25 files changed, 202 insertions, 246 deletions

diff --git a/fs/bad_inode.c b/fs/bad_inode.c
index 9ad2369d9e35..bfcb18feb1df 100644
--- a/fs/bad_inode.c
+++ b/fs/bad_inode.c
@@ -231,9 +231,6 @@ static int bad_inode_readlink(struct dentry *dentry, char __user *buffer,
 
 static int bad_inode_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
 	return -EIO;
 }
 
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 378b5b4443f3..300628795fdb 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -967,6 +967,12 @@ struct btrfs_fs_info {
 	struct srcu_struct subvol_srcu;
 
 	spinlock_t trans_lock;
+	/*
+	 * the reloc mutex goes with the trans lock, it is taken
+	 * during commit to protect us from the relocation code
+	 */
+	struct mutex reloc_mutex;
+
 	struct list_head trans_list;
 	struct list_head hashers;
 	struct list_head dead_roots;
@@ -1172,6 +1178,14 @@ struct btrfs_root {
 	u32 type;
 
 	u64 highest_objectid;
+
+	/* btrfs_record_root_in_trans is a multi-step process,
+	 * and it can race with the balancing code.   But the
+	 * race is very small, and only the first time the root
+	 * is added to each transaction.  So in_trans_setup
+	 * is used to tell us when more checks are required
+	 */
+	unsigned long in_trans_setup;
 	int ref_cows;
 	int track_dirty;
 	int in_radix;
@@ -1181,7 +1195,6 @@ struct btrfs_root {
 	struct btrfs_key defrag_max;
 	int defrag_running;
 	char *name;
-	int in_sysfs;
 
 	/* the dirty list is only used by non-reference counted roots */
 	struct list_head dirty_list;
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index 6462c29d2d37..f1cbd028f7b3 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -297,7 +297,6 @@ struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len)
 		item->data_len = data_len;
 		item->ins_or_del = 0;
 		item->bytes_reserved = 0;
-		item->block_rsv = NULL;
 		item->delayed_node = NULL;
 		atomic_set(&item->refs, 1);
 	}
@@ -593,10 +592,8 @@ static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans,
 
 	num_bytes = btrfs_calc_trans_metadata_size(root, 1);
 	ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes);
-	if (!ret) {
+	if (!ret)
 		item->bytes_reserved = num_bytes;
-		item->block_rsv = dst_rsv;
-	}
 
 	return ret;
 }
@@ -604,10 +601,13 @@ static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans,
 static void btrfs_delayed_item_release_metadata(struct btrfs_root *root,
 						struct btrfs_delayed_item *item)
 {
+	struct btrfs_block_rsv *rsv;
+
 	if (!item->bytes_reserved)
 		return;
 
-	btrfs_block_rsv_release(root, item->block_rsv,
+	rsv = &root->fs_info->global_block_rsv;
+	btrfs_block_rsv_release(root, rsv,
 				item->bytes_reserved);
 }
 
@@ -1014,6 +1014,7 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
 	struct btrfs_delayed_root *delayed_root;
 	struct btrfs_delayed_node *curr_node, *prev_node;
 	struct btrfs_path *path;
+	struct btrfs_block_rsv *block_rsv;
 	int ret = 0;
 
 	path = btrfs_alloc_path();
@@ -1021,6 +1022,9 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	path->leave_spinning = 1;
 
+	block_rsv = trans->block_rsv;
+	trans->block_rsv = &root->fs_info->global_block_rsv;
+
 	delayed_root = btrfs_get_delayed_root(root);
 
 	curr_node = btrfs_first_delayed_node(delayed_root);
@@ -1045,6 +1049,7 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
 	}
 
 	btrfs_free_path(path);
+	trans->block_rsv = block_rsv;
 	return ret;
 }
 
@@ -1052,6 +1057,7 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 					      struct btrfs_delayed_node *node)
 {
 	struct btrfs_path *path;
+	struct btrfs_block_rsv *block_rsv;
 	int ret;
 
 	path = btrfs_alloc_path();
@@ -1059,6 +1065,9 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	path->leave_spinning = 1;
 
+	block_rsv = trans->block_rsv;
+	trans->block_rsv = &node->root->fs_info->global_block_rsv;
+
 	ret = btrfs_insert_delayed_items(trans, path, node->root, node);
 	if (!ret)
 		ret = btrfs_delete_delayed_items(trans, path, node->root, node);
@@ -1066,6 +1075,7 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 		ret = btrfs_update_delayed_inode(trans, node->root, path, node);
 	btrfs_free_path(path);
 
