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

Fix fs/btrfs/super.c conflict around #includes

20 files changed, 426 insertions, 318 deletions

diff --git a/MAINTAINERS b/MAINTAINERS
index d992d407197b..8a7b0b36e8b4 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1021,6 +1021,14 @@ M:	mb@bu3sch.de
 W:	http://bu3sch.de/btgpio.php
 S:	Maintained
 
+BTRFS FILE SYSTEM
+P:	Chris Mason
+M:	chris.mason@oracle.com
+L:	linux-btrfs@vger.kernel.org
+W:	http://btrfs.wiki.kernel.org/
+T:	git kernel.org:/pub/scm/linux/kernel/git/mason/btrfs-unstable.git
+S:	Maintained
+
 BTTV VIDEO4LINUX DRIVER
 P:	Mauro Carvalho Chehab
 M:	mchehab@infradead.org
diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index 8e2fec05dbe0..d5f4e94f2ca2 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -16,7 +16,6 @@
  * Boston, MA 021110-1307, USA.
  */
 
-#include <linux/version.h>
 #include <linux/kthread.h>
 #include <linux/list.h>
 #include <linux/spinlock.h>
diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index ee848d8585d9..ab07627084f1 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -32,7 +32,6 @@
 #include <linux/swap.h>
 #include <linux/writeback.h>
 #include <linux/bit_spinlock.h>
-#include <linux/version.h>
 #include <linux/pagevec.h>
 #include "compat.h"
 #include "ctree.h"
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 9e46c0776816..2603ee539b7a 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -1210,8 +1210,7 @@ static noinline void reada_for_search(struct btrfs_root *root,
 	struct btrfs_disk_key disk_key;
 	u32 nritems;
 	u64 search;
-	u64 lowest_read;
-	u64 highest_read;
+	u64 target;
 	u64 nread = 0;
 	int direction = path->reada;
 	struct extent_buffer *eb;
@@ -1235,8 +1234,7 @@ static noinline void reada_for_search(struct btrfs_root *root,
 		return;
 	}
 
-	highest_read = search;
-	lowest_read = search;
+	target = search;
 
 	nritems = btrfs_header_nritems(node);
 	nr = slot;
@@ -1256,24 +1254,15 @@ static noinline void reada_for_search(struct btrfs_root *root,
 				break;
 		}
 		search = btrfs_node_blockptr(node, nr);
-		if ((search >= lowest_read && search <= highest_read) ||
-		    (search < lowest_read && lowest_read - search <= 16384) ||
-		    (search > highest_read && search - highest_read <= 16384)) {
+		if ((search <= target && target - search <= 65536) ||
+		    (search > target && search - target <= 65536)) {
 			readahead_tree_block(root, search, blocksize,
 				     btrfs_node_ptr_generation(node, nr));
 			nread += blocksize;
 		}
 		nscan++;
-		if (path->reada < 2 && (nread > (64 * 1024) || nscan > 32))
+		if ((nread > 65536 || nscan > 32))
 			break;
-
-		if (nread > (256 * 1024) || nscan > 128)
-			break;
-
-		if (search < lowest_read)
-			lowest_read = search;
-		if (search > highest_read)
-			highest_read = search;
 	}
 }
 
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index eee060f88113..de103a8a815e 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -454,17 +454,11 @@ struct btrfs_timespec {
 	__le32 nsec;
 } __attribute__ ((__packed__));
 
-typedef enum {
+enum btrfs_compression_type {
 	BTRFS_COMPRESS_NONE = 0,
 	BTRFS_COMPRESS_ZLIB = 1,
 	BTRFS_COMPRESS_LAST = 2,
-} btrfs_compression_type;
-
-/* we don't understand any encryption methods right now */
-typedef enum {
-	BTRFS_ENCRYPTION_NONE = 0,
-	BTRFS_ENCRYPTION_LAST = 1,
-} btrfs_encryption_type;
+};
 
 struct btrfs_inode_item {
 	/* nfs style generation number */
@@ -701,9 +695,7 @@ struct btrfs_fs_info {
 	struct btrfs_transaction *running_transaction;
 	wait_queue_head_t transaction_throttle;
 	wait_queue_head_t transaction_wait;
-
 	wait_queue_head_t async_submit_wait;
-	wait_queue_head_t tree_log_wait;
 
 	struct btrfs_super_block super_copy;
 	struct btrfs_super_block super_for_commit;
@@ -730,10 +722,6 @@ struct btrfs_fs_info {
 	atomic_t async_submit_draining;
 	atomic_t nr_async_bios;
 	atomic_t async_delalloc_pages;
-	atomic_t tree_log_writers;
-	atomic_t tree_log_commit;
-	unsigned long tree_log_batch;
-	u64 tree_log_transid;
 
 	/*
 	 * this is used by the balancing code to wait for all the pending
@@ -833,7 +821,14 @@ struct btrfs_root {
 	struct kobject root_kobj;
 	struct completion kobj_unregister;
 	struct mutex objectid_mutex;
+
 	struct mutex log_mutex;
+	wait_queue_head_t log_writer_wait;
+	wait_queue_head_t log_commit_wait[2];
+	atomic_t log_writers;
+	atomic_t log_commit[2];
+	unsigned long log_transid;
+	unsigned long log_batch;
 
 	u64 objectid;
 	u64 last_trans;
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 81a313874ae5..7feac5a475e9 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -16,7 +16,6 @@
  * Boston, MA 021110-1307, USA.
  */
 
-#include <linux/version.h>
 #include <linux/fs.h>
 #include <linux/blkdev.h>
 #include <linux/scatterlist.h>
@@ -850,6 +849,14 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	spin_lock_init(&root->list_lock);
 	mutex_init(&root->objectid_mutex);
 	mutex_init(&root->log_mutex);
+	init_waitqueue_head(&root->log_writer_wait);
+	init_waitqueue_head(&root->log_commit_wait[0]);
+	init_waitqueue_head(&root->log_commit_wait[1]);
+	atomic_set(&root->log_commit[0], 0);
+	atomic_set(&root->log_commit[1], 0);
+	atomic_set(&root->log_writers, 0);
+	root->log_batch = 0;
+	root->log_transid = 0;
 	extent_io_tree_init(&root->dirty_log_pages,
 			     fs_info->btree_inode->i_mapping, GFP_NOFS);
 
@@ -934,15 +941,16 @@ int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
-			     struct btrfs_fs_info *fs_info)
+static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
+					 struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_root *root;
 	struct btrfs_root *tree_root = fs_info->tree_root;
+	struct extent_buffer *leaf;
 
 	root = kzalloc(sizeof(*root), GFP_NOFS);
 	if (!root)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 
 	__setup_root(tree_root->nodesize, tree_root->leafsize,
 		     tree_root->sectorsize, tree_root->stripesize,
@@ -951,12 +959,23 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 	root->root_key.objectid = BTRFS_TREE_LOG_OBJECTID;
 	root->root_key.type = BTRFS_ROOT_ITEM_KEY;
 	root->root_key.offset = BTRFS_TREE_LOG_OBJECTID;
+	/*
+	 * log trees do not get reference counted because they go away
+	 * before a real commit is actually done.  They do store pointers
+	 * to file data extents, and those reference counts still get
+	 * updated (along with back refs to the log tree).
+	 */
 	root->ref_cows = 0;
 
