
 	path = btrfs_alloc_path();
@@ -273,7 +273,7 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size)
 		/* check to make sure this item is what we want */
 		if (found_key.objectid != key.objectid)
 			break;
-		if (btrfs_key_type(&found_key) != BTRFS_XATTR_ITEM_KEY)
+		if (found_key.type != BTRFS_XATTR_ITEM_KEY)
 			break;
 
 		di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);

fs/btrfs/zlib.c

@@ -33,8 +33,7 @@
 #include "compression.h"
 
 struct workspace {
-	z_stream inf_strm;
-	z_stream def_strm;
+	z_stream strm;
 	char *buf;
 	struct list_head list;
 };
@@ -43,8 +42,7 @@ static void zlib_free_workspace(struct list_head *ws)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
 
-	vfree(workspace->def_strm.workspace);
-	vfree(workspace->inf_strm.workspace);
+	vfree(workspace->strm.workspace);
 	kfree(workspace->buf);
 	kfree(workspace);
 }
@@ -52,17 +50,17 @@ static void zlib_free_workspace(struct list_head *ws)
 static struct list_head *zlib_alloc_workspace(void)
 {
 	struct workspace *workspace;
+	int workspacesize;
 
 	workspace = kzalloc(sizeof(*workspace), GFP_NOFS);
 	if (!workspace)
 		return ERR_PTR(-ENOMEM);
 
-	workspace->def_strm.workspace = vmalloc(zlib_deflate_workspacesize(
-						MAX_WBITS, MAX_MEM_LEVEL));
-	workspace->inf_strm.workspace = vmalloc(zlib_inflate_workspacesize());
+	workspacesize = max(zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL),
+			zlib_inflate_workspacesize());
+	workspace->strm.workspace = vmalloc(workspacesize);
 	workspace->buf = kmalloc(PAGE_CACHE_SIZE, GFP_NOFS);
-	if (!workspace->def_strm.workspace ||
-	    !workspace->inf_strm.workspace || !workspace->buf)
+	if (!workspace->strm.workspace || !workspace->buf)
 		goto fail;
 
 	INIT_LIST_HEAD(&workspace->list);
@@ -96,14 +94,14 @@ static int zlib_compress_pages(struct list_head *ws,
 	*total_out = 0;
 	*total_in = 0;
 
-	if (Z_OK != zlib_deflateInit(&workspace->def_strm, 3)) {
+	if (Z_OK != zlib_deflateInit(&workspace->strm, 3)) {
 		printk(KERN_WARNING "BTRFS: deflateInit failed\n");
 		ret = -EIO;
 		goto out;
 	}
 
-	workspace->def_strm.total_in = 0;
-	workspace->def_strm.total_out = 0;
+	workspace->strm.total_in = 0;
+	workspace->strm.total_out = 0;
 
 	in_page = find_get_page(mapping, start >> PAGE_CACHE_SHIFT);
 	data_in = kmap(in_page);
@@ -117,25 +115,25 @@ static int zlib_compress_pages(struct list_head *ws,
 	pages[0] = out_page;
 	nr_pages = 1;
 
-	workspace->def_strm.next_in = data_in;
-	workspace->def_strm.next_out = cpage_out;
-	workspace->def_strm.avail_out = PAGE_CACHE_SIZE;
-	workspace->def_strm.avail_in = min(len, PAGE_CACHE_SIZE);
+	workspace->strm.next_in = data_in;
+	workspace->strm.next_out = cpage_out;
+	workspace->strm.avail_out = PAGE_CACHE_SIZE;
+	workspace->strm.avail_in = min(len, PAGE_CACHE_SIZE);
 
-	while (workspace->def_strm.total_in < len) {
-		ret = zlib_deflate(&workspace->def_strm, Z_SYNC_FLUSH);
+	while (workspace->strm.total_in < len) {
+		ret = zlib_deflate(&workspace->strm, Z_SYNC_FLUSH);
 		if (ret != Z_OK) {
 			printk(KERN_DEBUG "BTRFS: deflate in loop returned %d\n",
 			       ret);
-			zlib_deflateEnd(&workspace->def_strm);
+			zlib_deflateEnd(&workspace->strm);
 			ret = -EIO;
 			goto out;
 		}
 