+	trans->block_rsv = block_rsv;
 	return ret;
 }
 
@@ -1116,6 +1126,7 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work)
 	struct btrfs_path *path;
 	struct btrfs_delayed_node *delayed_node = NULL;
 	struct btrfs_root *root;
+	struct btrfs_block_rsv *block_rsv;
 	unsigned long nr = 0;
 	int need_requeue = 0;
 	int ret;
@@ -1134,6 +1145,9 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work)
 	if (IS_ERR(trans))
 		goto free_path;
 
+	block_rsv = trans->block_rsv;
+	trans->block_rsv = &root->fs_info->global_block_rsv;
+
 	ret = btrfs_insert_delayed_items(trans, path, root, delayed_node);
 	if (!ret)
 		ret = btrfs_delete_delayed_items(trans, path, root,
@@ -1176,6 +1190,7 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work)
 
 	nr = trans->blocks_used;
 
+	trans->block_rsv = block_rsv;
 	btrfs_end_transaction_dmeta(trans, root);
 	__btrfs_btree_balance_dirty(root, nr);
 free_path:
@@ -1222,6 +1237,13 @@ again:
 	return 0;
 }
 
+void btrfs_assert_delayed_root_empty(struct btrfs_root *root)
+{
+	struct btrfs_delayed_root *delayed_root;
+	delayed_root = btrfs_get_delayed_root(root);
+	WARN_ON(btrfs_first_delayed_node(delayed_root));
+}
+
 void btrfs_balance_delayed_items(struct btrfs_root *root)
 {
 	struct btrfs_delayed_root *delayed_root;
diff --git a/fs/btrfs/delayed-inode.h b/fs/btrfs/delayed-inode.h
index eb7d240aa648..d1a6a2915c66 100644
--- a/fs/btrfs/delayed-inode.h
+++ b/fs/btrfs/delayed-inode.h
@@ -75,7 +75,6 @@ struct btrfs_delayed_item {
 	struct list_head tree_list;	/* used for batch insert/delete items */
 	struct list_head readdir_list;	/* used for readdir items */
 	u64 bytes_reserved;
-	struct btrfs_block_rsv *block_rsv;
 	struct btrfs_delayed_node *delayed_node;
 	atomic_t refs;
 	int ins_or_del;
@@ -138,4 +137,8 @@ int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent,
 /* for init */
 int __init btrfs_delayed_inode_init(void);
 void btrfs_delayed_inode_exit(void);
+
+/* for debugging */
+void btrfs_assert_delayed_root_empty(struct btrfs_root *root);
+
 #endif
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 9f68c6898653..1ac8db5dc0a3 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1044,7 +1044,6 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	root->last_trans = 0;
 	root->highest_objectid = 0;
 	root->name = NULL;
-	root->in_sysfs = 0;
 	root->inode_tree = RB_ROOT;
 	INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC);
 	root->block_rsv = NULL;
@@ -1300,19 +1299,21 @@ again:
 		return root;
 
 	root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
-	if (!root->free_ino_ctl)
-		goto fail;
 	root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
 					GFP_NOFS);
-	if (!root->free_ino_pinned)
+	if (!root->free_ino_pinned || !root->free_ino_ctl) {
+		ret = -ENOMEM;
 		goto fail;
+	}
 
 	btrfs_init_free_ino_ctl(root);
 	mutex_init(&root->fs_commit_mutex);
 	spin_lock_init(&root->cache_lock);
 	init_waitqueue_head(&root->cache_wait);
 
-	set_anon_super(&root->anon_super, NULL);
+	ret = set_anon_super(&root->anon_super, NULL);
+	if (ret)
+		goto fail;
 
 	if (btrfs_root_refs(&root->root_item) == 0) {
 		ret = -ENOENT;
@@ -1618,6 +1619,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	spin_lock_init(&fs_info->fs_roots_radix_lock);
 	spin_lock_init(&fs_info->delayed_iput_lock);
 	spin_lock_init(&fs_info->defrag_inodes_lock);
+	mutex_init(&fs_info->reloc_mutex);
 
 	init_completion(&fs_info->kobj_unregister);
 	fs_info->tree_root = tree_root;
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index b42efc2ded51..1f61bf5b4960 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3314,10 +3314,6 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 	if (reserved == 0)
 		return 0;
 
