Merge commit 'v3.2-rc5' into sched/core

Merge reason: Pick up the latest fixes.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

91 files changed, 2076 insertions, 1205 deletions

diff --git a/fs/bio.c b/fs/bio.c
index 41c93c72224..b1fe82cf88c 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -337,7 +337,7 @@ static void bio_fs_destructor(struct bio *bio)
  *	RETURNS:
  *	Pointer to new bio on success, NULL on failure.
  */
-struct bio *bio_alloc(gfp_t gfp_mask, int nr_iovecs)
+struct bio *bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs)
 {
 	struct bio *bio = bio_alloc_bioset(gfp_mask, nr_iovecs, fs_bio_set);
 
@@ -365,7 +365,7 @@ static void bio_kmalloc_destructor(struct bio *bio)
  *   %__GFP_WAIT, the allocation is guaranteed to succeed.
  *
  **/
-struct bio *bio_kmalloc(gfp_t gfp_mask, int nr_iovecs)
+struct bio *bio_kmalloc(gfp_t gfp_mask, unsigned int nr_iovecs)
 {
 	struct bio *bio;
 
@@ -696,7 +696,8 @@ static void bio_free_map_data(struct bio_map_data *bmd)
 	kfree(bmd);
 }
 
-static struct bio_map_data *bio_alloc_map_data(int nr_segs, int iov_count,
+static struct bio_map_data *bio_alloc_map_data(int nr_segs,
+					       unsigned int iov_count,
 					       gfp_t gfp_mask)
 {
 	struct bio_map_data *bmd;
diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index 8855aad3929..22c64fff1bd 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -683,7 +683,7 @@ static int inode_to_path(u64 inum, struct btrfs_inode_ref *iref,
 		return PTR_ERR(fspath);
 
 	if (fspath > fspath_min) {
-		ipath->fspath->val[i] = (u64)fspath;
+		ipath->fspath->val[i] = (u64)(unsigned long)fspath;
 		++ipath->fspath->elem_cnt;
 		ipath->fspath->bytes_left = fspath - fspath_min;
 	} else {
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 0fe615e4ea3..dede441bdee 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -514,10 +514,25 @@ static inline int should_cow_block(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   struct extent_buffer *buf)
 {
+	/* ensure we can see the force_cow */
+	smp_rmb();
+
+	/*
+	 * We do not need to cow a block if
+	 * 1) this block is not created or changed in this transaction;
+	 * 2) this block does not belong to TREE_RELOC tree;
+	 * 3) the root is not forced COW.
+	 *
+	 * What is forced COW:
+	 *    when we create snapshot during commiting the transaction,
+	 *    after we've finished coping src root, we must COW the shared
+	 *    block to ensure the metadata consistency.
+	 */
 	if (btrfs_header_generation(buf) == trans->transid &&
 	    !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN) &&
 	    !(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID &&
-	      btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)))
+	      btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)) &&
+	    !root->force_cow)
 		return 0;
 	return 1;
 }
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index b9ba59ff929..50634abef9b 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -848,7 +848,8 @@ struct btrfs_free_cluster {
 enum btrfs_caching_type {
 	BTRFS_CACHE_NO		= 0,
 	BTRFS_CACHE_STARTED	= 1,
-	BTRFS_CACHE_FINISHED	= 2,
+	BTRFS_CACHE_FAST	= 2,
+	BTRFS_CACHE_FINISHED	= 3,
 };
 
 enum btrfs_disk_cache_state {
@@ -1271,6 +1272,8 @@ struct btrfs_root {
 	 * for stat.  It may be used for more later
 	 */
 	dev_t anon_dev;
+
+	int force_cow;
 };
 
 struct btrfs_ioctl_defrag_range_args {
@@ -2366,6 +2369,9 @@ int btrfs_block_rsv_check(struct btrfs_root *root,
 int btrfs_block_rsv_refill(struct btrfs_root *root,
 			  struct btrfs_block_rsv *block_rsv,
 			  u64 min_reserved);
+int btrfs_block_rsv_refill_noflush(struct btrfs_root *root,
+				   struct btrfs_block_rsv *block_rsv,
+				   u64 min_reserved);
 int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv,
 			    struct btrfs_block_rsv *dst_rsv,
 			    u64 num_bytes);
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 62afe5c5694..632f8f3cc9d 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -620,7 +620,7 @@ out:
 
 static int btree_io_failed_hook(struct bio *failed_bio,
 			 struct page *page, u64 start, u64 end,
-			 u64 mirror_num, struct extent_state *state)
+			 int mirror_num, struct extent_state *state)
 {
 	struct extent_io_tree *tree;
 	unsigned long len;
@@ -2573,22 +2573,10 @@ static int write_dev_supers(struct btrfs_device *device,
 	int errors = 0;
 	u32 crc;
 	u64 bytenr;
-	int last_barrier = 0;
 
 	if (max_mirrors == 0)
 		max_mirrors = BTRFS_SUPER_MIRROR_MAX;
 
-	/* make sure only the last submit_bh does a barrier */
-	if (do_barriers) {
-		for (i = 0; i < max_mirrors; i++) {
-			bytenr = btrfs_sb_offset(i);
-			if (bytenr + BTRFS_SUPER_INFO_SIZE >=
-			    device->total_bytes)
-				break;
-			last_barrier = i;
-		}
-	}
-
 	for (i = 0; i < max_mirrors; i++) {
 		bytenr = btrfs_sb_offset(i);
 		if (bytenr + BTRFS_SUPER_INFO_SIZE >= device->total_bytes)
@@ -2634,17 +2622,136 @@ static int write_dev_supers(struct btrfs_device *device,
 			bh->b_end_io = btrfs_end_buffer_write_sync;
 		}
 
-		if (i == last_barrier && do_barriers)
-			ret = submit_bh(WRITE_FLUSH_FUA, bh);
-		else
-			ret = submit_bh(WRITE_SYNC, bh);
-
+		/*
+		 * we fua the first super.  The others we allow
+		 * to go down lazy.
+		 */
+		ret = submit_bh(WRITE_FUA, bh);
 		if (ret)
 			errors++;
 	}
 	return errors < i ? 0 : -1;
 }
 
+/*
+ * endio for the write_dev_flush, this will wake anyone waiting
+ * for the barrier when it is done
+ */
+static void btrfs_end_empty_barrier(struct bio *bio, int err)
+{
+	if (err) {
+		if (err == -EOPNOTSUPP)
+			set_bit(BIO_EOPNOTSUPP, &bio->bi_flags);
+		clear_bit(BIO_UPTODATE, &bio->bi_flags);
+	}
+	if (bio->bi_private)
+		complete(bio->bi_private);
+	bio_put(bio);
+}
+
+/*
+ * trigger flushes for one the devices.  If you pass wait == 0, the flushes are
+ * sent down.  With wait == 1, it waits for the previous flush.
+ *
+ * any device where the flush fails with eopnotsupp are flagged as not-barrier
+ * capable
+ */
+static int write_dev_flush(struct btrfs_device *device, int wait)
+{
+	struct bio *bio;
+	int ret = 0;
+
+	if (device->nobarriers)
+		return 0;
+
+	if (wait) {
+		bio = device->flush_bio;
+		if (!bio)
+			return 0;
+
+		wait_for_completion(&device->flush_wait);
+
+		if (bio_flagged(bio, BIO_EOPNOTSUPP)) {
+			printk("btrfs: disabling barriers on dev %s\n",
+			       device->name);
+			device->nobarriers = 1;
+		}
+		if (!bio_flagged(bio, BIO_UPTODATE)) {
+			ret = -EIO;
+		}
+
+		/* drop the reference from the wait == 0 run */
+		bio_put(bio);
+		device->flush_bio = NULL;
+
+		return ret;
+	}
+
+	/*
+	 * one reference for us, and we leave it for the
+	 * caller
+	 */
+	device->flush_bio = NULL;;
+	bio = bio_alloc(GFP_NOFS, 0);
+	if (!bio)
+		return -ENOMEM;
+
+	bio->bi_end_io = btrfs_end_empty_barrier;
+	bio->bi_bdev = device->bdev;
+	init_completion(&device->flush_wait);
+	bio->bi_private = &device->flush_wait;
+	device->flush_bio = bio;
+
+	bio_get(bio);
+	submit_bio(WRITE_FLUSH, bio);
+
+	return 0;
+}
+
+/*
+ * send an empty flush down to each device in parallel,
+ * then wait for them
+ */
+static int barrier_all_devices(struct btrfs_fs_info *info)
+{
+	struct list_head *head;
+	struct btrfs_device *dev;
+	int errors = 0;
+	int ret;
+
+	/* send down all the barriers */
+	head = &info->fs_devices->devices;
+	list_for_each_entry_rcu(dev, head, dev_list) {
+		if (!dev->bdev) {
+			errors++;
+			continue;
+		}
+		if (!dev->in_fs_metadata || !dev->writeable)
+			continue;
+
+		ret = write_dev_flush(dev, 0);
+		if (ret)
+			errors++;
+	}
+
+	/* wait for all the barriers */
+	list_for_each_entry_rcu(dev, head, dev_list) {
+		if (!dev->bdev) {
+			errors++;
+			continue;
+		}
+		if (!dev->in_fs_metadata || !dev->writeable)
+			continue;
+
+		ret = write_dev_flush(dev, 1);
+		if (ret)
+			errors++;
+	}
+	if (errors)
+		return -EIO;
+	return 0;
+}
+
 int write_all_supers(struct btrfs_root *root, int max_mirrors)
 {
 	struct list_head *head;
@@ -2666,6 +2773,10 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	head = &root->fs_info->fs_devices->devices;
+
+	if (do_barriers)
+		barrier_all_devices(root->fs_info);
+
 	list_for_each_entry_rcu(dev, head, dev_list) {
 		if (!dev->bdev) {
 			total_errors++;
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index b232150b5b6..2ad813674d7 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -467,13 +467,59 @@ static int cache_block_group(struct btrfs_block_group_cache *cache,
 			     struct btrfs_root *root,
 			     int load_cache_only)
 {
+	DEFINE_WAIT(wait);
 	struct btrfs_fs_info *fs_info = cache->fs_info;
 	struct btrfs_caching_control *caching_ctl;
 	int ret = 0;
 
-	smp_mb();
-	if (cache->cached != BTRFS_CACHE_NO)
+	caching_ctl = kzalloc(sizeof(*caching_ctl), GFP_NOFS);
+	BUG_ON(!caching_ctl);
+
+	INIT_LIST_HEAD(&caching_ctl->list);
+	mutex_init(&caching_ctl->mutex);
+	init_waitqueue_head(&caching_ctl->wait);
+	caching_ctl->block_group = cache;
+	caching_ctl->progress = cache->key.objectid;
+	atomic_set(&caching_ctl->count, 1);
+	caching_ctl->work.func = caching_thread;
+
+	spin_lock(&cache->lock);
+	/*
+	 * This should be a rare occasion, but this could happen I think in the
+	 * case where one thread starts to load the space cache info, and then
+	 * some other thread starts a transaction commit which tries to do an
+	 * allocation while the other thread is still loading the space cache
+	 * info.  The previous loop should have kept us from choosing this block
+	 * group, but if we've moved to the state where we will wait on caching
+	 * block groups we need to first check if we're doing a fast load here,
+	 * so we can wait for it to finish, otherwise we could end up allocating
+	 * from a block group who's cache gets evicted for one reason or
+	 * another.
+	 */
+	while (cache->cached == BTRFS_CACHE_FAST) {
+		struct btrfs_caching_control *ctl;
+
+		ctl = cache->caching_ctl;
+		atomic_inc(&ctl->count);
+		prepare_to_wait(&ctl->wait, &wait, TASK_UNINTERRUPTIBLE);
+		spin_unlock(&cache->lock);
+
+		schedule();
+
+		finish_wait(&ctl->wait, &wait);
+		put_caching_control(ctl);
+		spin_lock(&cache->lock);
+	}
+
+	if (cache->cached != BTRFS_CACHE_NO) {
+		spin_unlock(&cache->lock);
+		kfree(caching_ctl);
 		return 0;
+	}
+	WARN_ON(cache->caching_ctl);
+	cache->caching_ctl = caching_ctl;
+	cache->cached = BTRFS_CACHE_FAST;
+	spin_unlock(&cache->lock);
 
 	/*
 	 * We can't do the read from on-disk cache during a commit since we need
@@ -484,56 +530,51 @@ static int cache_block_group(struct btrfs_block_group_cache *cache,
 	if (trans && (!trans->transaction->in_commit) &&
 	    (root && root != root->fs_info->tree_root) &&
 	    btrfs_test_opt(root, SPACE_CACHE)) {
-		spin_lock(&cache->lock);
-		if (cache->cached != BTRFS_CACHE_NO) {
-			spin_unlock(&cache->lock);
-			return 0;
-		}
-		cache->cached = BTRFS_CACHE_STARTED;
-		spin_unlock(&cache->lock);
-
 		ret = load_free_space_cache(fs_info, cache);
 
 		spin_lock(&cache->lock);
 		if (ret == 1) {
+			cache->caching_ctl = NULL;
 			cache->cached = BTRFS_CACHE_FINISHED;
 			cache->last_byte_to_unpin = (u64)-1;
 		} else {
-			cache->cached = BTRFS_CACHE_NO;
+			if (load_cache_only) {
+				cache->caching_ctl = NULL;
+				cache->cached = BTRFS_CACHE_NO;
+			} else {
+				cache->cached = BTRFS_CACHE_STARTED;
+			}
 		}
 		spin_unlock(&cache->lock);
+		wake_up(&caching_ctl->wait);
 		if (ret == 1) {
+			put_caching_control(caching_ctl);
 			free_excluded_extents(fs_info->extent_root, cache);
 			return 0;
 		}
+	} else {
+		/*
+		 * We are not going to do the fast caching, set cached to the
+		 * appropriate value and wakeup any waiters.
+		 */
+		spin_lock(&cache->lock);
+		if (load_cache_only) {
+			cache->caching_ctl = NULL;
+			cache->cached = BTRFS_CACHE_NO;
+		} else {
+			cache->cached = BTRFS_CACHE_STARTED;
+		}
+		spin_unlock(&cache->lock);
+		wake_up(&caching_ctl->wait);
 	}
 
-	if (load_cache_only)
-		return 0;
-
-	caching_ctl = kzalloc(sizeof(*caching_ctl), GFP_NOFS);
-	BUG_ON(!caching_ctl);
-
-	INIT_LIST_HEAD(&caching_ctl->list);
-	mutex_init(&caching_ctl->mutex);
-	init_waitqueue_head(&caching_ctl->wait);
-	caching_ctl->block_group = cache;
-	caching_ctl->progress = cache->key.objectid;
-	/* one for caching kthread, one for caching block group list */
-	atomic_set(&caching_ctl->count, 2);
-	caching_ctl->work.func = caching_thread;
-
-	spin_lock(&cache->lock);
-	if (cache->cached != BTRFS_CACHE_NO) {
-		spin_unlock(&cache->lock);
-		kfree(caching_ctl);
+	if (load_cache_only) {
+		put_caching_control(caching_ctl);
 		return 0;
 	}
-	cache->caching_ctl = caching_ctl;
-	cache->cached = BTRFS_CACHE_STARTED;
-	spin_unlock(&cache->lock);
 
 	down_write(&fs_info->extent_commit_sem);
+	atomic_inc(&caching_ctl->count);
 	list_add_tail(&caching_ctl->list, &fs_info->caching_block_groups);
 	up_write(&fs_info->extent_commit_sem);
 
@@ -3847,9 +3888,9 @@ int btrfs_block_rsv_check(struct btrfs_root *root,
 	return ret;
 }
 
-int btrfs_block_rsv_refill(struct btrfs_root *root,
-			  struct btrfs_block_rsv *block_rsv,
-			  u64 min_reserved)
+static inline int __btrfs_block_rsv_refill(struct btrfs_root *root,
+					   struct btrfs_block_rsv *block_rsv,
+					   u64 min_reserved, int flush)
 {
 	u64 num_bytes = 0;
 	int ret = -ENOSPC;
@@ -3868,7 +3909,7 @@ int btrfs_block_rsv_refill(struct btrfs_root *root,
 	if (!ret)
 		return 0;
 
-	ret = reserve_metadata_bytes(root, block_rsv, num_bytes, 1);
+	ret = reserve_metadata_bytes(root, block_rsv, num_bytes, flush);
 	if (!ret) {
 		block_rsv_add_bytes(block_rsv, num_bytes, 0);
 		return 0;
@@ -3877,6 +3918,20 @@ int btrfs_block_rsv_refill(struct btrfs_root *root,
 	return ret;
 }
 
+int btrfs_block_rsv_refill(struct btrfs_root *root,
+			   struct btrfs_block_rsv *block_rsv,
+			   u64 min_reserved)
+{
+	return __btrfs_block_rsv_refill(root, block_rsv, min_reserved, 1);
+}
+
+int btrfs_block_rsv_refill_noflush(struct btrfs_root *root,
+				   struct btrfs_block_rsv *block_rsv,
+				   u64 min_reserved)
+{
+	return __btrfs_block_rsv_refill(root, block_rsv, min_reserved, 0);
+}
+
 int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv,
 			    struct btrfs_block_rsv *dst_rsv,
 			    u64 num_bytes)
@@ -5052,11 +5107,11 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 	struct btrfs_root *root = orig_root->fs_info->extent_root;
 	struct btrfs_free_cluster *last_ptr = NULL;
 	struct btrfs_block_group_cache *block_group = NULL;
+	struct btrfs_block_group_cache *used_block_group;
 	int empty_cluster = 2 * 1024 * 1024;
 	int allowed_chunk_alloc = 0;
 	int done_chunk_alloc = 0;
 	struct btrfs_space_info *space_info;
-	int last_ptr_loop = 0;
 	int loop = 0;
 	int index = 0;
 	int alloc_type = (data & BTRFS_BLOCK_GROUP_DATA) ?
@@ -5118,6 +5173,7 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 ideal_cache:
 		block_group = btrfs_lookup_block_group(root->fs_info,
 						       search_start);
+		used_block_group = block_group;
 		/*
 		 * we don't want to use the block group if it doesn't match our
 		 * allocation bits, or if its not cached.
@@ -5155,6 +5211,7 @@ search:
 		u64 offset;
 		int cached;
 
+		used_block_group = block_group;
 		btrfs_get_block_group(block_group);
 		search_start = block_group->key.objectid;
 
@@ -5178,13 +5235,15 @@ search:
 		}
 
 have_block_group:
-		if (unlikely(block_group->cached == BTRFS_CACHE_NO)) {
+		cached = block_group_cache_done(block_group);
+		if (unlikely(!cached)) {
 			u64 free_percent;
 
+			found_uncached_bg = true;
 			ret = cache_block_group(block_group, trans,
 						orig_root, 1);
 			if (block_group->cached == BTRFS_CACHE_FINISHED)
-				goto have_block_group;
+				goto alloc;
 
 			free_percent = btrfs_block_group_used(&block_group->item);
 			free_percent *= 100;
@@ -5206,7 +5265,6 @@ have_block_group:
 							orig_root, 0);
 				BUG_ON(ret);
 			}
-			found_uncached_bg = true;
 
 			/*
 			 * If loop is set for cached only, try the next block
@@ -5216,94 +5274,80 @@ have_block_group:
 				goto loop;
 		}
 
-		cached = block_group_cache_done(block_group);
-		if (unlikely(!cached))
-			found_uncached_bg = true;
-
+alloc:
 		if (unlikely(block_group->ro))
 			goto loop;
 
 		spin_lock(&block_group->free_space_ctl->tree_lock);
 		if (cached &&
 		    block_group->free_space_ctl->free_space <
-		    num_bytes + empty_size) {
+		    num_bytes + empty_cluster + empty_size) {
 			spin_unlock(&block_group->free_space_ctl->tree_lock);
 			goto loop;
 		}
 		spin_unlock(&block_group->free_space_ctl->tree_lock);
 
 		/*
-		 * Ok we want to try and use the cluster allocator, so lets look
-		 * there, unless we are on LOOP_NO_EMPTY_SIZE, since we will
-		 * have tried the cluster allocator plenty of times at this
-		 * point and not have found anything, so we are likely way too
-		 * fragmented for the clustering stuff to find anything, so lets
-		 * just skip it and let the allocator find whatever block it can
-		 * find
+		 * Ok we want to try and use the cluster allocator, so
+		 * lets look there
 		 */
-		if (last_ptr && loop < LOOP_NO_EMPTY_SIZE) {
+		if (last_ptr) {
 			/*
 			 * the refill lock keeps out other
 			 * people trying to start a new cluster
 			 */
 			spin_lock(&last_ptr->refill_lock);
-			if (last_ptr->block_group &&
-			    (last_ptr->block_group->ro ||
-			    !block_group_bits(last_ptr->block_group, data))) {
-				offset = 0;
+			used_block_group = last_ptr->block_group;
+			if (used_block_group != block_group &&
+			    (!used_block_group ||
+			     used_block_group->ro ||
+			     !block_group_bits(used_block_group, data))) {
+				used_block_group = block_group;
 				goto refill_cluster;
 			}
 
-			offset = btrfs_alloc_from_cluster(block_group, last_ptr,
-						 num_bytes, search_start);
+			if (used_block_group != block_group)
+				btrfs_get_block_group(used_block_group);
+
+			offset = btrfs_alloc_from_cluster(used_block_group,
+			  last_ptr, num_bytes, used_block_group->key.objectid);
 			if (offset) {
 				/* we have a block, we're done */
 				spin_unlock(&last_ptr->refill_lock);
 				goto checks;
 			}
 
-			spin_lock(&last_ptr->lock);
-			/*
-			 * whoops, this cluster doesn't actually point to
-			 * this block group.  Get a ref on the block
-			 * group is does point to and try again
-			 */
-			if (!last_ptr_loop && last_ptr->block_group &&
-			    last_ptr->block_group != block_group &&
-			    index <=
-				 get_block_group_index(last_ptr->block_group)) {
-
-				btrfs_put_block_group(block_group);
-				block_group = last_ptr->block_group;
-				btrfs_get_block_group(block_group);
-				spin_unlock(&last_ptr->lock);
-				spin_unlock(&last_ptr->refill_lock);
-
-				last_ptr_loop = 1;
-				search_start = block_group->key.objectid;
-				/*
-				 * we know this block group is properly
-				 * in the list because
-				 * btrfs_remove_block_group, drops the
-				 * cluster before it removes the block
-				 * group from the list
-				 */
-				goto have_block_group;
+			WARN_ON(last_ptr->block_group != used_block_group);
+			if (used_block_group != block_group) {
+				btrfs_put_block_group(used_block_group);
+				used_block_group = block_group;
 			}
-			spin_unlock(&last_ptr->lock);
 refill_cluster:
+			BUG_ON(used_block_group != block_group);
+			/* If we are on LOOP_NO_EMPTY_SIZE, we can't
+			 * set up a new clusters, so lets just skip it
+			 * and let the allocator find whatever block
+			 * it can find.  If we reach this point, we
+			 * will have tried the cluster allocator
+			 * plenty of times and not have found
+			 * anything, so we are likely way too
+			 * fragmented for the clustering stuff to find
+			 * anything.  */
+			if (loop >= LOOP_NO_EMPTY_SIZE) {
+				spin_unlock(&last_ptr->refill_lock);
+				goto unclustered_alloc;
+			}
+
 			/*
 			 * this cluster didn't work out, free it and
 			 * start over
 			 */
 			btrfs_return_cluster_to_free_space(NULL, last_ptr);
 
-			last_ptr_loop = 0;
-
 			/* allocate a cluster in this block group */
 			ret = btrfs_find_space_cluster(trans, root,
 					       block_group, last_ptr,
-					       offset, num_bytes,
+					       search_start, num_bytes,
 					       empty_cluster + empty_size);
 			if (ret == 0) {
 				/*
@@ -5339,6 +5383,7 @@ refill_cluster:
 			goto loop;
 		}
 
+unclustered_alloc:
 		offset = btrfs_find_space_for_alloc(block_group, search_start,
 						    num_bytes, empty_size);
 		/*
@@ -5365,14 +5410,14 @@ checks:
 		search_start = stripe_align(root, offset);
 		/* move on to the next group */
 		if (search_start + num_bytes >= search_end) {
-			btrfs_add_free_space(block_group, offset, num_bytes);
+			btrfs_add_free_space(used_block_group, offset, num_bytes);
 			goto loop;
 		}
 
 		/* move on to the next group */
 		if (search_start + num_bytes >
-		    block_group->key.objectid + block_group->key.offset) {
-			btrfs_add_free_space(block_group, offset, num_bytes);
+		    used_block_group->key.objectid + used_block_group->key.offset) {
+			btrfs_add_free_space(used_block_group, offset, num_bytes);
 			goto loop;
 		}
 
@@ -5380,14 +5425,14 @@ checks:
 		ins->offset = num_bytes;
 
 		if (offset < search_start)
-			btrfs_add_free_space(block_group, offset,
+			btrfs_add_free_space(used_block_group, offset,
 					     search_start - offset);
 		BUG_ON(offset > search_start);
 
-		ret = btrfs_update_reserved_bytes(block_group, num_bytes,
+		ret = btrfs_update_reserved_bytes(used_block_group, num_bytes,
 						  alloc_type);
 		if (ret == -EAGAIN) {
-			btrfs_add_free_space(block_group, offset, num_bytes);
+			btrfs_add_free_space(used_block_group, offset, num_bytes);
 			goto loop;
 		}
 
@@ -5396,15 +5441,19 @@ checks:
 		ins->offset = num_bytes;
 
 		if (offset < search_start)
-			btrfs_add_free_space(block_group, offset,
+			btrfs_add_free_space(used_block_group, offset,
 					     search_start - offset);
 		BUG_ON(offset > search_start);
+		if (used_block_group != block_group)
+			btrfs_put_block_group(used_block_group);
 		btrfs_put_block_group(block_group);
 		break;
 loop:
 		failed_cluster_refill = false;
 		failed_alloc = false;
 		BUG_ON(index != get_block_group_index(block_group));
+		if (used_block_group != block_group)
+			btrfs_put_block_group(used_block_group);
 		btrfs_put_block_group(block_group);
 	}
 	up_read(&space_info->groups_sem);
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 1f87c4d0e7a..49f3c9dc09f 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -935,8 +935,10 @@ again:
 	node = tree_search(tree, start);
 	if (!node) {
 		prealloc = alloc_extent_state_atomic(prealloc);
-		if (!prealloc)
-			return -ENOMEM;
+		if (!prealloc) {
+			err = -ENOMEM;
+			goto out;
+		}
 		err = insert_state(tree, prealloc, start, end, &bits);
 		prealloc = NULL;
 		BUG_ON(err == -EEXIST);
@@ -992,8 +994,10 @@ hit_next:
 	 */
 	if (state->start < start) {
 		prealloc = alloc_extent_state_atomic(prealloc);
-		if (!prealloc)
-			return -ENOMEM;
+		if (!prealloc) {
+			err = -ENOMEM;
+			goto out;
+		}
 		err = split_state(tree, state, prealloc, start);
 		BUG_ON(err == -EEXIST);
 		prealloc = NULL;
@@ -1024,8 +1028,10 @@ hit_next:
 			this_end = last_start - 1;
 
 		prealloc = alloc_extent_state_atomic(prealloc);
-		if (!prealloc)
-			return -ENOMEM;
+		if (!prealloc) {
+			err = -ENOMEM;
+			goto out;
+		}
 
 		/*
 		 * Avoid to free 'prealloc' if it can be merged with
@@ -1051,8 +1057,10 @@ hit_next:
 	 */
 	if (state->start <= end && state->end > end) {
 		prealloc = alloc_extent_state_atomic(prealloc);
-		if (!prealloc)
-			return -ENOMEM;
+		if (!prealloc) {
+			err = -ENOMEM;
+			goto out;
+		}
 
 		err = split_state(tree, state, prealloc, end + 1);
 		BUG_ON(err == -EEXIST);
@@ -2285,16 +2293,22 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 				clean_io_failure(start, page);
 		}
 		if (!uptodate) {
-			u64 failed_mirror;
-			failed_mirror = (u64)bio->bi_bdev;
-			if (tree->ops && tree->ops->readpage_io_failed_hook)
-				ret = tree->ops->readpage_io_failed_hook(
-						bio, page, start, end,
-						failed_mirror, state);
-			else
-				ret = bio_readpage_error(bio, page, start, end,
-							 failed_mirror, NULL);
+			int failed_mirror;
+			failed_mirror = (int)(unsigned long)bio->bi_bdev;
+			/*
+			 * The generic bio_readpage_error handles errors the
+			 * following way: If possible, new read requests are
+			 * created and submitted and will end up in
+			 * end_bio_extent_readpage as well (if we're lucky, not
+			 * in the !uptodate case). In that case it returns 0 and
+			 * we just go on with the next page in our bio. If it
+			 * can't handle the error it will return -EIO and we
+			 * remain responsible for that page.
+			 */
+			ret = bio_readpage_error(bio, page, start, end,
+							failed_mirror, NULL);
 			if (ret == 0) {
+error_handled:
 				uptodate =
 					test_bit(BIO_UPTODATE, &bio->bi_flags);
 				if (err)
@@ -2302,6 +2316,13 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 				uncache_state(&cached);
 				continue;
 			}
+			if (tree->ops && tree->ops->readpage_io_failed_hook) {
+				ret = tree->ops->readpage_io_failed_hook(
+							bio, page, start, end,
+							failed_mirror, state);
+				if (ret == 0)
+					goto error_handled;
+			}
 		}
 
 		if (uptodate) {
@@ -3366,6 +3387,9 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		return -ENOMEM;
 	path->leave_spinning = 1;
 
+	start = ALIGN(start, BTRFS_I(inode)->root->sectorsize);
+	len = ALIGN(len, BTRFS_I(inode)->root->sectorsize);
+
 	/*
 	 * lookup the last file extent.  We're not using i_size here
 	 * because there might be preallocation past i_size
@@ -3413,7 +3437,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, start, start + len, 0,
 			 &cached_state, GFP_NOFS);
 
-	em = get_extent_skip_holes(inode, off, last_for_get_extent,
+	em = get_extent_skip_holes(inode, start, last_for_get_extent,
 				   get_extent);
 	if (!em)
 		goto out;
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index feb9be0e23b..7604c300132 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -70,7 +70,7 @@ struct extent_io_ops {
 			      unsigned long bio_flags);
 	int (*readpage_io_hook)(struct page *page, u64 start, u64 end);
 	int (*readpage_io_failed_hook)(struct bio *bio, struct page *page,
-				       u64 start, u64 end, u64 failed_mirror,
+				       u64 start, u64 end, int failed_mirror,
 				       struct extent_state *state);
 	int (*writepage_io_failed_hook)(struct bio *bio, struct page *page,
 					u64 start, u64 end,
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 181760f9d2a..ec23d43d0c3 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -351,6 +351,11 @@ static int io_ctl_prepare_pages(struct io_ctl *io_ctl, struct inode *inode,
 		}
 	}
 
+	for (i = 0; i < io_ctl->num_pages; i++) {
+		clear_page_dirty_for_io(io_ctl->pages[i]);
+		set_page_extent_mapped(io_ctl->pages[i]);
+	}
+
 	return 0;
 }
 