 		/* we're making it bigger, give up */
-		if (workspace->def_strm.total_in > 8192 &&
-		    workspace->def_strm.total_in <
-		    workspace->def_strm.total_out) {
+		if (workspace->strm.total_in > 8192 &&
+		    workspace->strm.total_in <
+		    workspace->strm.total_out) {
 			ret = -E2BIG;
 			goto out;
 		}
@@ -143,7 +141,7 @@ static int zlib_compress_pages(struct list_head *ws,
 		 * before the total_in so we will pull in a new page for
 		 * the stream end if required
 		 */
-		if (workspace->def_strm.avail_out == 0) {
+		if (workspace->strm.avail_out == 0) {
 			kunmap(out_page);
 			if (nr_pages == nr_dest_pages) {
 				out_page = NULL;
@@ -158,19 +156,19 @@ static int zlib_compress_pages(struct list_head *ws,
 			cpage_out = kmap(out_page);
 			pages[nr_pages] = out_page;
 			nr_pages++;
-			workspace->def_strm.avail_out = PAGE_CACHE_SIZE;
-			workspace->def_strm.next_out = cpage_out;
+			workspace->strm.avail_out = PAGE_CACHE_SIZE;
+			workspace->strm.next_out = cpage_out;
 		}
 		/* we're all done */
-		if (workspace->def_strm.total_in >= len)
+		if (workspace->strm.total_in >= len)
 			break;
 
 		/* we've read in a full page, get a new one */
-		if (workspace->def_strm.avail_in == 0) {
-			if (workspace->def_strm.total_out > max_out)
+		if (workspace->strm.avail_in == 0) {
+			if (workspace->strm.total_out > max_out)
 				break;
 
-			bytes_left = len - workspace->def_strm.total_in;
+			bytes_left = len - workspace->strm.total_in;
 			kunmap(in_page);
 			page_cache_release(in_page);
 
@@ -178,28 +176,28 @@ static int zlib_compress_pages(struct list_head *ws,
 			in_page = find_get_page(mapping,
 						start >> PAGE_CACHE_SHIFT);
 			data_in = kmap(in_page);
-			workspace->def_strm.avail_in = min(bytes_left,
+			workspace->strm.avail_in = min(bytes_left,
 							   PAGE_CACHE_SIZE);
-			workspace->def_strm.next_in = data_in;
+			workspace->strm.next_in = data_in;
 		}
 	}
-	workspace->def_strm.avail_in = 0;
-	ret = zlib_deflate(&workspace->def_strm, Z_FINISH);
-	zlib_deflateEnd(&workspace->def_strm);
+	workspace->strm.avail_in = 0;
+	ret = zlib_deflate(&workspace->strm, Z_FINISH);
+	zlib_deflateEnd(&workspace->strm);
 
 	if (ret != Z_STREAM_END) {
 		ret = -EIO;
 		goto out;
 	}
 
-	if (workspace->def_strm.total_out >= workspace->def_strm.total_in) {
+	if (workspace->strm.total_out >= workspace->strm.total_in) {
 		ret = -E2BIG;
 		goto out;
 	}
 
 	ret = 0;
-	*total_out = workspace->def_strm.total_out;
-	*total_in = workspace->def_strm.total_in;
+	*total_out = workspace->strm.total_out;
+	*total_in = workspace->strm.total_in;
 out:
 	*out_pages = nr_pages;
 	if (out_page)
@@ -225,19 +223,18 @@ static int zlib_decompress_biovec(struct list_head *ws, struct page **pages_in,
 	size_t total_out = 0;
 	unsigned long page_in_index = 0;
 	unsigned long page_out_index = 0;
-	unsigned long total_pages_in = (srclen + PAGE_CACHE_SIZE - 1) /
-					PAGE_CACHE_SIZE;
+	unsigned long total_pages_in = DIV_ROUND_UP(srclen, PAGE_CACHE_SIZE);
 	unsigned long buf_start;
 	unsigned long pg_offset;
 
 	data_in = kmap(pages_in[page_in_index]);
-	workspace->inf_strm.next_in = data_in;
-	workspace->inf_strm.avail_in = min_t(size_t, srclen, PAGE_CACHE_SIZE);
-	workspace->inf_strm.total_in = 0;
+	workspace->strm.next_in = data_in;
+	workspace->strm.avail_in = min_t(size_t, srclen, PAGE_CACHE_SIZE);
+	workspace->strm.total_in = 0;
 