-	/* nothing to shrink - nothing to reclaim */
-	if (root->fs_info->delalloc_bytes == 0)
-		return 0;
-
 	max_reclaim = min(reserved, to_reclaim);
 
 	while (loops < 1024) {
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 751ddf8fc58a..0a9b10c5b0a7 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -3076,6 +3076,7 @@ int btrfs_unlink_subvol(struct btrfs_trans_handle *trans,
 	ret = btrfs_update_inode(trans, root, dir);
 	BUG_ON(ret);
 
+	btrfs_free_path(path);
 	return 0;
 }
 
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index b793d112d1f6..a3c4751e07db 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -482,8 +482,10 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 	ret = btrfs_snap_reserve_metadata(trans, pending_snapshot);
 	BUG_ON(ret);
 
+	spin_lock(&root->fs_info->trans_lock);
 	list_add(&pending_snapshot->list,
 		 &trans->transaction->pending_snapshots);
+	spin_unlock(&root->fs_info->trans_lock);
 	if (async_transid) {
 		*async_transid = trans->transid;
 		ret = btrfs_commit_transaction_async(trans,
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index b1ef27cc673b..5e0a3dc79a45 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -1368,7 +1368,7 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans,
 	int ret;
 
 	if (!root->reloc_root)
-		return 0;
+		goto out;
 
 	reloc_root = root->reloc_root;
 	root_item = &reloc_root->root_item;
@@ -1390,6 +1390,8 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans,
 	ret = btrfs_update_root(trans, root->fs_info->tree_root,
 				&reloc_root->root_key, root_item);
 	BUG_ON(ret);
+
+out:
 	return 0;
 }
 
@@ -2142,10 +2144,11 @@ int prepare_to_merge(struct reloc_control *rc, int err)
 	u64 num_bytes = 0;
 	int ret;
 
-	spin_lock(&root->fs_info->trans_lock);
+	mutex_lock(&root->fs_info->reloc_mutex);
 	rc->merging_rsv_size += root->nodesize * (BTRFS_MAX_LEVEL - 1) * 2;
 	rc->merging_rsv_size += rc->nodes_relocated * 2;
-	spin_unlock(&root->fs_info->trans_lock);
+	mutex_unlock(&root->fs_info->reloc_mutex);
+
 again:
 	if (!err) {
 		num_bytes = rc->merging_rsv_size;
@@ -2214,9 +2217,16 @@ int merge_reloc_roots(struct reloc_control *rc)
 	int ret;
 again:
 	root = rc->extent_root;
-	spin_lock(&root->fs_info->trans_lock);
+
+	/*
+	 * this serializes us with btrfs_record_root_in_transaction,
+	 * we have to make sure nobody is in the middle of
+	 * adding their roots to the list while we are
+	 * doing this splice
+	 */
+	mutex_lock(&root->fs_info->reloc_mutex);
 	list_splice_init(&rc->reloc_roots, &reloc_roots);
-	spin_unlock(&root->fs_info->trans_lock);
+	mutex_unlock(&root->fs_info->reloc_mutex);
 
 	while (!list_empty(&reloc_roots)) {
 		found = 1;
@@ -3590,17 +3600,19 @@ next:
 static void set_reloc_control(struct reloc_control *rc)
 {
 	struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
-	spin_lock(&fs_info->trans_lock);
+
+	mutex_lock(&fs_info->reloc_mutex);
 	fs_info->reloc_ctl = rc;
-	spin_unlock(&fs_info->trans_lock);
+	mutex_unlock(&fs_info->reloc_mutex);
 }
 
 static void unset_reloc_control(struct reloc_control *rc)
 {
 	struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
-	spin_lock(&fs_info->trans_lock);
+
+	mutex_lock(&fs_info->reloc_mutex);
 	fs_info->reloc_ctl = NULL;
-	spin_unlock(&fs_info->trans_lock);
+	mutex_unlock(&fs_info->reloc_mutex);
 }
 
 static int check_extent_flags(u64 flags)
diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c
index c3c223ae6691..daac9ae6d731 100644
--- a/fs/btrfs/sysfs.c
+++ b/fs/btrfs/sysfs.c
@@ -28,152 +28,6 @@
 #include "disk-io.h"
 #include "transaction.h"
 