-	root->node = btrfs_alloc_free_block(trans, root, root->leafsize,
-					    0, BTRFS_TREE_LOG_OBJECTID,
-					    trans->transid, 0, 0, 0);
+	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
+				      0, BTRFS_TREE_LOG_OBJECTID,
+				      trans->transid, 0, 0, 0);
+	if (IS_ERR(leaf)) {
+		kfree(root);
+		return ERR_CAST(leaf);
+	}
 
+	root->node = leaf;
 	btrfs_set_header_nritems(root->node, 0);
 	btrfs_set_header_level(root->node, 0);
 	btrfs_set_header_bytenr(root->node, root->node->start);
@@ -968,7 +987,48 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			    BTRFS_FSID_SIZE);
 	btrfs_mark_buffer_dirty(root->node);
 	btrfs_tree_unlock(root->node);
-	fs_info->log_root_tree = root;
+	return root;
+}
+
+int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
+			     struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_root *log_root;
+
+	log_root = alloc_log_tree(trans, fs_info);
+	if (IS_ERR(log_root))
+		return PTR_ERR(log_root);
+	WARN_ON(fs_info->log_root_tree);
+	fs_info->log_root_tree = log_root;
+	return 0;
+}
+
+int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root)
+{
+	struct btrfs_root *log_root;
+	struct btrfs_inode_item *inode_item;
+
+	log_root = alloc_log_tree(trans, root->fs_info);
+	if (IS_ERR(log_root))
+		return PTR_ERR(log_root);
+
+	log_root->last_trans = trans->transid;
+	log_root->root_key.offset = root->root_key.objectid;
+
+	inode_item = &log_root->root_item.inode;
+	inode_item->generation = cpu_to_le64(1);
+	inode_item->size = cpu_to_le64(3);
+	inode_item->nlink = cpu_to_le32(1);
+	inode_item->nbytes = cpu_to_le64(root->leafsize);
+	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);
+
+	btrfs_set_root_bytenr(&log_root->root_item, log_root->node->start);
+	btrfs_set_root_generation(&log_root->root_item, trans->transid);
+
+	WARN_ON(root->log_root);
+	root->log_root = log_root;
+	root->log_transid = 0;
 	return 0;
 }
 
@@ -1136,7 +1196,6 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 {
 	struct btrfs_fs_info *info = (struct btrfs_fs_info *)congested_data;
 	int ret = 0;
-	struct list_head *cur;
 	struct btrfs_device *device;
 	struct backing_dev_info *bdi;
 #if 0
@@ -1144,8 +1203,7 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 	    btrfs_congested_async(info, 0))
 		return 1;
 #endif
-	list_for_each(cur, &info->fs_devices->devices) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(device, &info->fs_devices->devices, dev_list) {
 		if (!device->bdev)
 			continue;
 		bdi = blk_get_backing_dev_info(device->bdev);
@@ -1163,13 +1221,11 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
  */
 static void __unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
 {
-	struct list_head *cur;
 	struct btrfs_device *device;
 	struct btrfs_fs_info *info;
 
 	info = (struct btrfs_fs_info *)bdi->unplug_io_data;
-	list_for_each(cur, &info->fs_devices->devices) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(device, &info->fs_devices->devices, dev_list) {
 		if (!device->bdev)
 			continue;
 
@@ -1535,10 +1591,6 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	init_waitqueue_head(&fs_info->transaction_throttle);
 	init_waitqueue_head(&fs_info->transaction_wait);
 	init_waitqueue_head(&fs_info->async_submit_wait);
-	init_waitqueue_head(&fs_info->tree_log_wait);
-	atomic_set(&fs_info->tree_log_commit, 0);
-	atomic_set(&fs_info->tree_log_writers, 0);
-	fs_info->tree_log_transid = 0;
 
 	__setup_root(4096, 4096, 4096, 4096, tree_root,
 		     fs_info, BTRFS_ROOT_TREE_OBJECTID);
@@ -1740,13 +1792,13 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->system_alloc_profile = fs_info->metadata_alloc_profile;
 	fs_info->cleaner_kthread = kthread_run(cleaner_kthread, tree_root,
 					       "btrfs-cleaner");
-	if (!fs_info->cleaner_kthread)
+	if (IS_ERR(fs_info->cleaner_kthread))
 		goto fail_csum_root;
 
 	fs_info->transaction_kthread = kthread_run(transaction_kthread,
 						   tree_root,
 						   "btrfs-transaction");
-	if (!fs_info->transaction_kthread)
+	if (IS_ERR(fs_info->transaction_kthread))
 		goto fail_cleaner;
 
 	if (btrfs_super_log_root(disk_super) != 0) {
@@ -1828,13 +1880,14 @@ fail_sb_buffer:
 fail_iput:
 	invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
 	iput(fs_info->btree_inode);
-fail:
+
 	btrfs_close_devices(fs_info->fs_devices);
 	btrfs_mapping_tree_free(&fs_info->mapping_tree);
+	bdi_destroy(&fs_info->bdi);
 
+fail:
 	kfree(extent_root);
 	kfree(tree_root);
-	bdi_destroy(&fs_info->bdi);
 	kfree(fs_info);
 	kfree(chunk_root);
 	kfree(dev_root);
@@ -1995,7 +2048,6 @@ static int write_dev_supers(struct btrfs_device *device,
 
 int write_all_supers(struct btrfs_root *root, int max_mirrors)
 {
-	struct list_head *cur;
 	struct list_head *head = &root->fs_info->fs_devices->devices;
 	struct btrfs_device *dev;
 	struct btrfs_super_block *sb;
@@ -2011,8 +2063,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 
 	sb = &root->fs_info->super_for_commit;
 	dev_item = &sb->dev_item;
-	list_for_each(cur, head) {
-		dev = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(dev, head, dev_list) {
 		if (!dev->bdev) {
 			total_errors++;
 			continue;
@@ -2045,8 +2096,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 	}
 
 	total_errors = 0;
-	list_for_each(cur, head) {
-		dev = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(dev, head, dev_list) {
 		if (!dev->bdev)
 			continue;
 		if (!dev->in_fs_metadata || !dev->writeable)
diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h
index c0ff404c31b7..494a56eb2986 100644
--- a/fs/btrfs/disk-io.h
+++ b/fs/btrfs/disk-io.h
@@ -98,5 +98,7 @@ int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info);
 int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info);
+int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root);
 int btree_lock_page_hook(struct page *page);
 #endif
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 293da650873f..3b26f0980946 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -19,7 +19,6 @@
 #include <linux/pagemap.h>
 #include <linux/writeback.h>
 #include <linux/blkdev.h>
-#include <linux/version.h>
 #include "compat.h"
 #include "hash.h"
 #include "crc32c.h"
@@ -30,7 +29,6 @@
 #include "volumes.h"
 #include "locking.h"
 #include "ref-cache.h"
-#include "compat.h"
 