-	workspace->inf_strm.total_out = 0;
-	workspace->inf_strm.next_out = workspace->buf;
-	workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+	workspace->strm.total_out = 0;
+	workspace->strm.next_out = workspace->buf;
+	workspace->strm.avail_out = PAGE_CACHE_SIZE;
 	pg_offset = 0;
 
 	/* If it's deflate, and it's got no preset dictionary, then
@@ -247,21 +244,21 @@ static int zlib_decompress_biovec(struct list_head *ws, struct page **pages_in,
 	    !(((data_in[0]<<8) + data_in[1]) % 31)) {
 
 		wbits = -((data_in[0] >> 4) + 8);
-		workspace->inf_strm.next_in += 2;
-		workspace->inf_strm.avail_in -= 2;
+		workspace->strm.next_in += 2;
+		workspace->strm.avail_in -= 2;
 	}
 
-	if (Z_OK != zlib_inflateInit2(&workspace->inf_strm, wbits)) {
+	if (Z_OK != zlib_inflateInit2(&workspace->strm, wbits)) {
 		printk(KERN_WARNING "BTRFS: inflateInit failed\n");
 		return -EIO;
 	}
-	while (workspace->inf_strm.total_in < srclen) {
-		ret = zlib_inflate(&workspace->inf_strm, Z_NO_FLUSH);
+	while (workspace->strm.total_in < srclen) {
+		ret = zlib_inflate(&workspace->strm, Z_NO_FLUSH);
 		if (ret != Z_OK && ret != Z_STREAM_END)
 			break;
 
 		buf_start = total_out;
-		total_out = workspace->inf_strm.total_out;
+		total_out = workspace->strm.total_out;
 
 		/* we didn't make progress in this inflate call, we're done */
 		if (buf_start == total_out)
@@ -276,10 +273,10 @@ static int zlib_decompress_biovec(struct list_head *ws, struct page **pages_in,
 			goto done;
 		}
 
-		workspace->inf_strm.next_out = workspace->buf;
-		workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+		workspace->strm.next_out = workspace->buf;
+		workspace->strm.avail_out = PAGE_CACHE_SIZE;
 
-		if (workspace->inf_strm.avail_in == 0) {
+		if (workspace->strm.avail_in == 0) {
 			unsigned long tmp;
 			kunmap(pages_in[page_in_index]);
 			page_in_index++;
@@ -288,9 +285,9 @@ static int zlib_decompress_biovec(struct list_head *ws, struct page **pages_in,
 				break;
 			}
 			data_in = kmap(pages_in[page_in_index]);
-			workspace->inf_strm.next_in = data_in;
-			tmp = srclen - workspace->inf_strm.total_in;
-			workspace->inf_strm.avail_in = min(tmp,
+			workspace->strm.next_in = data_in;
+			tmp = srclen - workspace->strm.total_in;
+			workspace->strm.avail_in = min(tmp,
 							   PAGE_CACHE_SIZE);
 		}
 	}
@@ -299,7 +296,7 @@ static int zlib_decompress_biovec(struct list_head *ws, struct page **pages_in,
 	else
 		ret = 0;
 done:
-	zlib_inflateEnd(&workspace->inf_strm);
+	zlib_inflateEnd(&workspace->strm);
 	if (data_in)
 		kunmap(pages_in[page_in_index]);
 	return ret;
@@ -317,13 +314,13 @@ static int zlib_decompress(struct list_head *ws, unsigned char *data_in,
 	unsigned long total_out = 0;
 	char *kaddr;
 
-	workspace->inf_strm.next_in = data_in;
-	workspace->inf_strm.avail_in = srclen;
-	workspace->inf_strm.total_in = 0;
+	workspace->strm.next_in = data_in;
+	workspace->strm.avail_in = srclen;
+	workspace->strm.total_in = 0;
 