-static ssize_t root_blocks_used_show(struct btrfs_root *root, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_root_used(&root->root_item));
-}
-
-static ssize_t root_block_limit_show(struct btrfs_root *root, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_root_limit(&root->root_item));
-}
-
-static ssize_t super_blocks_used_show(struct btrfs_fs_info *fs, char *buf)
-{
-
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_super_bytes_used(&fs->super_copy));
-}
-
-static ssize_t super_total_blocks_show(struct btrfs_fs_info *fs, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_super_total_bytes(&fs->super_copy));
-}
-
-static ssize_t super_blocksize_show(struct btrfs_fs_info *fs, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_super_sectorsize(&fs->super_copy));
-}
-
-/* this is for root attrs (subvols/snapshots) */
-struct btrfs_root_attr {
-	struct attribute attr;
-	ssize_t (*show)(struct btrfs_root *, char *);
-	ssize_t (*store)(struct btrfs_root *, const char *, size_t);
-};
-
-#define ROOT_ATTR(name, mode, show, store) \
-static struct btrfs_root_attr btrfs_root_attr_##name = __ATTR(name, mode, \
-							      show, store)
-
-ROOT_ATTR(blocks_used,	0444,	root_blocks_used_show,	NULL);
-ROOT_ATTR(block_limit,	0644,	root_block_limit_show,	NULL);
-
-static struct attribute *btrfs_root_attrs[] = {
-	&btrfs_root_attr_blocks_used.attr,
-	&btrfs_root_attr_block_limit.attr,
-	NULL,
-};
-
-/* this is for super attrs (actual full fs) */
-struct btrfs_super_attr {
-	struct attribute attr;
-	ssize_t (*show)(struct btrfs_fs_info *, char *);
-	ssize_t (*store)(struct btrfs_fs_info *, const char *, size_t);
-};
-
-#define SUPER_ATTR(name, mode, show, store) \
-static struct btrfs_super_attr btrfs_super_attr_##name = __ATTR(name, mode, \
-								show, store)
-
-SUPER_ATTR(blocks_used,		0444,	super_blocks_used_show,		NULL);
-SUPER_ATTR(total_blocks,	0444,	super_total_blocks_show,	NULL);
-SUPER_ATTR(blocksize,		0444,	super_blocksize_show,		NULL);
-
-static struct attribute *btrfs_super_attrs[] = {
-	&btrfs_super_attr_blocks_used.attr,
-	&btrfs_super_attr_total_blocks.attr,
-	&btrfs_super_attr_blocksize.attr,
-	NULL,
-};
-
-static ssize_t btrfs_super_attr_show(struct kobject *kobj,
-				    struct attribute *attr, char *buf)
-{
-	struct btrfs_fs_info *fs = container_of(kobj, struct btrfs_fs_info,
-						super_kobj);
-	struct btrfs_super_attr *a = container_of(attr,
-						  struct btrfs_super_attr,
-						  attr);
-
-	return a->show ? a->show(fs, buf) : 0;
-}
-
-static ssize_t btrfs_super_attr_store(struct kobject *kobj,
-				     struct attribute *attr,
-				     const char *buf, size_t len)
-{
-	struct btrfs_fs_info *fs = container_of(kobj, struct btrfs_fs_info,
-						super_kobj);
-	struct btrfs_super_attr *a = container_of(attr,
-						  struct btrfs_super_attr,
-						  attr);
-
-	return a->store ? a->store(fs, buf, len) : 0;
-}
-
-static ssize_t btrfs_root_attr_show(struct kobject *kobj,
-				    struct attribute *attr, char *buf)
-{
-	struct btrfs_root *root = container_of(kobj, struct btrfs_root,
-						root_kobj);
-	struct btrfs_root_attr *a = container_of(attr,
-						 struct btrfs_root_attr,
-						 attr);
-
-	return a->show ? a->show(root, buf) : 0;
-}
-
-static ssize_t btrfs_root_attr_store(struct kobject *kobj,
-				     struct attribute *attr,
-				     const char *buf, size_t len)
-{
-	struct btrfs_root *root = container_of(kobj, struct btrfs_root,
-						root_kobj);
-	struct btrfs_root_attr *a = container_of(attr,
-						 struct btrfs_root_attr,
-						 attr);
-	return a->store ? a->store(root, buf, len) : 0;
-}
-
-static void btrfs_super_release(struct kobject *kobj)
-{
-	struct btrfs_fs_info *fs = container_of(kobj, struct btrfs_fs_info,
-						super_kobj);
-	complete(&fs->kobj_unregister);
-}
-
-static void btrfs_root_release(struct kobject *kobj)
-{
-	struct btrfs_root *root = container_of(kobj, struct btrfs_root,
-						root_kobj);
-	complete(&root->kobj_unregister);
-}
-
-static const struct sysfs_ops btrfs_super_attr_ops = {
-	.show	= btrfs_super_attr_show,
-	.store	= btrfs_super_attr_store,
-};
-
-static const struct sysfs_ops btrfs_root_attr_ops = {
-	.show	= btrfs_root_attr_show,
-	.store	= btrfs_root_attr_store,
-};
-
 /* /sys/fs/btrfs/ entry */
 static struct kset *btrfs_kset;
 