 #define PENDING_EXTENT_INSERT 0
 #define PENDING_EXTENT_DELETE 1
@@ -326,10 +324,8 @@ static struct btrfs_space_info *__find_space_info(struct btrfs_fs_info *info,
 						  u64 flags)
 {
 	struct list_head *head = &info->space_info;
-	struct list_head *cur;
 	struct btrfs_space_info *found;
-	list_for_each(cur, head) {
-		found = list_entry(cur, struct btrfs_space_info, list);
+	list_for_each_entry(found, head, list) {
 		if (found->flags == flags)
 			return found;
 	}
@@ -2159,7 +2155,8 @@ again:
 		ret = find_first_extent_bit(&info->extent_ins, search, &start,
 					    &end, EXTENT_WRITEBACK);
 		if (ret) {
-			if (skipped && all && !num_inserts) {
+			if (skipped && all && !num_inserts &&
+			    list_empty(&update_list)) {
 				skipped = 0;
 				search = 0;
 				continue;
@@ -2547,6 +2544,7 @@ again:
 		if (ret) {
 			if (all && skipped && !nr) {
 				search = 0;
+				skipped = 0;
 				continue;
 			}
 			mutex_unlock(&info->extent_ins_mutex);
@@ -2700,13 +2698,9 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 	/* if metadata always pin */
 	if (owner_objectid < BTRFS_FIRST_FREE_OBJECTID) {
 		if (root->root_key.objectid == BTRFS_TREE_LOG_OBJECTID) {
-			struct btrfs_block_group_cache *cache;
-
-			/* btrfs_free_reserved_extent */
-			cache = btrfs_lookup_block_group(root->fs_info, bytenr);
-			BUG_ON(!cache);
-			btrfs_add_free_space(cache, bytenr, num_bytes);
-			put_block_group(cache);
+			mutex_lock(&root->fs_info->pinned_mutex);
+			btrfs_update_pinned_extents(root, bytenr, num_bytes, 1);
+			mutex_unlock(&root->fs_info->pinned_mutex);
 			update_reserved_extents(root, bytenr, num_bytes, 0);
 			return 0;
 		}
@@ -3014,7 +3008,6 @@ loop_check:
 static void dump_space_info(struct btrfs_space_info *info, u64 bytes)
 {
 	struct btrfs_block_group_cache *cache;
-	struct list_head *l;
 
 	printk(KERN_INFO "space_info has %llu free, is %sfull\n",
 	       (unsigned long long)(info->total_bytes - info->bytes_used -
@@ -3022,8 +3015,7 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes)
 	       (info->full) ? "" : "not ");
 
 	down_read(&info->groups_sem);
-	list_for_each(l, &info->block_groups) {
-		cache = list_entry(l, struct btrfs_block_group_cache, list);
+	list_for_each_entry(cache, &info->block_groups, list) {
 		spin_lock(&cache->lock);
 		printk(KERN_INFO "block group %llu has %llu bytes, %llu used "
 		       "%llu pinned %llu reserved\n",
@@ -4444,7 +4436,7 @@ static noinline int replace_one_extent(struct btrfs_trans_handle *trans,
 	u64 lock_end = 0;
 	u64 num_bytes;
 	u64 ext_offset;
-	u64 first_pos;
+	u64 search_end = (u64)-1;
 	u32 nritems;
 	int nr_scaned = 0;
 	int extent_locked = 0;
@@ -4452,7 +4444,6 @@ static noinline int replace_one_extent(struct btrfs_trans_handle *trans,
 	int ret;
 
 	memcpy(&key, leaf_key, sizeof(key));
-	first_pos = INT_LIMIT(loff_t) - extent_key->offset;
 	if (ref_path->owner_objectid != BTRFS_MULTIPLE_OBJECTIDS) {
 		if (key.objectid < ref_path->owner_objectid ||
 		    (key.objectid == ref_path->owner_objectid &&
@@ -4501,7 +4492,7 @@ next:
 			if ((key.objectid > ref_path->owner_objectid) ||
 			    (key.objectid == ref_path->owner_objectid &&
 			     key.type > BTRFS_EXTENT_DATA_KEY) ||
-			    (key.offset >= first_pos + extent_key->offset))
+			    key.offset >= search_end)
 				break;
 		}
 
@@ -4534,8 +4525,10 @@ next:
 		num_bytes = btrfs_file_extent_num_bytes(leaf, fi);
 		ext_offset = btrfs_file_extent_offset(leaf, fi);
 
-		if (first_pos > key.offset - ext_offset)
-			first_pos = key.offset - ext_offset;
+		if (search_end == (u64)-1) {
+			search_end = key.offset - ext_offset +
+				btrfs_file_extent_ram_bytes(leaf, fi);
+		}
 
 		if (!extent_locked) {
 			lock_start = key.offset;
@@ -4724,7 +4717,7 @@ next:
 		}
 skip:
 		if (ref_path->owner_objectid != BTRFS_MULTIPLE_OBJECTIDS &&
-		    key.offset >= first_pos + extent_key->offset)
+		    key.offset >= search_end)
 			break;
 
 		cond_resched();
@@ -5957,9 +5950,11 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 	path = btrfs_alloc_path();
 	BUG_ON(!path);
 
-	btrfs_remove_free_space_cache(block_group);
+	spin_lock(&root->fs_info->block_group_cache_lock);
 	rb_erase(&block_group->cache_node,
 		 &root->fs_info->block_group_cache_tree);
+	spin_unlock(&root->fs_info->block_group_cache_lock);
+	btrfs_remove_free_space_cache(block_group);
 	down_write(&block_group->space_info->groups_sem);
 	list_del(&block_group->list);
 	up_write(&block_group->space_info->groups_sem);
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index e086d407f1fa..a3b0676403f7 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -9,7 +9,6 @@
 #include <linux/spinlock.h>
 #include <linux/blkdev.h>
 #include <linux/swap.h>
-#include <linux/version.h>
 #include <linux/writeback.h>
 #include <linux/pagevec.h>
 #include "extent_io.h"
@@ -2855,6 +2854,98 @@ out:
 	return sector;
 }
 