@@ -1465,6 +1470,7 @@ static void add_new_bitmap(struct btrfs_free_space_ctl *ctl,
 {
 	info->offset = offset_to_bitmap(ctl, offset);
 	info->bytes = 0;
+	INIT_LIST_HEAD(&info->list);
 	link_free_space(ctl, info);
 	ctl->total_bitmaps++;
 
@@ -1844,7 +1850,13 @@ again:
 		info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
 					  1, 0);
 		if (!info) {
-			WARN_ON(1);
+			/* the tree logging code might be calling us before we
+			 * have fully loaded the free space rbtree for this
+			 * block group.  So it is possible the entry won't
+			 * be in the rbtree yet at all.  The caching code
+			 * will make sure not to put it in the rbtree if
+			 * the logging code has pinned it.
+			 */
 			goto out_lock;
 		}
 	}
@@ -2308,6 +2320,7 @@ again:
 
 	if (!found) {
 		start = i;
+		cluster->max_size = 0;
 		found = true;
 	}
 
@@ -2451,16 +2464,23 @@ setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
 {
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
 	struct btrfs_free_space *entry;
-	struct rb_node *node;
 	int ret = -ENOSPC;
+	u64 bitmap_offset = offset_to_bitmap(ctl, offset);
 
 	if (ctl->total_bitmaps == 0)
 		return -ENOSPC;
 
 	/*
-	 * First check our cached list of bitmaps and see if there is an entry
-	 * here that will work.
+	 * The bitmap that covers offset won't be in the list unless offset
+	 * is just its start offset.
 	 */
+	entry = list_first_entry(bitmaps, struct btrfs_free_space, list);
+	if (entry->offset != bitmap_offset) {
+		entry = tree_search_offset(ctl, bitmap_offset, 1, 0);
+		if (entry && list_empty(&entry->list))
+			list_add(&entry->list, bitmaps);
+	}
+
 	list_for_each_entry(entry, bitmaps, list) {
 		if (entry->bytes < min_bytes)
 			continue;
@@ -2471,38 +2491,10 @@ setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
 	}
 
 	/*
-	 * If we do have entries on our list and we are here then we didn't find
-	 * anything, so go ahead and get the next entry after the last entry in
-	 * this list and start the search from there.
+	 * The bitmaps list has all the bitmaps that record free space
+	 * starting after offset, so no more search is required.
 	 */
-	if (!list_empty(bitmaps)) {
-		entry = list_entry(bitmaps->prev, struct btrfs_free_space,
-				   list);
-		node = rb_next(&entry->offset_index);
-		if (!node)
-			return -ENOSPC;
-		entry = rb_entry(node, struct btrfs_free_space, offset_index);
-		goto search;
-	}
-
-	entry = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), 0, 1);
-	if (!entry)
-		return -ENOSPC;
-
-search:
-	node = &entry->offset_index;
-	do {
-		entry = rb_entry(node, struct btrfs_free_space, offset_index);
-		node = rb_next(&entry->offset_index);
-		if (!entry->bitmap)
-			continue;
-		if (entry->bytes < min_bytes)
-			continue;
-		ret = btrfs_bitmap_cluster(block_group, entry, cluster, offset,
-					   bytes, min_bytes);
-	} while (ret && node);
-
-	return ret;
+	return -ENOSPC;
 }
 
 /*
@@ -2520,8 +2512,8 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
 			     u64 offset, u64 bytes, u64 empty_size)
 {
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
-	struct list_head bitmaps;
 	struct btrfs_free_space *entry, *tmp;
+	LIST_HEAD(bitmaps);
 	u64 min_bytes;
 	int ret;
 
@@ -2560,7 +2552,6 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
 		goto out;
 	}
 
-	INIT_LIST_HEAD(&bitmaps);
 	ret = setup_cluster_no_bitmap(block_group, cluster, &bitmaps, offset,
 				      bytes, min_bytes);
 	if (ret)
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 116ab67a06d..2c984f7d4c2 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -3490,7 +3490,7 @@ void btrfs_evict_inode(struct inode *inode)
 	 * doing the truncate.
 	 */
 	while (1) {
-		ret = btrfs_block_rsv_refill(root, rsv, min_size);
+		ret = btrfs_block_rsv_refill_noflush(root, rsv, min_size);
 
 		/*
 		 * Try and steal from the global reserve since we will
@@ -6794,11 +6794,13 @@ static int btrfs_getattr(struct vfsmount *mnt,
 			 struct dentry *dentry, struct kstat *stat)
 {
 	struct inode *inode = dentry->d_inode;
+	u32 blocksize = inode->i_sb->s_blocksize;
+
 	generic_fillattr(inode, stat);
 	stat->dev = BTRFS_I(inode)->root->anon_dev;
 	stat->blksize = PAGE_CACHE_SIZE;
-	stat->blocks = (inode_get_bytes(inode) +
-			BTRFS_I(inode)->delalloc_bytes) >> 9;
+	stat->blocks = (ALIGN(inode_get_bytes(inode), blocksize) +
+		ALIGN(BTRFS_I(inode)->delalloc_bytes, blocksize)) >> 9;
 	return 0;
 }
 
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 4a34c472f12..72d461656f6 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1216,12 +1216,12 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root,
 		*devstr = '\0';
 		devstr = vol_args->name;
 		devid = simple_strtoull(devstr, &end, 10);
-		printk(KERN_INFO "resizing devid %llu\n",
+		printk(KERN_INFO "btrfs: resizing devid %llu\n",
 		       (unsigned long long)devid);
 	}
 	device = btrfs_find_device(root, devid, NULL, NULL);
 	if (!device) {
-		printk(KERN_INFO "resizer unable to find device %llu\n",
+		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
 		       (unsigned long long)devid);
 		ret = -EINVAL;
 		goto out_unlock;
@@ -1267,7 +1267,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root,
 	do_div(new_size, root->sectorsize);
 	new_size *= root->sectorsize;
 
-	printk(KERN_INFO "new size for %s is %llu\n",
+	printk(KERN_INFO "btrfs: new size for %s is %llu\n",
 		device->name, (unsigned long long)new_size);
 
 	if (new_size > old_size) {
@@ -1278,7 +1278,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root,
 		}
 		ret = btrfs_grow_device(trans, device, new_size);
 		btrfs_commit_transaction(trans, root);
-	} else {
+	} else if (new_size < old_size) {
 		ret = btrfs_shrink_device(device, new_size);
 	}
 
@@ -2930,11 +2930,13 @@ static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
 		goto out;
 
 	for (i = 0; i < ipath->fspath->elem_cnt; ++i) {
-		rel_ptr = ipath->fspath->val[i] - (u64)ipath->fspath->val;
+		rel_ptr = ipath->fspath->val[i] -
+			  (u64)(unsigned long)ipath->fspath->val;
 		ipath->fspath->val[i] = rel_ptr;
 	}
 
-	ret = copy_to_user((void *)ipa->fspath, (void *)ipath->fspath, size);
+	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
+			   (void *)(unsigned long)ipath->fspath, size);
 	if (ret) {
 		ret = -EFAULT;
 		goto out;
@@ -3017,7 +3019,8 @@ static long btrfs_ioctl_logical_to_ino(struct btrfs_root *root,
 	if (ret < 0)
 		goto out;
 
-	ret = copy_to_user((void *)loi->inodes, (void *)inodes, size);
+	ret = copy_to_user((void *)(unsigned long)loi->inodes,
+			   (void *)(unsigned long)inodes, size);
 	if (ret)
 		ret = -EFAULT;
 
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index f4190f22edf..c27bcb67f33 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -256,6 +256,11 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root, void *ctx)
 	btrfs_release_path(swarn->path);
 
 	ipath = init_ipath(4096, local_root, swarn->path);
+	if (IS_ERR(ipath)) {
+		ret = PTR_ERR(ipath);
+		ipath = NULL;
+		goto err;
+	}
 	ret = paths_from_inode(inum, ipath);
 
 	if (ret < 0)
@@ -272,7 +277,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root, void *ctx)
 			swarn->logical, swarn->dev->name,
 			(unsigned long long)swarn->sector, root, inum, offset,
 			min(isize - offset, (u64)PAGE_SIZE), nlink,
-			(char *)ipath->fspath->val[i]);
+			(char *)(unsigned long)ipath->fspath->val[i]);
 
 	free_ipath(ipath);
 	return 0;
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 8bd9d6d0e07..e28ad4baf48 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -825,13 +825,9 @@ static char *setup_root_args(char *args)
 static struct dentry *mount_subvol(const char *subvol_name, int flags,
 				   const char *device_name, char *data)
 {
-	struct super_block *s;
 	struct dentry *root;
 	struct vfsmount *mnt;
-	struct mnt_namespace *ns_private;
 	char *newargs;
-	struct path path;
-	int error;
 
 	newargs = setup_root_args(data);
 	if (!newargs)
@@ -842,39 +838,17 @@ static struct dentry *mount_subvol(const char *subvol_name, int flags,
 	if (IS_ERR(mnt))
 		return ERR_CAST(mnt);
 
-	ns_private = create_mnt_ns(mnt);
-	if (IS_ERR(ns_private)) {
-		mntput(mnt);
-		return ERR_CAST(ns_private);
-	}
+	root = mount_subtree(mnt, subvol_name);
 
-	/*
-	 * This will trigger the automount of the subvol so we can just
-	 * drop the mnt we have here and return the dentry that we
-	 * found.
-	 */
-	error = vfs_path_lookup(mnt->mnt_root, mnt, subvol_name,
-				LOOKUP_FOLLOW, &path);
-	put_mnt_ns(ns_private);
-	if (error)
-		return ERR_PTR(error);
-
-	if (!is_subvolume_inode(path.dentry->d_inode)) {
-		path_put(&path);
-		mntput(mnt);
-		error = -EINVAL;
+	if (!IS_ERR(root) && !is_subvolume_inode(root->d_inode)) {
+		struct super_block *s = root->d_sb;
+		dput(root);
+		root = ERR_PTR(-EINVAL);
+		deactivate_locked_super(s);
 		printk(KERN_ERR "btrfs: '%s' is not a valid subvolume\n",
 				subvol_name);
-		return ERR_PTR(-EINVAL);
 	}
 
-	/* Get a ref to the sb and the dentry we found and return it */
-	s = path.mnt->mnt_sb;
-	atomic_inc(&s->s_active);
-	root = dget(path.dentry);
-	path_put(&path);
-	down_write(&s->s_umount);
-
 	return root;
 }
 
@@ -1083,7 +1057,7 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 	int i = 0, nr_devices;
 	int ret;
 
-	nr_devices = fs_info->fs_devices->rw_devices;
+	nr_devices = fs_info->fs_devices->open_devices;
 	BUG_ON(!nr_devices);
 
 	devices_info = kmalloc(sizeof(*devices_info) * nr_devices,
@@ -1105,8 +1079,8 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 	else
 		min_stripe_size = BTRFS_STRIPE_LEN;
 
-	list_for_each_entry(device, &fs_devices->alloc_list, dev_alloc_list) {
-		if (!device->in_fs_metadata)
+	list_for_each_entry(device, &fs_devices->devices, dev_list) {
+		if (!device->in_fs_metadata || !device->bdev)
 			continue;
 
 		avail_space = device->total_bytes - device->bytes_used;
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 6a0574e923b..81376d94cd3 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -785,6 +785,10 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans,
 
 			btrfs_save_ino_cache(root, trans);
 
+			/* see comments in should_cow_block() */
+			root->force_cow = 0;
+			smp_wmb();
+
 			if (root->commit_root != root->node) {
 				mutex_lock(&root->fs_commit_mutex);
 				switch_commit_root(root);
@@ -947,6 +951,10 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	btrfs_tree_unlock(old);
 	free_extent_buffer(old);
 
+	/* see comments in should_cow_block() */
+	root->force_cow = 1;
+	smp_wmb();
+
 	btrfs_set_root_node(new_root_item, tmp);
 	/* record when the snapshot was created in key.offset */
 	key.offset = trans->transid;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index c37433d3cd8..0a8c8f8304b 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1611,7 +1611,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	if ((sb->s_flags & MS_RDONLY) && !root->fs_info->fs_devices->seeding)
 		return -EINVAL;
 
-	bdev = blkdev_get_by_path(device_path, FMODE_EXCL,
+	bdev = blkdev_get_by_path(device_path, FMODE_WRITE | FMODE_EXCL,
 				  root->fs_info->bdev_holder);
 	if (IS_ERR(bdev))
 		return PTR_ERR(bdev);
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index ab5b1c49f35..78f2d4d4f37 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -100,6 +100,12 @@ struct btrfs_device {
 	struct reada_zone *reada_curr_zone;
 	struct radix_tree_root reada_zones;
 	struct radix_tree_root reada_extents;
+
+	/* for sending down flush barriers */
+	struct bio *flush_bio;
+	struct completion flush_wait;
+	int nobarriers;
+
 };
 
 struct btrfs_fs_devices {
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 2abd0dfad7f..bca3948e9db 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -1143,7 +1143,7 @@ static void ceph_d_prune(struct dentry *dentry)
 {
 	struct ceph_dentry_info *di;
 
-	dout("d_release %p\n", dentry);
+	dout("ceph_d_prune %p\n", dentry);
 
 	/* do we have a valid parent? */
 	if (!dentry->d_parent || IS_ROOT(dentry))
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index e392bfce84a..116f36502f1 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -1328,12 +1328,13 @@ int ceph_inode_set_size(struct inode *inode, loff_t size)
  */
 void ceph_queue_writeback(struct inode *inode)
 {
+	ihold(inode);
 	if (queue_work(ceph_inode_to_client(inode)->wb_wq,
 		       &ceph_inode(inode)->i_wb_work)) {
 		dout("ceph_queue_writeback %p\n", inode);
-		ihold(inode);
 	} else {
 		dout("ceph_queue_writeback %p failed\n", inode);
+		iput(inode);
 	}
 }
 
@@ -1353,12 +1354,13 @@ static void ceph_writeback_work(struct work_struct *work)
  */
 void ceph_queue_invalidate(struct inode *inode)
 {
+	ihold(inode);
 	if (queue_work(ceph_inode_to_client(inode)->pg_inv_wq,
 		       &ceph_inode(inode)->i_pg_inv_work)) {
 		dout("ceph_queue_invalidate %p\n", inode);
-		ihold(inode);
 	} else {
 		dout("ceph_queue_invalidate %p failed\n", inode);
+		iput(inode);
 	}
 }
 
@@ -1434,13 +1436,14 @@ void ceph_queue_vmtruncate(struct inode *inode)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 
+	ihold(inode);
 	if (queue_work(ceph_sb_to_client(inode->i_sb)->trunc_wq,
 		       &ci->i_vmtruncate_work)) {
 		dout("ceph_queue_vmtruncate %p\n", inode);
-		ihold(inode);
 	} else {
 		dout("ceph_queue_vmtruncate %p failed, pending=%d\n",
 		     inode, ci->i_truncate_pending);
+		iput(inode);
 	}
 }
 
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index a90846fac75..8dc73a594a9 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -638,10 +638,12 @@ static struct dentry *open_root_dentry(struct ceph_fs_client *fsc,
 	if (err == 0) {
 		dout("open_root_inode success\n");
 		if (ceph_ino(req->r_target_inode) == CEPH_INO_ROOT &&
-		    fsc->sb->s_root == NULL)
+		    fsc->sb->s_root == NULL) {
 			root = d_alloc_root(req->r_target_inode);
-		else
+			ceph_init_dentry(root);
+		} else {
 			root = d_obtain_alias(req->r_target_inode);
+		}
 		req->r_target_inode = NULL;
 		dout("open_root_inode success, root dentry is %p\n", root);
 	} else {
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index d6a972df033..8cd4b52d421 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -441,6 +441,8 @@ cifs_readv_from_socket(struct TCP_Server_Info *server, struct kvec *iov_orig,
 	smb_msg.msg_controllen = 0;
 
 	for (total_read = 0; to_read; total_read += length, to_read -= length) {
+		try_to_freeze();
+
 		if (server_unresponsive(server)) {
 			total_read = -EAGAIN;
 			break;
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index cf0b1539b32..4dd9283885e 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -702,6 +702,13 @@ cifs_find_lock_conflict(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock,
 					 lock->type, lock->netfid, conf_lock);
 }
 
+/*
+ * Check if there is another lock that prevents us to set the lock (mandatory
+ * style). If such a lock exists, update the flock structure with its
+ * properties. Otherwise, set the flock type to F_UNLCK if we can cache brlocks
+ * or leave it the same if we can't. Returns 0 if we don't need to request to
+ * the server or 1 otherwise.
+ */
 static int
 cifs_lock_test(struct cifsInodeInfo *cinode, __u64 offset, __u64 length,
 	       __u8 type, __u16 netfid, struct file_lock *flock)
@@ -739,6 +746,12 @@ cifs_lock_add(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock)
 	mutex_unlock(&cinode->lock_mutex);
 }
 
+/*
+ * Set the byte-range lock (mandatory style). Returns:
+ * 1) 0, if we set the lock and don't need to request to the server;
+ * 2) 1, if no locks prevent us but we need to request to the server;
+ * 3) -EACCESS, if there is a lock that prevents us and wait is false.
+ */
 static int
 cifs_lock_add_if(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock,
 		 bool wait)
@@ -778,6 +791,13 @@ try_again:
 	return rc;
 }
 
+/*
+ * Check if there is another lock that prevents us to set the lock (posix
+ * style). If such a lock exists, update the flock structure with its
+ * properties. Otherwise, set the flock type to F_UNLCK if we can cache brlocks
+ * or leave it the same if we can't. Returns 0 if we don't need to request to
+ * the server or 1 otherwise.
+ */
 static int
 cifs_posix_lock_test(struct file *file, struct file_lock *flock)
 {
@@ -800,6 +820,12 @@ cifs_posix_lock_test(struct file *file, struct file_lock *flock)
 	return rc;
 }
 
+/*
+ * Set the byte-range lock (posix style). Returns:
+ * 1) 0, if we set the lock and don't need to request to the server;
+ * 2) 1, if we need to request to the server;
+ * 3) <0, if the error occurs while setting the lock.
+ */
 static int
 cifs_posix_lock_set(struct file *file, struct file_lock *flock)
 {
diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c
index 5de03ec2014..a090bbe6ee2 100644
--- a/fs/cifs/readdir.c
+++ b/fs/cifs/readdir.c
@@ -554,7 +554,10 @@ static int find_cifs_entry(const int xid, struct cifs_tcon *pTcon,
 				 rc);
 			return rc;
 		}
-		cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile);
+		/* FindFirst/Next set last_entry to NULL on malformed reply */
+		if (cifsFile->srch_inf.last_entry)
+			cifs_save_resume_key(cifsFile->srch_inf.last_entry,
+						cifsFile);
 	}
 
 	while ((index_to_find >= cifsFile->srch_inf.index_of_last_entry) &&
@@ -562,7 +565,10 @@ static int find_cifs_entry(const int xid, struct cifs_tcon *pTcon,
 		cFYI(1, "calling findnext2");
 		rc = CIFSFindNext(xid, pTcon, cifsFile->netfid,
 				  &cifsFile->srch_inf);
-		cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile);
+		/* FindFirst/Next set last_entry to NULL on malformed reply */
+		if (cifsFile->srch_inf.last_entry)
+			cifs_save_resume_key(cifsFile->srch_inf.last_entry,
+						cifsFile);
 		if (rc)
 			return -ENOENT;
 	}
diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c
index 7cacba12b8f..80d85088193 100644
--- a/fs/cifs/smbencrypt.c
+++ b/fs/cifs/smbencrypt.c
@@ -209,7 +209,7 @@ E_md4hash(const unsigned char *passwd, unsigned char *p16,
 {
 	int rc;
 	int len;
-	__u16 wpwd[129];
+	__le16 wpwd[129];
 
 	/* Password cannot be longer than 128 characters */
 	if (passwd) /* Password must be converted to NT unicode */
@@ -219,8 +219,8 @@ E_md4hash(const unsigned char *passwd, unsigned char *p16,
 		*wpwd = 0; /* Ensure string is null terminated */
 	}
 
-	rc = mdfour(p16, (unsigned char *) wpwd, len * sizeof(__u16));
-	memset(wpwd, 0, 129 * sizeof(__u16));
+	rc = mdfour(p16, (unsigned char *) wpwd, len * sizeof(__le16));
+	memset(wpwd, 0, 129 * sizeof(__le16));
 
 	return rc;
 }
diff --git a/fs/dcache.c b/fs/dcache.c
index a901c6901bc..89509b5a090 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -36,6 +36,7 @@
 #include <linux/bit_spinlock.h>
 #include <linux/rculist_bl.h>
 #include <linux/prefetch.h>
+#include <linux/ratelimit.h>
 #include "internal.h"
 
 /*
@@ -2383,8 +2384,16 @@ struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode)
 				actual = __d_unalias(inode, dentry, alias);
 			}
 			write_sequnlock(&rename_lock);
-			if (IS_ERR(actual))
+			if (IS_ERR(actual)) {
+				if (PTR_ERR(actual) == -ELOOP)
+					pr_warn_ratelimited(
+						"VFS: Lookup of '%s' in %s %s"
+						" would have caused loop\n",
+						dentry->d_name.name,
+						inode->i_sb->s_type->name,
+						inode->i_sb->s_id);
 				dput(alias);
+			}
 			goto out_nolock;
 		}
 	}
@@ -2430,16 +2439,14 @@ static int prepend_name(char **buffer, int *buflen, struct qstr *name)
 /**
  * prepend_path - Prepend path string to a buffer
  * @path: the dentry/vfsmount to report
- * @root: root vfsmnt/dentry (may be modified by this function)
+ * @root: root vfsmnt/dentry
  * @buffer: pointer to the end of the buffer
  * @buflen: pointer to buffer length
  *
  * Caller holds the rename_lock.
- *
- * If path is not reachable from the supplied root, then the value of
- * root is changed (without modifying refcounts).
  */
-static int prepend_path(const struct path *path, struct path *root,
+static int prepend_path(const struct path *path,
+			const struct path *root,
 			char **buffer, int *buflen)
 {
 	struct dentry *dentry = path->dentry;
@@ -2474,10 +2481,10 @@ static int prepend_path(const struct path *path, struct path *root,
 		dentry = parent;
 	}
 
-out:
 	if (!error && !slash)
 		error = prepend(buffer, buflen, "/", 1);
 
+out:
 	br_read_unlock(vfsmount_lock);
 	return error;
 
@@ -2491,15 +2498,17 @@ global_root:
 		WARN(1, "Root dentry has weird name <%.*s>\n",
 		     (int) dentry->d_name.len, dentry->d_name.name);
 	}
-	root->mnt = vfsmnt;
-	root->dentry = dentry;
+	if (!slash)
+		error = prepend(buffer, buflen, "/", 1);
+	if (!error)
+		error = vfsmnt->mnt_ns ? 1 : 2;
 	goto out;
 }
 
 /**
  * __d_path - return the path of a dentry
  * @path: the dentry/vfsmount to report
- * @root: root vfsmnt/dentry (may be modified by this function)
+ * @root: root vfsmnt/dentry
  * @buf: buffer to return value in
  * @buflen: buffer length
  *
@@ -2510,10 +2519,10 @@ global_root:
  *
  * "buflen" should be positive.
  *
- * If path is not reachable from the supplied root, then the value of
- * root is changed (without modifying refcounts).
+ * If the path is not reachable from the supplied root, return %NULL.
  */
-char *__d_path(const struct path *path, struct path *root,
+char *__d_path(const struct path *path,
+	       const struct path *root,
 	       char *buf, int buflen)
 {
 	char *res = buf + buflen;
@@ -2524,7 +2533,28 @@ char *__d_path(const struct path *path, struct path *root,
 	error = prepend_path(path, root, &res, &buflen);
 	write_sequnlock(&rename_lock);
 
-	if (error)
+	if (error < 0)
+		return ERR_PTR(error);
+	if (error > 0)
+		return NULL;
+	return res;
+}
+
+char *d_absolute_path(const struct path *path,
+	       char *buf, int buflen)
+{
+	struct path root = {};
+	char *res = buf + buflen;
+	int error;
+
+	prepend(&res, &buflen, "\0", 1);
+	write_seqlock(&rename_lock);
+	error = prepend_path(path, &root, &res, &buflen);
+	write_sequnlock(&rename_lock);
+
+	if (error > 1)
+		error = -EINVAL;
+	if (error < 0)
 		return ERR_PTR(error);
 	return res;
 }
@@ -2532,8 +2562,9 @@ char *__d_path(const struct path *path, struct path *root,
 /*
  * same as __d_path but appends "(deleted)" for unlinked files.
  */
-static int path_with_deleted(const struct path *path, struct path *root,
-				 char **buf, int *buflen)
+static int path_with_deleted(const struct path *path,
+			     const struct path *root,
+			     char **buf, int *buflen)
 {
 	prepend(buf, buflen, "\0", 1);
 	if (d_unlinked(path->dentry)) {
@@ -2570,7 +2601,6 @@ char *d_path(const struct path *path, char *buf, int buflen)
 {
 	char *res = buf + buflen;
 	struct path root;
-	struct path tmp;
 	int error;
 
 	/*
@@ -2585,9 +2615,8 @@ char *d_path(const struct path *path, char *buf, int buflen)
 
 	get_fs_root(current->fs, &root);
 	write_seqlock(&rename_lock);
-	tmp = root;
-	error = path_with_deleted(path, &tmp, &res, &buflen);
-	if (error)
+	error = path_with_deleted(path, &root, &res, &buflen);
+	if (error < 0)
 		res = ERR_PTR(error);
 	write_sequnlock(&rename_lock);
 	path_put(&root);
@@ -2608,7 +2637,6 @@ char *d_path_with_unreachable(const struct path *path, char *buf, int buflen)
 {
 	char *res = buf + buflen;
 	struct path root;
-	struct path tmp;
 	int error;
 
 	if (path->dentry->d_op && path->dentry->d_op->d_dname)
@@ -2616,9 +2644,8 @@ char *d_path_with_unreachable(const struct path *path, char *buf, int buflen)
 
 	get_fs_root(current->fs, &root);
 	write_seqlock(&rename_lock);
-	tmp = root;
-	error = path_with_deleted(path, &tmp, &res, &buflen);
-	if (!error && !path_equal(&tmp, &root))
+	error = path_with_deleted(path, &root, &res, &buflen);
+	if (error > 0)
 		error = prepend_unreachable(&res, &buflen);
 	write_sequnlock(&rename_lock);
 	path_put(&root);
@@ -2749,19 +2776,18 @@ SYSCALL_DEFINE2(getcwd, char __user *, buf, unsigned long, size)
 	write_seqlock(&rename_lock);
 	if (!d_unlinked(pwd.dentry)) {
 		unsigned long len;
-		struct path tmp = root;
 		char *cwd = page + PAGE_SIZE;
 		int buflen = PAGE_SIZE;
 
 		prepend(&cwd, &buflen, "\0", 1);
-		error = prepend_path(&pwd, &tmp, &cwd, &buflen);
+		error = prepend_path(&pwd, &root, &cwd, &buflen);
 		write_sequnlock(&rename_lock);
 
-		if (error)
+		if (error < 0)
 			goto out;
 
 		/* Unreachable from current root */
-		if (!path_equal(&tmp, &root)) {
+		if (error > 0) {
 			error = prepend_unreachable(&cwd, &buflen);
 			if (error)
 				goto out;
diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c
index 58609bde3b9..2a834255c75 100644
--- a/fs/ecryptfs/crypto.c
+++ b/fs/ecryptfs/crypto.c
@@ -967,7 +967,7 @@ static void ecryptfs_set_default_crypt_stat_vals(
 
 /**
  * ecryptfs_new_file_context
- * @ecryptfs_dentry: The eCryptfs dentry
+ * @ecryptfs_inode: The eCryptfs inode
  *
  * If the crypto context for the file has not yet been established,
  * this is where we do that.  Establishing a new crypto context
@@ -984,13 +984,13 @@ static void ecryptfs_set_default_crypt_stat_vals(
  *
  * Returns zero on success; non-zero otherwise
  */
-int ecryptfs_new_file_context(struct dentry *ecryptfs_dentry)
+int ecryptfs_new_file_context(struct inode *ecryptfs_inode)
 {
 	struct ecryptfs_crypt_stat *crypt_stat =
-	    &ecryptfs_inode_to_private(ecryptfs_dentry->d_inode)->crypt_stat;
+	    &ecryptfs_inode_to_private(ecryptfs_inode)->crypt_stat;
 	struct ecryptfs_mount_crypt_stat *mount_crypt_stat =
 	    &ecryptfs_superblock_to_private(
-		    ecryptfs_dentry->d_sb)->mount_crypt_stat;
+		    ecryptfs_inode->i_sb)->mount_crypt_stat;
 	int cipher_name_len;
 	int rc = 0;
 
@@ -1299,12 +1299,12 @@ static int ecryptfs_write_headers_virt(char *page_virt, size_t max,
 }
 
 static int
-ecryptfs_write_metadata_to_contents(struct dentry *ecryptfs_dentry,
+ecryptfs_write_metadata_to_contents(struct inode *ecryptfs_inode,
 				    char *virt, size_t virt_len)
 {
 	int rc;
 
-	rc = ecryptfs_write_lower(ecryptfs_dentry->d_inode, virt,
+	rc = ecryptfs_write_lower(ecryptfs_inode, virt,
 				  0, virt_len);
 	if (rc < 0)
 		printk(KERN_ERR "%s: Error attempting to write header "
@@ -1338,7 +1338,8 @@ static unsigned long ecryptfs_get_zeroed_pages(gfp_t gfp_mask,
 
 /**
  * ecryptfs_write_metadata
- * @ecryptfs_dentry: The eCryptfs dentry
+ * @ecryptfs_dentry: The eCryptfs dentry, which should be negative
+ * @ecryptfs_inode: The newly created eCryptfs inode
  *
  * Write the file headers out.  This will likely involve a userspace
  * callout, in which the session key is encrypted with one or more
@@ -1348,10 +1349,11 @@ static unsigned long ecryptfs_get_zeroed_pages(gfp_t gfp_mask,
  *
  * Returns zero on success; non-zero on error
  */
-int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry)
+int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry,
+			    struct inode *ecryptfs_inode)
 {
 	struct ecryptfs_crypt_stat *crypt_stat =
-		&ecryptfs_inode_to_private(ecryptfs_dentry->d_inode)->crypt_stat;
+		&ecryptfs_inode_to_private(ecryptfs_inode)->crypt_stat;
 	unsigned int order;
 	char *virt;
 	size_t virt_len;
@@ -1391,7 +1393,7 @@ int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry)
 		rc = ecryptfs_write_metadata_to_xattr(ecryptfs_dentry, virt,
 						      size);
 	else
-		rc = ecryptfs_write_metadata_to_contents(ecryptfs_dentry, virt,
+		rc = ecryptfs_write_metadata_to_contents(ecryptfs_inode, virt,
 							 virt_len);
 	if (rc) {
 		printk(KERN_ERR "%s: Error writing metadata out to lower file; "
@@ -1943,7 +1945,7 @@ static unsigned char *portable_filename_chars = ("-.0123456789ABCD"
 