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 2b3590b9fe98..51dcec86757f 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -126,28 +126,85 @@ static noinline int join_transaction(struct btrfs_root *root, int nofail)
  * to make sure the old root from before we joined the transaction is deleted
  * when the transaction commits
  */
-int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans,
+static int record_root_in_trans(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root)
 {
 	if (root->ref_cows && root->last_trans < trans->transid) {
 		WARN_ON(root == root->fs_info->extent_root);
 		WARN_ON(root->commit_root != root->node);
 
+		/*
+		 * see below for in_trans_setup usage rules
+		 * we have the reloc mutex held now, so there
+		 * is only one writer in this function
+		 */
+		root->in_trans_setup = 1;
+
+		/* make sure readers find in_trans_setup before
+		 * they find our root->last_trans update
+		 */
+		smp_wmb();
+
 		spin_lock(&root->fs_info->fs_roots_radix_lock);
 		if (root->last_trans == trans->transid) {
 			spin_unlock(&root->fs_info->fs_roots_radix_lock);
 			return 0;
 		}
-		root->last_trans = trans->transid;
 		radix_tree_tag_set(&root->fs_info->fs_roots_radix,
 			   (unsigned long)root->root_key.objectid,
 			   BTRFS_ROOT_TRANS_TAG);
 		spin_unlock(&root->fs_info->fs_roots_radix_lock);
+		root->last_trans = trans->transid;
+
+		/* this is pretty tricky.  We don't want to
+		 * take the relocation lock in btrfs_record_root_in_trans
+		 * unless we're really doing the first setup for this root in
+		 * this transaction.
+		 *
+		 * Normally we'd use root->last_trans as a flag to decide
+		 * if we want to take the expensive mutex.
+		 *
+		 * But, we have to set root->last_trans before we
+		 * init the relocation root, otherwise, we trip over warnings
+		 * in ctree.c.  The solution used here is to flag ourselves
+		 * with root->in_trans_setup.  When this is 1, we're still
+		 * fixing up the reloc trees and everyone must wait.
+		 *
+		 * When this is zero, they can trust root->last_trans and fly
+		 * through btrfs_record_root_in_trans without having to take the
+		 * lock.  smp_wmb() makes sure that all the writes above are
+		 * done before we pop in the zero below
+		 */
 		btrfs_init_reloc_root(trans, root);
+		smp_wmb();
+		root->in_trans_setup = 0;
 	}
 	return 0;
 }
 
+
+int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *root)
+{
+	if (!root->ref_cows)
+		return 0;
+
+	/*
+	 * see record_root_in_trans for comments about in_trans_setup usage
+	 * and barriers
+	 */
+	smp_rmb();
+	if (root->last_trans == trans->transid &&
+	    !root->in_trans_setup)
+		return 0;
+
+	mutex_lock(&root->fs_info->reloc_mutex);
+	record_root_in_trans(trans, root);
+	mutex_unlock(&root->fs_info->reloc_mutex);
+
+	return 0;
+}
+
 /* wait for commit against the current transaction to become unblocked
  * when this is done, it is safe to start a new transaction, but the current
  * transaction might not be fully on disk.
@@ -882,7 +939,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	parent = dget_parent(dentry);
 	parent_inode = parent->d_inode;
 	parent_root = BTRFS_I(parent_inode)->root;
-	btrfs_record_root_in_trans(trans, parent_root);
+	record_root_in_trans(trans, parent_root);
 
 	/*
 	 * insert the directory item
@@ -900,7 +957,16 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	ret = btrfs_update_inode(trans, parent_root, parent_inode);
 	BUG_ON(ret);
 