-	workspace->inf_strm.next_out = workspace->buf;
-	workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
-	workspace->inf_strm.total_out = 0;
+	workspace->strm.next_out = workspace->buf;
+	workspace->strm.avail_out = PAGE_CACHE_SIZE;
+	workspace->strm.total_out = 0;
 	/* If it's deflate, and it's got no preset dictionary, then
 	   we can tell zlib to skip the adler32 check. */
 	if (srclen > 2 && !(data_in[1] & PRESET_DICT) &&
@@ -331,11 +328,11 @@ static int zlib_decompress(struct list_head *ws, unsigned char *data_in,
 	    !(((data_in[0]<<8) + data_in[1]) % 31)) {
 
 		wbits = -((data_in[0] >> 4) + 8);
-		workspace->inf_strm.next_in += 2;
-		workspace->inf_strm.avail_in -= 2;
+		workspace->strm.next_in += 2;
+		workspace->strm.avail_in -= 2;
 	}
 
-	if (Z_OK != zlib_inflateInit2(&workspace->inf_strm, wbits)) {
+	if (Z_OK != zlib_inflateInit2(&workspace->strm, wbits)) {
 		printk(KERN_WARNING "BTRFS: inflateInit failed\n");
 		return -EIO;
 	}
@@ -346,12 +343,12 @@ static int zlib_decompress(struct list_head *ws, unsigned char *data_in,
 		unsigned long bytes;
 		unsigned long pg_offset = 0;
 
-		ret = zlib_inflate(&workspace->inf_strm, Z_NO_FLUSH);
+		ret = zlib_inflate(&workspace->strm, Z_NO_FLUSH);
 		if (ret != Z_OK && ret != Z_STREAM_END)
 			break;
 
 		buf_start = total_out;
-		total_out = workspace->inf_strm.total_out;
+		total_out = workspace->strm.total_out;
 
 		if (total_out == buf_start) {
 			ret = -EIO;
@@ -377,8 +374,8 @@ static int zlib_decompress(struct list_head *ws, unsigned char *data_in,
 		pg_offset += bytes;
 		bytes_left -= bytes;
 next:
-		workspace->inf_strm.next_out = workspace->buf;
-		workspace->inf_strm.avail_out = PAGE_CACHE_SIZE;
+		workspace->strm.next_out = workspace->buf;
+		workspace->strm.avail_out = PAGE_CACHE_SIZE;
 	}
 
 	if (ret != Z_STREAM_END && bytes_left != 0)
@@ -386,7 +383,7 @@ next:
 	else
 		ret = 0;
 
-	zlib_inflateEnd(&workspace->inf_strm);
+	zlib_inflateEnd(&workspace->strm);
 	return ret;
 }
 

include/trace/events/btrfs.h

@@ -23,6 +23,7 @@ struct map_lookup;
 struct extent_buffer;
 struct btrfs_work;
 struct __btrfs_workqueue;
+struct btrfs_qgroup_operation;
 
 #define show_ref_type(type)						\
 	__print_symbolic(type,						\
@@ -157,12 +158,13 @@ DEFINE_EVENT(btrfs__inode, btrfs_inode_evict,
 
 #define show_map_flags(flag)						\
 	__print_flags(flag, "|",					\
-		{ EXTENT_FLAG_PINNED, 		"PINNED" 	},	\
-		{ EXTENT_FLAG_COMPRESSED, 	"COMPRESSED" 	},	\
-		{ EXTENT_FLAG_VACANCY, 		"VACANCY" 	},	\
-		{ EXTENT_FLAG_PREALLOC, 	"PREALLOC" 	},	\
-		{ EXTENT_FLAG_LOGGING,	 	"LOGGING" 	},	\
-		{ EXTENT_FLAG_FILLING,	 	"FILLING" 	})
+		{ (1 << EXTENT_FLAG_PINNED), 		"PINNED" 	},\
+		{ (1 << EXTENT_FLAG_COMPRESSED), 	"COMPRESSED" 	},\
+		{ (1 << EXTENT_FLAG_VACANCY), 		"VACANCY" 	},\
+		{ (1 << EXTENT_FLAG_PREALLOC), 		"PREALLOC" 	},\
+		{ (1 << EXTENT_FLAG_LOGGING),	 	"LOGGING" 	},\
+		{ (1 << EXTENT_FLAG_FILLING),	 	"FILLING" 	},\
+		{ (1 << EXTENT_FLAG_FS_MAPPING),	"FS_MAPPING"	})
 