 /* We could either offset on every reverse map or just pad some 0x00's
  * at the front here */
-static const unsigned char filename_rev_map[] = {
+static const unsigned char filename_rev_map[256] = {
 	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 7 */
 	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 15 */
 	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 23 */
@@ -1959,7 +1961,7 @@ static const unsigned char filename_rev_map[] = {
 	0x00, 0x26, 0x27, 0x28, 0x29, 0x2A, 0x2B, 0x2C, /* 103 */
 	0x2D, 0x2E, 0x2F, 0x30, 0x31, 0x32, 0x33, 0x34, /* 111 */
 	0x35, 0x36, 0x37, 0x38, 0x39, 0x3A, 0x3B, 0x3C, /* 119 */
-	0x3D, 0x3E, 0x3F
+	0x3D, 0x3E, 0x3F /* 123 - 255 initialized to 0x00 */
 };
 
 /**
diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h
index 54481a3b2c7..a9f29b12fbf 100644
--- a/fs/ecryptfs/ecryptfs_kernel.h
+++ b/fs/ecryptfs/ecryptfs_kernel.h
@@ -584,9 +584,10 @@ int ecryptfs_init_crypt_ctx(struct ecryptfs_crypt_stat *crypt_stat);
 int ecryptfs_write_inode_size_to_metadata(struct inode *ecryptfs_inode);
 int ecryptfs_encrypt_page(struct page *page);
 int ecryptfs_decrypt_page(struct page *page);
-int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry);
+int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry,
+			    struct inode *ecryptfs_inode);
 int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry);
-int ecryptfs_new_file_context(struct dentry *ecryptfs_dentry);
+int ecryptfs_new_file_context(struct inode *ecryptfs_inode);
 void ecryptfs_write_crypt_stat_flags(char *page_virt,
 				     struct ecryptfs_crypt_stat *crypt_stat,
 				     size_t *written);
diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c
index c6ac98cf9ba..d3f95f941c4 100644
--- a/fs/ecryptfs/file.c
+++ b/fs/ecryptfs/file.c
@@ -139,6 +139,27 @@ out:
 	return rc;
 }
 
+static void ecryptfs_vma_close(struct vm_area_struct *vma)
+{
+	filemap_write_and_wait(vma->vm_file->f_mapping);
+}
+
+static const struct vm_operations_struct ecryptfs_file_vm_ops = {
+	.close		= ecryptfs_vma_close,
+	.fault		= filemap_fault,
+};
+
+static int ecryptfs_file_mmap(struct file *file, struct vm_area_struct *vma)
+{
+	int rc;
+
+	rc = generic_file_mmap(file, vma);
+	if (!rc)
+		vma->vm_ops = &ecryptfs_file_vm_ops;
+
+	return rc;
+}
+
 struct kmem_cache *ecryptfs_file_info_cache;
 
 /**
@@ -349,7 +370,7 @@ const struct file_operations ecryptfs_main_fops = {
 #ifdef CONFIG_COMPAT
 	.compat_ioctl = ecryptfs_compat_ioctl,
 #endif
-	.mmap = generic_file_mmap,
+	.mmap = ecryptfs_file_mmap,
 	.open = ecryptfs_open,
 	.flush = ecryptfs_flush,
 	.release = ecryptfs_release,
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index a36d327f152..32f90a3ae63 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -172,22 +172,23 @@ ecryptfs_create_underlying_file(struct inode *lower_dir_inode,
  * it. It will also update the eCryptfs directory inode to mimic the
  * stat of the lower directory inode.
  *
- * Returns zero on success; non-zero on error condition
+ * Returns the new eCryptfs inode on success; an ERR_PTR on error condition
  */
-static int
+static struct inode *
 ecryptfs_do_create(struct inode *directory_inode,
 		   struct dentry *ecryptfs_dentry, int mode)
 {
 	int rc;
 	struct dentry *lower_dentry;
 	struct dentry *lower_dir_dentry;
+	struct inode *inode;
 
 	lower_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry);
 	lower_dir_dentry = lock_parent(lower_dentry);
 	if (IS_ERR(lower_dir_dentry)) {
 		ecryptfs_printk(KERN_ERR, "Error locking directory of "
 				"dentry\n");
-		rc = PTR_ERR(lower_dir_dentry);
+		inode = ERR_CAST(lower_dir_dentry);
 		goto out;
 	}
 	rc = ecryptfs_create_underlying_file(lower_dir_dentry->d_inode,
@@ -195,20 +196,19 @@ ecryptfs_do_create(struct inode *directory_inode,
 	if (rc) {
 		printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
 		       "rc = [%d]\n", __func__, rc);
+		inode = ERR_PTR(rc);
 		goto out_lock;
 	}
-	rc = ecryptfs_interpose(lower_dentry, ecryptfs_dentry,
-				directory_inode->i_sb);
-	if (rc) {
-		ecryptfs_printk(KERN_ERR, "Failure in ecryptfs_interpose\n");
+	inode = __ecryptfs_get_inode(lower_dentry->d_inode,
+				     directory_inode->i_sb);
+	if (IS_ERR(inode))
 		goto out_lock;
-	}
 	fsstack_copy_attr_times(directory_inode, lower_dir_dentry->d_inode);
 	fsstack_copy_inode_size(directory_inode, lower_dir_dentry->d_inode);
 out_lock:
 	unlock_dir(lower_dir_dentry);
 out:
-	return rc;
+	return inode;
 }
 
 /**
@@ -219,26 +219,26 @@ out:
  *
  * Returns zero on success
  */
-static int ecryptfs_initialize_file(struct dentry *ecryptfs_dentry)
+static int ecryptfs_initialize_file(struct dentry *ecryptfs_dentry,
+				    struct inode *ecryptfs_inode)
 {
 	struct ecryptfs_crypt_stat *crypt_stat =
-		&ecryptfs_inode_to_private(ecryptfs_dentry->d_inode)->crypt_stat;
+		&ecryptfs_inode_to_private(ecryptfs_inode)->crypt_stat;
 	int rc = 0;
 
-	if (S_ISDIR(ecryptfs_dentry->d_inode->i_mode)) {
+	if (S_ISDIR(ecryptfs_inode->i_mode)) {
 		ecryptfs_printk(KERN_DEBUG, "This is a directory\n");
 		crypt_stat->flags &= ~(ECRYPTFS_ENCRYPTED);
 		goto out;
 	}
 	ecryptfs_printk(KERN_DEBUG, "Initializing crypto context\n");
-	rc = ecryptfs_new_file_context(ecryptfs_dentry);
+	rc = ecryptfs_new_file_context(ecryptfs_inode);
 	if (rc) {
 		ecryptfs_printk(KERN_ERR, "Error creating new file "
 				"context; rc = [%d]\n", rc);
 		goto out;
 	}
-	rc = ecryptfs_get_lower_file(ecryptfs_dentry,
-				     ecryptfs_dentry->d_inode);
+	rc = ecryptfs_get_lower_file(ecryptfs_dentry, ecryptfs_inode);
 	if (rc) {
 		printk(KERN_ERR "%s: Error attempting to initialize "
 			"the lower file for the dentry with name "
@@ -246,10 +246,10 @@ static int ecryptfs_initialize_file(struct dentry *ecryptfs_dentry)
 			ecryptfs_dentry->d_name.name, rc);
 		goto out;
 	}
-	rc = ecryptfs_write_metadata(ecryptfs_dentry);
+	rc = ecryptfs_write_metadata(ecryptfs_dentry, ecryptfs_inode);
 	if (rc)
 		printk(KERN_ERR "Error writing headers; rc = [%d]\n", rc);
-	ecryptfs_put_lower_file(ecryptfs_dentry->d_inode);
+	ecryptfs_put_lower_file(ecryptfs_inode);
 out:
 	return rc;
 }
@@ -269,18 +269,28 @@ static int
 ecryptfs_create(struct inode *directory_inode, struct dentry *ecryptfs_dentry,
 		int mode, struct nameidata *nd)
 {
+	struct inode *ecryptfs_inode;
 	int rc;
 
-	/* ecryptfs_do_create() calls ecryptfs_interpose() */
-	rc = ecryptfs_do_create(directory_inode, ecryptfs_dentry, mode);
-	if (unlikely(rc)) {
+	ecryptfs_inode = ecryptfs_do_create(directory_inode, ecryptfs_dentry,
+					    mode);
+	if (unlikely(IS_ERR(ecryptfs_inode))) {
 		ecryptfs_printk(KERN_WARNING, "Failed to create file in"
 				"lower filesystem\n");
+		rc = PTR_ERR(ecryptfs_inode);
 		goto out;
 	}
 	/* At this point, a file exists on "disk"; we need to make sure
 	 * that this on disk file is prepared to be an ecryptfs file */
-	rc = ecryptfs_initialize_file(ecryptfs_dentry);
+	rc = ecryptfs_initialize_file(ecryptfs_dentry, ecryptfs_inode);
+	if (rc) {
+		drop_nlink(ecryptfs_inode);
+		unlock_new_inode(ecryptfs_inode);
+		iput(ecryptfs_inode);
+		goto out;
+	}
+	d_instantiate(ecryptfs_dentry, ecryptfs_inode);
+	unlock_new_inode(ecryptfs_inode);
 out:
 	return rc;
 }
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index f6dba4505f1..12ccacda44e 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -565,7 +565,7 @@ ext4_fsblk_t ext4_count_free_clusters(struct super_block *sb)
 	brelse(bitmap_bh);
 	printk(KERN_DEBUG "ext4_count_free_clusters: stored = %llu"
 	       ", computed = %llu, %llu\n",
-	       EXT4_B2C(sbi, ext4_free_blocks_count(es)),
+	       EXT4_B2C(EXT4_SB(sb), ext4_free_blocks_count(es)),
 	       desc_count, bitmap_count);
 	return bitmap_count;
 #else
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 240f6e2dc7e..848f436df29 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2270,6 +2270,7 @@ retry:
 			ext4_msg(inode->i_sb, KERN_CRIT, "%s: jbd2_start: "
 			       "%ld pages, ino %lu; err %d", __func__,
 				wbc->nr_to_write, inode->i_ino, ret);
+			blk_finish_plug(&plug);
 			goto out_writepages;
 		}
 
@@ -2806,8 +2807,8 @@ out:
 	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
 
 	/* queue the work to convert unwritten extents to written */
-	queue_work(wq, &io_end->work);
 	iocb->private = NULL;
+	queue_work(wq, &io_end->work);
 
 	/* XXX: probably should move into the real I/O completion handler */
 	inode_dio_done(inode);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 9953d80145a..3858767ec67 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1683,7 +1683,9 @@ static int parse_options(char *options, struct super_block *sb,
 			data_opt = EXT4_MOUNT_WRITEBACK_DATA;
 		datacheck:
 			if (is_remount) {
-				if (test_opt(sb, DATA_FLAGS) != data_opt) {
+				if (!sbi->s_journal)
+					ext4_msg(sb, KERN_WARNING, "Remounting file system with no journal so ignoring journalled data option");
+				else if (test_opt(sb, DATA_FLAGS) != data_opt) {
 					ext4_msg(sb, KERN_ERR,
 						"Cannot change data mode on remount");
 					return 0;
@@ -3099,8 +3101,6 @@ static void ext4_destroy_lazyinit_thread(void)
 }
 
 static int ext4_fill_super(struct super_block *sb, void *data, int silent)
-				__releases(kernel_lock)
-				__acquires(kernel_lock)
 {
 	char *orig_data = kstrdup(data, GFP_KERNEL);
 	struct buffer_head *bh;
diff --git a/fs/hfs/trans.c b/fs/hfs/trans.c
index e673a88b8ae..b1ce4c7ad3f 100644
--- a/fs/hfs/trans.c
+++ b/fs/hfs/trans.c
@@ -40,6 +40,8 @@ int hfs_mac2asc(struct super_block *sb, char *out, const struct hfs_name *in)
 
 	src = in->name;
 	srclen = in->len;
+	if (srclen > HFS_NAMELEN)
+		srclen = HFS_NAMELEN;
 	dst = out;
 	dstlen = HFS_MAX_NAMELEN;
 	if (nls_io) {
diff --git a/fs/minix/bitmap.c b/fs/minix/bitmap.c
index 3f32bcb0d9b..ef175cb8cfd 100644
--- a/fs/minix/bitmap.c
+++ b/fs/minix/bitmap.c
@@ -16,38 +16,26 @@
 #include <linux/bitops.h>
 #include <linux/sched.h>
 
-static const int nibblemap[] = { 4,3,3,2,3,2,2,1,3,2,2,1,2,1,1,0 };
-
 static DEFINE_SPINLOCK(bitmap_lock);
 
-static unsigned long count_free(struct buffer_head *map[], unsigned numblocks, __u32 numbits)
+/*
+ * bitmap consists of blocks filled with 16bit words
+ * bit set == busy, bit clear == free
+ * endianness is a mess, but for counting zero bits it really doesn't matter...
+ */
+static __u32 count_free(struct buffer_head *map[], unsigned blocksize, __u32 numbits)
 {
-	unsigned i, j, sum = 0;
-	struct buffer_head *bh;
-  
-	for (i=0; i<numblocks-1; i++) {
-		if (!(bh=map[i])) 
-			return(0);
-		for (j=0; j<bh->b_size; j++)
-			sum += nibblemap[bh->b_data[j] & 0xf]
-				+ nibblemap[(bh->b_data[j]>>4) & 0xf];
-	}
+	__u32 sum = 0;
+	unsigned blocks = DIV_ROUND_UP(numbits, blocksize * 8);
 
-	if (numblocks==0 || !(bh=map[numblocks-1]))
-		return(0);
-	i = ((numbits - (numblocks-1) * bh->b_size * 8) / 16) * 2;
-	for (j=0; j<i; j++) {
-		sum += nibblemap[bh->b_data[j] & 0xf]
-			+ nibblemap[(bh->b_data[j]>>4) & 0xf];
+	while (blocks--) {
+		unsigned words = blocksize / 2;
+		__u16 *p = (__u16 *)(*map++)->b_data;
+		while (words--)
+			sum += 16 - hweight16(*p++);
 	}
 
-	i = numbits%16;
-	if (i!=0) {
-		i = *(__u16 *)(&bh->b_data[j]) | ~((1<<i) - 1);
-		sum += nibblemap[i & 0xf] + nibblemap[(i>>4) & 0xf];
-		sum += nibblemap[(i>>8) & 0xf] + nibblemap[(i>>12) & 0xf];
-	}
-	return(sum);
+	return sum;
 }
 
 void minix_free_block(struct inode *inode, unsigned long block)
@@ -105,10 +93,12 @@ int minix_new_block(struct inode * inode)
 	return 0;
 }
 
-unsigned long minix_count_free_blocks(struct minix_sb_info *sbi)
+unsigned long minix_count_free_blocks(struct super_block *sb)
 {
-	return (count_free(sbi->s_zmap, sbi->s_zmap_blocks,
-		sbi->s_nzones - sbi->s_firstdatazone + 1)
+	struct minix_sb_info *sbi = minix_sb(sb);
+	u32 bits = sbi->s_nzones - (sbi->s_firstdatazone + 1);
+
+	return (count_free(sbi->s_zmap, sb->s_blocksize, bits)
 		<< sbi->s_log_zone_size);
 }
 
@@ -273,7 +263,10 @@ struct inode *minix_new_inode(const struct inode *dir, int mode, int *error)
 	return inode;
 }
 
-unsigned long minix_count_free_inodes(struct minix_sb_info *sbi)
+unsigned long minix_count_free_inodes(struct super_block *sb)
 {
-	return count_free(sbi->s_imap, sbi->s_imap_blocks, sbi->s_ninodes + 1);
+	struct minix_sb_info *sbi = minix_sb(sb);
+	u32 bits = sbi->s_ninodes + 1;
+
+	return count_free(sbi->s_imap, sb->s_blocksize, bits);
 }
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index 64cdcd662ff..1d9e33966db 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -279,6 +279,27 @@ static int minix_fill_super(struct super_block *s, void *data, int silent)
  	else if (sbi->s_mount_state & MINIX_ERROR_FS)
 		printk("MINIX-fs: mounting file system with errors, "
 			"running fsck is recommended\n");
+
+	/* Apparently minix can create filesystems that allocate more blocks for
+	 * the bitmaps than needed.  We simply ignore that, but verify it didn't
+	 * create one with not enough blocks and bail out if so.
+	 */
+	block = minix_blocks_needed(sbi->s_ninodes, s->s_blocksize);
+	if (sbi->s_imap_blocks < block) {
+		printk("MINIX-fs: file system does not have enough "
+				"imap blocks allocated.  Refusing to mount\n");
+		goto out_iput;
+	}
+
+	block = minix_blocks_needed(
+			(sbi->s_nzones - (sbi->s_firstdatazone + 1)),
+			s->s_blocksize);
+	if (sbi->s_zmap_blocks < block) {
+		printk("MINIX-fs: file system does not have enough "
+				"zmap blocks allocated.  Refusing to mount.\n");
+		goto out_iput;
+	}
+
 	return 0;
 
 out_iput:
@@ -339,10 +360,10 @@ static int minix_statfs(struct dentry *dentry, struct kstatfs *buf)
 	buf->f_type = sb->s_magic;
 	buf->f_bsize = sb->s_blocksize;
 	buf->f_blocks = (sbi->s_nzones - sbi->s_firstdatazone) << sbi->s_log_zone_size;
-	buf->f_bfree = minix_count_free_blocks(sbi);
+	buf->f_bfree = minix_count_free_blocks(sb);
 	buf->f_bavail = buf->f_bfree;
 	buf->f_files = sbi->s_ninodes;
-	buf->f_ffree = minix_count_free_inodes(sbi);
+	buf->f_ffree = minix_count_free_inodes(sb);
 	buf->f_namelen = sbi->s_namelen;
 	buf->f_fsid.val[0] = (u32)id;
 	buf->f_fsid.val[1] = (u32)(id >> 32);
diff --git a/fs/minix/minix.h b/fs/minix/minix.h
index 341e2122879..26bbd55e82e 100644
--- a/fs/minix/minix.h
+++ b/fs/minix/minix.h
@@ -48,10 +48,10 @@ extern struct minix_inode * minix_V1_raw_inode(struct super_block *, ino_t, stru
 extern struct minix2_inode * minix_V2_raw_inode(struct super_block *, ino_t, struct buffer_head **);
 extern struct inode * minix_new_inode(const struct inode *, int, int *);
 extern void minix_free_inode(struct inode * inode);
-extern unsigned long minix_count_free_inodes(struct minix_sb_info *sbi);
+extern unsigned long minix_count_free_inodes(struct super_block *sb);
 extern int minix_new_block(struct inode * inode);
 extern void minix_free_block(struct inode *inode, unsigned long block);
-extern unsigned long minix_count_free_blocks(struct minix_sb_info *sbi);
+extern unsigned long minix_count_free_blocks(struct super_block *sb);
 extern int minix_getattr(struct vfsmount *, struct dentry *, struct kstat *);
 extern int minix_prepare_chunk(struct page *page, loff_t pos, unsigned len);
 
@@ -88,6 +88,11 @@ static inline struct minix_inode_info *minix_i(struct inode *inode)
 	return list_entry(inode, struct minix_inode_info, vfs_inode);
 }
 
+static inline unsigned minix_blocks_needed(unsigned bits, unsigned blocksize)
+{
+	return DIV_ROUND_UP(bits, blocksize * 8);
+}
+
 #if defined(CONFIG_MINIX_FS_NATIVE_ENDIAN) && \
 	defined(CONFIG_MINIX_FS_BIG_ENDIAN_16BIT_INDEXED)
 
@@ -125,7 +130,7 @@ static inline int minix_find_first_zero_bit(const void *vaddr, unsigned size)
 	if (!size)
 		return 0;
 
-	size = (size >> 4) + ((size & 15) > 0);
+	size >>= 4;
 	while (*p++ == 0xffff) {
 		if (--size == 0)
 			return (p - addr) << 4;
diff --git a/fs/namespace.c b/fs/namespace.c
index e5e1c7d1839..cfc6d4448aa 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1048,15 +1048,12 @@ static int show_mountinfo(struct seq_file *m, void *v)
 	if (err)
 		goto out;
 	seq_putc(m, ' ');
-	seq_path_root(m, &mnt_path, &root, " \t\n\\");
-	if (root.mnt != p->root.mnt || root.dentry != p->root.dentry) {
-		/*
-		 * Mountpoint is outside root, discard that one.  Ugly,
-		 * but less so than trying to do that in iterator in a
-		 * race-free way (due to renames).
-		 */
-		return SEQ_SKIP;
-	}
+
+	/* mountpoints outside of chroot jail will give SEQ_SKIP on this */
+	err = seq_path_root(m, &mnt_path, &root, " \t\n\\");
+	if (err)
+		goto out;
+
 	seq_puts(m, mnt->mnt_flags & MNT_READONLY ? " ro" : " rw");
 	show_mnt_opts(m, mnt);
 
@@ -2483,11 +2480,43 @@ struct mnt_namespace *create_mnt_ns(struct vfsmount *mnt)
 		__mnt_make_longterm(mnt);
 		new_ns->root = mnt;
 		list_add(&new_ns->list, &new_ns->root->mnt_list);
+	} else {
+		mntput(mnt);
 	}
 	return new_ns;
 }
 EXPORT_SYMBOL(create_mnt_ns);
 
+struct dentry *mount_subtree(struct vfsmount *mnt, const char *name)
+{
+	struct mnt_namespace *ns;
+	struct super_block *s;
+	struct path path;
+	int err;
+
+	ns = create_mnt_ns(mnt);
+	if (IS_ERR(ns))
+		return ERR_CAST(ns);
+
+	err = vfs_path_lookup(mnt->mnt_root, mnt,
+			name, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &path);
+
+	put_mnt_ns(ns);
+
+	if (err)
+		return ERR_PTR(err);
+
+	/* trade a vfsmount reference for active sb one */
+	s = path.mnt->mnt_sb;
+	atomic_inc(&s->s_active);
+	mntput(path.mnt);
+	/* lock the sucker */
+	down_write(&s->s_umount);
+	/* ... and return the root of (sub)tree on it */
+	return path.dentry;
+}
+EXPORT_SYMBOL(mount_subtree);
+
 SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name,
 		char __user *, type, unsigned long, flags, void __user *, data)
 {
@@ -2744,3 +2773,8 @@ void kern_unmount(struct vfsmount *mnt)
 	}
 }
 EXPORT_SYMBOL(kern_unmount);
+
+bool our_mnt(struct vfsmount *mnt)
+{
+	return check_mnt(mnt);
+}
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index b238d95ac48..ac289909814 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1468,12 +1468,12 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 				res = NULL;
 				goto out;
 			/* This turned out not to be a regular file */
+			case -EISDIR:
 			case -ENOTDIR:
 				goto no_open;
 			case -ELOOP:
 				if (!(nd->intent.open.flags & O_NOFOLLOW))
 					goto no_open;
-			/* case -EISDIR: */
 			/* case -EINVAL: */
 			default:
 				res = ERR_CAST(inode);
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 0a1f8312b4d..eca56d4b39c 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -40,48 +40,8 @@
 
 #define NFSDBG_FACILITY		NFSDBG_FILE
 
-static int nfs_file_open(struct inode *, struct file *);
-static int nfs_file_release(struct inode *, struct file *);
-static loff_t nfs_file_llseek(struct file *file, loff_t offset, int origin);
-static int  nfs_file_mmap(struct file *, struct vm_area_struct *);
-static ssize_t nfs_file_splice_read(struct file *filp, loff_t *ppos,
-					struct pipe_inode_info *pipe,
-					size_t count, unsigned int flags);
-static ssize_t nfs_file_read(struct kiocb *, const struct iovec *iov,
-				unsigned long nr_segs, loff_t pos);
-static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
-					struct file *filp, loff_t *ppos,
-					size_t count, unsigned int flags);
-static ssize_t nfs_file_write(struct kiocb *, const struct iovec *iov,
-				unsigned long nr_segs, loff_t pos);
-static int  nfs_file_flush(struct file *, fl_owner_t id);
-static int  nfs_file_fsync(struct file *, loff_t, loff_t, int datasync);
-static int nfs_check_flags(int flags);
-static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl);
-static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl);
-static int nfs_setlease(struct file *file, long arg, struct file_lock **fl);
-
 static const struct vm_operations_struct nfs_file_vm_ops;
 
-const struct file_operations nfs_file_operations = {
-	.llseek		= nfs_file_llseek,
-	.read		= do_sync_read,
-	.write		= do_sync_write,
-	.aio_read	= nfs_file_read,
-	.aio_write	= nfs_file_write,
-	.mmap		= nfs_file_mmap,
-	.open		= nfs_file_open,
-	.flush		= nfs_file_flush,
-	.release	= nfs_file_release,
-	.fsync		= nfs_file_fsync,
-	.lock		= nfs_lock,
-	.flock		= nfs_flock,
-	.splice_read	= nfs_file_splice_read,
-	.splice_write	= nfs_file_splice_write,
-	.check_flags	= nfs_check_flags,
-	.setlease	= nfs_setlease,
-};
-
 const struct inode_operations nfs_file_inode_operations = {
 	.permission	= nfs_permission,
 	.getattr	= nfs_getattr,
@@ -886,3 +846,54 @@ static int nfs_setlease(struct file *file, long arg, struct file_lock **fl)
 			file->f_path.dentry->d_name.name, arg);
 	return -EINVAL;
 }
+
+const struct file_operations nfs_file_operations = {
+	.llseek		= nfs_file_llseek,
+	.read		= do_sync_read,
+	.write		= do_sync_write,
+	.aio_read	= nfs_file_read,
+	.aio_write	= nfs_file_write,
+	.mmap		= nfs_file_mmap,
+	.open		= nfs_file_open,
+	.flush		= nfs_file_flush,
+	.release	= nfs_file_release,
+	.fsync		= nfs_file_fsync,
+	.lock		= nfs_lock,
+	.flock		= nfs_flock,
+	.splice_read	= nfs_file_splice_read,
+	.splice_write	= nfs_file_splice_write,
+	.check_flags	= nfs_check_flags,
+	.setlease	= nfs_setlease,
+};
+
+#ifdef CONFIG_NFS_V4
+static int
+nfs4_file_open(struct inode *inode, struct file *filp)
+{
+	/*
+	 * NFSv4 opens are handled in d_lookup and d_revalidate. If we get to
+	 * this point, then something is very wrong
+	 */
+	dprintk("NFS: %s called! inode=%p filp=%p\n", __func__, inode, filp);
+	return -ENOTDIR;
+}
+
+const struct file_operations nfs4_file_operations = {
+	.llseek		= nfs_file_llseek,
+	.read		= do_sync_read,
+	.write		= do_sync_write,
+	.aio_read	= nfs_file_read,
+	.aio_write	= nfs_file_write,
+	.mmap		= nfs_file_mmap,
+	.open		= nfs4_file_open,
+	.flush		= nfs_file_flush,
+	.release	= nfs_file_release,
+	.fsync		= nfs_file_fsync,
+	.lock		= nfs_lock,
+	.flock		= nfs_flock,
+	.splice_read	= nfs_file_splice_read,
+	.splice_write	= nfs_file_splice_write,
+	.check_flags	= nfs_check_flags,
+	.setlease	= nfs_setlease,
+};
+#endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index c07a55aec83..50a15fa8cf9 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -291,7 +291,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 		 */
 		inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->file_inode_ops;
 		if (S_ISREG(inode->i_mode)) {
-			inode->i_fop = &nfs_file_operations;
+			inode->i_fop = NFS_SB(sb)->nfs_client->rpc_ops->file_ops;
 			inode->i_data.a_ops = &nfs_file_aops;
 			inode->i_data.backing_dev_info = &NFS_SB(sb)->backing_dev_info;
 		} else if (S_ISDIR(inode->i_mode)) {
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index c1a1bd8ddf1..3f4d95751d5 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -299,6 +299,8 @@ extern void nfs_read_prepare(struct rpc_task *task, void *calldata);
 extern int nfs_generic_pagein(struct nfs_pageio_descriptor *desc,
 		struct list_head *head);
 
+extern void nfs_pageio_init_read_mds(struct nfs_pageio_descriptor *pgio,
+		struct inode *inode);
 extern void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio);
 extern void nfs_readdata_release(struct nfs_read_data *rdata);
 
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 85f1690ca08..d4bc9ed9174 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -853,6 +853,7 @@ const struct nfs_rpc_ops nfs_v3_clientops = {
 	.dentry_ops	= &nfs_dentry_operations,
 	.dir_inode_ops	= &nfs3_dir_inode_operations,
 	.file_inode_ops	= &nfs3_file_inode_operations,
+	.file_ops	= &nfs_file_operations,
 	.getroot	= nfs3_proc_get_root,
 	.getattr	= nfs3_proc_getattr,
 	.setattr	= nfs3_proc_setattr,
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b60fddf606f..be2bbac1381 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2464,8 +2464,7 @@ static int nfs4_proc_lookup(struct rpc_clnt *clnt, struct inode *dir, struct qst
 		case -NFS4ERR_BADNAME:
 			return -ENOENT;
 		case -NFS4ERR_MOVED:
-			err = nfs4_get_referral(dir, name, fattr, fhandle);
-			break;
+			return nfs4_get_referral(dir, name, fattr, fhandle);
 		case -NFS4ERR_WRONGSEC:
 			nfs_fixup_secinfo_attributes(fattr, fhandle);
 		}
@@ -6253,6 +6252,7 @@ const struct nfs_rpc_ops nfs_v4_clientops = {
 	.dentry_ops	= &nfs4_dentry_operations,
 	.dir_inode_ops	= &nfs4_dir_inode_operations,
 	.file_inode_ops	= &nfs4_file_inode_operations,
+	.file_ops	= &nfs4_file_operations,
 	.getroot	= nfs4_proc_get_root,
 	.getattr	= nfs4_proc_getattr,
 	.setattr	= nfs4_proc_setattr,
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index baf73536bc0..8e672a2b2d6 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -1260,6 +1260,25 @@ pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
 }
 EXPORT_SYMBOL_GPL(pnfs_generic_pg_writepages);
 