-	btrfs_record_root_in_trans(trans, root);
+	/*
+	 * pull in the delayed directory update
+	 * and the delayed inode item
+	 * otherwise we corrupt the FS during
+	 * snapshot
+	 */
+	ret = btrfs_run_delayed_items(trans, root);
+	BUG_ON(ret);
+
+	record_root_in_trans(trans, root);
 	btrfs_set_root_last_snapshot(&root->root_item, trans->transid);
 	memcpy(new_root_item, &root->root_item, sizeof(*new_root_item));
 	btrfs_check_and_init_root_item(new_root_item);
@@ -961,14 +1027,6 @@ static noinline int create_pending_snapshots(struct btrfs_trans_handle *trans,
 	int ret;
 
 	list_for_each_entry(pending, head, list) {
-		/*
-		 * We must deal with the delayed items before creating
-		 * snapshots, or we will create a snapthot with inconsistent
-		 * information.
-		*/
-		ret = btrfs_run_delayed_items(trans, fs_info->fs_root);
-		BUG_ON(ret);
-
 		ret = create_pending_snapshot(trans, fs_info, pending);
 		BUG_ON(ret);
 	}
@@ -1241,21 +1299,42 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 			schedule_timeout(1);
 
 		finish_wait(&cur_trans->writer_wait, &wait);
-		spin_lock(&root->fs_info->trans_lock);
-		root->fs_info->trans_no_join = 1;
-		spin_unlock(&root->fs_info->trans_lock);
 	} while (atomic_read(&cur_trans->num_writers) > 1 ||
 		 (should_grow && cur_trans->num_joined != joined));
 
-	ret = create_pending_snapshots(trans, root->fs_info);
-	BUG_ON(ret);
+	/*
+	 * Ok now we need to make sure to block out any other joins while we
+	 * commit the transaction.  We could have started a join before setting
+	 * no_join so make sure to wait for num_writers to == 1 again.
+	 */
+	spin_lock(&root->fs_info->trans_lock);
+	root->fs_info->trans_no_join = 1;
+	spin_unlock(&root->fs_info->trans_lock);
+	wait_event(cur_trans->writer_wait,
+		   atomic_read(&cur_trans->num_writers) == 1);
+
+	/*
+	 * the reloc mutex makes sure that we stop
+	 * the balancing code from coming in and moving
+	 * extents around in the middle of the commit
+	 */
+	mutex_lock(&root->fs_info->reloc_mutex);
 
 	ret = btrfs_run_delayed_items(trans, root);
 	BUG_ON(ret);
 
+	ret = create_pending_snapshots(trans, root->fs_info);
+	BUG_ON(ret);
+
 	ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
 	BUG_ON(ret);
 
+	/*
+	 * make sure none of the code above managed to slip in a
+	 * delayed item
+	 */
+	btrfs_assert_delayed_root_empty(root);
+
 	WARN_ON(cur_trans != trans->transaction);
 
 	btrfs_scrub_pause(root);
@@ -1312,6 +1391,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	root->fs_info->running_transaction = NULL;
 	root->fs_info->trans_no_join = 0;
 	spin_unlock(&root->fs_info->trans_lock);
+	mutex_unlock(&root->fs_info->reloc_mutex);
 
 	wake_up(&root->fs_info->transaction_wait);
 
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 592396c6dc47..4ce8a9f41d1e 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3177,7 +3177,7 @@ again:
 		tmp_key.offset = (u64)-1;
 
 		wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key);
-		BUG_ON(!wc.replay_dest);
+		BUG_ON(IS_ERR_OR_NULL(wc.replay_dest));
 
 		wc.replay_dest->log_root = log;
 		btrfs_record_root_in_trans(trans, wc.replay_dest);
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index e9def996e383..2f0c58646c10 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -257,9 +257,6 @@ static int cifs_permission(struct inode *inode, int mask, unsigned int flags)
 {
 	struct cifs_sb_info *cifs_sb;
 
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
 	cifs_sb = CIFS_SB(inode->i_sb);
 
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM) {
diff --git a/fs/coda/pioctl.c b/fs/coda/pioctl.c
index 6cbb3afb36dc..cb140ef293e4 100644
--- a/fs/coda/pioctl.c
+++ b/fs/coda/pioctl.c
@@ -43,8 +43,6 @@ const struct file_operations coda_ioctl_operations = {
 /* the coda pioctl inode ops */
 static int coda_ioctl_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
 	return (mask & MAY_EXEC) ? -EACCES : 0;
 }
 
diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c
index 9ed89d1663f8..1afae26cf236 100644
--- a/fs/logfs/dir.c
+++ b/fs/logfs/dir.c
@@ -555,13 +555,6 @@ static int logfs_symlink(struct inode *dir, struct dentry *dentry,
 	return __logfs_create(dir, dentry, inode, target, destlen);
 }
 