+int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		__u64 start, __u64 len, get_extent_t *get_extent)
+{
+	int ret;
+	u64 off = start;
+	u64 max = start + len;
+	u32 flags = 0;
+	u64 disko = 0;
+	struct extent_map *em = NULL;
+	int end = 0;
+	u64 em_start = 0, em_len = 0;
+	unsigned long emflags;
+	ret = 0;
+
+	if (len == 0)
+		return -EINVAL;
+
+	lock_extent(&BTRFS_I(inode)->io_tree, start, start + len,
+		GFP_NOFS);
+	em = get_extent(inode, NULL, 0, off, max - off, 0);
+	if (!em)
+		goto out;
+	if (IS_ERR(em)) {
+		ret = PTR_ERR(em);
+		goto out;
+	}
+	while (!end) {
+		off = em->start + em->len;
+		if (off >= max)
+			end = 1;
+
+		em_start = em->start;
+		em_len = em->len;
+
+		disko = 0;
+		flags = 0;
+
+		switch (em->block_start) {
+		case EXTENT_MAP_LAST_BYTE:
+			end = 1;
+			flags |= FIEMAP_EXTENT_LAST;
+			break;
+		case EXTENT_MAP_HOLE:
+			flags |= FIEMAP_EXTENT_UNWRITTEN;
+			break;
+		case EXTENT_MAP_INLINE:
+			flags |= (FIEMAP_EXTENT_DATA_INLINE |
+				  FIEMAP_EXTENT_NOT_ALIGNED);
+			break;
+		case EXTENT_MAP_DELALLOC:
+			flags |= (FIEMAP_EXTENT_DELALLOC |
+				  FIEMAP_EXTENT_UNKNOWN);
+			break;
+		default:
+			disko = em->block_start;
+			break;
+		}
+		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags))
+			flags |= FIEMAP_EXTENT_ENCODED;
+
+		emflags = em->flags;
+		free_extent_map(em);
+		em = NULL;
+
+		if (!end) {
+			em = get_extent(inode, NULL, 0, off, max - off, 0);
+			if (!em)
+				goto out;
+			if (IS_ERR(em)) {
+				ret = PTR_ERR(em);
+				goto out;
+			}
+			emflags = em->flags;
+		}
+		if (test_bit(EXTENT_FLAG_VACANCY, &emflags)) {
+			flags |= FIEMAP_EXTENT_LAST;
+			end = 1;
+		}
+
+		ret = fiemap_fill_next_extent(fieinfo, em_start, disko,
+					em_len, flags);
+		if (ret)
+			goto out_free;
+	}
+out_free:
+	free_extent_map(em);
+out:
+	unlock_extent(&BTRFS_I(inode)->io_tree, start, start + len,
+			GFP_NOFS);
+	return ret;
+}
+
 static inline struct page *extent_buffer_page(struct extent_buffer *eb,
 					      unsigned long i)
 {
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index c5b483a79137..e80c6d96b318 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -193,6 +193,8 @@ int extent_commit_write(struct extent_io_tree *tree,
 			unsigned from, unsigned to);
 sector_t extent_bmap(struct address_space *mapping, sector_t iblock,
 		get_extent_t *get_extent);
+int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		__u64 start, __u64 len, get_extent_t *get_extent);
 int set_range_dirty(struct extent_io_tree *tree, u64 start, u64 end);
 int set_state_private(struct extent_io_tree *tree, u64 start, u64 private);
 int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private);
diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
index 4a83e33ada32..50da69da20ce 100644
--- a/fs/btrfs/extent_map.c
+++ b/fs/btrfs/extent_map.c
@@ -3,7 +3,6 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/spinlock.h>
-#include <linux/version.h>
 #include <linux/hardirq.h>
 #include "extent_map.h"
 
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 90268334145e..3e8023efaff7 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -29,7 +29,6 @@
 #include <linux/writeback.h>
 #include <linux/statfs.h>
 #include <linux/compat.h>
-#include <linux/version.h>
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -1215,10 +1214,10 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	}
 	mutex_unlock(&root->fs_info->trans_mutex);
 
-	root->fs_info->tree_log_batch++;
+	root->log_batch++;
 	filemap_fdatawrite(inode->i_mapping);
 	btrfs_wait_ordered_range(inode, 0, (u64)-1);
-	root->fs_info->tree_log_batch++;
+	root->log_batch++;
 
 	/*
 	 * ok we haven't committed the transaction yet, lets do a commit
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 8adfe059ab41..288c2cdc7543 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -34,7 +34,6 @@
 #include <linux/statfs.h>
 #include <linux/compat.h>
 #include <linux/bit_spinlock.h>
-#include <linux/version.h>
 #include <linux/xattr.h>
 #include <linux/posix_acl.h>
 #include <linux/falloc.h>
@@ -1324,12 +1323,11 @@ static noinline int add_pending_csums(struct btrfs_trans_handle *trans,
 			     struct inode *inode, u64 file_offset,
 			     struct list_head *list)
 {
-	struct list_head *cur;
 	struct btrfs_ordered_sum *sum;
 
 	btrfs_set_trans_block_group(trans, inode);
-	list_for_each(cur, list) {
-		sum = list_entry(cur, struct btrfs_ordered_sum, list);
+
+	list_for_each_entry(sum, list, list) {
 		btrfs_csum_file_blocks(trans,
 		       BTRFS_I(inode)->root->fs_info->csum_root, sum);
 	}
@@ -4158,9 +4156,10 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 	return -EINVAL;
 }
 
-static sector_t btrfs_bmap(struct address_space *mapping, sector_t iblock)
+static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		__u64 start, __u64 len)
 {
-	return extent_bmap(mapping, iblock, btrfs_get_extent);
+	return extent_fiemap(inode, fieinfo, start, len, btrfs_get_extent);
 }
 
 int btrfs_readpage(struct file *file, struct page *page)
@@ -4987,13 +4986,24 @@ static struct extent_io_ops btrfs_extent_io_ops = {
 	.clear_bit_hook = btrfs_clear_bit_hook,
 };
 