+static void pnfs_ld_handle_read_error(struct nfs_read_data *data)
+{
+	struct nfs_pageio_descriptor pgio;
+
+	put_lseg(data->lseg);
+	data->lseg = NULL;
+	dprintk("pnfs write error = %d\n", data->pnfs_error);
+
+	nfs_pageio_init_read_mds(&pgio, data->inode);
+
+	while (!list_empty(&data->pages)) {
+		struct nfs_page *req = nfs_list_entry(data->pages.next);
+
+		nfs_list_remove_request(req);
+		nfs_pageio_add_request(&pgio, req);
+	}
+	nfs_pageio_complete(&pgio);
+}
+
 /*
  * Called by non rpc-based layout drivers
  */
@@ -1268,11 +1287,8 @@ void pnfs_ld_read_done(struct nfs_read_data *data)
 	if (likely(!data->pnfs_error)) {
 		__nfs4_read_done_cb(data);
 		data->mds_ops->rpc_call_done(&data->task, data);
-	} else {
-		put_lseg(data->lseg);
-		data->lseg = NULL;
-		dprintk("pnfs write error = %d\n", data->pnfs_error);
-	}
+	} else
+		pnfs_ld_handle_read_error(data);
 	data->mds_ops->rpc_release(data);
 }
 EXPORT_SYMBOL_GPL(pnfs_ld_read_done);
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index ac40b8535d7..f48125da198 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -710,6 +710,7 @@ const struct nfs_rpc_ops nfs_v2_clientops = {
 	.dentry_ops	= &nfs_dentry_operations,
 	.dir_inode_ops	= &nfs_dir_inode_operations,
 	.file_inode_ops	= &nfs_file_inode_operations,
+	.file_ops	= &nfs_file_operations,
 	.getroot	= nfs_proc_get_root,
 	.getattr	= nfs_proc_getattr,
 	.setattr	= nfs_proc_setattr,
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index 8b48ec63f72..cfa175c223d 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -109,7 +109,7 @@ static void nfs_readpage_truncate_uninitialised_page(struct nfs_read_data *data)
 	}
 }
 
-static void nfs_pageio_init_read_mds(struct nfs_pageio_descriptor *pgio,
+void nfs_pageio_init_read_mds(struct nfs_pageio_descriptor *pgio,
 		struct inode *inode)
 {
 	nfs_pageio_init(pgio, inode, &nfs_pageio_read_ops,
@@ -534,23 +534,13 @@ static void nfs_readpage_result_full(struct rpc_task *task, void *calldata)
 static void nfs_readpage_release_full(void *calldata)
 {
 	struct nfs_read_data *data = calldata;
-	struct nfs_pageio_descriptor pgio;
 
-	if (data->pnfs_error) {
-		nfs_pageio_init_read_mds(&pgio, data->inode);
-		pgio.pg_recoalesce = 1;
-	}
 	while (!list_empty(&data->pages)) {
 		struct nfs_page *req = nfs_list_entry(data->pages.next);
 
 		nfs_list_remove_request(req);
-		if (!data->pnfs_error)
-			nfs_readpage_release(req);
-		else
-			nfs_pageio_add_request(&pgio, req);
+		nfs_readpage_release(req);
 	}
-	if (data->pnfs_error)
-		nfs_pageio_complete(&pgio);
 	nfs_readdata_release(calldata);
 }
 
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 480b3b6bf71..134777406ee 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2787,43 +2787,18 @@ static void nfs_referral_loop_unprotect(void)
 static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 		const char *export_path)
 {
-	struct mnt_namespace *ns_private;
-	struct super_block *s;
 	struct dentry *dentry;
-	struct path path;
-	int ret;
-
-	ns_private = create_mnt_ns(root_mnt);
-	ret = PTR_ERR(ns_private);
-	if (IS_ERR(ns_private))
-		goto out_mntput;
-
-	ret = nfs_referral_loop_protect();
-	if (ret != 0)
-		goto out_put_mnt_ns;
+	int ret = nfs_referral_loop_protect();
 
-	ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt,
-			export_path, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &path);
+	if (ret) {
+		mntput(root_mnt);
+		return ERR_PTR(ret);
+	}
 
+	dentry = mount_subtree(root_mnt, export_path);
 	nfs_referral_loop_unprotect();
-	put_mnt_ns(ns_private);
-
-	if (ret != 0)
-		goto out_err;
-
-	s = path.mnt->mnt_sb;
-	atomic_inc(&s->s_active);
-	dentry = dget(path.dentry);
 
-	path_put(&path);
-	down_write(&s->s_umount);
 	return dentry;
-out_put_mnt_ns:
-	put_mnt_ns(ns_private);
-out_mntput:
-	mntput(root_mnt);
-out_err:
-	return ERR_PTR(ret);
 }
 
 static struct dentry *nfs4_try_mount(int flags, const char *dev_name,
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index ed553c60de8..3165aebb43c 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -5699,7 +5699,7 @@ int ocfs2_remove_btree_range(struct inode *inode,
 					   OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out;
+		goto out_commit;
 	}
 
 	dquot_free_space_nodirty(inode,
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index c1efe939c77..78b68af3b0e 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -290,7 +290,15 @@ static int ocfs2_readpage(struct file *file, struct page *page)
 	}
 
 	if (down_read_trylock(&oi->ip_alloc_sem) == 0) {
+		/*
+		 * Unlock the page and cycle ip_alloc_sem so that we don't
+		 * busyloop waiting for ip_alloc_sem to unlock
+		 */
 		ret = AOP_TRUNCATED_PAGE;
+		unlock_page(page);
+		unlock = 0;
+		down_read(&oi->ip_alloc_sem);
+		up_read(&oi->ip_alloc_sem);
 		goto out_inode_unlock;
 	}
 
@@ -563,6 +571,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 {
 	struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
 	int level;
+	wait_queue_head_t *wq = ocfs2_ioend_wq(inode);
 
 	/* this io's submitter should not have unlocked this before we could */
 	BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
@@ -570,6 +579,15 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 	if (ocfs2_iocb_is_sem_locked(iocb))
 		ocfs2_iocb_clear_sem_locked(iocb);
 
+	if (ocfs2_iocb_is_unaligned_aio(iocb)) {
+		ocfs2_iocb_clear_unaligned_aio(iocb);
+
+		if (atomic_dec_and_test(&OCFS2_I(inode)->ip_unaligned_aio) &&
+		    waitqueue_active(wq)) {
+			wake_up_all(wq);
+		}
+	}
+
 	ocfs2_iocb_clear_rw_locked(iocb);
 
 	level = ocfs2_iocb_rw_locked_level(iocb);
@@ -863,6 +881,12 @@ struct ocfs2_write_ctxt {
 	struct page			*w_target_page;
 
 	/*
+	 * w_target_locked is used for page_mkwrite path indicating no unlocking
+	 * against w_target_page in ocfs2_write_end_nolock.
+	 */
+	unsigned int			w_target_locked:1;
+
+	/*
 	 * ocfs2_write_end() uses this to know what the real range to
 	 * write in the target should be.
 	 */
@@ -895,6 +919,24 @@ void ocfs2_unlock_and_free_pages(struct page **pages, int num_pages)
 
 static void ocfs2_free_write_ctxt(struct ocfs2_write_ctxt *wc)
 {
+	int i;
+
+	/*
+	 * w_target_locked is only set to true in the page_mkwrite() case.
+	 * The intent is to allow us to lock the target page from write_begin()
+	 * to write_end(). The caller must hold a ref on w_target_page.
+	 */
+	if (wc->w_target_locked) {
+		BUG_ON(!wc->w_target_page);
+		for (i = 0; i < wc->w_num_pages; i++) {
+			if (wc->w_target_page == wc->w_pages[i]) {
+				wc->w_pages[i] = NULL;
+				break;
+			}
+		}
+		mark_page_accessed(wc->w_target_page);
+		page_cache_release(wc->w_target_page);
+	}
 	ocfs2_unlock_and_free_pages(wc->w_pages, wc->w_num_pages);
 
 	brelse(wc->w_di_bh);
@@ -1132,20 +1174,17 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping,
 			 */
 			lock_page(mmap_page);
 
+			/* Exit and let the caller retry */
 			if (mmap_page->mapping != mapping) {
+				WARN_ON(mmap_page->mapping);
 				unlock_page(mmap_page);
-				/*
-				 * Sanity check - the locking in
-				 * ocfs2_pagemkwrite() should ensure
-				 * that this code doesn't trigger.
-				 */
-				ret = -EINVAL;
-				mlog_errno(ret);
+				ret = -EAGAIN;
 				goto out;
 			}
 
 			page_cache_get(mmap_page);
 			wc->w_pages[i] = mmap_page;
+			wc->w_target_locked = true;
 		} else {
 			wc->w_pages[i] = find_or_create_page(mapping, index,
 							     GFP_NOFS);
@@ -1160,6 +1199,8 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping,
 			wc->w_target_page = wc->w_pages[i];
 	}
 out:
+	if (ret)
+		wc->w_target_locked = false;
 	return ret;
 }
 
@@ -1817,11 +1858,23 @@ try_again:
 	 */
 	ret = ocfs2_grab_pages_for_write(mapping, wc, wc->w_cpos, pos, len,
 					 cluster_of_pages, mmap_page);
-	if (ret) {
+	if (ret && ret != -EAGAIN) {
 		mlog_errno(ret);
 		goto out_quota;
 	}
 
+	/*
+	 * ocfs2_grab_pages_for_write() returns -EAGAIN if it could not lock
+	 * the target page. In this case, we exit with no error and no target
+	 * page. This will trigger the caller, page_mkwrite(), to re-try
+	 * the operation.
+	 */
+	if (ret == -EAGAIN) {
+		BUG_ON(wc->w_target_page);
+		ret = 0;
+		goto out_quota;
+	}
+
 	ret = ocfs2_write_cluster_by_desc(mapping, data_ac, meta_ac, wc, pos,
 					  len);
 	if (ret) {
diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h
index 75cf3ad987a..ffb2da370a9 100644
--- a/fs/ocfs2/aops.h
+++ b/fs/ocfs2/aops.h
@@ -78,6 +78,7 @@ enum ocfs2_iocb_lock_bits {
 	OCFS2_IOCB_RW_LOCK = 0,
 	OCFS2_IOCB_RW_LOCK_LEVEL,
 	OCFS2_IOCB_SEM,
+	OCFS2_IOCB_UNALIGNED_IO,
 	OCFS2_IOCB_NUM_LOCKS
 };
 
@@ -91,4 +92,17 @@ enum ocfs2_iocb_lock_bits {
 	clear_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private)
 #define ocfs2_iocb_is_sem_locked(iocb) \
 	test_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private)
+
+#define ocfs2_iocb_set_unaligned_aio(iocb) \
+	set_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+#define ocfs2_iocb_clear_unaligned_aio(iocb) \
+	clear_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+#define ocfs2_iocb_is_unaligned_aio(iocb) \
+	test_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+
+#define OCFS2_IOEND_WQ_HASH_SZ	37
+#define ocfs2_ioend_wq(v)   (&ocfs2__ioend_wq[((unsigned long)(v)) %\
+					    OCFS2_IOEND_WQ_HASH_SZ])
+extern wait_queue_head_t ocfs2__ioend_wq[OCFS2_IOEND_WQ_HASH_SZ];
+
 #endif /* OCFS2_FILE_H */
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 9a3e6bbff27..a4e855e3690 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -216,6 +216,7 @@ struct o2hb_region {
 
 	struct list_head	hr_all_item;
 	unsigned		hr_unclean_stop:1,
+				hr_aborted_start:1,
 				hr_item_pinned:1,
 				hr_item_dropped:1;
 
@@ -254,6 +255,10 @@ struct o2hb_region {
 	 * a more complete api that doesn't lead to this sort of fragility. */
 	atomic_t		hr_steady_iterations;
 
+	/* terminate o2hb thread if it does not reach steady state
+	 * (hr_steady_iterations == 0) within hr_unsteady_iterations */
+	atomic_t		hr_unsteady_iterations;
+
 	char			hr_dev_name[BDEVNAME_SIZE];
 
 	unsigned int		hr_timeout_ms;
@@ -324,6 +329,10 @@ static void o2hb_write_timeout(struct work_struct *work)
 
 static void o2hb_arm_write_timeout(struct o2hb_region *reg)
 {
+	/* Arm writeout only after thread reaches steady state */
+	if (atomic_read(&reg->hr_steady_iterations) != 0)
+		return;
+
 	mlog(ML_HEARTBEAT, "Queue write timeout for %u ms\n",
 	     O2HB_MAX_WRITE_TIMEOUT_MS);
 
@@ -537,9 +546,14 @@ static int o2hb_verify_crc(struct o2hb_region *reg,
 	return read == computed;
 }
 
-/* We want to make sure that nobody is heartbeating on top of us --
- * this will help detect an invalid configuration. */
-static void o2hb_check_last_timestamp(struct o2hb_region *reg)
+/*
+ * Compare the slot data with what we wrote in the last iteration.
+ * If the match fails, print an appropriate error message. This is to
+ * detect errors like... another node hearting on the same slot,
+ * flaky device that is losing writes, etc.
+ * Returns 1 if check succeeds, 0 otherwise.
+ */
+static int o2hb_check_own_slot(struct o2hb_region *reg)
 {
 	struct o2hb_disk_slot *slot;
 	struct o2hb_disk_heartbeat_block *hb_block;
@@ -548,13 +562,13 @@ static void o2hb_check_last_timestamp(struct o2hb_region *reg)
 	slot = &reg->hr_slots[o2nm_this_node()];
 	/* Don't check on our 1st timestamp */
 	if (!slot->ds_last_time)
-		return;
+		return 0;
 
 	hb_block = slot->ds_raw_block;
 	if (le64_to_cpu(hb_block->hb_seq) == slot->ds_last_time &&
 	    le64_to_cpu(hb_block->hb_generation) == slot->ds_last_generation &&
 	    hb_block->hb_node == slot->ds_node_num)
-		return;
+		return 1;
 
 #define ERRSTR1		"Another node is heartbeating on device"
 #define ERRSTR2		"Heartbeat generation mismatch on device"
@@ -574,6 +588,8 @@ static void o2hb_check_last_timestamp(struct o2hb_region *reg)
 	     (unsigned long long)slot->ds_last_time, hb_block->hb_node,
 	     (unsigned long long)le64_to_cpu(hb_block->hb_generation),
 	     (unsigned long long)le64_to_cpu(hb_block->hb_seq));
+
+	return 0;
 }
 
 static inline void o2hb_prepare_block(struct o2hb_region *reg,
@@ -719,17 +735,24 @@ static void o2hb_shutdown_slot(struct o2hb_disk_slot *slot)
 	o2nm_node_put(node);
 }
 
-static void o2hb_set_quorum_device(struct o2hb_region *reg,
-				   struct o2hb_disk_slot *slot)
+static void o2hb_set_quorum_device(struct o2hb_region *reg)
 {
-	assert_spin_locked(&o2hb_live_lock);
-
 	if (!o2hb_global_heartbeat_active())
 		return;
 
-	if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+	/* Prevent race with o2hb_heartbeat_group_drop_item() */
+	if (kthread_should_stop())
+		return;
+
+	/* Tag region as quorum only after thread reaches steady state */
+	if (atomic_read(&reg->hr_steady_iterations) != 0)
 		return;
 
+	spin_lock(&o2hb_live_lock);
+
+	if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+		goto unlock;
+
 	/*
 	 * A region can be added to the quorum only when it sees all
 	 * live nodes heartbeat on it. In other words, the region has been
@@ -737,13 +760,10 @@ static void o2hb_set_quorum_device(struct o2hb_region *reg,
 	 */
 	if (memcmp(reg->hr_live_node_bitmap, o2hb_live_node_bitmap,
 		   sizeof(o2hb_live_node_bitmap)))
-		return;
-
-	if (slot->ds_changed_samples < O2HB_LIVE_THRESHOLD)
-		return;
+		goto unlock;
 
-	printk(KERN_NOTICE "o2hb: Region %s is now a quorum device\n",
-	       config_item_name(&reg->hr_item));
+	printk(KERN_NOTICE "o2hb: Region %s (%s) is now a quorum device\n",
+	       config_item_name(&reg->hr_item), reg->hr_dev_name);
 
 	set_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
 
@@ -754,6 +774,8 @@ static void o2hb_set_quorum_device(struct o2hb_region *reg,
 	if (o2hb_pop_count(&o2hb_quorum_region_bitmap,
 			   O2NM_MAX_REGIONS) > O2HB_PIN_CUT_OFF)
 		o2hb_region_unpin(NULL);
+unlock:
+	spin_unlock(&o2hb_live_lock);
 }
 
 static int o2hb_check_slot(struct o2hb_region *reg,
@@ -925,8 +947,6 @@ fire_callbacks:
 		slot->ds_equal_samples = 0;
 	}
 out:
-	o2hb_set_quorum_device(reg, slot);
-
 	spin_unlock(&o2hb_live_lock);
 
 	o2hb_run_event_list(&event);
@@ -957,7 +977,8 @@ static int o2hb_highest_node(unsigned long *nodes,
 
 static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 {
-	int i, ret, highest_node, change = 0;
+	int i, ret, highest_node;
+	int membership_change = 0, own_slot_ok = 0;
 	unsigned long configured_nodes[BITS_TO_LONGS(O2NM_MAX_NODES)];
 	unsigned long live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
 	struct o2hb_bio_wait_ctxt write_wc;
@@ -966,7 +987,7 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 				       sizeof(configured_nodes));
 	if (ret) {
 		mlog_errno(ret);
-		return ret;
+		goto bail;
 	}
 
 	/*
@@ -982,8 +1003,9 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 
 	highest_node = o2hb_highest_node(configured_nodes, O2NM_MAX_NODES);
 	if (highest_node >= O2NM_MAX_NODES) {
-		mlog(ML_NOTICE, "ocfs2_heartbeat: no configured nodes found!\n");
-		return -EINVAL;
+		mlog(ML_NOTICE, "o2hb: No configured nodes found!\n");
+		ret = -EINVAL;
+		goto bail;
 	}
 
 	/* No sense in reading the slots of nodes that don't exist
@@ -993,29 +1015,27 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 	ret = o2hb_read_slots(reg, highest_node + 1);
 	if (ret < 0) {
 		mlog_errno(ret);
-		return ret;
+		goto bail;
 	}
 
 	/* With an up to date view of the slots, we can check that no
 	 * other node has been improperly configured to heartbeat in
 	 * our slot. */
-	o2hb_check_last_timestamp(reg);
+	own_slot_ok = o2hb_check_own_slot(reg);
 
 	/* fill in the proper info for our next heartbeat */
 	o2hb_prepare_block(reg, reg->hr_generation);
 
-	/* And fire off the write. Note that we don't wait on this I/O
-	 * until later. */
 	ret = o2hb_issue_node_write(reg, &write_wc);
 	if (ret < 0) {
 		mlog_errno(ret);
-		return ret;
+		goto bail;
 	}
 
 	i = -1;
 	while((i = find_next_bit(configured_nodes,
 				 O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) {
-		change |= o2hb_check_slot(reg, &reg->hr_slots[i]);
+		membership_change |= o2hb_check_slot(reg, &reg->hr_slots[i]);
 	}
 
 	/*
@@ -1030,18 +1050,39 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 		 * disk */
 		mlog(ML_ERROR, "Write error %d on device \"%s\"\n",
 		     write_wc.wc_error, reg->hr_dev_name);
-		return write_wc.wc_error;
+		ret = write_wc.wc_error;
+		goto bail;
 	}
 
-	o2hb_arm_write_timeout(reg);
+	/* Skip disarming the timeout if own slot has stale/bad data */
+	if (own_slot_ok) {
+		o2hb_set_quorum_device(reg);
+		o2hb_arm_write_timeout(reg);
+	}
 
+bail:
 	/* let the person who launched us know when things are steady */
-	if (!change && (atomic_read(&reg->hr_steady_iterations) != 0)) {
-		if (atomic_dec_and_test(&reg->hr_steady_iterations))
+	if (atomic_read(&reg->hr_steady_iterations) != 0) {
+		if (!ret && own_slot_ok && !membership_change) {
+			if (atomic_dec_and_test(&reg->hr_steady_iterations))
+				wake_up(&o2hb_steady_queue);
+		}
+	}
+
+	if (atomic_read(&reg->hr_steady_iterations) != 0) {
+		if (atomic_dec_and_test(&reg->hr_unsteady_iterations)) {
+			printk(KERN_NOTICE "o2hb: Unable to stabilize "
+			       "heartbeart on region %s (%s)\n",
+			       config_item_name(&reg->hr_item),
+			       reg->hr_dev_name);
+			atomic_set(&reg->hr_steady_iterations, 0);
+			reg->hr_aborted_start = 1;
 			wake_up(&o2hb_steady_queue);
+			ret = -EIO;
+		}
 	}
 
-	return 0;
+	return ret;
 }
 
 /* Subtract b from a, storing the result in a. a *must* have a larger
@@ -1095,7 +1136,8 @@ static int o2hb_thread(void *data)
 	/* Pin node */
 	o2nm_depend_this_node();
 
-	while (!kthread_should_stop() && !reg->hr_unclean_stop) {
+	while (!kthread_should_stop() &&
+	       !reg->hr_unclean_stop && !reg->hr_aborted_start) {
 		/* We track the time spent inside
 		 * o2hb_do_disk_heartbeat so that we avoid more than
 		 * hr_timeout_ms between disk writes. On busy systems
@@ -1103,10 +1145,7 @@ static int o2hb_thread(void *data)
 		 * likely to time itself out. */
 		do_gettimeofday(&before_hb);
 
-		i = 0;
-		do {
-			ret = o2hb_do_disk_heartbeat(reg);
-		} while (ret && ++i < 2);
+		ret = o2hb_do_disk_heartbeat(reg);
 
 		do_gettimeofday(&after_hb);
 		elapsed_msec = o2hb_elapsed_msecs(&before_hb, &after_hb);
@@ -1117,7 +1156,8 @@ static int o2hb_thread(void *data)
 		     after_hb.tv_sec, (unsigned long) after_hb.tv_usec,
 		     elapsed_msec);
 
-		if (elapsed_msec < reg->hr_timeout_ms) {
+		if (!kthread_should_stop() &&
+		    elapsed_msec < reg->hr_timeout_ms) {
 			/* the kthread api has blocked signals for us so no
 			 * need to record the return value. */
 			msleep_interruptible(reg->hr_timeout_ms - elapsed_msec);
@@ -1134,20 +1174,20 @@ static int o2hb_thread(void *data)
 	 * to timeout on this region when we could just as easily
 	 * write a clear generation - thus indicating to them that
 	 * this node has left this region.
-	 *
-	 * XXX: Should we skip this on unclean_stop? */
-	o2hb_prepare_block(reg, 0);
-	ret = o2hb_issue_node_write(reg, &write_wc);
-	if (ret == 0) {
-		o2hb_wait_on_io(reg, &write_wc);
-	} else {
-		mlog_errno(ret);
+	 */
+	if (!reg->hr_unclean_stop && !reg->hr_aborted_start) {
+		o2hb_prepare_block(reg, 0);
+		ret = o2hb_issue_node_write(reg, &write_wc);
+		if (ret == 0)
+			o2hb_wait_on_io(reg, &write_wc);
+		else
+			mlog_errno(ret);
 	}
 
 	/* Unpin node */
 	o2nm_undepend_this_node();
 
-	mlog(ML_HEARTBEAT|ML_KTHREAD, "hb thread exiting\n");
+	mlog(ML_HEARTBEAT|ML_KTHREAD, "o2hb thread exiting\n");
 
 	return 0;
 }
@@ -1158,6 +1198,7 @@ static int o2hb_debug_open(struct inode *inode, struct file *file)
 	struct o2hb_debug_buf *db = inode->i_private;
 	struct o2hb_region *reg;
 	unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	unsigned long lts;
 	char *buf = NULL;
 	int i = -1;
 	int out = 0;
@@ -1194,9 +1235,11 @@ static int o2hb_debug_open(struct inode *inode, struct file *file)
 
 	case O2HB_DB_TYPE_REGION_ELAPSED_TIME:
 		reg = (struct o2hb_region *)db->db_data;
-		out += snprintf(buf + out, PAGE_SIZE - out, "%u\n",
-				jiffies_to_msecs(jiffies -
-						 reg->hr_last_timeout_start));
+		lts = reg->hr_last_timeout_start;
+		/* If 0, it has never been set before */
+		if (lts)
+			lts = jiffies_to_msecs(jiffies - lts);
+		out += snprintf(buf + out, PAGE_SIZE - out, "%lu\n", lts);
 		goto done;
 
 	case O2HB_DB_TYPE_REGION_PINNED:
@@ -1426,6 +1469,8 @@ static void o2hb_region_release(struct config_item *item)
 	struct page *page;
 	struct o2hb_region *reg = to_o2hb_region(item);
 
+	mlog(ML_HEARTBEAT, "hb region release (%s)\n", reg->hr_dev_name);
+
 	if (reg->hr_tmp_block)
 		kfree(reg->hr_tmp_block);
 
@@ -1792,7 +1837,10 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 			live_threshold <<= 1;
 		spin_unlock(&o2hb_live_lock);
 	}
-	atomic_set(&reg->hr_steady_iterations, live_threshold + 1);
+	++live_threshold;
+	atomic_set(&reg->hr_steady_iterations, live_threshold);
+	/* unsteady_iterations is double the steady_iterations */
+	atomic_set(&reg->hr_unsteady_iterations, (live_threshold << 1));
 
 	hb_task = kthread_run(o2hb_thread, reg, "o2hb-%s",
 			      reg->hr_item.ci_name);
@@ -1809,14 +1857,12 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 	ret = wait_event_interruptible(o2hb_steady_queue,
 				atomic_read(&reg->hr_steady_iterations) == 0);
 	if (ret) {
-		/* We got interrupted (hello ptrace!).  Clean up */
-		spin_lock(&o2hb_live_lock);
-		hb_task = reg->hr_task;
-		reg->hr_task = NULL;
-		spin_unlock(&o2hb_live_lock);
+		atomic_set(&reg->hr_steady_iterations, 0);
+		reg->hr_aborted_start = 1;
+	}
 
-		if (hb_task)
-			kthread_stop(hb_task);
+	if (reg->hr_aborted_start) {
+		ret = -EIO;
 		goto out;
 	}
 
@@ -1833,8 +1879,8 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 		ret = -EIO;
 
 	if (hb_task && o2hb_global_heartbeat_active())
-		printk(KERN_NOTICE "o2hb: Heartbeat started on region %s\n",
-		       config_item_name(&reg->hr_item));
+		printk(KERN_NOTICE "o2hb: Heartbeat started on region %s (%s)\n",
+		       config_item_name(&reg->hr_item), reg->hr_dev_name);
 
 out:
 	if (filp)
@@ -2092,13 +2138,6 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group,
 
 	/* stop the thread when the user removes the region dir */
 	spin_lock(&o2hb_live_lock);
-	if (o2hb_global_heartbeat_active()) {
-		clear_bit(reg->hr_region_num, o2hb_region_bitmap);
-		clear_bit(reg->hr_region_num, o2hb_live_region_bitmap);
-		if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
-			quorum_region = 1;
-		clear_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
-	}
 	hb_task = reg->hr_task;
 	reg->hr_task = NULL;
 	reg->hr_item_dropped = 1;
@@ -2107,19 +2146,30 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group,
 	if (hb_task)
 		kthread_stop(hb_task);
 
+	if (o2hb_global_heartbeat_active()) {
+		spin_lock(&o2hb_live_lock);
+		clear_bit(reg->hr_region_num, o2hb_region_bitmap);
+		clear_bit(reg->hr_region_num, o2hb_live_region_bitmap);
+		if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+			quorum_region = 1;
+		clear_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
+		spin_unlock(&o2hb_live_lock);
+		printk(KERN_NOTICE "o2hb: Heartbeat %s on region %s (%s)\n",
+		       ((atomic_read(&reg->hr_steady_iterations) == 0) ?
+			"stopped" : "start aborted"), config_item_name(item),
+		       reg->hr_dev_name);
+	}
+
 	/*
 	 * If we're racing a dev_write(), we need to wake them.  They will
 	 * check reg->hr_task
 	 */
 	if (atomic_read(&reg->hr_steady_iterations) != 0) {
+		reg->hr_aborted_start = 1;
 		atomic_set(&reg->hr_steady_iterations, 0);
 		wake_up(&o2hb_steady_queue);
 	}
 
-	if (o2hb_global_heartbeat_active())
-		printk(KERN_NOTICE "o2hb: Heartbeat stopped on region %s\n",
-		       config_item_name(&reg->hr_item));
-
 	config_item_put(item);
 
 	if (!o2hb_global_heartbeat_active() || !quorum_region)
diff --git a/fs/ocfs2/cluster/netdebug.c b/fs/ocfs2/cluster/netdebug.c
index 3a5835904b3..dc45deb19e6 100644
--- a/fs/ocfs2/cluster/netdebug.c
+++ b/fs/ocfs2/cluster/netdebug.c
@@ -47,6 +47,7 @@
 #define SC_DEBUG_NAME		"sock_containers"
 #define NST_DEBUG_NAME		"send_tracking"
 #define STATS_DEBUG_NAME	"stats"
+#define NODES_DEBUG_NAME	"connected_nodes"
 
 #define SHOW_SOCK_CONTAINERS	0
 #define SHOW_SOCK_STATS		1
@@ -55,6 +56,7 @@ static struct dentry *o2net_dentry;
 static struct dentry *sc_dentry;
 static struct dentry *nst_dentry;
 static struct dentry *stats_dentry;
+static struct dentry *nodes_dentry;
 
 static DEFINE_SPINLOCK(o2net_debug_lock);
 
@@ -491,53 +493,87 @@ static const struct file_operations sc_seq_fops = {
 	.release = sc_fop_release,
 };
 
-int o2net_debugfs_init(void)
+static int o2net_fill_bitmap(char *buf, int len)
 {
-	o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL);
-	if (!o2net_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	int i = -1, out = 0;
 
-	nst_dentry = debugfs_create_file(NST_DEBUG_NAME, S_IFREG|S_IRUSR,
-					 o2net_dentry, NULL,
-					 &nst_seq_fops);
-	if (!nst_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	o2net_fill_node_map(map, sizeof(map));
 
-	sc_dentry = debugfs_create_file(SC_DEBUG_NAME, S_IFREG|S_IRUSR,
-					o2net_dentry, NULL,
-					&sc_seq_fops);
-	if (!sc_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	while ((i = find_next_bit(map, O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES)
+		out += snprintf(buf + out, PAGE_SIZE - out, "%d ", i);
+	out += snprintf(buf + out, PAGE_SIZE - out, "\n");
 
-	stats_dentry = debugfs_create_file(STATS_DEBUG_NAME, S_IFREG|S_IRUSR,
-					   o2net_dentry, NULL,
-					   &stats_seq_fops);
-	if (!stats_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	return out;
+}
+
+static int nodes_fop_open(struct inode *inode, struct file *file)
+{
+	char *buf;
+
+	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	i_size_write(inode, o2net_fill_bitmap(buf, PAGE_SIZE));
+
+	file->private_data = buf;
 
 	return 0;
-bail:
-	debugfs_remove(stats_dentry);
-	debugfs_remove(sc_dentry);
-	debugfs_remove(nst_dentry);
-	debugfs_remove(o2net_dentry);
-	return -ENOMEM;
 }
 
+static int o2net_debug_release(struct inode *inode, struct file *file)
+{
+	kfree(file->private_data);
+	return 0;
+}
+
+static ssize_t o2net_debug_read(struct file *file, char __user *buf,
+				size_t nbytes, loff_t *ppos)
+{
+	return simple_read_from_buffer(buf, nbytes, ppos, file->private_data,
+				       i_size_read(file->f_mapping->host));
+}
+
+static const struct file_operations nodes_fops = {
+	.open		= nodes_fop_open,
+	.release	= o2net_debug_release,
+	.read		= o2net_debug_read,
+	.llseek		= generic_file_llseek,
+};
+
 void o2net_debugfs_exit(void)
 {
+	debugfs_remove(nodes_dentry);
 	debugfs_remove(stats_dentry);
 	debugfs_remove(sc_dentry);
 	debugfs_remove(nst_dentry);
 	debugfs_remove(o2net_dentry);
 }
 