 TRACE_EVENT_CONDITION(btrfs_get_extent,
 
@@ -996,6 +998,7 @@ DECLARE_EVENT_CLASS(btrfs__work,
 		__field(	void *,	func			)
 		__field(	void *,	ordered_func		)
 		__field(	void *,	ordered_free		)
+		__field(	void *,	normal_work		)
 	),
 
 	TP_fast_assign(
@@ -1004,11 +1007,13 @@ DECLARE_EVENT_CLASS(btrfs__work,
 		__entry->func		= work->func;
 		__entry->ordered_func	= work->ordered_func;
 		__entry->ordered_free	= work->ordered_free;
+		__entry->normal_work	= &work->normal_work;
 	),
 
-	TP_printk("work=%p, wq=%p, func=%p, ordered_func=%p, ordered_free=%p",
-		  __entry->work, __entry->wq, __entry->func,
-		  __entry->ordered_func, __entry->ordered_free)
+	TP_printk("work=%p (normal_work=%p), wq=%p, func=%pf, ordered_func=%p,"
+		  " ordered_free=%p",
+		  __entry->work, __entry->normal_work, __entry->wq,
+		   __entry->func, __entry->ordered_func, __entry->ordered_free)
 );
 
 /* For situiations that the work is freed */
@@ -1043,13 +1048,6 @@ DEFINE_EVENT(btrfs__work, btrfs_work_sched,
 	TP_ARGS(work)
 );
 
-DEFINE_EVENT(btrfs__work, btrfs_normal_work_done,
-
-	TP_PROTO(struct btrfs_work *work),
-
-	TP_ARGS(work)
-);
-
 DEFINE_EVENT(btrfs__work__done, btrfs_all_work_done,
 
 	TP_PROTO(struct btrfs_work *work),
@@ -1119,6 +1117,61 @@ DEFINE_EVENT(btrfs__workqueue_done, btrfs_workqueue_destroy,
 	TP_ARGS(wq)
 );
 
+#define show_oper_type(type)						\
+	__print_symbolic(type,						\
+		{ BTRFS_QGROUP_OPER_ADD_EXCL, 	"OPER_ADD_EXCL" },	\
+		{ BTRFS_QGROUP_OPER_ADD_SHARED, "OPER_ADD_SHARED" },	\
+		{ BTRFS_QGROUP_OPER_SUB_EXCL, 	"OPER_SUB_EXCL" },	\
+		{ BTRFS_QGROUP_OPER_SUB_SHARED,	"OPER_SUB_SHARED" })
+
+DECLARE_EVENT_CLASS(btrfs_qgroup_oper,
+
+	TP_PROTO(struct btrfs_qgroup_operation *oper),
+
+	TP_ARGS(oper),
+
+	TP_STRUCT__entry(
+		__field(	u64,  ref_root		)
+		__field(	u64,  bytenr		)
+		__field(	u64,  num_bytes		)
+		__field(	u64,  seq		)
+		__field(	int,  type		)
+		__field(	u64,  elem_seq		)
+	),
+
+	TP_fast_assign(
+		__entry->ref_root	= oper->ref_root;
+		__entry->bytenr		= oper->bytenr,
+		__entry->num_bytes	= oper->num_bytes;
+		__entry->seq 		= oper->seq;
+		__entry->type		= oper->type;
+		__entry->elem_seq	= oper->elem.seq;
+	),
+
+	TP_printk("ref_root = %llu, bytenr = %llu, num_bytes = %llu, "
+		  "seq = %llu, elem.seq = %llu, type = %s",
+		  (unsigned long long)__entry->ref_root,
+		  (unsigned long long)__entry->bytenr,
+		  (unsigned long long)__entry->num_bytes,
+		  (unsigned long long)__entry->seq,
+		  (unsigned long long)__entry->elem_seq,
+		  show_oper_type(__entry->type))
+);
+
+DEFINE_EVENT(btrfs_qgroup_oper, btrfs_qgroup_account,
+
+	TP_PROTO(struct btrfs_qgroup_operation *oper),
+
+	TP_ARGS(oper)
+);
+
+DEFINE_EVENT(btrfs_qgroup_oper, btrfs_qgroup_record_ref,
+
+	TP_PROTO(struct btrfs_qgroup_operation *oper),
+
+	TP_ARGS(oper)
+);
+
 #endif /* _TRACE_BTRFS_H */
 
 /* This part must be outside protection */