+/*
+ * btrfs doesn't support the bmap operation because swapfiles
+ * use bmap to make a mapping of extents in the file.  They assume
+ * these extents won't change over the life of the file and they
+ * use the bmap result to do IO directly to the drive.
+ *
+ * the btrfs bmap call would return logical addresses that aren't
+ * suitable for IO and they also will change frequently as COW
+ * operations happen.  So, swapfile + btrfs == corruption.
+ *
+ * For now we're avoiding this by dropping bmap.
+ */
 static struct address_space_operations btrfs_aops = {
 	.readpage	= btrfs_readpage,
 	.writepage	= btrfs_writepage,
 	.writepages	= btrfs_writepages,
 	.readpages	= btrfs_readpages,
 	.sync_page	= block_sync_page,
-	.bmap		= btrfs_bmap,
 	.direct_IO	= btrfs_direct_IO,
 	.invalidatepage = btrfs_invalidatepage,
 	.releasepage	= btrfs_releasepage,
@@ -5017,6 +5027,7 @@ static struct inode_operations btrfs_file_inode_operations = {
 	.removexattr	= btrfs_removexattr,
 	.permission	= btrfs_permission,
 	.fallocate	= btrfs_fallocate,
+	.fiemap		= btrfs_fiemap,
 };
 static struct inode_operations btrfs_special_inode_operations = {
 	.getattr	= btrfs_getattr,
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index c2aa33e3feb5..988fdc8b49eb 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -38,7 +38,6 @@
 #include <linux/compat.h>
 #include <linux/bit_spinlock.h>
 #include <linux/security.h>
-#include <linux/version.h>
 #include <linux/xattr.h>
 #include <linux/vmalloc.h>
 #include "compat.h"
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index a20940170274..77c2411a5f0f 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -613,7 +613,6 @@ int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
 	struct btrfs_sector_sum *sector_sums;
 	struct btrfs_ordered_extent *ordered;
 	struct btrfs_ordered_inode_tree *tree = &BTRFS_I(inode)->ordered_tree;
-	struct list_head *cur;
 	unsigned long num_sectors;
 	unsigned long i;
 	u32 sectorsize = BTRFS_I(inode)->root->sectorsize;
@@ -624,8 +623,7 @@ int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
 		return 1;
 
 	mutex_lock(&tree->mutex);
-	list_for_each_prev(cur, &ordered->list) {
-		ordered_sum = list_entry(cur, struct btrfs_ordered_sum, list);
+	list_for_each_entry_reverse(ordered_sum, &ordered->list, list) {
 		if (disk_bytenr >= ordered_sum->bytenr) {
 			num_sectors = ordered_sum->len / sectorsize;
 			sector_sums = ordered_sum->sums;
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index db9fb3bc1e33..f3fd7e2cbc38 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -37,7 +37,6 @@
 #include <linux/ctype.h>
 #include <linux/namei.h>
 #include <linux/miscdevice.h>
-#include <linux/version.h>
 #include <linux/magic.h>
 #include "compat.h"
 #include "ctree.h"
@@ -583,17 +582,18 @@ static long btrfs_control_ioctl(struct file *file, unsigned int cmd,
 	struct btrfs_ioctl_vol_args *vol;
 	struct btrfs_fs_devices *fs_devices;
 	int ret = -ENOTTY;
-	int len;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
 	vol = kmalloc(sizeof(*vol), GFP_KERNEL);
+	if (!vol)
+		return -ENOMEM;
+
 	if (copy_from_user(vol, (void __user *)arg, sizeof(*vol))) {
 		ret = -EFAULT;
 		goto out;
 	}
-	len = strnlen(vol->name, BTRFS_PATH_NAME_MAX);
 
 	switch (cmd) {
 	case BTRFS_IOC_SCAN_DEV:
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 8a08f9443340..919172de5c9a 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -852,11 +852,9 @@ static noinline int create_pending_snapshots(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_pending_snapshot *pending;
 	struct list_head *head = &trans->transaction->pending_snapshots;
-	struct list_head *cur;
 	int ret;
 
-	list_for_each(cur, head) {
-		pending = list_entry(cur, struct btrfs_pending_snapshot, list);
+	list_for_each_entry(pending, head, list) {
 		ret = create_pending_snapshot(trans, fs_info, pending);
 		BUG_ON(ret);
 	}
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index d81cda2e077c..4f26f3ed0c87 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -78,104 +78,6 @@ static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
  */
 
 /*
- * btrfs_add_log_tree adds a new per-subvolume log tree into the
- * tree of log tree roots.  This must be called with a tree log transaction
- * running (see start_log_trans).
- */
-static int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
-		      struct btrfs_root *root)
-{
-	struct btrfs_key key;
-	struct btrfs_root_item root_item;
-	struct btrfs_inode_item *inode_item;
-	struct extent_buffer *leaf;
-	struct btrfs_root *new_root = root;
-	int ret;
-	u64 objectid = root->root_key.objectid;
-
-	leaf = btrfs_alloc_free_block(trans, root, root->leafsize, 0,
-				      BTRFS_TREE_LOG_OBJECTID,
-				      trans->transid, 0, 0, 0);
-	if (IS_ERR(leaf)) {
-		ret = PTR_ERR(leaf);
-		return ret;
-	}
-
-	btrfs_set_header_nritems(leaf, 0);
-	btrfs_set_header_level(leaf, 0);
-	btrfs_set_header_bytenr(leaf, leaf->start);
-	btrfs_set_header_generation(leaf, trans->transid);
-	btrfs_set_header_owner(leaf, BTRFS_TREE_LOG_OBJECTID);
-
-	write_extent_buffer(leaf, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(leaf),
-			    BTRFS_FSID_SIZE);
-	btrfs_mark_buffer_dirty(leaf);
-
-	inode_item = &root_item.inode;
-	memset(inode_item, 0, sizeof(*inode_item));
-	inode_item->generation = cpu_to_le64(1);
-	inode_item->size = cpu_to_le64(3);
-	inode_item->nlink = cpu_to_le32(1);
-	inode_item->nbytes = cpu_to_le64(root->leafsize);
-	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);
-
-	btrfs_set_root_bytenr(&root_item, leaf->start);
-	btrfs_set_root_generation(&root_item, trans->transid);
-	btrfs_set_root_level(&root_item, 0);
-	btrfs_set_root_refs(&root_item, 0);
-	btrfs_set_root_used(&root_item, 0);
-
-	memset(&root_item.drop_progress, 0, sizeof(root_item.drop_progress));
-	root_item.drop_level = 0;
-
-	btrfs_tree_unlock(leaf);
-	free_extent_buffer(leaf);
-	leaf = NULL;
-
-	btrfs_set_root_dirid(&root_item, 0);
-
-	key.objectid = BTRFS_TREE_LOG_OBJECTID;
-	key.offset = objectid;
-	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
-	ret = btrfs_insert_root(trans, root->fs_info->log_root_tree, &key,
-				&root_item);
-	if (ret)
-		goto fail;
-
-	new_root = btrfs_read_fs_root_no_radix(root->fs_info->log_root_tree,
-					       &key);
-	BUG_ON(!new_root);
-
-	WARN_ON(root->log_root);
-	root->log_root = new_root;
-
-	/*
-	 * log trees do not get reference counted because they go away
-	 * before a real commit is actually done.  They do store pointers
-	 * to file data extents, and those reference counts still get
-	 * updated (along with back refs to the log tree).
-	 */
-	new_root->ref_cows = 0;
-	new_root->last_trans = trans->transid;
-
-	/*
-	 * we need to make sure the root block for this new tree
-	 * is marked as dirty in the dirty_log_pages tree.  This
-	 * is how it gets flushed down to disk at tree log commit time.
-	 *
-	 * the tree logging mutex keeps others from coming in and changing
-	 * the new_root->node, so we can safely access it here
-	 */
-	set_extent_dirty(&new_root->dirty_log_pages, new_root->node->start,
-			 new_root->node->start + new_root->node->len - 1,
-			 GFP_NOFS);
-
-fail:
-	return ret;
-}
-
-/*
  * start a sub transaction and setup the log tree
  * this increments the log tree writer count to make the people
  * syncing the tree wait for us to finish
@@ -184,6 +86,14 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root)
 {
 	int ret;
+
+	mutex_lock(&root->log_mutex);
+	if (root->log_root) {
+		root->log_batch++;
+		atomic_inc(&root->log_writers);
+		mutex_unlock(&root->log_mutex);
+		return 0;
+	}
 	mutex_lock(&root->fs_info->tree_log_mutex);
 	if (!root->fs_info->log_root_tree) {
 		ret = btrfs_init_log_root_tree(trans, root->fs_info);
@@ -193,9 +103,10 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
 		ret = btrfs_add_log_tree(trans, root);
 		BUG_ON(ret);
 	}
-	atomic_inc(&root->fs_info->tree_log_writers);
-	root->fs_info->tree_log_batch++;
 	mutex_unlock(&root->fs_info->tree_log_mutex);
+	root->log_batch++;
+	atomic_inc(&root->log_writers);
+	mutex_unlock(&root->log_mutex);
 	return 0;
 }
 