+int o2net_debugfs_init(void)
+{
+	mode_t mode = S_IFREG|S_IRUSR;
+
+	o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL);
+	if (o2net_dentry)
+		nst_dentry = debugfs_create_file(NST_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &nst_seq_fops);
+	if (nst_dentry)
+		sc_dentry = debugfs_create_file(SC_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &sc_seq_fops);
+	if (sc_dentry)
+		stats_dentry = debugfs_create_file(STATS_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &stats_seq_fops);
+	if (stats_dentry)
+		nodes_dentry = debugfs_create_file(NODES_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &nodes_fops);
+	if (nodes_dentry)
+		return 0;
+
+	o2net_debugfs_exit();
+	mlog_errno(-ENOMEM);
+	return -ENOMEM;
+}
+
 #endif	/* CONFIG_DEBUG_FS */
diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c
index ad7d0c155de..044e7b58d31 100644
--- a/fs/ocfs2/cluster/tcp.c
+++ b/fs/ocfs2/cluster/tcp.c
@@ -546,7 +546,7 @@ static void o2net_set_nn_state(struct o2net_node *nn,
 	}
 
 	if (was_valid && !valid) {
-		printk(KERN_NOTICE "o2net: no longer connected to "
+		printk(KERN_NOTICE "o2net: No longer connected to "
 		       SC_NODEF_FMT "\n", SC_NODEF_ARGS(old_sc));
 		o2net_complete_nodes_nsw(nn);
 	}
@@ -556,7 +556,7 @@ static void o2net_set_nn_state(struct o2net_node *nn,
 		cancel_delayed_work(&nn->nn_connect_expired);
 		printk(KERN_NOTICE "o2net: %s " SC_NODEF_FMT "\n",
 		       o2nm_this_node() > sc->sc_node->nd_num ?
-		       		"connected to" : "accepted connection from",
+		       "Connected to" : "Accepted connection from",
 		       SC_NODEF_ARGS(sc));
 	}
 
@@ -644,7 +644,7 @@ static void o2net_state_change(struct sock *sk)
 			o2net_sc_queue_work(sc, &sc->sc_connect_work);
 			break;
 		default:
-			printk(KERN_INFO "o2net: connection to " SC_NODEF_FMT
+			printk(KERN_INFO "o2net: Connection to " SC_NODEF_FMT
 			      " shutdown, state %d\n",
 			      SC_NODEF_ARGS(sc), sk->sk_state);
 			o2net_sc_queue_work(sc, &sc->sc_shutdown_work);
@@ -1035,6 +1035,25 @@ static int o2net_tx_can_proceed(struct o2net_node *nn,
 	return ret;
 }
 
+/* Get a map of all nodes to which this node is currently connected to */
+void o2net_fill_node_map(unsigned long *map, unsigned bytes)
+{
+	struct o2net_sock_container *sc;
+	int node, ret;
+
+	BUG_ON(bytes < (BITS_TO_LONGS(O2NM_MAX_NODES) * sizeof(unsigned long)));
+
+	memset(map, 0, bytes);
+	for (node = 0; node < O2NM_MAX_NODES; ++node) {
+		o2net_tx_can_proceed(o2net_nn_from_num(node), &sc, &ret);
+		if (!ret) {
+			set_bit(node, map);
+			sc_put(sc);
+		}
+	}
+}
+EXPORT_SYMBOL_GPL(o2net_fill_node_map);
+
 int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec,
 			   size_t caller_veclen, u8 target_node, int *status)
 {
@@ -1285,11 +1304,11 @@ static int o2net_check_handshake(struct o2net_sock_container *sc)
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
 
 	if (hand->protocol_version != cpu_to_be64(O2NET_PROTOCOL_VERSION)) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " advertised net protocol "
-		     "version %llu but %llu is required, disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     (unsigned long long)be64_to_cpu(hand->protocol_version),
-		     O2NET_PROTOCOL_VERSION);
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " Advertised net "
+		       "protocol version %llu but %llu is required. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       (unsigned long long)be64_to_cpu(hand->protocol_version),
+		       O2NET_PROTOCOL_VERSION);
 
 		/* don't bother reconnecting if its the wrong version. */
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
@@ -1303,33 +1322,33 @@ static int o2net_check_handshake(struct o2net_sock_container *sc)
 	 */
 	if (be32_to_cpu(hand->o2net_idle_timeout_ms) !=
 				o2net_idle_timeout()) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " uses a network idle timeout of "
-		     "%u ms, but we use %u ms locally.  disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     be32_to_cpu(hand->o2net_idle_timeout_ms),
-		     o2net_idle_timeout());
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a network "
+		       "idle timeout of %u ms, but we use %u ms locally. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       be32_to_cpu(hand->o2net_idle_timeout_ms),
+		       o2net_idle_timeout());
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
 		return -1;
 	}
 
 	if (be32_to_cpu(hand->o2net_keepalive_delay_ms) !=
 			o2net_keepalive_delay()) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " uses a keepalive delay of "
-		     "%u ms, but we use %u ms locally.  disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     be32_to_cpu(hand->o2net_keepalive_delay_ms),
-		     o2net_keepalive_delay());
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a keepalive "
+		       "delay of %u ms, but we use %u ms locally. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       be32_to_cpu(hand->o2net_keepalive_delay_ms),
+		       o2net_keepalive_delay());
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
 		return -1;
 	}
 
 	if (be32_to_cpu(hand->o2hb_heartbeat_timeout_ms) !=
 			O2HB_MAX_WRITE_TIMEOUT_MS) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " uses a heartbeat timeout of "
-		     "%u ms, but we use %u ms locally.  disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
-		     O2HB_MAX_WRITE_TIMEOUT_MS);
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a heartbeat "
+		       "timeout of %u ms, but we use %u ms locally. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
+		       O2HB_MAX_WRITE_TIMEOUT_MS);
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
 		return -1;
 	}
@@ -1540,28 +1559,16 @@ static void o2net_idle_timer(unsigned long data)
 {
 	struct o2net_sock_container *sc = (struct o2net_sock_container *)data;
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
-
 #ifdef CONFIG_DEBUG_FS
-	ktime_t now = ktime_get();
+	unsigned long msecs = ktime_to_ms(ktime_get()) -
+		ktime_to_ms(sc->sc_tv_timer);
+#else
+	unsigned long msecs = o2net_idle_timeout();
 #endif
 
-	printk(KERN_NOTICE "o2net: connection to " SC_NODEF_FMT " has been idle for %u.%u "
-	     "seconds, shutting it down.\n", SC_NODEF_ARGS(sc),
-		     o2net_idle_timeout() / 1000,
-		     o2net_idle_timeout() % 1000);
-
-#ifdef CONFIG_DEBUG_FS
-	mlog(ML_NOTICE, "Here are some times that might help debug the "
-	     "situation: (Timer: %lld, Now %lld, DataReady %lld, Advance %lld-%lld, "
-	     "Key 0x%08x, Func %u, FuncTime %lld-%lld)\n",
-	     (long long)ktime_to_us(sc->sc_tv_timer), (long long)ktime_to_us(now),
-	     (long long)ktime_to_us(sc->sc_tv_data_ready),
-	     (long long)ktime_to_us(sc->sc_tv_advance_start),
-	     (long long)ktime_to_us(sc->sc_tv_advance_stop),
-	     sc->sc_msg_key, sc->sc_msg_type,
-	     (long long)ktime_to_us(sc->sc_tv_func_start),
-	     (long long)ktime_to_us(sc->sc_tv_func_stop));
-#endif
+	printk(KERN_NOTICE "o2net: Connection to " SC_NODEF_FMT " has been "
+	       "idle for %lu.%lu secs, shutting it down.\n", SC_NODEF_ARGS(sc),
+	       msecs / 1000, msecs % 1000);
 
 	/*
 	 * Initialize the nn_timeout so that the next connection attempt
@@ -1694,8 +1701,8 @@ static void o2net_start_connect(struct work_struct *work)
 
 out:
 	if (ret) {
-		mlog(ML_NOTICE, "connect attempt to " SC_NODEF_FMT " failed "
-		     "with errno %d\n", SC_NODEF_ARGS(sc), ret);
+		printk(KERN_NOTICE "o2net: Connect attempt to " SC_NODEF_FMT
+		       " failed with errno %d\n", SC_NODEF_ARGS(sc), ret);
 		/* 0 err so that another will be queued and attempted
 		 * from set_nn_state */
 		if (sc)
@@ -1718,8 +1725,8 @@ static void o2net_connect_expired(struct work_struct *work)
 
 	spin_lock(&nn->nn_lock);
 	if (!nn->nn_sc_valid) {
-		mlog(ML_ERROR, "no connection established with node %u after "
-		     "%u.%u seconds, giving up and returning errors.\n",
+		printk(KERN_NOTICE "o2net: No connection established with "
+		       "node %u after %u.%u seconds, giving up.\n",
 		     o2net_num_from_nn(nn),
 		     o2net_idle_timeout() / 1000,
 		     o2net_idle_timeout() % 1000);
@@ -1862,21 +1869,21 @@ static int o2net_accept_one(struct socket *sock)
 
 	node = o2nm_get_node_by_ip(sin.sin_addr.s_addr);
 	if (node == NULL) {
-		mlog(ML_NOTICE, "attempt to connect from unknown node at %pI4:%d\n",
-		     &sin.sin_addr.s_addr, ntohs(sin.sin_port));
+		printk(KERN_NOTICE "o2net: Attempt to connect from unknown "
+		       "node at %pI4:%d\n", &sin.sin_addr.s_addr,
+		       ntohs(sin.sin_port));
 		ret = -EINVAL;
 		goto out;
 	}
 
 	if (o2nm_this_node() >= node->nd_num) {
 		local_node = o2nm_get_node_by_num(o2nm_this_node());
-		mlog(ML_NOTICE, "unexpected connect attempt seen at node '%s' ("
-		     "%u, %pI4:%d) from node '%s' (%u, %pI4:%d)\n",
-		     local_node->nd_name, local_node->nd_num,
-		     &(local_node->nd_ipv4_address),
-		     ntohs(local_node->nd_ipv4_port),
-		     node->nd_name, node->nd_num, &sin.sin_addr.s_addr,
-		     ntohs(sin.sin_port));
+		printk(KERN_NOTICE "o2net: Unexpected connect attempt seen "
+		       "at node '%s' (%u, %pI4:%d) from node '%s' (%u, "
+		       "%pI4:%d)\n", local_node->nd_name, local_node->nd_num,
+		       &(local_node->nd_ipv4_address),
+		       ntohs(local_node->nd_ipv4_port), node->nd_name,
+		       node->nd_num, &sin.sin_addr.s_addr, ntohs(sin.sin_port));
 		ret = -EINVAL;
 		goto out;
 	}
@@ -1901,10 +1908,10 @@ static int o2net_accept_one(struct socket *sock)
 		ret = 0;
 	spin_unlock(&nn->nn_lock);
 	if (ret) {
-		mlog(ML_NOTICE, "attempt to connect from node '%s' at "
-		     "%pI4:%d but it already has an open connection\n",
-		     node->nd_name, &sin.sin_addr.s_addr,
-		     ntohs(sin.sin_port));
+		printk(KERN_NOTICE "o2net: Attempt to connect from node '%s' "
+		       "at %pI4:%d but it already has an open connection\n",
+		       node->nd_name, &sin.sin_addr.s_addr,
+		       ntohs(sin.sin_port));
 		goto out;
 	}
 
@@ -1984,7 +1991,7 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
 
 	ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
 	if (ret < 0) {
-		mlog(ML_ERROR, "unable to create socket, ret=%d\n", ret);
+		printk(KERN_ERR "o2net: Error %d while creating socket\n", ret);
 		goto out;
 	}
 
@@ -2001,16 +2008,15 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
 	sock->sk->sk_reuse = 1;
 	ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin));
 	if (ret < 0) {
-		mlog(ML_ERROR, "unable to bind socket at %pI4:%u, "
-		     "ret=%d\n", &addr, ntohs(port), ret);
+		printk(KERN_ERR "o2net: Error %d while binding socket at "
+		       "%pI4:%u\n", ret, &addr, ntohs(port)); 
 		goto out;
 	}
 
 	ret = sock->ops->listen(sock, 64);
-	if (ret < 0) {
-		mlog(ML_ERROR, "unable to listen on %pI4:%u, ret=%d\n",
-		     &addr, ntohs(port), ret);
-	}
+	if (ret < 0)
+		printk(KERN_ERR "o2net: Error %d while listening on %pI4:%u\n",
+		       ret, &addr, ntohs(port));
 
 out:
 	if (ret) {
diff --git a/fs/ocfs2/cluster/tcp.h b/fs/ocfs2/cluster/tcp.h
index fd6179eb26d..5bada2a69b5 100644
--- a/fs/ocfs2/cluster/tcp.h
+++ b/fs/ocfs2/cluster/tcp.h
@@ -106,6 +106,8 @@ int o2net_register_handler(u32 msg_type, u32 key, u32 max_len,
 			   struct list_head *unreg_list);
 void o2net_unregister_handler_list(struct list_head *list);
 
+void o2net_fill_node_map(unsigned long *map, unsigned bytes);
+
 struct o2nm_node;
 int o2net_register_hb_callbacks(void);
 void o2net_unregister_hb_callbacks(void);
diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
index e2878b5895f..8fe4e2892ab 100644
--- a/fs/ocfs2/dir.c
+++ b/fs/ocfs2/dir.c
@@ -1184,8 +1184,7 @@ static int __ocfs2_delete_entry(handle_t *handle, struct inode *dir,
 			if (pde)
 				le16_add_cpu(&pde->rec_len,
 						le16_to_cpu(de->rec_len));
-			else
-				de->inode = 0;
+			de->inode = 0;
 			dir->i_version++;
 			ocfs2_journal_dirty(handle, bh);
 			goto bail;
diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h
index d602abb51b6..a5952ceecba 100644
--- a/fs/ocfs2/dlm/dlmcommon.h
+++ b/fs/ocfs2/dlm/dlmcommon.h
@@ -859,8 +859,8 @@ void dlm_complete_recovery_thread(struct dlm_ctxt *dlm);
 void dlm_wait_for_recovery(struct dlm_ctxt *dlm);
 void dlm_kick_recovery_thread(struct dlm_ctxt *dlm);
 int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node);
-int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout);
-int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout);
+void dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout);
+void dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout);
 
 void dlm_put(struct dlm_ctxt *dlm);
 struct dlm_ctxt *dlm_grab(struct dlm_ctxt *dlm);
@@ -877,9 +877,8 @@ static inline void dlm_lockres_get(struct dlm_lock_resource *res)
 	kref_get(&res->refs);
 }
 void dlm_lockres_put(struct dlm_lock_resource *res);
-void __dlm_unhash_lockres(struct dlm_lock_resource *res);
-void __dlm_insert_lockres(struct dlm_ctxt *dlm,
-			  struct dlm_lock_resource *res);
+void __dlm_unhash_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res);
+void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res);
 struct dlm_lock_resource * __dlm_lookup_lockres_full(struct dlm_ctxt *dlm,
 						     const char *name,
 						     unsigned int len,
@@ -902,46 +901,15 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
 					  const char *name,
 					  unsigned int namelen);
 
-#define dlm_lockres_set_refmap_bit(bit,res)  \
-	__dlm_lockres_set_refmap_bit(bit,res,__FILE__,__LINE__)
-#define dlm_lockres_clear_refmap_bit(bit,res)  \
-	__dlm_lockres_clear_refmap_bit(bit,res,__FILE__,__LINE__)
+void dlm_lockres_set_refmap_bit(struct dlm_ctxt *dlm,
+				struct dlm_lock_resource *res, int bit);
+void dlm_lockres_clear_refmap_bit(struct dlm_ctxt *dlm,
+				  struct dlm_lock_resource *res, int bit);
 
-static inline void __dlm_lockres_set_refmap_bit(int bit,
-						struct dlm_lock_resource *res,
-						const char *file,
-						int line)
-{
-	//printk("%s:%d:%.*s: setting bit %d\n", file, line,
-	//     res->lockname.len, res->lockname.name, bit);
-	set_bit(bit, res->refmap);
-}
-
-static inline void __dlm_lockres_clear_refmap_bit(int bit,
-						  struct dlm_lock_resource *res,
-						  const char *file,
-						  int line)
-{
-	//printk("%s:%d:%.*s: clearing bit %d\n", file, line,
-	//     res->lockname.len, res->lockname.name, bit);
-	clear_bit(bit, res->refmap);
-}
-
-void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   const char *file,
-				   int line);
-void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   int new_lockres,
-				   const char *file,
-				   int line);
-#define dlm_lockres_drop_inflight_ref(d,r)  \
-	__dlm_lockres_drop_inflight_ref(d,r,__FILE__,__LINE__)
-#define dlm_lockres_grab_inflight_ref(d,r)  \
-	__dlm_lockres_grab_inflight_ref(d,r,0,__FILE__,__LINE__)
-#define dlm_lockres_grab_inflight_ref_new(d,r)  \
-	__dlm_lockres_grab_inflight_ref(d,r,1,__FILE__,__LINE__)
+void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res);
+void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res);
 
 void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
 void dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
index 6ed6b95dcf9..92f2ead0fab 100644
--- a/fs/ocfs2/dlm/dlmdomain.c
+++ b/fs/ocfs2/dlm/dlmdomain.c
@@ -157,16 +157,18 @@ static int dlm_protocol_compare(struct dlm_protocol_version *existing,
 
 static void dlm_unregister_domain_handlers(struct dlm_ctxt *dlm);
 
-void __dlm_unhash_lockres(struct dlm_lock_resource *lockres)
+void __dlm_unhash_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 {
-	if (!hlist_unhashed(&lockres->hash_node)) {
-		hlist_del_init(&lockres->hash_node);
-		dlm_lockres_put(lockres);
-	}
+	if (hlist_unhashed(&res->hash_node))
+		return;
+
+	mlog(0, "%s: Unhash res %.*s\n", dlm->name, res->lockname.len,
+	     res->lockname.name);
+	hlist_del_init(&res->hash_node);
+	dlm_lockres_put(res);
 }
 
-void __dlm_insert_lockres(struct dlm_ctxt *dlm,
-		       struct dlm_lock_resource *res)
+void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 {
 	struct hlist_head *bucket;
 	struct qstr *q;
@@ -180,6 +182,9 @@ void __dlm_insert_lockres(struct dlm_ctxt *dlm,
 	dlm_lockres_get(res);
 
 	hlist_add_head(&res->hash_node, bucket);
+
+	mlog(0, "%s: Hash res %.*s\n", dlm->name, res->lockname.len,
+	     res->lockname.name);
 }
 
 struct dlm_lock_resource * __dlm_lookup_lockres_full(struct dlm_ctxt *dlm,
@@ -539,17 +544,17 @@ again:
 
 static void __dlm_print_nodes(struct dlm_ctxt *dlm)
 {
-	int node = -1;
+	int node = -1, num = 0;
 
 	assert_spin_locked(&dlm->spinlock);
 
-	printk(KERN_NOTICE "o2dlm: Nodes in domain %s: ", dlm->name);
-
+	printk("( ");
 	while ((node = find_next_bit(dlm->domain_map, O2NM_MAX_NODES,
 				     node + 1)) < O2NM_MAX_NODES) {
 		printk("%d ", node);
+		++num;
 	}
-	printk("\n");
+	printk(") %u nodes\n", num);
 }
 
 static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
@@ -566,11 +571,10 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
 
 	node = exit_msg->node_idx;
 
-	printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s\n", node, dlm->name);
-
 	spin_lock(&dlm->spinlock);
 	clear_bit(node, dlm->domain_map);
 	clear_bit(node, dlm->exit_domain_map);
+	printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s ", node, dlm->name);
 	__dlm_print_nodes(dlm);
 
 	/* notify anything attached to the heartbeat events */
@@ -755,6 +759,7 @@ void dlm_unregister_domain(struct dlm_ctxt *dlm)
 
 		dlm_mark_domain_leaving(dlm);
 		dlm_leave_domain(dlm);
+		printk(KERN_NOTICE "o2dlm: Leaving domain %s\n", dlm->name);
 		dlm_force_free_mles(dlm);
 		dlm_complete_dlm_shutdown(dlm);
 	}
@@ -970,7 +975,7 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data,
 		clear_bit(assert->node_idx, dlm->exit_domain_map);
 		__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
 
-		printk(KERN_NOTICE "o2dlm: Node %u joins domain %s\n",
+		printk(KERN_NOTICE "o2dlm: Node %u joins domain %s ",
 		       assert->node_idx, dlm->name);
 		__dlm_print_nodes(dlm);
 
@@ -1701,8 +1706,10 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm)
 bail:
 	spin_lock(&dlm->spinlock);
 	__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
-	if (!status)
+	if (!status) {
+		printk(KERN_NOTICE "o2dlm: Joining domain %s ", dlm->name);
 		__dlm_print_nodes(dlm);
+	}
 	spin_unlock(&dlm->spinlock);
 
 	if (ctxt) {
@@ -2131,13 +2138,6 @@ struct dlm_ctxt * dlm_register_domain(const char *domain,
 		goto leave;
 	}
 
-	if (!o2hb_check_local_node_heartbeating()) {
-		mlog(ML_ERROR, "the local node has not been configured, or is "
-		     "not heartbeating\n");
-		ret = -EPROTO;
-		goto leave;
-	}
-
 	mlog(0, "register called for domain \"%s\"\n", domain);
 
 retry:
diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c
index 8d39e0fd66f..975810b9849 100644
--- a/fs/ocfs2/dlm/dlmlock.c
+++ b/fs/ocfs2/dlm/dlmlock.c
@@ -183,10 +183,6 @@ static enum dlm_status dlmlock_master(struct dlm_ctxt *dlm,
 			kick_thread = 1;
 		}
 	}
-	/* reduce the inflight count, this may result in the lockres
-	 * being purged below during calc_usage */
-	if (lock->ml.node == dlm->node_num)
-		dlm_lockres_drop_inflight_ref(dlm, res);
 
 	spin_unlock(&res->spinlock);
 	wake_up(&res->wq);
@@ -231,10 +227,16 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm,
 	     lock->ml.type, res->lockname.len,
 	     res->lockname.name, flags);
 
+	/*
+	 * Wait if resource is getting recovered, remastered, etc.
+	 * If the resource was remastered and new owner is self, then exit.
+	 */
 	spin_lock(&res->spinlock);
-
-	/* will exit this call with spinlock held */
 	__dlm_wait_on_lockres(res);
+	if (res->owner == dlm->node_num) {
+		spin_unlock(&res->spinlock);
+		return DLM_RECOVERING;
+	}
 	res->state |= DLM_LOCK_RES_IN_PROGRESS;
 
 	/* add lock to local (secondary) queue */
@@ -319,27 +321,23 @@ static enum dlm_status dlm_send_remote_lock_request(struct dlm_ctxt *dlm,
 	tmpret = o2net_send_message(DLM_CREATE_LOCK_MSG, dlm->key, &create,
 				    sizeof(create), res->owner, &status);
 	if (tmpret >= 0) {
-		// successfully sent and received
-		ret = status;  // this is already a dlm_status
+		ret = status;
 		if (ret == DLM_REJECTED) {
-			mlog(ML_ERROR, "%s:%.*s: BUG.  this is a stale lockres "
-			     "no longer owned by %u.  that node is coming back "
-			     "up currently.\n", dlm->name, create.namelen,
+			mlog(ML_ERROR, "%s: res %.*s, Stale lockres no longer "
+			     "owned by node %u. That node is coming back up "
+			     "currently.\n", dlm->name, create.namelen,
 			     create.name, res->owner);
 			dlm_print_one_lock_resource(res);
 			BUG();
 		}
 	} else {
-		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
-		     "node %u\n", tmpret, DLM_CREATE_LOCK_MSG, dlm->key,
-		     res->owner);
-		if (dlm_is_host_down(tmpret)) {
+		mlog(ML_ERROR, "%s: res %.*s, Error %d send CREATE LOCK to "
+		     "node %u\n", dlm->name, create.namelen, create.name,
+		     tmpret, res->owner);
+		if (dlm_is_host_down(tmpret))
 			ret = DLM_RECOVERING;
-			mlog(0, "node %u died so returning DLM_RECOVERING "
-			     "from lock message!\n", res->owner);
-		} else {
+		else
 			ret = dlm_err_to_dlm_status(tmpret);
-		}
 	}
 
 	return ret;
@@ -440,7 +438,7 @@ struct dlm_lock * dlm_new_lock(int type, u8 node, u64 cookie,
 		/* zero memory only if kernel-allocated */
 		lksb = kzalloc(sizeof(*lksb), GFP_NOFS);
 		if (!lksb) {
-			kfree(lock);
+			kmem_cache_free(dlm_lock_cache, lock);
 			return NULL;
 		}
 		kernel_allocated = 1;
@@ -718,18 +716,10 @@ retry_lock:
 
 		if (status == DLM_RECOVERING || status == DLM_MIGRATING ||
 		    status == DLM_FORWARD) {
-			mlog(0, "retrying lock with migration/"
-			     "recovery/in progress\n");
 			msleep(100);
-			/* no waiting for dlm_reco_thread */
 			if (recovery) {
 				if (status != DLM_RECOVERING)
 					goto retry_lock;
-
-				mlog(0, "%s: got RECOVERING "
-				     "for $RECOVERY lock, master "
-				     "was %u\n", dlm->name,
-				     res->owner);
 				/* wait to see the node go down, then
 				 * drop down and allow the lockres to
 				 * get cleaned up.  need to remaster. */
@@ -741,6 +731,14 @@ retry_lock:
 			}
 		}
 
+		/* Inflight taken in dlm_get_lock_resource() is dropped here */
+		spin_lock(&res->spinlock);
+		dlm_lockres_drop_inflight_ref(dlm, res);
+		spin_unlock(&res->spinlock);
+
+		dlm_lockres_calc_usage(dlm, res);
+		dlm_kick_thread(dlm, res);
+
 		if (status != DLM_NORMAL) {
 			lock->lksb->flags &= ~DLM_LKSB_GET_LVB;
 			if (status != DLM_NOTQUEUED)
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index 11eefb8c12e..005261c333b 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -631,39 +631,54 @@ error:
 	return NULL;
 }
 
-void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   int new_lockres,
-				   const char *file,
-				   int line)
+void dlm_lockres_set_refmap_bit(struct dlm_ctxt *dlm,
+				struct dlm_lock_resource *res, int bit)
 {
-	if (!new_lockres)
-		assert_spin_locked(&res->spinlock);
+	assert_spin_locked(&res->spinlock);
+
+	mlog(0, "res %.*s, set node %u, %ps()\n", res->lockname.len,
+	     res->lockname.name, bit, __builtin_return_address(0));
+
+	set_bit(bit, res->refmap);
+}
+
+void dlm_lockres_clear_refmap_bit(struct dlm_ctxt *dlm,
+				  struct dlm_lock_resource *res, int bit)
+{
+	assert_spin_locked(&res->spinlock);
+
+	mlog(0, "res %.*s, clr node %u, %ps()\n", res->lockname.len,
+	     res->lockname.name, bit, __builtin_return_address(0));
+
+	clear_bit(bit, res->refmap);
+}
+
+
+void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res)
+{
+	assert_spin_locked(&res->spinlock);
 
-	if (!test_bit(dlm->node_num, res->refmap)) {
-		BUG_ON(res->inflight_locks != 0);
-		dlm_lockres_set_refmap_bit(dlm->node_num, res);
-	}
 	res->inflight_locks++;
-	mlog(0, "%s:%.*s: inflight++: now %u\n",
-	     dlm->name, res->lockname.len, res->lockname.name,
-	     res->inflight_locks);
+
+	mlog(0, "%s: res %.*s, inflight++: now %u, %ps()\n", dlm->name,
+	     res->lockname.len, res->lockname.name, res->inflight_locks,
+	     __builtin_return_address(0));
 }
 
-void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   const char *file,
-				   int line)
+void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res)
 {
 	assert_spin_locked(&res->spinlock);
 
 	BUG_ON(res->inflight_locks == 0);
+
 	res->inflight_locks--;
-	mlog(0, "%s:%.*s: inflight--: now %u\n",
-	     dlm->name, res->lockname.len, res->lockname.name,
-	     res->inflight_locks);
-	if (res->inflight_locks == 0)
-		dlm_lockres_clear_refmap_bit(dlm->node_num, res);
+
+	mlog(0, "%s: res %.*s, inflight--: now %u, %ps()\n", dlm->name,
+	     res->lockname.len, res->lockname.name, res->inflight_locks,
+	     __builtin_return_address(0));
+
 	wake_up(&res->wq);
 }
 
@@ -697,7 +712,6 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
 	unsigned int hash;
 	int tries = 0;
 	int bit, wait_on_recovery = 0;
-	int drop_inflight_if_nonlocal = 0;
 
 	BUG_ON(!lockid);
 
@@ -709,36 +723,33 @@ lookup:
 	spin_lock(&dlm->spinlock);
 	tmpres = __dlm_lookup_lockres_full(dlm, lockid, namelen, hash);
 	if (tmpres) {
-		int dropping_ref = 0;
-
 		spin_unlock(&dlm->spinlock);
-
 		spin_lock(&tmpres->spinlock);
-		/* We wait for the other thread that is mastering the resource */
+		/* Wait on the thread that is mastering the resource */
 		if (tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN) {
 			__dlm_wait_on_lockres(tmpres);
 			BUG_ON(tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN);
+			spin_unlock(&tmpres->spinlock);
+			dlm_lockres_put(tmpres);
+			tmpres = NULL;
+			goto lookup;
 		}
 
-		if (tmpres->owner == dlm->node_num) {
-			BUG_ON(tmpres->state & DLM_LOCK_RES_DROPPING_REF);
-			dlm_lockres_grab_inflight_ref(dlm, tmpres);
-		} else if (tmpres->state & DLM_LOCK_RES_DROPPING_REF)
-			dropping_ref = 1;
-		spin_unlock(&tmpres->spinlock);
-
-		/* wait until done messaging the master, drop our ref to allow
-		 * the lockres to be purged, start over. */
-		if (dropping_ref) {
-			spin_lock(&tmpres->spinlock);
-			__dlm_wait_on_lockres_flags(tmpres, DLM_LOCK_RES_DROPPING_REF);
+		/* Wait on the resource purge to complete before continuing */
+		if (tmpres->state & DLM_LOCK_RES_DROPPING_REF) {
+			BUG_ON(tmpres->owner == dlm->node_num);
+			__dlm_wait_on_lockres_flags(tmpres,
+						    DLM_LOCK_RES_DROPPING_REF);
 			spin_unlock(&tmpres->spinlock);
 			dlm_lockres_put(tmpres);
 			tmpres = NULL;
 			goto lookup;
 		}
 