-static int logfs_permission(struct inode *inode, int mask, unsigned int flags)
-{
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-	return generic_permission(inode, mask, flags, NULL);
-}
-
 static int logfs_link(struct dentry *old_dentry, struct inode *dir,
 		struct dentry *dentry)
 {
@@ -820,7 +813,6 @@ const struct inode_operations logfs_dir_iops = {
 	.mknod		= logfs_mknod,
 	.rename		= logfs_rename,
 	.rmdir		= logfs_rmdir,
-	.permission	= logfs_permission,
 	.symlink	= logfs_symlink,
 	.unlink		= logfs_unlink,
 };
diff --git a/fs/namei.c b/fs/namei.c
index 9e425e7e6c8f..0223c41fb114 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -238,7 +238,8 @@ int generic_permission(struct inode *inode, int mask, unsigned int flags,
 
 	/*
 	 * Read/write DACs are always overridable.
-	 * Executable DACs are overridable if at least one exec bit is set.
+	 * Executable DACs are overridable for all directories and
+	 * for non-directories that have least one exec bit set.
 	 */
 	if (!(mask & MAY_EXEC) || execute_ok(inode))
 		if (ns_capable(inode_userns(inode), CAP_DAC_OVERRIDE))
@@ -1011,9 +1012,6 @@ failed:
  * Follow down to the covering mount currently visible to userspace.  At each
  * point, the filesystem owning that dentry may be queried as to whether the
  * caller is permitted to proceed or not.
- *
- * Care must be taken as namespace_sem may be held (indicated by mounting_here
- * being true).
  */
 int follow_down(struct path *path)
 {
diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index 18b3e8975fe0..fbb2a5ef5817 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -82,6 +82,7 @@ config NFSD_V4
 	select NFSD_V3
 	select FS_POSIX_ACL
 	select SUNRPC_GSS
+	select CRYPTO
 	help
 	  This option enables support in your system's NFS server for
 	  version 4 of the NFS protocol (RFC 3530).
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 1f5eae40f34e..2b1449dd2f49 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -13,6 +13,7 @@
 #include <linux/lockd/lockd.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/gss_api.h>
+#include <linux/sunrpc/gss_krb5_enctypes.h>
 
 #include "idmap.h"
 #include "nfsd.h"
@@ -189,18 +190,10 @@ static struct file_operations export_features_operations = {
 	.release	= single_release,
 };
 
-#ifdef CONFIG_SUNRPC_GSS
+#if defined(CONFIG_SUNRPC_GSS) || defined(CONFIG_SUNRPC_GSS_MODULE)
 static int supported_enctypes_show(struct seq_file *m, void *v)
 {
-	struct gss_api_mech *k5mech;
-
-	k5mech = gss_mech_get_by_name("krb5");
-	if (k5mech == NULL)
-		goto out;
-	if (k5mech->gm_upcall_enctypes != NULL)
-		seq_printf(m, k5mech->gm_upcall_enctypes);
-	gss_mech_put(k5mech);
-out:
+	seq_printf(m, KRB5_SUPPORTED_ENCTYPES);
 	return 0;
 }
 
@@ -215,7 +208,7 @@ static struct file_operations supported_enctypes_ops = {
 	.llseek		= seq_lseek,
 	.release	= single_release,
 };
-#endif /* CONFIG_SUNRPC_GSS */
+#endif /* CONFIG_SUNRPC_GSS or CONFIG_SUNRPC_GSS_MODULE */
 
 extern int nfsd_pool_stats_open(struct inode *inode, struct file *file);
 extern int nfsd_pool_stats_release(struct inode *inode, struct file *file);
@@ -1427,9 +1420,9 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 		[NFSD_Versions] = {"versions", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Ports] = {"portlist", &transaction_ops, S_IWUSR|S_IRUGO},
 		[NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO},
-#ifdef CONFIG_SUNRPC_GSS
+#if defined(CONFIG_SUNRPC_GSS) || defined(CONFIG_SUNRPC_GSS_MODULE)
 		[NFSD_SupportedEnctypes] = {"supported_krb5_enctypes", &supported_enctypes_ops, S_IRUGO},
-#endif /* CONFIG_SUNRPC_GSS */
+#endif /* CONFIG_SUNRPC_GSS or CONFIG_SUNRPC_GSS_MODULE */
 #ifdef CONFIG_NFSD_V4
 		[NFSD_Leasetime] = {"nfsv4leasetime", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Gracetime] = {"nfsv4gracetime", &transaction_ops, S_IWUSR|S_IRUSR},
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index d5718273bb32..fd0acca5370a 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -696,7 +696,15 @@ nfsd_access(struct svc_rqst *rqstp, struct svc_fh *fhp, u32 *access, u32 *suppor
 }
 #endif /* CONFIG_NFSD_V3 */
 