@@ -212,13 +123,12 @@ static int join_running_log_trans(struct btrfs_root *root)
 	if (!root->log_root)
 		return -ENOENT;
 
-	mutex_lock(&root->fs_info->tree_log_mutex);
+	mutex_lock(&root->log_mutex);
 	if (root->log_root) {
 		ret = 0;
-		atomic_inc(&root->fs_info->tree_log_writers);
-		root->fs_info->tree_log_batch++;
+		atomic_inc(&root->log_writers);
 	}
-	mutex_unlock(&root->fs_info->tree_log_mutex);
+	mutex_unlock(&root->log_mutex);
 	return ret;
 }
 
@@ -228,10 +138,11 @@ static int join_running_log_trans(struct btrfs_root *root)
  */
 static int end_log_trans(struct btrfs_root *root)
 {
-	atomic_dec(&root->fs_info->tree_log_writers);
-	smp_mb();
-	if (waitqueue_active(&root->fs_info->tree_log_wait))
-		wake_up(&root->fs_info->tree_log_wait);
+	if (atomic_dec_and_test(&root->log_writers)) {
+		smp_mb();
+		if (waitqueue_active(&root->log_writer_wait))
+			wake_up(&root->log_writer_wait);
+	}
 	return 0;
 }
 
@@ -1902,26 +1813,65 @@ static int walk_log_tree(struct btrfs_trans_handle *trans,
 		}
 	}
 	btrfs_free_path(path);
-	if (wc->free)
-		free_extent_buffer(log->node);
 	return ret;
 }
 
-static int wait_log_commit(struct btrfs_root *log)
+/*
+ * helper function to update the item for a given subvolumes log root
+ * in the tree of log roots
+ */
+static int update_log_root(struct btrfs_trans_handle *trans,
+			   struct btrfs_root *log)
+{
+	int ret;
+
+	if (log->log_transid == 1) {
+		/* insert root item on the first sync */
+		ret = btrfs_insert_root(trans, log->fs_info->log_root_tree,
+				&log->root_key, &log->root_item);
+	} else {
+		ret = btrfs_update_root(trans, log->fs_info->log_root_tree,
+				&log->root_key, &log->root_item);
+	}
+	return ret;
+}
+
+static int wait_log_commit(struct btrfs_root *root, unsigned long transid)
 {
 	DEFINE_WAIT(wait);
-	u64 transid = log->fs_info->tree_log_transid;
+	int index = transid % 2;
 
+	/*
+	 * we only allow two pending log transactions at a time,
+	 * so we know that if ours is more than 2 older than the
+	 * current transaction, we're done
+	 */
 	do {
-		prepare_to_wait(&log->fs_info->tree_log_wait, &wait,
-				TASK_UNINTERRUPTIBLE);
-		mutex_unlock(&log->fs_info->tree_log_mutex);
-		if (atomic_read(&log->fs_info->tree_log_commit))
+		prepare_to_wait(&root->log_commit_wait[index],
+				&wait, TASK_UNINTERRUPTIBLE);
+		mutex_unlock(&root->log_mutex);
+		if (root->log_transid < transid + 2 &&
+		    atomic_read(&root->log_commit[index]))
 			schedule();
-		finish_wait(&log->fs_info->tree_log_wait, &wait);
-		mutex_lock(&log->fs_info->tree_log_mutex);
-	} while (transid == log->fs_info->tree_log_transid &&
-		atomic_read(&log->fs_info->tree_log_commit));
+		finish_wait(&root->log_commit_wait[index], &wait);
+		mutex_lock(&root->log_mutex);
+	} while (root->log_transid < transid + 2 &&
+		 atomic_read(&root->log_commit[index]));
+	return 0;
+}
+
+static int wait_for_writer(struct btrfs_root *root)
+{
+	DEFINE_WAIT(wait);
+	while (atomic_read(&root->log_writers)) {
+		prepare_to_wait(&root->log_writer_wait,
+				&wait, TASK_UNINTERRUPTIBLE);
+		mutex_unlock(&root->log_mutex);
+		if (atomic_read(&root->log_writers))
+			schedule();
+		mutex_lock(&root->log_mutex);
+		finish_wait(&root->log_writer_wait, &wait);
+	}
 	return 0;
 }
 
@@ -1933,57 +1883,114 @@ static int wait_log_commit(struct btrfs_root *log)
 int btrfs_sync_log(struct btrfs_trans_handle *trans,
 		   struct btrfs_root *root)
 {
+	int index1;
+	int index2;
 	int ret;
-	unsigned long batch;
 	struct btrfs_root *log = root->log_root;
+	struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
 