-		mlog(0, "found in hash!\n");
+		/* Grab inflight ref to pin the resource */
+		dlm_lockres_grab_inflight_ref(dlm, tmpres);
+
+		spin_unlock(&tmpres->spinlock);
 		if (res)
 			dlm_lockres_put(res);
 		res = tmpres;
@@ -829,8 +840,8 @@ lookup:
 		 * but they might own this lockres.  wait on them. */
 		bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
 		if (bit < O2NM_MAX_NODES) {
-			mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
-			     "recover before lock mastery can begin\n",
+			mlog(0, "%s: res %.*s, At least one node (%d) "
+			     "to recover before lock mastery can begin\n",
 			     dlm->name, namelen, (char *)lockid, bit);
 			wait_on_recovery = 1;
 		}
@@ -843,12 +854,11 @@ lookup:
 
 	/* finally add the lockres to its hash bucket */
 	__dlm_insert_lockres(dlm, res);
-	/* since this lockres is new it doesn't not require the spinlock */
-	dlm_lockres_grab_inflight_ref_new(dlm, res);
 
-	/* if this node does not become the master make sure to drop
-	 * this inflight reference below */
-	drop_inflight_if_nonlocal = 1;
+	/* Grab inflight ref to pin the resource */
+	spin_lock(&res->spinlock);
+	dlm_lockres_grab_inflight_ref(dlm, res);
+	spin_unlock(&res->spinlock);
 
 	/* get an extra ref on the mle in case this is a BLOCK
 	 * if so, the creator of the BLOCK may try to put the last
@@ -864,8 +874,8 @@ redo_request:
 		 * dlm spinlock would be detectable be a change on the mle,
 		 * so we only need to clear out the recovery map once. */
 		if (dlm_is_recovery_lock(lockid, namelen)) {
-			mlog(ML_NOTICE, "%s: recovery map is not empty, but "
-			     "must master $RECOVERY lock now\n", dlm->name);
+			mlog(0, "%s: Recovery map is not empty, but must "
+			     "master $RECOVERY lock now\n", dlm->name);
 			if (!dlm_pre_master_reco_lockres(dlm, res))
 				wait_on_recovery = 0;
 			else {
@@ -883,8 +893,8 @@ redo_request:
 		spin_lock(&dlm->spinlock);
 		bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
 		if (bit < O2NM_MAX_NODES) {
-			mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
-			     "recover before lock mastery can begin\n",
+			mlog(0, "%s: res %.*s, At least one node (%d) "
+			     "to recover before lock mastery can begin\n",
 			     dlm->name, namelen, (char *)lockid, bit);
 			wait_on_recovery = 1;
 		} else
@@ -913,8 +923,8 @@ redo_request:
 			 * yet, keep going until it does.  this is how the
 			 * master will know that asserts are needed back to
 			 * the lower nodes. */
-			mlog(0, "%s:%.*s: requests only up to %u but master "
-			     "is %u, keep going\n", dlm->name, namelen,
+			mlog(0, "%s: res %.*s, Requests only up to %u but "
+			     "master is %u, keep going\n", dlm->name, namelen,
 			     lockid, nodenum, mle->master);
 		}
 	}
@@ -924,13 +934,12 @@ wait:
 	ret = dlm_wait_for_lock_mastery(dlm, res, mle, &blocked);
 	if (ret < 0) {
 		wait_on_recovery = 1;
-		mlog(0, "%s:%.*s: node map changed, redo the "
-		     "master request now, blocked=%d\n",
-		     dlm->name, res->lockname.len,
+		mlog(0, "%s: res %.*s, Node map changed, redo the master "
+		     "request now, blocked=%d\n", dlm->name, res->lockname.len,
 		     res->lockname.name, blocked);
 		if (++tries > 20) {
-			mlog(ML_ERROR, "%s:%.*s: spinning on "
-			     "dlm_wait_for_lock_mastery, blocked=%d\n",
+			mlog(ML_ERROR, "%s: res %.*s, Spinning on "
+			     "dlm_wait_for_lock_mastery, blocked = %d\n",
 			     dlm->name, res->lockname.len,
 			     res->lockname.name, blocked);
 			dlm_print_one_lock_resource(res);
@@ -940,7 +949,8 @@ wait:
 		goto redo_request;
 	}
 
-	mlog(0, "lockres mastered by %u\n", res->owner);
+	mlog(0, "%s: res %.*s, Mastered by %u\n", dlm->name, res->lockname.len,
+	     res->lockname.name, res->owner);
 	/* make sure we never continue without this */
 	BUG_ON(res->owner == O2NM_MAX_NODES);
 
@@ -952,8 +962,6 @@ wait:
 
 wake_waiters:
 	spin_lock(&res->spinlock);
-	if (res->owner != dlm->node_num && drop_inflight_if_nonlocal)
-		dlm_lockres_drop_inflight_ref(dlm, res);
 	res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
 	spin_unlock(&res->spinlock);
 	wake_up(&res->wq);
@@ -1426,9 +1434,7 @@ way_up_top:
 		}
 
 		if (res->owner == dlm->node_num) {
-			mlog(0, "%s:%.*s: setting bit %u in refmap\n",
-			     dlm->name, namelen, name, request->node_idx);
-			dlm_lockres_set_refmap_bit(request->node_idx, res);
+			dlm_lockres_set_refmap_bit(dlm, res, request->node_idx);
 			spin_unlock(&res->spinlock);
 			response = DLM_MASTER_RESP_YES;
 			if (mle)
@@ -1493,10 +1499,8 @@ way_up_top:
 				 * go back and clean the mles on any
 				 * other nodes */
 				dispatch_assert = 1;
-				dlm_lockres_set_refmap_bit(request->node_idx, res);
-				mlog(0, "%s:%.*s: setting bit %u in refmap\n",
-				     dlm->name, namelen, name,
-				     request->node_idx);
+				dlm_lockres_set_refmap_bit(dlm, res,
+							   request->node_idx);
 			} else
 				response = DLM_MASTER_RESP_NO;
 		} else {
@@ -1702,7 +1706,7 @@ again:
 			     "lockres, set the bit in the refmap\n",
 			     namelen, lockname, to);
 			spin_lock(&res->spinlock);
-			dlm_lockres_set_refmap_bit(to, res);
+			dlm_lockres_set_refmap_bit(dlm, res, to);
 			spin_unlock(&res->spinlock);
 		}
 	}
@@ -2187,8 +2191,6 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 	namelen = res->lockname.len;
 	BUG_ON(namelen > O2NM_MAX_NAME_LEN);
 
-	mlog(0, "%s:%.*s: sending deref to %d\n",
-	     dlm->name, namelen, lockname, res->owner);
 	memset(&deref, 0, sizeof(deref));
 	deref.node_idx = dlm->node_num;
 	deref.namelen = namelen;
@@ -2197,14 +2199,12 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 	ret = o2net_send_message(DLM_DEREF_LOCKRES_MSG, dlm->key,
 				 &deref, sizeof(deref), res->owner, &r);
 	if (ret < 0)
-		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
-		     "node %u\n", ret, DLM_DEREF_LOCKRES_MSG, dlm->key,
-		     res->owner);
+		mlog(ML_ERROR, "%s: res %.*s, error %d send DEREF to node %u\n",
+		     dlm->name, namelen, lockname, ret, res->owner);
 	else if (r < 0) {
 		/* BAD.  other node says I did not have a ref. */
-		mlog(ML_ERROR,"while dropping ref on %s:%.*s "
-		    "(master=%u) got %d.\n", dlm->name, namelen,
-		    lockname, res->owner, r);
+		mlog(ML_ERROR, "%s: res %.*s, DEREF to node %u got %d\n",
+		     dlm->name, namelen, lockname, res->owner, r);
 		dlm_print_one_lock_resource(res);
 		BUG();
 	}
@@ -2260,7 +2260,7 @@ int dlm_deref_lockres_handler(struct o2net_msg *msg, u32 len, void *data,
 	else {
 		BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
 		if (test_bit(node, res->refmap)) {
-			dlm_lockres_clear_refmap_bit(node, res);
+			dlm_lockres_clear_refmap_bit(dlm, res, node);
 			cleared = 1;
 		}
 	}
@@ -2320,7 +2320,7 @@ static void dlm_deref_lockres_worker(struct dlm_work_item *item, void *data)
 	BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
 	if (test_bit(node, res->refmap)) {
 		__dlm_wait_on_lockres_flags(res, DLM_LOCK_RES_SETREF_INPROG);
-		dlm_lockres_clear_refmap_bit(node, res);
+		dlm_lockres_clear_refmap_bit(dlm, res, node);
 		cleared = 1;
 	}
 	spin_unlock(&res->spinlock);
@@ -2802,7 +2802,8 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
 				BUG_ON(!list_empty(&lock->bast_list));
 				BUG_ON(lock->ast_pending);
 				BUG_ON(lock->bast_pending);
-				dlm_lockres_clear_refmap_bit(lock->ml.node, res);
+				dlm_lockres_clear_refmap_bit(dlm, res,
+							     lock->ml.node);
 				list_del_init(&lock->list);
 				dlm_lock_put(lock);
 				/* In a normal unlock, we would have added a
@@ -2823,7 +2824,7 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
 			mlog(0, "%s:%.*s: node %u had a ref to this "
 			     "migrating lockres, clearing\n", dlm->name,
 			     res->lockname.len, res->lockname.name, bit);
-			dlm_lockres_clear_refmap_bit(bit, res);
+			dlm_lockres_clear_refmap_bit(dlm, res, bit);
 		}
 		bit++;
 	}
@@ -2916,9 +2917,9 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 					 &migrate, sizeof(migrate), nodenum,
 					 &status);
 		if (ret < 0) {
-			mlog(ML_ERROR, "Error %d when sending message %u (key "
-			     "0x%x) to node %u\n", ret, DLM_MIGRATE_REQUEST_MSG,
-			     dlm->key, nodenum);
+			mlog(ML_ERROR, "%s: res %.*s, Error %d send "
+			     "MIGRATE_REQUEST to node %u\n", dlm->name,
+			     migrate.namelen, migrate.name, ret, nodenum);
 			if (!dlm_is_host_down(ret)) {
 				mlog(ML_ERROR, "unhandled error=%d!\n", ret);
 				BUG();
@@ -2937,7 +2938,7 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 			     dlm->name, res->lockname.len, res->lockname.name,
 			     nodenum);
 			spin_lock(&res->spinlock);
-			dlm_lockres_set_refmap_bit(nodenum, res);
+			dlm_lockres_set_refmap_bit(dlm, res, nodenum);
 			spin_unlock(&res->spinlock);
 		}
 	}
@@ -3271,7 +3272,7 @@ int dlm_finish_migration(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	 * mastery reference here since old_master will briefly have
 	 * a reference after the migration completes */
 	spin_lock(&res->spinlock);
-	dlm_lockres_set_refmap_bit(old_master, res);
+	dlm_lockres_set_refmap_bit(dlm, res, old_master);
 	spin_unlock(&res->spinlock);
 
 	mlog(0, "now time to do a migrate request to other nodes\n");
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 7efab6d28a2..01ebfd0bdad 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -362,40 +362,38 @@ static int dlm_is_node_recovered(struct dlm_ctxt *dlm, u8 node)
 }
 
 
-int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout)
+void dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout)
 {
-	if (timeout) {
-		mlog(ML_NOTICE, "%s: waiting %dms for notification of "
-		     "death of node %u\n", dlm->name, timeout, node);
+	if (dlm_is_node_dead(dlm, node))
+		return;
+
+	printk(KERN_NOTICE "o2dlm: Waiting on the death of node %u in "
+	       "domain %s\n", node, dlm->name);
+
+	if (timeout)
 		wait_event_timeout(dlm->dlm_reco_thread_wq,
-			   dlm_is_node_dead(dlm, node),
-			   msecs_to_jiffies(timeout));
-	} else {
-		mlog(ML_NOTICE, "%s: waiting indefinitely for notification "
-		     "of death of node %u\n", dlm->name, node);
+				   dlm_is_node_dead(dlm, node),
+				   msecs_to_jiffies(timeout));
+	else
 		wait_event(dlm->dlm_reco_thread_wq,
 			   dlm_is_node_dead(dlm, node));
-	}
-	/* for now, return 0 */
-	return 0;
 }
 
-int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout)
+void dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout)
 {
-	if (timeout) {
-		mlog(0, "%s: waiting %dms for notification of "
-		     "recovery of node %u\n", dlm->name, timeout, node);
+	if (dlm_is_node_recovered(dlm, node))
+		return;
+
+	printk(KERN_NOTICE "o2dlm: Waiting on the recovery of node %u in "
+	       "domain %s\n", node, dlm->name);
+
+	if (timeout)
 		wait_event_timeout(dlm->dlm_reco_thread_wq,
-			   dlm_is_node_recovered(dlm, node),
-			   msecs_to_jiffies(timeout));
-	} else {
-		mlog(0, "%s: waiting indefinitely for notification "
-		     "of recovery of node %u\n", dlm->name, node);
+				   dlm_is_node_recovered(dlm, node),
+				   msecs_to_jiffies(timeout));
+	else
 		wait_event(dlm->dlm_reco_thread_wq,
 			   dlm_is_node_recovered(dlm, node));
-	}
-	/* for now, return 0 */
-	return 0;
 }
 
 /* callers of the top-level api calls (dlmlock/dlmunlock) should
@@ -430,6 +428,8 @@ static void dlm_begin_recovery(struct dlm_ctxt *dlm)
 {
 	spin_lock(&dlm->spinlock);
 	BUG_ON(dlm->reco.state & DLM_RECO_STATE_ACTIVE);
+	printk(KERN_NOTICE "o2dlm: Begin recovery on domain %s for node %u\n",
+	       dlm->name, dlm->reco.dead_node);
 	dlm->reco.state |= DLM_RECO_STATE_ACTIVE;
 	spin_unlock(&dlm->spinlock);
 }
@@ -440,9 +440,18 @@ static void dlm_end_recovery(struct dlm_ctxt *dlm)
 	BUG_ON(!(dlm->reco.state & DLM_RECO_STATE_ACTIVE));
 	dlm->reco.state &= ~DLM_RECO_STATE_ACTIVE;
 	spin_unlock(&dlm->spinlock);
+	printk(KERN_NOTICE "o2dlm: End recovery on domain %s\n", dlm->name);
 	wake_up(&dlm->reco.event);
 }
 
+static void dlm_print_recovery_master(struct dlm_ctxt *dlm)
+{
+	printk(KERN_NOTICE "o2dlm: Node %u (%s) is the Recovery Master for the "
+	       "dead node %u in domain %s\n", dlm->reco.new_master,
+	       (dlm->node_num == dlm->reco.new_master ? "me" : "he"),
+	       dlm->reco.dead_node, dlm->name);
+}
+
 static int dlm_do_recovery(struct dlm_ctxt *dlm)
 {
 	int status = 0;
@@ -505,9 +514,8 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
 		}
 		mlog(0, "another node will master this recovery session.\n");
 	}
-	mlog(0, "dlm=%s (%d), new_master=%u, this node=%u, dead_node=%u\n",
-	     dlm->name, task_pid_nr(dlm->dlm_reco_thread_task), dlm->reco.new_master,
-	     dlm->node_num, dlm->reco.dead_node);
+
+	dlm_print_recovery_master(dlm);
 
 	/* it is safe to start everything back up here
 	 * because all of the dead node's lock resources
@@ -518,15 +526,13 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
 	return 0;
 
 master_here:
-	mlog(ML_NOTICE, "(%d) Node %u is the Recovery Master for the Dead Node "
-	     "%u for Domain %s\n", task_pid_nr(dlm->dlm_reco_thread_task),
-	     dlm->node_num, dlm->reco.dead_node, dlm->name);
+	dlm_print_recovery_master(dlm);
 
 	status = dlm_remaster_locks(dlm, dlm->reco.dead_node);
 	if (status < 0) {
 		/* we should never hit this anymore */
-		mlog(ML_ERROR, "error %d remastering locks for node %u, "
-		     "retrying.\n", status, dlm->reco.dead_node);
+		mlog(ML_ERROR, "%s: Error %d remastering locks for node %u, "
+		     "retrying.\n", dlm->name, status, dlm->reco.dead_node);
 		/* yield a bit to allow any final network messages
 		 * to get handled on remaining nodes */
 		msleep(100);
@@ -567,7 +573,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
 		BUG_ON(ndata->state != DLM_RECO_NODE_DATA_INIT);
 		ndata->state = DLM_RECO_NODE_DATA_REQUESTING;
 
-		mlog(0, "requesting lock info from node %u\n",
+		mlog(0, "%s: Requesting lock info from node %u\n", dlm->name,
 		     ndata->node_num);
 
 		if (ndata->node_num == dlm->node_num) {
@@ -640,7 +646,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
 		spin_unlock(&dlm_reco_state_lock);
 	}
 
-	mlog(0, "done requesting all lock info\n");
+	mlog(0, "%s: Done requesting all lock info\n", dlm->name);
 
 	/* nodes should be sending reco data now
 	 * just need to wait */
@@ -802,10 +808,9 @@ static int dlm_request_all_locks(struct dlm_ctxt *dlm, u8 request_from,
 
 	/* negative status is handled by caller */
 	if (ret < 0)
-		mlog(ML_ERROR, "Error %d when sending message %u (key "
-		     "0x%x) to node %u\n", ret, DLM_LOCK_REQUEST_MSG,
-		     dlm->key, request_from);
-
+		mlog(ML_ERROR, "%s: Error %d send LOCK_REQUEST to node %u "
+		     "to recover dead node %u\n", dlm->name, ret,
+		     request_from, dead_node);
 	// return from here, then
 	// sleep until all received or error
 	return ret;
@@ -956,9 +961,9 @@ static int dlm_send_all_done_msg(struct dlm_ctxt *dlm, u8 dead_node, u8 send_to)
 	ret = o2net_send_message(DLM_RECO_DATA_DONE_MSG, dlm->key, &done_msg,
 				 sizeof(done_msg), send_to, &tmpret);
 	if (ret < 0) {
-		mlog(ML_ERROR, "Error %d when sending message %u (key "
-		     "0x%x) to node %u\n", ret, DLM_RECO_DATA_DONE_MSG,
-		     dlm->key, send_to);
+		mlog(ML_ERROR, "%s: Error %d send RECO_DATA_DONE to node %u "
+		     "to recover dead node %u\n", dlm->name, ret, send_to,
+		     dead_node);
 		if (!dlm_is_host_down(ret)) {
 			BUG();
 		}
@@ -1127,9 +1132,11 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm,
 	if (ret < 0) {
 		/* XXX: negative status is not handled.
 		 * this will end up killing this node. */
-		mlog(ML_ERROR, "Error %d when sending message %u (key "
-		     "0x%x) to node %u\n", ret, DLM_MIG_LOCKRES_MSG,
-		     dlm->key, send_to);
+		mlog(ML_ERROR, "%s: res %.*s, Error %d send MIG_LOCKRES to "
+		     "node %u (%s)\n", dlm->name, mres->lockname_len,
+		     mres->lockname, ret, send_to,
+		     (orig_flags & DLM_MRES_MIGRATION ?
+		      "migration" : "recovery"));
 	} else {
 		/* might get an -ENOMEM back here */
 		ret = status;
@@ -1767,7 +1774,7 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm,
 			     dlm->name, mres->lockname_len, mres->lockname,
 			     from);
 			spin_lock(&res->spinlock);
-			dlm_lockres_set_refmap_bit(from, res);
+			dlm_lockres_set_refmap_bit(dlm, res, from);
 			spin_unlock(&res->spinlock);
 			added++;
 			break;
@@ -1965,7 +1972,7 @@ skip_lvb:
 			mlog(0, "%s:%.*s: added lock for node %u, "
 			     "setting refmap bit\n", dlm->name,
 			     res->lockname.len, res->lockname.name, ml->node);
-			dlm_lockres_set_refmap_bit(ml->node, res);
+			dlm_lockres_set_refmap_bit(dlm, res, ml->node);
 			added++;
 		}
 		spin_unlock(&res->spinlock);
@@ -2084,6 +2091,9 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
 
 	list_for_each_entry_safe(res, next, &dlm->reco.resources, recovering) {
 		if (res->owner == dead_node) {
+			mlog(0, "%s: res %.*s, Changing owner from %u to %u\n",
+			     dlm->name, res->lockname.len, res->lockname.name,
+			     res->owner, new_master);
 			list_del_init(&res->recovering);
 			spin_lock(&res->spinlock);
 			/* new_master has our reference from
@@ -2105,40 +2115,30 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
 	for (i = 0; i < DLM_HASH_BUCKETS; i++) {
 		bucket = dlm_lockres_hash(dlm, i);
 		hlist_for_each_entry(res, hash_iter, bucket, hash_node) {
-			if (res->state & DLM_LOCK_RES_RECOVERING) {
-				if (res->owner == dead_node) {
-					mlog(0, "(this=%u) res %.*s owner=%u "
-					     "was not on recovering list, but "
-					     "clearing state anyway\n",
-					     dlm->node_num, res->lockname.len,
-					     res->lockname.name, new_master);
-				} else if (res->owner == dlm->node_num) {
-					mlog(0, "(this=%u) res %.*s owner=%u "
-					     "was not on recovering list, "
-					     "owner is THIS node, clearing\n",
-					     dlm->node_num, res->lockname.len,
-					     res->lockname.name, new_master);
-				} else
-					continue;
+			if (!(res->state & DLM_LOCK_RES_RECOVERING))
+				continue;
 
-				if (!list_empty(&res->recovering)) {
-					mlog(0, "%s:%.*s: lockres was "
-					     "marked RECOVERING, owner=%u\n",
-					     dlm->name, res->lockname.len,
-					     res->lockname.name, res->owner);
-					list_del_init(&res->recovering);
-					dlm_lockres_put(res);
-				}
-				spin_lock(&res->spinlock);
-				/* new_master has our reference from
-				 * the lock state sent during recovery */
-				dlm_change_lockres_owner(dlm, res, new_master);
-				res->state &= ~DLM_LOCK_RES_RECOVERING;
-				if (__dlm_lockres_has_locks(res))
-					__dlm_dirty_lockres(dlm, res);
-				spin_unlock(&res->spinlock);
-				wake_up(&res->wq);
+			if (res->owner != dead_node &&
+			    res->owner != dlm->node_num)
+				continue;
+
+			if (!list_empty(&res->recovering)) {
+				list_del_init(&res->recovering);
+				dlm_lockres_put(res);
 			}
+
+			/* new_master has our reference from
+			 * the lock state sent during recovery */
+			mlog(0, "%s: res %.*s, Changing owner from %u to %u\n",
+			     dlm->name, res->lockname.len, res->lockname.name,
+			     res->owner, new_master);
+			spin_lock(&res->spinlock);
+			dlm_change_lockres_owner(dlm, res, new_master);
+			res->state &= ~DLM_LOCK_RES_RECOVERING;
+			if (__dlm_lockres_has_locks(res))
+				__dlm_dirty_lockres(dlm, res);
+			spin_unlock(&res->spinlock);
+			wake_up(&res->wq);
 		}
 	}
 }
@@ -2252,12 +2252,12 @@ static void dlm_free_dead_locks(struct dlm_ctxt *dlm,
 			     res->lockname.len, res->lockname.name, freed, dead_node);
 			__dlm_print_one_lock_resource(res);
 		}
-		dlm_lockres_clear_refmap_bit(dead_node, res);
+		dlm_lockres_clear_refmap_bit(dlm, res, dead_node);
 	} else if (test_bit(dead_node, res->refmap)) {
 		mlog(0, "%s:%.*s: dead node %u had a ref, but had "
 		     "no locks and had not purged before dying\n", dlm->name,
 		     res->lockname.len, res->lockname.name, dead_node);
-		dlm_lockres_clear_refmap_bit(dead_node, res);
+		dlm_lockres_clear_refmap_bit(dlm, res, dead_node);
 	}
 
 	/* do not kick thread yet */
@@ -2324,9 +2324,9 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node)
 			dlm_revalidate_lvb(dlm, res, dead_node);
 			if (res->owner == dead_node) {
 				if (res->state & DLM_LOCK_RES_DROPPING_REF) {
-					mlog(ML_NOTICE, "Ignore %.*s for "
+					mlog(ML_NOTICE, "%s: res %.*s, Skip "
 					     "recovery as it is being freed\n",
-					     res->lockname.len,
+					     dlm->name, res->lockname.len,
 					     res->lockname.name);
 				} else
 					dlm_move_lockres_to_recovery_list(dlm,
diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c
index 1d6d1d22c47..e73c833fc2a 100644
--- a/fs/ocfs2/dlm/dlmthread.c
+++ b/fs/ocfs2/dlm/dlmthread.c
@@ -94,24 +94,26 @@ int __dlm_lockres_unused(struct dlm_lock_resource *res)
 {
 	int bit;
 
+	assert_spin_locked(&res->spinlock);
+
 	if (__dlm_lockres_has_locks(res))
 		return 0;
 
+	/* Locks are in the process of being created */
+	if (res->inflight_locks)
+		return 0;
+
 	if (!list_empty(&res->dirty) || res->state & DLM_LOCK_RES_DIRTY)
 		return 0;
 
 	if (res->state & DLM_LOCK_RES_RECOVERING)
 		return 0;
 
+	/* Another node has this resource with this node as the master */
 	bit = find_next_bit(res->refmap, O2NM_MAX_NODES, 0);
 	if (bit < O2NM_MAX_NODES)
 		return 0;
 
-	/*
-	 * since the bit for dlm->node_num is not set, inflight_locks better
-	 * be zero
-	 */
-	BUG_ON(res->inflight_locks != 0);
 	return 1;
 }
 
@@ -185,8 +187,6 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm,
 		/* clear our bit from the master's refmap, ignore errors */
 		ret = dlm_drop_lockres_ref(dlm, res);
 		if (ret < 0) {
-			mlog(ML_ERROR, "%s: deref %.*s failed %d\n", dlm->name,
-			     res->lockname.len, res->lockname.name, ret);
 			if (!dlm_is_host_down(ret))
 				BUG();
 		}
@@ -209,7 +209,7 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm,
 		BUG();
 	}
 
-	__dlm_unhash_lockres(res);
+	__dlm_unhash_lockres(dlm, res);
 
 	/* lockres is not in the hash now.  drop the flag and wake up
 	 * any processes waiting in dlm_get_lock_resource. */
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index e1ed5e502ff..81a4cd22f80 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -1692,7 +1692,7 @@ int ocfs2_open_lock(struct inode *inode)
 	mlog(0, "inode %llu take PRMODE open lock\n",
 	     (unsigned long long)OCFS2_I(inode)->ip_blkno);
 
-	if (ocfs2_mount_local(osb))
+	if (ocfs2_is_hard_readonly(osb) || ocfs2_mount_local(osb))
 		goto out;
 
 	lockres = &OCFS2_I(inode)->ip_open_lockres;
@@ -1718,6 +1718,12 @@ int ocfs2_try_open_lock(struct inode *inode, int write)
 	     (unsigned long long)OCFS2_I(inode)->ip_blkno,
 	     write ? "EXMODE" : "PRMODE");
 
+	if (ocfs2_is_hard_readonly(osb)) {
+		if (write)
+			status = -EROFS;
+		goto out;
+	}
+
 	if (ocfs2_mount_local(osb))
 		goto out;
 
@@ -2298,7 +2304,7 @@ int ocfs2_inode_lock_full_nested(struct inode *inode,
 	if (ocfs2_is_hard_readonly(osb)) {
 		if (ex)
 			status = -EROFS;
-		goto bail;
+		goto getbh;
 	}
 
 	if (ocfs2_mount_local(osb))
@@ -2356,7 +2362,7 @@ local:
 			mlog_errno(status);
 		goto bail;
 	}
-
+getbh:
 	if (ret_bh) {
 		status = ocfs2_assign_bh(inode, ret_bh, local_bh);
 		if (status < 0) {
@@ -2628,8 +2634,11 @@ int ocfs2_dentry_lock(struct dentry *dentry, int ex)
 
 	BUG_ON(!dl);
 
-	if (ocfs2_is_hard_readonly(osb))
-		return -EROFS;
+	if (ocfs2_is_hard_readonly(osb)) {
+		if (ex)
+			return -EROFS;
+		return 0;
+	}
 
 	if (ocfs2_mount_local(osb))
 		return 0;
@@ -2647,7 +2656,7 @@ void ocfs2_dentry_unlock(struct dentry *dentry, int ex)
 	struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
 	struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
 
-	if (!ocfs2_mount_local(osb))
+	if (!ocfs2_is_hard_readonly(osb) && !ocfs2_mount_local(osb))
 		ocfs2_cluster_unlock(osb, &dl->dl_lockres, level);
 }
 
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 23457b491e8..2f5b92ef0e5 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -832,6 +832,102 @@ out:
 	return ret;
 }
 
+int ocfs2_seek_data_hole_offset(struct file *file, loff_t *offset, int origin)
+{
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+	unsigned int is_last = 0, is_data = 0;
+	u16 cs_bits = OCFS2_SB(inode->i_sb)->s_clustersize_bits;
+	u32 cpos, cend, clen, hole_size;
+	u64 extoff, extlen;
+	struct buffer_head *di_bh = NULL;
+	struct ocfs2_extent_rec rec;
+
+	BUG_ON(origin != SEEK_DATA && origin != SEEK_HOLE);
+
+	ret = ocfs2_inode_lock(inode, &di_bh, 0);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
+	down_read(&OCFS2_I(inode)->ip_alloc_sem);
+
+	if (*offset >= inode->i_size) {
+		ret = -ENXIO;
+		goto out_unlock;
+	}
+
+	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
+		if (origin == SEEK_HOLE)
+			*offset = inode->i_size;
+		goto out_unlock;
+	}
+
+	clen = 0;
+	cpos = *offset >> cs_bits;
+	cend = ocfs2_clusters_for_bytes(inode->i_sb, inode->i_size);
+
+	while (cpos < cend && !is_last) {
+		ret = ocfs2_get_clusters_nocache(inode, di_bh, cpos, &hole_size,
+						 &rec, &is_last);
+		if (ret) {
+			mlog_errno(ret);
+			goto out_unlock;
+		}
+
+		extoff = cpos;
+		extoff <<= cs_bits;
+
+		if (rec.e_blkno == 0ULL) {
+			clen = hole_size;
+			is_data = 0;
+		} else {
+			clen = le16_to_cpu(rec.e_leaf_clusters) -
+				(cpos - le32_to_cpu(rec.e_cpos));
+			is_data = (rec.e_flags & OCFS2_EXT_UNWRITTEN) ?  0 : 1;
+		}
+
+		if ((!is_data && origin == SEEK_HOLE) ||
+		    (is_data && origin == SEEK_DATA)) {
+			if (extoff > *offset)
+				*offset = extoff;
+			goto out_unlock;
+		}
+
+		if (!is_last)
+			cpos += clen;
+	}
+
+	if (origin == SEEK_HOLE) {
+		extoff = cpos;
+		extoff <<= cs_bits;
+		extlen = clen;
+		extlen <<=  cs_bits;
+
+		if ((extoff + extlen) > inode->i_size)
+			extlen = inode->i_size - extoff;
+		extoff += extlen;
+		if (extoff > *offset)
+			*offset = extoff;
+		goto out_unlock;
+	}
+
+	ret = -ENXIO;
+
+out_unlock:
+
+	brelse(di_bh);
+
+	up_read(&OCFS2_I(inode)->ip_alloc_sem);
+
+	ocfs2_inode_unlock(inode, 0);
+out:
+	if (ret && ret != -ENXIO)
+		ret = -ENXIO;
+	return ret;
+}
+
 int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
 			   struct buffer_head *bhs[], int flags,
 			   int (*validate)(struct super_block *sb,
diff --git a/fs/ocfs2/extent_map.h b/fs/ocfs2/extent_map.h
index e79d41c2c90..67ea57d2fd5 100644
--- a/fs/ocfs2/extent_map.h
+++ b/fs/ocfs2/extent_map.h
@@ -53,6 +53,8 @@ int ocfs2_extent_map_get_blocks(struct inode *inode, u64 v_blkno, u64 *p_blkno,
 int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		 u64 map_start, u64 map_len);
 