+static int nfsd_open_break_lease(struct inode *inode, int access)
+{
+	unsigned int mode;
 
+	if (access & NFSD_MAY_NOT_BREAK_LEASE)
+		return 0;
+	mode = (access & NFSD_MAY_WRITE) ? O_WRONLY : O_RDONLY;
+	return break_lease(inode, mode | O_NONBLOCK);
+}
 
 /*
  * Open an existing file or directory.
@@ -744,12 +752,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	if (!inode->i_fop)
 		goto out;
 
-	/*
-	 * Check to see if there are any leases on this file.
-	 * This may block while leases are broken.
-	 */
-	if (!(access & NFSD_MAY_NOT_BREAK_LEASE))
-		host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0));
+	host_err = nfsd_open_break_lease(inode, access);
 	if (host_err) /* NOMEM or WOULDBLOCK */
 		goto out_nfserr;
 
@@ -1660,8 +1663,10 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 	if (!dold->d_inode)
 		goto out_drop_write;
 	host_err = nfsd_break_lease(dold->d_inode);
-	if (host_err)
+	if (host_err) {
+		err = nfserrno(host_err);
 		goto out_drop_write;
+	}
 	host_err = vfs_link(dold, dirp, dnew);
 	if (!host_err) {
 		err = nfserrno(commit_metadata(ffhp));
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index b954878ad6ce..b9b45fc2903e 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -801,12 +801,7 @@ out_err:
 
 int nilfs_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	struct nilfs_root *root;
-
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
-	root = NILFS_I(inode)->i_root;
+	struct nilfs_root *root = NILFS_I(inode)->i_root;
 	if ((mask & MAY_WRITE) && root &&
 	    root->cno != NILFS_CPTREE_CURRENT_CNO)
 		return -EROFS; /* snapshot is not writable */
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 14def991d9dd..8a84210ca080 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2169,11 +2169,7 @@ static const struct file_operations proc_fd_operations = {
  */
 static int proc_fd_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	int rv;
-
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-	rv = generic_permission(inode, mask, flags, NULL);
+	int rv = generic_permission(inode, mask, flags, NULL);
 	if (rv == 0)
 		return 0;
 	if (task_pid(current) == proc_pid(inode))
diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index f50133c11c24..d167de365a8d 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -304,9 +304,6 @@ static int proc_sys_permission(struct inode *inode, int mask,unsigned int flags)
 	struct ctl_table *table;
 	int error;
 
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
 	/* Executable files are not allowed under /proc/sys/ */
 	if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode))
 		return -EACCES;
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index e8a62f41b458..d78089690965 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -954,8 +954,6 @@ static int xattr_mount_check(struct super_block *s)
 
 int reiserfs_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
 	/*
 	 * We don't do permission checks on the internal objects.
 	 * Permissions are determined by the "owning" object.
diff --git a/fs/timerfd.c b/fs/timerfd.c
index f67acbdda5e8..dffeb3795af1 100644
--- a/fs/timerfd.c
+++ b/fs/timerfd.c
@@ -61,7 +61,9 @@ static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr)
 
 /*
  * Called when the clock was set to cancel the timers in the cancel
- * list.
+ * list. This will wake up processes waiting on these timers. The
+ * wake-up requires ctx->ticks to be non zero, therefore we increment
+ * it before calling wake_up_locked().
  */
 void timerfd_clock_was_set(void)
 {
@@ -76,6 +78,7 @@ void timerfd_clock_was_set(void)
 		spin_lock_irqsave(&ctx->wqh.lock, flags);
 		if (ctx->moffs.tv64 != moffs.tv64) {
 			ctx->moffs.tv64 = KTIME_MAX;
+			ctx->ticks++;
 			wake_up_locked(&ctx->wqh);
 		}
 		spin_unlock_irqrestore(&ctx->wqh.lock, flags);
diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index 8c892c2d5300..529be0582029 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -2146,6 +2146,7 @@ static struct dentry *ubifs_mount(struct file_system_type *fs_type, int flags,
 	if (IS_ERR(sb)) {
 		err = PTR_ERR(sb);
 		kfree(c);
+		goto out_close;
 	}
 
 	if (sb->s_root) {