-	mutex_lock(&log->fs_info->tree_log_mutex);
-	if (atomic_read(&log->fs_info->tree_log_commit)) {
-		wait_log_commit(log);
-		goto out;
+	mutex_lock(&root->log_mutex);
+	index1 = root->log_transid % 2;
+	if (atomic_read(&root->log_commit[index1])) {
+		wait_log_commit(root, root->log_transid);
+		mutex_unlock(&root->log_mutex);
+		return 0;
 	}
-	atomic_set(&log->fs_info->tree_log_commit, 1);
+	atomic_set(&root->log_commit[index1], 1);
+
+	/* wait for previous tree log sync to complete */
+	if (atomic_read(&root->log_commit[(index1 + 1) % 2]))
+		wait_log_commit(root, root->log_transid - 1);
 
 	while (1) {
-		batch = log->fs_info->tree_log_batch;
-		mutex_unlock(&log->fs_info->tree_log_mutex);
+		unsigned long batch = root->log_batch;
+		mutex_unlock(&root->log_mutex);
 		schedule_timeout_uninterruptible(1);
-		mutex_lock(&log->fs_info->tree_log_mutex);
-
-		while (atomic_read(&log->fs_info->tree_log_writers)) {
-			DEFINE_WAIT(wait);
-			prepare_to_wait(&log->fs_info->tree_log_wait, &wait,
-					TASK_UNINTERRUPTIBLE);
-			mutex_unlock(&log->fs_info->tree_log_mutex);
-			if (atomic_read(&log->fs_info->tree_log_writers))
-				schedule();
-			mutex_lock(&log->fs_info->tree_log_mutex);
-			finish_wait(&log->fs_info->tree_log_wait, &wait);
-		}
-		if (batch == log->fs_info->tree_log_batch)
+		mutex_lock(&root->log_mutex);
+		wait_for_writer(root);
+		if (batch == root->log_batch)
 			break;
 	}
 
 	ret = btrfs_write_and_wait_marked_extents(log, &log->dirty_log_pages);
 	BUG_ON(ret);
-	ret = btrfs_write_and_wait_marked_extents(root->fs_info->log_root_tree,
-			       &root->fs_info->log_root_tree->dirty_log_pages);
+
+	btrfs_set_root_bytenr(&log->root_item, log->node->start);
+	btrfs_set_root_generation(&log->root_item, trans->transid);
+	btrfs_set_root_level(&log->root_item, btrfs_header_level(log->node));
+
+	root->log_batch = 0;
+	root->log_transid++;
+	log->log_transid = root->log_transid;
+	smp_mb();
+	/*
+	 * log tree has been flushed to disk, new modifications of
+	 * the log will be written to new positions. so it's safe to
+	 * allow log writers to go in.
+	 */
+	mutex_unlock(&root->log_mutex);
+
+	mutex_lock(&log_root_tree->log_mutex);
+	log_root_tree->log_batch++;
+	atomic_inc(&log_root_tree->log_writers);
+	mutex_unlock(&log_root_tree->log_mutex);
+
+	ret = update_log_root(trans, log);
+	BUG_ON(ret);
+
+	mutex_lock(&log_root_tree->log_mutex);
+	if (atomic_dec_and_test(&log_root_tree->log_writers)) {
+		smp_mb();
+		if (waitqueue_active(&log_root_tree->log_writer_wait))
+			wake_up(&log_root_tree->log_writer_wait);
+	}
+
+	index2 = log_root_tree->log_transid % 2;
+	if (atomic_read(&log_root_tree->log_commit[index2])) {
+		wait_log_commit(log_root_tree, log_root_tree->log_transid);
+		mutex_unlock(&log_root_tree->log_mutex);
+		goto out;
+	}
+	atomic_set(&log_root_tree->log_commit[index2], 1);
+
+	if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2]))
+		wait_log_commit(log_root_tree, log_root_tree->log_transid - 1);
+
+	wait_for_writer(log_root_tree);
+
+	ret = btrfs_write_and_wait_marked_extents(log_root_tree,
+				&log_root_tree->dirty_log_pages);
 	BUG_ON(ret);
 
 	btrfs_set_super_log_root(&root->fs_info->super_for_commit,
-				 log->fs_info->log_root_tree->node->start);
+				log_root_tree->node->start);
 	btrfs_set_super_log_root_level(&root->fs_info->super_for_commit,
-		       btrfs_header_level(log->fs_info->log_root_tree->node));
+				btrfs_header_level(log_root_tree->node));
+
+	log_root_tree->log_batch = 0;
+	log_root_tree->log_transid++;
+	smp_mb();
+
+	mutex_unlock(&log_root_tree->log_mutex);
+
+	/*
+	 * nobody else is going to jump in and write the the ctree
+	 * super here because the log_commit atomic below is protecting
+	 * us.  We must be called with a transaction handle pinning
+	 * the running transaction open, so a full commit can't hop
+	 * in and cause problems either.
+	 */
+	write_ctree_super(trans, root->fs_info->tree_root, 2);
 
-	write_ctree_super(trans, log->fs_info->tree_root, 2);
-	log->fs_info->tree_log_transid++;
-	log->fs_info->tree_log_batch = 0;
-	atomic_set(&log->fs_info->tree_log_commit, 0);
+	atomic_set(&log_root_tree->log_commit[index2], 0);
 	smp_mb();
-	if (waitqueue_active(&log->fs_info->tree_log_wait))
-		wake_up(&log->fs_info->tree_log_wait);
+	if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
+		wake_up(&log_root_tree->log_commit_wait[index2]);
 out:
-	mutex_unlock(&log->fs_info->tree_log_mutex);
+	atomic_set(&root->log_commit[index1], 0);
+	smp_mb();
+	if (waitqueue_active(&root->log_commit_wait[index1]))
+		wake_up(&root->log_commit_wait[index1]);
 	return 0;
 }
 
@@ -2019,38 +2026,18 @@ int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root)
 				   start, end, GFP_NOFS);
 	}
 
-	log = root->log_root;
-	ret = btrfs_del_root(trans, root->fs_info->log_root_tree,
-			     &log->root_key);
-	BUG_ON(ret);
+	if (log->log_transid > 0) {
+		ret = btrfs_del_root(trans, root->fs_info->log_root_tree,
+				     &log->root_key);
+		BUG_ON(ret);
+	}
 	root->log_root = NULL;
-	kfree(root->log_root);
+	free_extent_buffer(log->node);
+	kfree(log);
 	return 0;
 }
 
 /*
- * helper function to update the item for a given subvolumes log root
- * in the tree of log roots
- */
-static int update_log_root(struct btrfs_trans_handle *trans,
-			   struct btrfs_root *log)
-{
-	u64 bytenr = btrfs_root_bytenr(&log->root_item);
-	int ret;
-
-	if (log->node->start == bytenr)
-		return 0;
-
-	btrfs_set_root_bytenr(&log->root_item, log->node->start);
-	btrfs_set_root_generation(&log->root_item, trans->transid);
-	btrfs_set_root_level(&log->root_item, btrfs_header_level(log->node));
-	ret = btrfs_update_root(trans, log->fs_info->log_root_tree,
-				&log->root_key, &log->root_item);
-	BUG_ON(ret);
-	return ret;
-}
-
-/*
  * If both a file and directory are logged, and unlinks or renames are
  * mixed in, we have a few interesting corners:
  *
@@ -2711,11 +2698,6 @@ next_slot:
 
 	btrfs_free_path(path);
 	btrfs_free_path(dst_path);
-
-	mutex_lock(&root->fs_info->tree_log_mutex);
-	ret = update_log_root(trans, log);
-	BUG_ON(ret);
-	mutex_unlock(&root->fs_info->tree_log_mutex);
 out:
 	return 0;
 }
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 3451e1cca2b5..fd0bedb07a64 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -20,7 +20,6 @@
 #include <linux/buffer_head.h>
 #include <linux/blkdev.h>
 #include <linux/random.h>
-#include <linux/version.h>
 #include <asm/div64.h>
 #include "compat.h"
 #include "ctree.h"
@@ -104,10 +103,8 @@ static noinline struct btrfs_device *__find_device(struct list_head *head,
 						   u64 devid, u8 *uuid)
 {
 	struct btrfs_device *dev;
-	struct list_head *cur;
 