+int ocfs2_seek_data_hole_offset(struct file *file, loff_t *offset, int origin);
+
 int ocfs2_xattr_get_clusters(struct inode *inode, u32 v_cluster,
 			     u32 *p_cluster, u32 *num_clusters,
 			     struct ocfs2_extent_list *el,
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index de4ea1af041..6e396683c3d 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1950,6 +1950,9 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode,
 	if (ret < 0)
 		mlog_errno(ret);
 
+	if (file->f_flags & O_SYNC)
+		handle->h_sync = 1;
+
 	ocfs2_commit_trans(osb, handle);
 
 out_inode_unlock:
@@ -2052,6 +2055,23 @@ out:
 	return ret;
 }
 
+static void ocfs2_aiodio_wait(struct inode *inode)
+{
+	wait_queue_head_t *wq = ocfs2_ioend_wq(inode);
+
+	wait_event(*wq, (atomic_read(&OCFS2_I(inode)->ip_unaligned_aio) == 0));
+}
+
+static int ocfs2_is_io_unaligned(struct inode *inode, size_t count, loff_t pos)
+{
+	int blockmask = inode->i_sb->s_blocksize - 1;
+	loff_t final_size = pos + count;
+
+	if ((pos & blockmask) || (final_size & blockmask))
+		return 1;
+	return 0;
+}
+
 static int ocfs2_prepare_inode_for_refcount(struct inode *inode,
 					    struct file *file,
 					    loff_t pos, size_t count,
@@ -2230,6 +2250,7 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	int full_coherency = !(osb->s_mount_opt &
 			       OCFS2_MOUNT_COHERENCY_BUFFERED);
+	int unaligned_dio = 0;
 
 	trace_ocfs2_file_aio_write(inode, file, file->f_path.dentry,
 		(unsigned long long)OCFS2_I(inode)->ip_blkno,
@@ -2297,6 +2318,10 @@ relock:
 		goto out;
 	}
 
+	if (direct_io && !is_sync_kiocb(iocb))
+		unaligned_dio = ocfs2_is_io_unaligned(inode, iocb->ki_left,
+						      *ppos);
+
 	/*
 	 * We can't complete the direct I/O as requested, fall back to
 	 * buffered I/O.
@@ -2311,6 +2336,18 @@ relock:
 		goto relock;
 	}
 
+	if (unaligned_dio) {
+		/*
+		 * Wait on previous unaligned aio to complete before
+		 * proceeding.
+		 */
+		ocfs2_aiodio_wait(inode);
+
+		/* Mark the iocb as needing a decrement in ocfs2_dio_end_io */
+		atomic_inc(&OCFS2_I(inode)->ip_unaligned_aio);
+		ocfs2_iocb_set_unaligned_aio(iocb);
+	}
+
 	/*
 	 * To later detect whether a journal commit for sync writes is
 	 * necessary, we sample i_size, and cluster count here.
@@ -2382,8 +2419,12 @@ out_dio:
 	if ((ret == -EIOCBQUEUED) || (!ocfs2_iocb_is_rw_locked(iocb))) {
 		rw_level = -1;
 		have_alloc_sem = 0;
+		unaligned_dio = 0;
 	}
 
+	if (unaligned_dio)
+		atomic_dec(&OCFS2_I(inode)->ip_unaligned_aio);
+
 out:
 	if (rw_level != -1)
 		ocfs2_rw_unlock(inode, rw_level);
@@ -2591,6 +2632,57 @@ bail:
 	return ret;
 }
 
+/* Refer generic_file_llseek_unlocked() */
+static loff_t ocfs2_file_llseek(struct file *file, loff_t offset, int origin)
+{
+	struct inode *inode = file->f_mapping->host;
+	int ret = 0;
+
+	mutex_lock(&inode->i_mutex);
+
+	switch (origin) {
+	case SEEK_SET:
+		break;
+	case SEEK_END:
+		offset += inode->i_size;
+		break;
+	case SEEK_CUR:
+		if (offset == 0) {
+			offset = file->f_pos;
+			goto out;
+		}
+		offset += file->f_pos;
+		break;
+	case SEEK_DATA:
+	case SEEK_HOLE:
+		ret = ocfs2_seek_data_hole_offset(file, &offset, origin);
+		if (ret)
+			goto out;
+		break;
+	default:
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET))
+		ret = -EINVAL;
+	if (!ret && offset > inode->i_sb->s_maxbytes)
+		ret = -EINVAL;
+	if (ret)
+		goto out;
+
+	if (offset != file->f_pos) {
+		file->f_pos = offset;
+		file->f_version = 0;
+	}
+
+out:
+	mutex_unlock(&inode->i_mutex);
+	if (ret)
+		return ret;
+	return offset;
+}
+
 const struct inode_operations ocfs2_file_iops = {
 	.setattr	= ocfs2_setattr,
 	.getattr	= ocfs2_getattr,
@@ -2615,7 +2707,7 @@ const struct inode_operations ocfs2_special_file_iops = {
  * ocfs2_fops_no_plocks and ocfs2_dops_no_plocks!
  */
 const struct file_operations ocfs2_fops = {
-	.llseek		= generic_file_llseek,
+	.llseek		= ocfs2_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
 	.mmap		= ocfs2_mmap,
@@ -2663,7 +2755,7 @@ const struct file_operations ocfs2_dops = {
  * the cluster.
  */
 const struct file_operations ocfs2_fops_no_plocks = {
-	.llseek		= generic_file_llseek,
+	.llseek		= ocfs2_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
 	.mmap		= ocfs2_mmap,
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index a22d2c09889..17454a904d7 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -951,7 +951,7 @@ static void ocfs2_cleanup_delete_inode(struct inode *inode,
 	trace_ocfs2_cleanup_delete_inode(
 		(unsigned long long)OCFS2_I(inode)->ip_blkno, sync_data);
 	if (sync_data)
-		write_inode_now(inode, 1);
+		filemap_write_and_wait(inode->i_mapping);
 	truncate_inode_pages(&inode->i_data, 0);
 }
 
diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h
index 1c508b149b3..88924a3133f 100644
--- a/fs/ocfs2/inode.h
+++ b/fs/ocfs2/inode.h
@@ -43,6 +43,9 @@ struct ocfs2_inode_info
 	/* protects extended attribute changes on this inode */
 	struct rw_semaphore		ip_xattr_sem;
 
+	/* Number of outstanding AIO's which are not page aligned */
+	atomic_t			ip_unaligned_aio;
+
 	/* These fields are protected by ip_lock */
 	spinlock_t			ip_lock;
 	u32				ip_open_count;
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index bc91072b721..726ff265b29 100644
--- a/fs/ocfs2/ioctl.c
+++ b/fs/ocfs2/ioctl.c
@@ -122,7 +122,7 @@ static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
 	if ((oldflags & OCFS2_IMMUTABLE_FL) || ((flags ^ oldflags) &
 		(OCFS2_APPEND_FL | OCFS2_IMMUTABLE_FL))) {
 		if (!capable(CAP_LINUX_IMMUTABLE))
-			goto bail_unlock;
+			goto bail_commit;
 	}
 
 	ocfs2_inode->ip_attr = flags;
@@ -132,6 +132,7 @@ static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
 	if (status < 0)
 		mlog_errno(status);
 
+bail_commit:
 	ocfs2_commit_trans(osb, handle);
 bail_unlock:
 	ocfs2_inode_unlock(inode, 1);
@@ -381,7 +382,7 @@ int ocfs2_info_handle_freeinode(struct inode *inode,
 	if (!oifi) {
 		status = -ENOMEM;
 		mlog_errno(status);
-		goto bail;
+		goto out_err;
 	}
 
 	if (o2info_from_user(*oifi, req))
@@ -431,7 +432,7 @@ bail:
 		o2info_set_request_error(&oifi->ifi_req, req);
 
 	kfree(oifi);
-
+out_err:
 	return status;
 }
 
@@ -666,7 +667,7 @@ int ocfs2_info_handle_freefrag(struct inode *inode,
 	if (!oiff) {
 		status = -ENOMEM;
 		mlog_errno(status);
-		goto bail;
+		goto out_err;
 	}
 
 	if (o2info_from_user(*oiff, req))
@@ -716,7 +717,7 @@ bail:
 		o2info_set_request_error(&oiff->iff_req, req);
 
 	kfree(oiff);
-
+out_err:
 	return status;
 }
 
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index 295d56454e8..0a42ae96dca 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -1544,9 +1544,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
 	/* we need to run complete recovery for offline orphan slots */
 	ocfs2_replay_map_set_state(osb, REPLAY_NEEDED);
 
-	mlog(ML_NOTICE, "Recovering node %d from slot %d on device (%u,%u)\n",
-	     node_num, slot_num,
-	     MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));
+	printk(KERN_NOTICE "ocfs2: Begin replay journal (node %d, slot %d) on "\
+	       "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+	       MINOR(osb->sb->s_dev));
 
 	OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
 
@@ -1601,6 +1601,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
 
 	jbd2_journal_destroy(journal);
 
+	printk(KERN_NOTICE "ocfs2: End replay journal (node %d, slot %d) on "\
+	       "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+	       MINOR(osb->sb->s_dev));
 done:
 	/* drop the lock on this nodes journal */
 	if (got_lock)
@@ -1808,6 +1811,20 @@ static inline unsigned long ocfs2_orphan_scan_timeout(void)
  * every slot, queuing a recovery of the slot on the ocfs2_wq thread. This
  * is done to catch any orphans that are left over in orphan directories.
  *
+ * It scans all slots, even ones that are in use. It does so to handle the
+ * case described below:
+ *
+ *   Node 1 has an inode it was using. The dentry went away due to memory
+ *   pressure.  Node 1 closes the inode, but it's on the free list. The node
+ *   has the open lock.
+ *   Node 2 unlinks the inode. It grabs the dentry lock to notify others,
+ *   but node 1 has no dentry and doesn't get the message. It trylocks the
+ *   open lock, sees that another node has a PR, and does nothing.
+ *   Later node 2 runs its orphan dir. It igets the inode, trylocks the
+ *   open lock, sees the PR still, and does nothing.
+ *   Basically, we have to trigger an orphan iput on node 1. The only way
+ *   for this to happen is if node 1 runs node 2's orphan dir.
+ *
  * ocfs2_queue_orphan_scan gets called every ORPHAN_SCAN_SCHEDULE_TIMEOUT
  * seconds.  It gets an EX lock on os_lockres and checks sequence number
  * stored in LVB. If the sequence number has changed, it means some other
diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h
index 68cf2f6d3c6..a3385b63ff5 100644
--- a/fs/ocfs2/journal.h
+++ b/fs/ocfs2/journal.h
@@ -441,10 +441,11 @@ static inline int ocfs2_mknod_credits(struct super_block *sb, int is_dir,
 #define OCFS2_SIMPLE_DIR_EXTEND_CREDITS (2)
 
 /* file update (nlink, etc) + directory mtime/ctime + dir entry block + quota
- * update on dir + index leaf + dx root update for free list */
+ * update on dir + index leaf + dx root update for free list +
+ * previous dirblock update in the free list */
 static inline int ocfs2_link_credits(struct super_block *sb)
 {
-	return 2*OCFS2_INODE_UPDATE_CREDITS + 3 +
+	return 2*OCFS2_INODE_UPDATE_CREDITS + 4 +
 	       ocfs2_quota_trans_credits(sb);
 }
 
diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c
index 3e9393ca39e..9cd41083e99 100644
--- a/fs/ocfs2/mmap.c
+++ b/fs/ocfs2/mmap.c
@@ -61,7 +61,7 @@ static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
 static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
 				struct page *page)
 {
-	int ret;
+	int ret = VM_FAULT_NOPAGE;
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct address_space *mapping = inode->i_mapping;
 	loff_t pos = page_offset(page);
@@ -71,32 +71,25 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
 	void *fsdata;
 	loff_t size = i_size_read(inode);
 
-	/*
-	 * Another node might have truncated while we were waiting on
-	 * cluster locks.
-	 * We don't check size == 0 before the shift. This is borrowed
-	 * from do_generic_file_read.
-	 */
 	last_index = (size - 1) >> PAGE_CACHE_SHIFT;
-	if (unlikely(!size || page->index > last_index)) {
-		ret = -EINVAL;
-		goto out;
-	}
 
 	/*
-	 * The i_size check above doesn't catch the case where nodes
-	 * truncated and then re-extended the file. We'll re-check the
-	 * page mapping after taking the page lock inside of
-	 * ocfs2_write_begin_nolock().
+	 * There are cases that lead to the page no longer bebongs to the
+	 * mapping.
+	 * 1) pagecache truncates locally due to memory pressure.
+	 * 2) pagecache truncates when another is taking EX lock against 
+	 * inode lock. see ocfs2_data_convert_worker.
+	 * 
+	 * The i_size check doesn't catch the case where nodes truncated and
+	 * then re-extended the file. We'll re-check the page mapping after
+	 * taking the page lock inside of ocfs2_write_begin_nolock().
+	 *
+	 * Let VM retry with these cases.
 	 */
-	if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
-		/*
-		 * the page has been umapped in ocfs2_data_downconvert_worker.
-		 * So return 0 here and let VFS retry.
-		 */
-		ret = 0;
+	if ((page->mapping != inode->i_mapping) ||
+	    (!PageUptodate(page)) ||
+	    (page_offset(page) >= size))
 		goto out;
-	}
 
 	/*
 	 * Call ocfs2_write_begin() and ocfs2_write_end() to take
@@ -116,17 +109,21 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
 	if (ret) {
 		if (ret != -ENOSPC)
 			mlog_errno(ret);
+		if (ret == -ENOMEM)
+			ret = VM_FAULT_OOM;
+		else
+			ret = VM_FAULT_SIGBUS;
 		goto out;
 	}
 
-	ret = ocfs2_write_end_nolock(mapping, pos, len, len, locked_page,
-				     fsdata);
-	if (ret < 0) {
-		mlog_errno(ret);
+	if (!locked_page) {
+		ret = VM_FAULT_NOPAGE;
 		goto out;
 	}
+	ret = ocfs2_write_end_nolock(mapping, pos, len, len, locked_page,
+				     fsdata);
 	BUG_ON(ret != len);
-	ret = 0;
+	ret = VM_FAULT_LOCKED;
 out:
 	return ret;
 }
@@ -168,8 +165,6 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 
 out:
 	ocfs2_unblock_signals(&oldset);
-	if (ret)
-		ret = VM_FAULT_SIGBUS;
 	return ret;
 }
 
diff --git a/fs/ocfs2/move_extents.c b/fs/ocfs2/move_extents.c
index d53cb706f14..184c76b8c29 100644
--- a/fs/ocfs2/move_extents.c
+++ b/fs/ocfs2/move_extents.c
@@ -745,7 +745,7 @@ static int ocfs2_move_extent(struct ocfs2_move_extents_context *context,
 	 */
 	ocfs2_probe_alloc_group(inode, gd_bh, &goal_bit, len, move_max_hop,
 				new_phys_cpos);
-	if (!new_phys_cpos) {
+	if (!*new_phys_cpos) {
 		ret = -ENOSPC;
 		goto out_commit;
 	}
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 409285854f6..d355e6e36b3 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -836,18 +836,65 @@ static inline unsigned int ocfs2_clusters_to_megabytes(struct super_block *sb,
 
 static inline void _ocfs2_set_bit(unsigned int bit, unsigned long *bitmap)
 {
-	__test_and_set_bit_le(bit, bitmap);
+	__set_bit_le(bit, bitmap);
 }
 #define ocfs2_set_bit(bit, addr) _ocfs2_set_bit((bit), (unsigned long *)(addr))
 
 static inline void _ocfs2_clear_bit(unsigned int bit, unsigned long *bitmap)
 {
-	__test_and_clear_bit_le(bit, bitmap);
+	__clear_bit_le(bit, bitmap);
 }
 #define ocfs2_clear_bit(bit, addr) _ocfs2_clear_bit((bit), (unsigned long *)(addr))
 
 #define ocfs2_test_bit test_bit_le
 #define ocfs2_find_next_zero_bit find_next_zero_bit_le
 #define ocfs2_find_next_bit find_next_bit_le
+
+static inline void *correct_addr_and_bit_unaligned(int *bit, void *addr)
+{
+#if BITS_PER_LONG == 64
+	*bit += ((unsigned long) addr & 7UL) << 3;
+	addr = (void *) ((unsigned long) addr & ~7UL);
+#elif BITS_PER_LONG == 32
+	*bit += ((unsigned long) addr & 3UL) << 3;
+	addr = (void *) ((unsigned long) addr & ~3UL);
+#else
+#error "how many bits you are?!"
+#endif
+	return addr;
+}
+
+static inline void ocfs2_set_bit_unaligned(int bit, void *bitmap)
+{
+	bitmap = correct_addr_and_bit_unaligned(&bit, bitmap);
+	ocfs2_set_bit(bit, bitmap);
+}
+
+static inline void ocfs2_clear_bit_unaligned(int bit, void *bitmap)
+{
+	bitmap = correct_addr_and_bit_unaligned(&bit, bitmap);
+	ocfs2_clear_bit(bit, bitmap);
+}
+
+static inline int ocfs2_test_bit_unaligned(int bit, void *bitmap)
+{
+	bitmap = correct_addr_and_bit_unaligned(&bit, bitmap);
+	return ocfs2_test_bit(bit, bitmap);
+}
+
+static inline int ocfs2_find_next_zero_bit_unaligned(void *bitmap, int max,
+							int start)
+{
+	int fix = 0, ret, tmpmax;
+	bitmap = correct_addr_and_bit_unaligned(&fix, bitmap);
+	tmpmax = max + fix;
+	start += fix;
+
+	ret = ocfs2_find_next_zero_bit(bitmap, tmpmax, start) - fix;
+	if (ret > max)
+		return max;
+	return ret;
+}
+
 #endif  /* OCFS2_H */
 
diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c
index dc8007fc924..f100bf70a90 100644
--- a/fs/ocfs2/quota_local.c
+++ b/fs/ocfs2/quota_local.c
@@ -404,7 +404,9 @@ struct ocfs2_quota_recovery *ocfs2_begin_quota_recovery(
 	int status = 0;
 	struct ocfs2_quota_recovery *rec;
 
-	mlog(ML_NOTICE, "Beginning quota recovery in slot %u\n", slot_num);
+	printk(KERN_NOTICE "ocfs2: Beginning quota recovery on device (%s) for "
+	       "slot %u\n", osb->dev_str, slot_num);
+
 	rec = ocfs2_alloc_quota_recovery();
 	if (!rec)
 		return ERR_PTR(-ENOMEM);
@@ -549,8 +551,8 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode,
 				goto out_commit;
 			}
 			lock_buffer(qbh);
-			WARN_ON(!ocfs2_test_bit(bit, dchunk->dqc_bitmap));
-			ocfs2_clear_bit(bit, dchunk->dqc_bitmap);
+			WARN_ON(!ocfs2_test_bit_unaligned(bit, dchunk->dqc_bitmap));
+			ocfs2_clear_bit_unaligned(bit, dchunk->dqc_bitmap);
 			le32_add_cpu(&dchunk->dqc_free, 1);
 			unlock_buffer(qbh);
 			ocfs2_journal_dirty(handle, qbh);
@@ -596,7 +598,9 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
 	struct inode *lqinode;
 	unsigned int flags;
 
-	mlog(ML_NOTICE, "Finishing quota recovery in slot %u\n", slot_num);
+	printk(KERN_NOTICE "ocfs2: Finishing quota recovery on device (%s) for "
+	       "slot %u\n", osb->dev_str, slot_num);
+
 	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
 	for (type = 0; type < MAXQUOTAS; type++) {
 		if (list_empty(&(rec->r_list[type])))
@@ -612,8 +616,9 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
 		/* Someone else is holding the lock? Then he must be
 		 * doing the recovery. Just skip the file... */
 		if (status == -EAGAIN) {
-			mlog(ML_NOTICE, "skipping quota recovery for slot %d "
-			     "because quota file is locked.\n", slot_num);
+			printk(KERN_NOTICE "ocfs2: Skipping quota recovery on "
+			       "device (%s) for slot %d because quota file is "
+			       "locked.\n", osb->dev_str, slot_num);
 			status = 0;
 			goto out_put;
 		} else if (status < 0) {
@@ -944,7 +949,7 @@ static struct ocfs2_quota_chunk *ocfs2_find_free_entry(struct super_block *sb,
 		      * ol_quota_entries_per_block(sb);
 	}
 
-	found = ocfs2_find_next_zero_bit(dchunk->dqc_bitmap, len, 0);
+	found = ocfs2_find_next_zero_bit_unaligned(dchunk->dqc_bitmap, len, 0);
 	/* We failed? */
 	if (found == len) {
 		mlog(ML_ERROR, "Did not find empty entry in chunk %d with %u"
@@ -1208,7 +1213,7 @@ static void olq_alloc_dquot(struct buffer_head *bh, void *private)
 	struct ocfs2_local_disk_chunk *dchunk;
 
 	dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data;
-	ocfs2_set_bit(*offset, dchunk->dqc_bitmap);
+	ocfs2_set_bit_unaligned(*offset, dchunk->dqc_bitmap);
 	le32_add_cpu(&dchunk->dqc_free, -1);
 }
 
@@ -1289,7 +1294,7 @@ int ocfs2_local_release_dquot(handle_t *handle, struct dquot *dquot)
 			(od->dq_chunk->qc_headerbh->b_data);
 	/* Mark structure as freed */
 	lock_buffer(od->dq_chunk->qc_headerbh);
-	ocfs2_clear_bit(offset, dchunk->dqc_bitmap);
+	ocfs2_clear_bit_unaligned(offset, dchunk->dqc_bitmap);
 	le32_add_cpu(&dchunk->dqc_free, 1);
 	unlock_buffer(od->dq_chunk->qc_headerbh);
 	ocfs2_journal_dirty(handle, od->dq_chunk->qc_headerbh);
diff --git a/fs/ocfs2/slot_map.c b/fs/ocfs2/slot_map.c
index 26fc0014d50..1424c151ccc 100644
--- a/fs/ocfs2/slot_map.c
+++ b/fs/ocfs2/slot_map.c
@@ -493,8 +493,8 @@ int ocfs2_find_slot(struct ocfs2_super *osb)
 			goto bail;
 		}
 	} else
-		mlog(ML_NOTICE, "slot %d is already allocated to this node!\n",
-		     slot);
+		printk(KERN_INFO "ocfs2: Slot %d on device (%s) was already "
+		       "allocated to this node!\n", slot, osb->dev_str);
 
 	ocfs2_set_slot(si, slot, osb->node_num);
 	osb->slot_num = slot;
diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c
index 19965b00c43..94368017edb 100644
--- a/fs/ocfs2/stack_o2cb.c
+++ b/fs/ocfs2/stack_o2cb.c
@@ -28,6 +28,7 @@
 #include "cluster/masklog.h"
 #include "cluster/nodemanager.h"
 #include "cluster/heartbeat.h"
+#include "cluster/tcp.h"
 
 #include "stackglue.h"
 
@@ -256,6 +257,61 @@ static void o2cb_dump_lksb(struct ocfs2_dlm_lksb *lksb)
 }
 
 /*
+ * Check if this node is heartbeating and is connected to all other
+ * heartbeating nodes.
+ */
+static int o2cb_cluster_check(void)
+{
+	u8 node_num;
+	int i;
+	unsigned long hbmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	unsigned long netmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
+
+	node_num = o2nm_this_node();
+	if (node_num == O2NM_MAX_NODES) {
+		printk(KERN_ERR "o2cb: This node has not been configured.\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * o2dlm expects o2net sockets to be created. If not, then
+	 * dlm_join_domain() fails with a stack of errors which are both cryptic
+	 * and incomplete. The idea here is to detect upfront whether we have
+	 * managed to connect to all nodes or not. If not, then list the nodes
+	 * to allow the user to check the configuration (incorrect IP, firewall,
+	 * etc.) Yes, this is racy. But its not the end of the world.
+	 */
+#define	O2CB_MAP_STABILIZE_COUNT	60
+	for (i = 0; i < O2CB_MAP_STABILIZE_COUNT; ++i) {
+		o2hb_fill_node_map(hbmap, sizeof(hbmap));
+		if (!test_bit(node_num, hbmap)) {
+			printk(KERN_ERR "o2cb: %s heartbeat has not been "
+			       "started.\n", (o2hb_global_heartbeat_active() ?
+					      "Global" : "Local"));
+			return -EINVAL;
+		}
+		o2net_fill_node_map(netmap, sizeof(netmap));
+		/* Force set the current node to allow easy compare */
+		set_bit(node_num, netmap);
+		if (!memcmp(hbmap, netmap, sizeof(hbmap)))
+			return 0;
+		if (i < O2CB_MAP_STABILIZE_COUNT)
+			msleep(1000);
+	}
+
+	printk(KERN_ERR "o2cb: This node could not connect to nodes:");
+	i = -1;
+	while ((i = find_next_bit(hbmap, O2NM_MAX_NODES,
+				  i + 1)) < O2NM_MAX_NODES) {
+		if (!test_bit(i, netmap))
+			printk(" %u", i);
+	}
+	printk(".\n");
+
+	return -ENOTCONN;
+}
+
+/*
  * Called from the dlm when it's about to evict a node. This is how the
  * classic stack signals node death.
  */
@@ -263,8 +319,8 @@ static void o2dlm_eviction_cb(int node_num, void *data)
 {
 	struct ocfs2_cluster_connection *conn = data;
 
-	mlog(ML_NOTICE, "o2dlm has evicted node %d from group %.*s\n",
-	     node_num, conn->cc_namelen, conn->cc_name);
+	printk(KERN_NOTICE "o2cb: o2dlm has evicted node %d from domain %.*s\n",
+	       node_num, conn->cc_namelen, conn->cc_name);
 
 	conn->cc_recovery_handler(node_num, conn->cc_recovery_data);
 }
@@ -280,12 +336,11 @@ static int o2cb_cluster_connect(struct ocfs2_cluster_connection *conn)
 	BUG_ON(conn == NULL);
 	BUG_ON(conn->cc_proto == NULL);
 
-	/* for now we only have one cluster/node, make sure we see it
-	 * in the heartbeat universe */
-	if (!o2hb_check_local_node_heartbeating()) {
-		if (o2hb_global_heartbeat_active())
-			mlog(ML_ERROR, "Global heartbeat not started\n");
-		rc = -EINVAL;
+	/* Ensure cluster stack is up and all nodes are connected */
+	rc = o2cb_cluster_check();
+	if (rc) {
+		printk(KERN_ERR "o2cb: Cluster check failed. Fix errors "
+		       "before retrying.\n");
 		goto out;
 	}
 
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 56f61027236..4994f8b0e60 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -54,6 +54,7 @@
 #include "ocfs1_fs_compat.h"
 
 #include "alloc.h"
+#include "aops.h"
 #include "blockcheck.h"
 #include "dlmglue.h"
 #include "export.h"
@@ -1107,9 +1108,9 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 
 		ocfs2_set_ro_flag(osb, 1);
 
-		printk(KERN_NOTICE "Readonly device detected. No cluster "
-		       "services will be utilized for this mount. Recovery "
-		       "will be skipped.\n");
+		printk(KERN_NOTICE "ocfs2: Readonly device (%s) detected. "
+		       "Cluster services will not be used for this mount. "
+		       "Recovery will be skipped.\n", osb->dev_str);
 	}
 
 	if (!ocfs2_is_hard_readonly(osb)) {
@@ -1616,12 +1617,17 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt)
 	return 0;
 }
 
+wait_queue_head_t ocfs2__ioend_wq[OCFS2_IOEND_WQ_HASH_SZ];
+
 static int __init ocfs2_init(void)
 {
-	int status;
+	int status, i;
 
 	ocfs2_print_version();
 