-	list_for_each(cur, head) {
-		dev = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(dev, head, dev_list) {
 		if (dev->devid == devid &&
 		    (!uuid || !memcmp(dev->uuid, uuid, BTRFS_UUID_SIZE))) {
 			return dev;
@@ -118,11 +115,9 @@ static noinline struct btrfs_device *__find_device(struct list_head *head,
 
 static noinline struct btrfs_fs_devices *find_fsid(u8 *fsid)
 {
-	struct list_head *cur;
 	struct btrfs_fs_devices *fs_devices;
 
-	list_for_each(cur, &fs_uuids) {
-		fs_devices = list_entry(cur, struct btrfs_fs_devices, list);
+	list_for_each_entry(fs_devices, &fs_uuids, list) {
 		if (memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE) == 0)
 			return fs_devices;
 	}
@@ -345,14 +340,11 @@ error:
 
 int btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices)
 {
-	struct list_head *tmp;
-	struct list_head *cur;
-	struct btrfs_device *device;
+	struct btrfs_device *device, *next;
 
 	mutex_lock(&uuid_mutex);
 again:
-	list_for_each_safe(cur, tmp, &fs_devices->devices) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
 		if (device->in_fs_metadata)
 			continue;
 
@@ -383,14 +375,12 @@ again:
 
 static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 {
-	struct list_head *cur;
 	struct btrfs_device *device;
 
 	if (--fs_devices->opened > 0)
 		return 0;
 
-	list_for_each(cur, &fs_devices->devices) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(device, &fs_devices->devices, dev_list) {
 		if (device->bdev) {
 			close_bdev_exclusive(device->bdev, device->mode);
 			fs_devices->open_devices--;
@@ -439,7 +429,6 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 {
 	struct block_device *bdev;
 	struct list_head *head = &fs_devices->devices;
-	struct list_head *cur;
 	struct btrfs_device *device;
 	struct block_device *latest_bdev = NULL;
 	struct buffer_head *bh;
@@ -450,8 +439,7 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 	int seeding = 1;
 	int ret = 0;
 
-	list_for_each(cur, head) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(device, head, dev_list) {
 		if (device->bdev)
 			continue;
 		if (!device->name)
@@ -578,7 +566,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 		       *(unsigned long long *)disk_super->fsid,
 		       *(unsigned long long *)(disk_super->fsid + 8));
 	}
-	printk(KERN_INFO "devid %llu transid %llu %s\n",
+	printk(KERN_CONT "devid %llu transid %llu %s\n",
 	       (unsigned long long)devid, (unsigned long long)transid, path);
 	ret = device_list_add(path, disk_super, devid, fs_devices_ret);
 
@@ -1017,14 +1005,12 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	}
 
 	if (strcmp(device_path, "missing") == 0) {
-		struct list_head *cur;
 		struct list_head *devices;
 		struct btrfs_device *tmp;
 
 		device = NULL;
 		devices = &root->fs_info->fs_devices->devices;
-		list_for_each(cur, devices) {
-			tmp = list_entry(cur, struct btrfs_device, dev_list);
+		list_for_each_entry(tmp, devices, dev_list) {
 			if (tmp->in_fs_metadata && !tmp->bdev) {
 				device = tmp;
 				break;
@@ -1280,7 +1266,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	struct btrfs_trans_handle *trans;
 	struct btrfs_device *device;
 	struct block_device *bdev;
-	struct list_head *cur;
 	struct list_head *devices;
 	struct super_block *sb = root->fs_info->sb;
 	u64 total_bytes;
@@ -1304,8 +1289,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	mutex_lock(&root->fs_info->volume_mutex);
 
 	devices = &root->fs_info->fs_devices->devices;
-	list_for_each(cur, devices) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(device, devices, dev_list) {
 		if (device->bdev == bdev) {
 			ret = -EEXIST;
 			goto error;
@@ -1704,7 +1688,6 @@ static u64 div_factor(u64 num, int factor)
 int btrfs_balance(struct btrfs_root *dev_root)
 {
 	int ret;
-	struct list_head *cur;
 	struct list_head *devices = &dev_root->fs_info->fs_devices->devices;
 	struct btrfs_device *device;
 	u64 old_size;
@@ -1723,8 +1706,7 @@ int btrfs_balance(struct btrfs_root *dev_root)
 	dev_root = dev_root->fs_info->dev_root;
 
 	/* step one make some room on all the devices */
-	list_for_each(cur, devices) {
-		device = list_entry(cur, struct btrfs_device, dev_list);
+	list_for_each_entry(device, devices, dev_list) {
 		old_size = device->total_bytes;
 		size_to_free = div_factor(old_size, 1);
 		size_to_free = min(size_to_free, (u64)1 * 1024 * 1024);
diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c
index 7f332e270894..b4fa5f4b6ad1 100644
--- a/fs/btrfs/xattr.c
+++ b/fs/btrfs/xattr.c
@@ -45,9 +45,12 @@ ssize_t __btrfs_getxattr(struct inode *inode, const char *name,
 	/* lookup the xattr by name */
 	di = btrfs_lookup_xattr(NULL, root, path, inode->i_ino, name,
 				strlen(name), 0);
-	if (!di || IS_ERR(di)) {
+	if (!di) {
 		ret = -ENODATA;
 		goto out;
+	} else if (IS_ERR(di)) {
+		ret = PTR_ERR(di);
+		goto out;
 	}
 
 	leaf = path->nodes[0];
@@ -62,6 +65,14 @@ ssize_t __btrfs_getxattr(struct inode *inode, const char *name,
 		ret = -ERANGE;
 		goto out;
 	}
+
+	/*
+	 * The way things are packed into the leaf is like this
+	 * |struct btrfs_dir_item|name|data|
+	 * where name is the xattr name, so security.foo, and data is the
+	 * content of the xattr.  data_ptr points to the location in memory
+	 * where the data starts in the in memory leaf
+	 */
 	data_ptr = (unsigned long)((char *)(di + 1) +
 				   btrfs_dir_name_len(leaf, di));
 	read_extent_buffer(leaf, buffer, data_ptr,
@@ -176,7 +187,6 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size)
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 	if (ret < 0)
 		goto err;
-	ret = 0;
 	advance = 0;
 	while (1) {
 		leaf = path->nodes[0];