+	for (i = 0; i < OCFS2_IOEND_WQ_HASH_SZ; i++)
+		init_waitqueue_head(&ocfs2__ioend_wq[i]);
+
 	status = init_ocfs2_uptodate_cache();
 	if (status < 0) {
 		mlog_errno(status);
@@ -1760,7 +1766,7 @@ static void ocfs2_inode_init_once(void *data)
 	ocfs2_extent_map_init(&oi->vfs_inode);
 	INIT_LIST_HEAD(&oi->ip_io_markers);
 	oi->ip_dir_start_lookup = 0;
-
+	atomic_set(&oi->ip_unaligned_aio, 0);
 	init_rwsem(&oi->ip_alloc_sem);
 	init_rwsem(&oi->ip_xattr_sem);
 	mutex_init(&oi->ip_io_mutex);
@@ -1974,7 +1980,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
 	 * If we failed before we got a uuid_str yet, we can't stop
 	 * heartbeat.  Otherwise, do it.
 	 */
-	if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str)
+	if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str &&
+	    !ocfs2_is_hard_readonly(osb))
 		hangup_needed = 1;
 
 	if (osb->cconn)
@@ -2353,7 +2360,7 @@ static int ocfs2_initialize_super(struct super_block *sb,
 		mlog_errno(status);
 		goto bail;
 	}
-	cleancache_init_shared_fs((char *)&uuid_net_key, sb);
+	cleancache_init_shared_fs((char *)&di->id2.i_super.s_uuid, sb);
 
 bail:
 	return status;
@@ -2462,8 +2469,8 @@ static int ocfs2_check_volume(struct ocfs2_super *osb)
 			goto finally;
 		}
 	} else {
-		mlog(ML_NOTICE, "File system was not unmounted cleanly, "
-		     "recovering volume.\n");
+		printk(KERN_NOTICE "ocfs2: File system on device (%s) was not "
+		       "unmounted cleanly, recovering it.\n", osb->dev_str);
 	}
 
 	local = ocfs2_mount_local(osb);
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index 194fb22ef79..aa9e8777b09 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -2376,16 +2376,18 @@ static int ocfs2_remove_value_outside(struct inode*inode,
 		}
 
 		ret = ocfs2_xattr_value_truncate(inode, vb, 0, &ctxt);
-		if (ret < 0) {
-			mlog_errno(ret);
-			break;
-		}
 
 		ocfs2_commit_trans(osb, ctxt.handle);
 		if (ctxt.meta_ac) {
 			ocfs2_free_alloc_context(ctxt.meta_ac);
 			ctxt.meta_ac = NULL;
 		}
+
+		if (ret < 0) {
+			mlog_errno(ret);
+			break;
+		}
+
 	}
 
 	if (ctxt.meta_ac)
diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c
index 586174168e2..80e4645f799 100644
--- a/fs/proc/meminfo.c
+++ b/fs/proc/meminfo.c
@@ -131,12 +131,13 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
 		K(i.freeswap),
 		K(global_page_state(NR_FILE_DIRTY)),
 		K(global_page_state(NR_WRITEBACK)),
-		K(global_page_state(NR_ANON_PAGES)
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
+		K(global_page_state(NR_ANON_PAGES)
 		  + global_page_state(NR_ANON_TRANSPARENT_HUGEPAGES) *
-		  HPAGE_PMD_NR
+		  HPAGE_PMD_NR),
+#else
+		K(global_page_state(NR_ANON_PAGES)),
 #endif
-		  ),
 		K(global_page_state(NR_FILE_MAPPED)),
 		K(global_page_state(NR_SHMEM)),
 		K(global_page_state(NR_SLAB_RECLAIMABLE) +
diff --git a/fs/proc/stat.c b/fs/proc/stat.c
index 8a6ab666e9f..2527a68057f 100644
--- a/fs/proc/stat.c
+++ b/fs/proc/stat.c
@@ -31,7 +31,7 @@ static u64 get_idle_time(int cpu)
 		idle = kcpustat_cpu(cpu).cpustat[CPUTIME_IDLE];
 		idle += arch_idle_time(cpu);
 	} else
-		idle = usecs_to_cputime(idle_time);
+		idle = nsecs_to_jiffies64(1000 * idle_time);
 
 	return idle;
 }
@@ -44,7 +44,7 @@ static u64 get_iowait_time(int cpu)
 		/* !NO_HZ so we can rely on cpustat.iowait */
 		iowait = kcpustat_cpu(cpu).cpustat[CPUTIME_IOWAIT];
 	else
-		iowait = usecs_to_cputime(iowait_time);
+		iowait = nsecs_to_jiffies64(1000 * iowait_time);
 
 	return iowait;
 }
diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c
index 2bd620f0d79..57bbf9078ac 100644
--- a/fs/pstore/platform.c
+++ b/fs/pstore/platform.c
@@ -167,6 +167,7 @@ int pstore_register(struct pstore_info *psi)
 	}
 
 	psinfo = psi;
+	mutex_init(&psinfo->read_mutex);
 	spin_unlock(&pstore_lock);
 
 	if (owner && !try_module_get(owner)) {
@@ -195,30 +196,32 @@ EXPORT_SYMBOL_GPL(pstore_register);
 void pstore_get_records(int quiet)
 {
 	struct pstore_info *psi = psinfo;
+	char			*buf = NULL;
 	ssize_t			size;
 	u64			id;
 	enum pstore_type_id	type;
 	struct timespec		time;
 	int			failed = 0, rc;
-	unsigned long		flags;
 
 	if (!psi)
 		return;
 
-	spin_lock_irqsave(&psinfo->buf_lock, flags);
+	mutex_lock(&psi->read_mutex);
 	rc = psi->open(psi);
 	if (rc)
 		goto out;
 
-	while ((size = psi->read(&id, &type, &time, psi)) > 0) {
-		rc = pstore_mkfile(type, psi->name, id, psi->buf, (size_t)size,
+	while ((size = psi->read(&id, &type, &time, &buf, psi)) > 0) {
+		rc = pstore_mkfile(type, psi->name, id, buf, (size_t)size,
 				  time, psi);
+		kfree(buf);
+		buf = NULL;
 		if (rc && (rc != -EEXIST || !quiet))
 			failed++;
 	}
 	psi->close(psi);
 out:
-	spin_unlock_irqrestore(&psinfo->buf_lock, flags);
+	mutex_unlock(&psi->read_mutex);
 
 	if (failed)
 		printk(KERN_WARNING "pstore: failed to load %d record(s) from '%s'\n",
diff --git a/fs/seq_file.c b/fs/seq_file.c
index 05d6b0e78c9..dba43c3ea3a 100644
--- a/fs/seq_file.c
+++ b/fs/seq_file.c
@@ -449,8 +449,6 @@ EXPORT_SYMBOL(seq_path);
 
 /*
  * Same as seq_path, but relative to supplied root.
- *
- * root may be changed, see __d_path().
  */
 int seq_path_root(struct seq_file *m, struct path *path, struct path *root,
 		  char *esc)
@@ -463,6 +461,8 @@ int seq_path_root(struct seq_file *m, struct path *path, struct path *root,
 		char *p;
 
 		p = __d_path(path, root, buf, size);
+		if (!p)
+			return SEQ_SKIP;
 		res = PTR_ERR(p);
 		if (!IS_ERR(p)) {
 			char *end = mangle_path(buf, p, esc);
@@ -474,7 +474,7 @@ int seq_path_root(struct seq_file *m, struct path *path, struct path *root,
 	}
 	seq_commit(m, res);
 
-	return res < 0 ? res : 0;
+	return res < 0 && res != -ENAMETOOLONG ? res : 0;
 }
 
 /*
diff --git a/fs/xfs/xfs_acl.c b/fs/xfs/xfs_acl.c
index b6c4b3795c4..76e4266d2e7 100644
--- a/fs/xfs/xfs_acl.c
+++ b/fs/xfs/xfs_acl.c
@@ -42,6 +42,8 @@ xfs_acl_from_disk(struct xfs_acl *aclp)
 	int count, i;
 
 	count = be32_to_cpu(aclp->acl_cnt);
+	if (count > XFS_ACL_MAX_ENTRIES)
+		return ERR_PTR(-EFSCORRUPTED);
 
 	acl = posix_acl_alloc(count, GFP_KERNEL);
 	if (!acl)
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index d4906e7c978..c1b55e59655 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -110,6 +110,7 @@ xfs_attr_namesp_match(int arg_flags, int ondisk_flags)
 /*
  * Query whether the requested number of additional bytes of extended
  * attribute space will be able to fit inline.
+ *
  * Returns zero if not, else the di_forkoff fork offset to be used in the
  * literal area for attribute data once the new bytes have been added.
  *
@@ -122,7 +123,7 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
 	int offset;
 	int minforkoff;	/* lower limit on valid forkoff locations */
 	int maxforkoff;	/* upper limit on valid forkoff locations */
-	int dsize;	
+	int dsize;
 	xfs_mount_t *mp = dp->i_mount;
 
 	offset = (XFS_LITINO(mp) - bytes) >> 3; /* rounded down */
@@ -136,47 +137,60 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
 		return (offset >= minforkoff) ? minforkoff : 0;
 	}
 
-	if (!(mp->m_flags & XFS_MOUNT_ATTR2)) {
-		if (bytes <= XFS_IFORK_ASIZE(dp))
-			return dp->i_d.di_forkoff;
+	/*
+	 * If the requested numbers of bytes is smaller or equal to the
+	 * current attribute fork size we can always proceed.
+	 *
+	 * Note that if_bytes in the data fork might actually be larger than
+	 * the current data fork size is due to delalloc extents. In that
+	 * case either the extent count will go down when they are converted
+	 * to real extents, or the delalloc conversion will take care of the
+	 * literal area rebalancing.
+	 */
+	if (bytes <= XFS_IFORK_ASIZE(dp))
+		return dp->i_d.di_forkoff;
+
+	/*
+	 * For attr2 we can try to move the forkoff if there is space in the
+	 * literal area, but for the old format we are done if there is no
+	 * space in the fixed attribute fork.
+	 */
+	if (!(mp->m_flags & XFS_MOUNT_ATTR2))
 		return 0;
-	}
 
 	dsize = dp->i_df.if_bytes;
-	
+
 	switch (dp->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
-		/* 
+		/*
 		 * If there is no attr fork and the data fork is extents, 
-		 * determine if creating the default attr fork will result 
-		 * in the extents form migrating to btree. If so, the 
-		 * minimum offset only needs to be the space required for 
+		 * determine if creating the default attr fork will result
+		 * in the extents form migrating to btree. If so, the
+		 * minimum offset only needs to be the space required for
 		 * the btree root.
-		 */ 
+		 */
 		if (!dp->i_d.di_forkoff && dp->i_df.if_bytes >
 		    xfs_default_attroffset(dp))
 			dsize = XFS_BMDR_SPACE_CALC(MINDBTPTRS);
 		break;
-		
 	case XFS_DINODE_FMT_BTREE:
 		/*
-		 * If have data btree then keep forkoff if we have one,
-		 * otherwise we are adding a new attr, so then we set 
-		 * minforkoff to where the btree root can finish so we have 
+		 * If we have a data btree then keep forkoff if we have one,
+		 * otherwise we are adding a new attr, so then we set
+		 * minforkoff to where the btree root can finish so we have
 		 * plenty of room for attrs
 		 */
 		if (dp->i_d.di_forkoff) {
-			if (offset < dp->i_d.di_forkoff) 
+			if (offset < dp->i_d.di_forkoff)
 				return 0;
-			else 
-				return dp->i_d.di_forkoff;
-		} else
-			dsize = XFS_BMAP_BROOT_SPACE(dp->i_df.if_broot);
+			return dp->i_d.di_forkoff;
+		}
+		dsize = XFS_BMAP_BROOT_SPACE(dp->i_df.if_broot);
 		break;
 	}
-	
-	/* 
-	 * A data fork btree root must have space for at least 
+
+	/*
+	 * A data fork btree root must have space for at least
 	 * MINDBTPTRS key/ptr pairs if the data fork is small or empty.
 	 */
 	minforkoff = MAX(dsize, XFS_BMDR_SPACE_CALC(MINDBTPTRS));
@@ -186,10 +200,10 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
 	maxforkoff = XFS_LITINO(mp) - XFS_BMDR_SPACE_CALC(MINABTPTRS);
 	maxforkoff = maxforkoff >> 3;	/* rounded down */
 
-	if (offset >= minforkoff && offset < maxforkoff)
-		return offset;
 	if (offset >= maxforkoff)
 		return maxforkoff;
+	if (offset >= minforkoff)
+		return offset;
 	return 0;
 }
 
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index c68baeb0974..d0ab7883705 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -2383,6 +2383,8 @@ xfs_bmap_btalloc(
 	int		tryagain;
 	int		error;
 
+	ASSERT(ap->length);
+
 	mp = ap->ip->i_mount;
 	align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
 	if (unlikely(align)) {
@@ -4629,6 +4631,8 @@ xfs_bmapi_allocate(
 	int			error;
 	int			rt;
 
+	ASSERT(bma->length > 0);
+
 	rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(bma->ip);
 
 	/*
@@ -4849,6 +4853,7 @@ xfs_bmapi_write(
 	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
 	ASSERT(!(flags & XFS_BMAPI_IGSTATE));
 	ASSERT(tp != NULL);
+	ASSERT(len > 0);
 
 	whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
 		XFS_ATTR_FORK : XFS_DATA_FORK;
@@ -4918,9 +4923,22 @@ xfs_bmapi_write(
 			bma.eof = eof;
 			bma.conv = !!(flags & XFS_BMAPI_CONVERT);
 			bma.wasdel = wasdelay;
-			bma.length = len;
 			bma.offset = bno;
 
+			/*
+			 * There's a 32/64 bit type mismatch between the
+			 * allocation length request (which can be 64 bits in
+			 * length) and the bma length request, which is
+			 * xfs_extlen_t and therefore 32 bits. Hence we have to
+			 * check for 32-bit overflows and handle them here.
+			 */
+			if (len > (xfs_filblks_t)MAXEXTLEN)
+				bma.length = MAXEXTLEN;
+			else
+				bma.length = len;
+
+			ASSERT(len > 0);
+			ASSERT(bma.length > 0);
 			error = xfs_bmapi_allocate(&bma, flags);
 			if (error)
 				goto error0;
diff --git a/fs/xfs/xfs_export.c b/fs/xfs/xfs_export.c
index da108977b21..558910f5e3c 100644
--- a/fs/xfs/xfs_export.c
+++ b/fs/xfs/xfs_export.c
@@ -98,22 +98,22 @@ xfs_fs_encode_fh(
 	switch (fileid_type) {
 	case FILEID_INO32_GEN_PARENT:
 		spin_lock(&dentry->d_lock);
-		fid->i32.parent_ino = dentry->d_parent->d_inode->i_ino;
+		fid->i32.parent_ino = XFS_I(dentry->d_parent->d_inode)->i_ino;
 		fid->i32.parent_gen = dentry->d_parent->d_inode->i_generation;
 		spin_unlock(&dentry->d_lock);
 		/*FALLTHRU*/
 	case FILEID_INO32_GEN:
-		fid->i32.ino = inode->i_ino;
+		fid->i32.ino = XFS_I(inode)->i_ino;
 		fid->i32.gen = inode->i_generation;
 		break;
 	case FILEID_INO32_GEN_PARENT | XFS_FILEID_TYPE_64FLAG:
 		spin_lock(&dentry->d_lock);
-		fid64->parent_ino = dentry->d_parent->d_inode->i_ino;
+		fid64->parent_ino = XFS_I(dentry->d_parent->d_inode)->i_ino;
 		fid64->parent_gen = dentry->d_parent->d_inode->i_generation;
 		spin_unlock(&dentry->d_lock);
 		/*FALLTHRU*/
 	case FILEID_INO32_GEN | XFS_FILEID_TYPE_64FLAG:
-		fid64->ino = inode->i_ino;
+		fid64->ino = XFS_I(inode)->i_ino;
 		fid64->gen = inode->i_generation;
 		break;
 	}
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index c0237c602f1..755ee816488 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2835,6 +2835,27 @@ corrupt_out:
 	return XFS_ERROR(EFSCORRUPTED);
 }
 
+void
+xfs_promote_inode(
+	struct xfs_inode	*ip)
+{
+	struct xfs_buf		*bp;
+
+	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
+
+	bp = xfs_incore(ip->i_mount->m_ddev_targp, ip->i_imap.im_blkno,
+			ip->i_imap.im_len, XBF_TRYLOCK);
+	if (!bp)
+		return;
+
+	if (XFS_BUF_ISDELAYWRITE(bp)) {
+		xfs_buf_delwri_promote(bp);
+		wake_up_process(ip->i_mount->m_ddev_targp->bt_task);
+	}
+
+	xfs_buf_relse(bp);
+}
+
 /*
  * Return a pointer to the extent record at file index idx.
  */
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 760140d1dd6..b4cd4739f98 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -498,6 +498,7 @@ int		xfs_iunlink(struct xfs_trans *, xfs_inode_t *);
 void		xfs_iext_realloc(xfs_inode_t *, int, int);
 void		xfs_iunpin_wait(xfs_inode_t *);
 int		xfs_iflush(xfs_inode_t *, uint);
+void		xfs_promote_inode(struct xfs_inode *);
 void		xfs_lock_inodes(xfs_inode_t **, int, uint);
 void		xfs_lock_two_inodes(xfs_inode_t *, xfs_inode_t *, uint);
 
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index a14cd89fe46..34817adf4b9 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -150,6 +150,117 @@ xlog_grant_add_space(
 	} while (head_val != old);
 }
 
+STATIC bool
+xlog_reserveq_wake(
+	struct log		*log,
+	int			*free_bytes)
+{
+	struct xlog_ticket	*tic;
+	int			need_bytes;
+
+	list_for_each_entry(tic, &log->l_reserveq, t_queue) {
+		if (tic->t_flags & XLOG_TIC_PERM_RESERV)
+			need_bytes = tic->t_unit_res * tic->t_cnt;
+		else
+			need_bytes = tic->t_unit_res;
+
+		if (*free_bytes < need_bytes)
+			return false;
+		*free_bytes -= need_bytes;
+
+		trace_xfs_log_grant_wake_up(log, tic);
+		wake_up(&tic->t_wait);
+	}
+
+	return true;
+}
+
+STATIC bool
+xlog_writeq_wake(
+	struct log		*log,
+	int			*free_bytes)
+{
+	struct xlog_ticket	*tic;
+	int			need_bytes;
+
+	list_for_each_entry(tic, &log->l_writeq, t_queue) {
+		ASSERT(tic->t_flags & XLOG_TIC_PERM_RESERV);
+
+		need_bytes = tic->t_unit_res;
+
+		if (*free_bytes < need_bytes)
+			return false;
+		*free_bytes -= need_bytes;
+
+		trace_xfs_log_regrant_write_wake_up(log, tic);
+		wake_up(&tic->t_wait);
+	}
+
+	return true;
+}
+
+STATIC int
+xlog_reserveq_wait(
+	struct log		*log,
+	struct xlog_ticket	*tic,
+	int			need_bytes)
+{
+	list_add_tail(&tic->t_queue, &log->l_reserveq);
+
+	do {
+		if (XLOG_FORCED_SHUTDOWN(log))
+			goto shutdown;
+		xlog_grant_push_ail(log, need_bytes);
+
+		XFS_STATS_INC(xs_sleep_logspace);
+		trace_xfs_log_grant_sleep(log, tic);
+
+		xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
+		trace_xfs_log_grant_wake(log, tic);
+
+		spin_lock(&log->l_grant_reserve_lock);
+		if (XLOG_FORCED_SHUTDOWN(log))
+			goto shutdown;
+	} while (xlog_space_left(log, &log->l_grant_reserve_head) < need_bytes);
+
+	list_del_init(&tic->t_queue);
+	return 0;
+shutdown:
+	list_del_init(&tic->t_queue);
+	return XFS_ERROR(EIO);
+}
+
+STATIC int
+xlog_writeq_wait(
+	struct log		*log,
+	struct xlog_ticket	*tic,
+	int			need_bytes)
+{
+	list_add_tail(&tic->t_queue, &log->l_writeq);
+
+	do {
+		if (XLOG_FORCED_SHUTDOWN(log))
+			goto shutdown;
+		xlog_grant_push_ail(log, need_bytes);
+
+		XFS_STATS_INC(xs_sleep_logspace);
+		trace_xfs_log_regrant_write_sleep(log, tic);
+
+		xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
+		trace_xfs_log_regrant_write_wake(log, tic);
+
+		spin_lock(&log->l_grant_write_lock);
+		if (XLOG_FORCED_SHUTDOWN(log))
+			goto shutdown;
+	} while (xlog_space_left(log, &log->l_grant_write_head) < need_bytes);
+
+	list_del_init(&tic->t_queue);
+	return 0;
+shutdown:
+	list_del_init(&tic->t_queue);
+	return XFS_ERROR(EIO);
+}
+
 static void
 xlog_tic_reset_res(xlog_ticket_t *tic)
 {
@@ -350,8 +461,19 @@ xfs_log_reserve(
 		retval = xlog_grant_log_space(log, internal_ticket);
 	}
 
+	if (unlikely(retval)) {
+		/*
+		 * If we are failing, make sure the ticket doesn't have any
+		 * current reservations.  We don't want to add this back
+		 * when the ticket/ transaction gets cancelled.
+		 */
+		internal_ticket->t_curr_res = 0;
+		/* ungrant will give back unit_res * t_cnt. */
+		internal_ticket->t_cnt = 0;
+	}
+
 	return retval;
-}	/* xfs_log_reserve */
+}
 
 
 /*
@@ -2481,8 +2603,8 @@ restart:
 /*
  * Atomically get the log space required for a log ticket.
  *
- * Once a ticket gets put onto the reserveq, it will only return after
- * the needed reservation is satisfied.
+ * Once a ticket gets put onto the reserveq, it will only return after the
+ * needed reservation is satisfied.
  *
  * This function is structured so that it has a lock free fast path. This is
  * necessary because every new transaction reservation will come through this
@@ -2490,113 +2612,53 @@ restart:
  * every pass.
  *
  * As tickets are only ever moved on and off the reserveq under the
- * l_grant_reserve_lock, we only need to take that lock if we are going
- * to add the ticket to the queue and sleep. We can avoid taking the lock if the
- * ticket was never added to the reserveq because the t_queue list head will be
- * empty and we hold the only reference to it so it can safely be checked
- * unlocked.
+ * l_grant_reserve_lock, we only need to take that lock if we are going to add
+ * the ticket to the queue and sleep. We can avoid taking the lock if the ticket
+ * was never added to the reserveq because the t_queue list head will be empty
+ * and we hold the only reference to it so it can safely be checked unlocked.
  */
 STATIC int
-xlog_grant_log_space(xlog_t	   *log,
-		     xlog_ticket_t *tic)
+xlog_grant_log_space(
+	struct log		*log,
+	struct xlog_ticket	*tic)
 {
-	int		 free_bytes;
-	int		 need_bytes;
+	int			free_bytes, need_bytes;
+	int			error = 0;
 
-#ifdef DEBUG
-	if (log->l_flags & XLOG_ACTIVE_RECOVERY)
-		panic("grant Recovery problem");
-#endif
+	ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
 
 	trace_xfs_log_grant_enter(log, tic);
 
+	/*
+	 * If there are other waiters on the queue then give them a chance at
+	 * logspace before us.  Wake up the first waiters, if we do not wake
+	 * up all the waiters then go to sleep waiting for more free space,
+	 * otherwise try to get some space for this transaction.
+	 */
 	need_bytes = tic->t_unit_res;
 	if (tic->t_flags & XFS_LOG_PERM_RESERV)
 		need_bytes *= tic->t_ocnt;
-
-	/* something is already sleeping; insert new transaction at end */
-	if (!list_empty_careful(&log->l_reserveq)) {
-		spin_lock(&log->l_grant_reserve_lock);
-		/* recheck the queue now we are locked */
-		if (list_empty(&log->l_reserveq)) {
-			spin_unlock(&log->l_grant_reserve_lock);
-			goto redo;
-		}
-		list_add_tail(&tic->t_queue, &log->l_reserveq);
-
-		trace_xfs_log_grant_sleep1(log, tic);
-
-		/*
-		 * Gotta check this before going to sleep, while we're
-		 * holding the grant lock.
-		 */
-		if (XLOG_FORCED_SHUTDOWN(log))
-			goto error_return;
-
-		XFS_STATS_INC(xs_sleep_logspace);
-		xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
-
-		/*
-		 * If we got an error, and the filesystem is shutting down,
-		 * we'll catch it down below. So just continue...
-		 */
-		trace_xfs_log_grant_wake1(log, tic);
-	}
-
-redo:
-	if (XLOG_FORCED_SHUTDOWN(log))
-		goto error_return_unlocked;
-
 	free_bytes = xlog_space_left(log, &log->l_grant_reserve_head);
-	if (free_bytes < need_bytes) {
+	if (!list_empty_careful(&log->l_reserveq)) {
 		spin_lock(&log->l_grant_reserve_lock);
-		if (list_empty(&tic->t_queue))
-			list_add_tail(&tic->t_queue, &log->l_reserveq);
-
-		trace_xfs_log_grant_sleep2(log, tic);
-
-		if (XLOG_FORCED_SHUTDOWN(log))
-			goto error_return;
-
-		xlog_grant_push_ail(log, need_bytes);
-
-		XFS_STATS_INC(xs_sleep_logspace);
-		xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
-
-		trace_xfs_log_grant_wake2(log, tic);
-		goto redo;
-	}
-
-	if (!list_empty(&tic->t_queue)) {
+		if (!xlog_reserveq_wake(log, &free_bytes) ||
+		    free_bytes < need_bytes)
+			error = xlog_reserveq_wait(log, tic, need_bytes);
+		spin_unlock(&log->l_grant_reserve_lock);
+	} else if (free_bytes < need_bytes) {
 		spin_lock(&log->l_grant_reserve_lock);
-		list_del_init(&tic->t_queue);
+		error = xlog_reserveq_wait(log, tic, need_bytes);
 		spin_unlock(&log->l_grant_reserve_lock);
 	}
+	if (error)
+		return error;
 
-	/* we've got enough space */
 	xlog_grant_add_space(log, &log->l_grant_reserve_head, need_bytes);
 	xlog_grant_add_space(log, &log->l_grant_write_head, need_bytes);
 	trace_xfs_log_grant_exit(log, tic);
 	xlog_verify_grant_tail(log);
 	return 0;
-
-error_return_unlocked:
-	spin_lock(&log->l_grant_reserve_lock);
-error_return:
-	list_del_init(&tic->t_queue);
-	spin_unlock(&log->l_grant_reserve_lock);
-	trace_xfs_log_grant_error(log, tic);
-
-	/*
-	 * If we are failing, make sure the ticket doesn't have any
-	 * current reservations. We don't want to add this back when
-	 * the ticket/transaction gets cancelled.
-	 */
-	tic->t_curr_res = 0;
-	tic->t_cnt = 0; /* ungrant will give back unit_res * t_cnt. */
-	return XFS_ERROR(EIO);
-}	/* xlog_grant_log_space */
-
+}
 
 /*
  * Replenish the byte reservation required by moving the grant write head.
@@ -2605,10 +2667,12 @@ error_return:
  * free fast path.
  */
 STATIC int
-xlog_regrant_write_log_space(xlog_t	   *log,
-			     xlog_ticket_t *tic)
+xlog_regrant_write_log_space(
+	struct log		*log,
+	struct xlog_ticket	*tic)
 {
-	int		free_bytes, need_bytes;
+	int			free_bytes, need_bytes;
+	int			error = 0;
 
 	tic->t_curr_res = tic->t_unit_res;
 	xlog_tic_reset_res(tic);
@@ -2616,104 +2680,38 @@ xlog_regrant_write_log_space(xlog_t	   *log,
 	if (tic->t_cnt > 0)
 		return 0;
 
-#ifdef DEBUG
-	if (log->l_flags & XLOG_ACTIVE_RECOVERY)
-		panic("regrant Recovery problem");
-#endif
+	ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
 
 	trace_xfs_log_regrant_write_enter(log, tic);
-	if (XLOG_FORCED_SHUTDOWN(log))
-		goto error_return_unlocked;
 
-	/* If there are other waiters on the queue then give them a
-	 * chance at logspace before us. Wake up the first waiters,
-	 * if we do not wake up all the waiters then go to sleep waiting
-	 * for more free space, otherwise try to get some space for
-	 * this transaction.
+	/*
+	 * If there are other waiters on the queue then give them a chance at
+	 * logspace before us.  Wake up the first waiters, if we do not wake
+	 * up all the waiters then go to sleep waiting for more free space,
+	 * otherwise try to get some space for this transaction.
 	 */
 	need_bytes = tic->t_unit_res;
-	if (!list_empty_careful(&log->l_writeq)) {
-		struct xlog_ticket *ntic;
-
-		spin_lock(&log->l_grant_write_lock);
-		free_bytes = xlog_space_left(log, &log->l_grant_write_head);
-		list_for_each_entry(ntic, &log->l_writeq, t_queue) {
-			ASSERT(ntic->t_flags & XLOG_TIC_PERM_RESERV);
-
-			if (free_bytes < ntic->t_unit_res)
-				break;
-			free_bytes -= ntic->t_unit_res;
-			wake_up(&ntic->t_wait);
-		}
-
-		if (ntic != list_first_entry(&log->l_writeq,
-						struct xlog_ticket, t_queue)) {
-			if (list_empty(&tic->t_queue))
-				list_add_tail(&tic->t_queue, &log->l_writeq);
-			trace_xfs_log_regrant_write_sleep1(log, tic);
-
-			xlog_grant_push_ail(log, need_bytes);
-
-			XFS_STATS_INC(xs_sleep_logspace);
-			xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
-			trace_xfs_log_regrant_write_wake1(log, tic);
-		} else
-			spin_unlock(&log->l_grant_write_lock);
-	}
-
-redo:
-	if (XLOG_FORCED_SHUTDOWN(log))
-		goto error_return_unlocked;
-
 	free_bytes = xlog_space_left(log, &log->l_grant_write_head);
-	if (free_bytes < need_bytes) {
+	if (!list_empty_careful(&log->l_writeq)) {
 		spin_lock(&log->l_grant_write_lock);
-		if (list_empty(&tic->t_queue))
-			list_add_tail(&tic->t_queue, &log->l_writeq);
-
-		if (XLOG_FORCED_SHUTDOWN(log))
-			goto error_return;
-
-		xlog_grant_push_ail(log, need_bytes);
-
-		XFS_STATS_INC(xs_sleep_logspace);
-		trace_xfs_log_regrant_write_sleep2(log, tic);
-		xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
-
-		trace_xfs_log_regrant_write_wake2(log, tic);
-		goto redo;
-	}
-
-	if (!list_empty(&tic->t_queue)) {
+		if (!xlog_writeq_wake(log, &free_bytes) ||
+		    free_bytes < need_bytes)
+			error = xlog_writeq_wait(log, tic, need_bytes);
+		spin_unlock(&log->l_grant_write_lock);
+	} else if (free_bytes < need_bytes) {
 		spin_lock(&log->l_grant_write_lock);
-		list_del_init(&tic->t_queue);
+		error = xlog_writeq_wait(log, tic, need_bytes);
 		spin_unlock(&log->l_grant_write_lock);
 	}
 
-	/* we've got enough space */
+	if (error)
+		return error;
+
 	xlog_grant_add_space(log, &log->l_grant_write_head, need_bytes);
 	trace_xfs_log_regrant_write_exit(log, tic);
 	xlog_verify_grant_tail(log);
 	return 0;
-
-
- error_return_unlocked:
-	spin_lock(&log->l_grant_write_lock);
- error_return:
-	list_del_init(&tic->t_queue);
-	spin_unlock(&log->l_grant_write_lock);
-	trace_xfs_log_regrant_write_error(log, tic);
-
-	/*
-	 * If we are failing, make sure the ticket doesn't have any
-	 * current reservations. We don't want to add this back when
-	 * the ticket/transaction gets cancelled.
-	 */
-	tic->t_curr_res = 0;
-	tic->t_cnt = 0; /* ungrant will give back unit_res * t_cnt. */
-	return XFS_ERROR(EIO);
-}	/* xlog_regrant_write_log_space */
-
+}
 
 /* The first cnt-1 times through here we don't need to
  * move the grant write head because the permanent
diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c
index 5cff443f6cd..0bbb1a41998 100644
--- a/fs/xfs/xfs_qm.c
+++ b/fs/xfs/xfs_qm.c
@@ -674,7 +674,8 @@ xfs_qm_dqattach_one(
 	 * disk and we didn't ask it to allocate;
 	 * ESRCH if quotas got turned off suddenly.
 	 */
-	error = xfs_qm_dqget(ip->i_mount, ip, id, type, XFS_QMOPT_DOWARN, &dqp);
+	error = xfs_qm_dqget(ip->i_mount, ip, id, type,
+			     doalloc | XFS_QMOPT_DOWARN, &dqp);
 	if (error)
 		return error;
 
diff --git a/fs/xfs/xfs_sync.c b/fs/xfs/xfs_sync.c
index aa3dc1a4d53..be5c51d8f75 100644
--- a/fs/xfs/xfs_sync.c
+++ b/fs/xfs/xfs_sync.c
@@ -770,6 +770,17 @@ restart:
 	if (!xfs_iflock_nowait(ip)) {
 		if (!(sync_mode & SYNC_WAIT))
 			goto out;
+
+		/*
+		 * If we only have a single dirty inode in a cluster there is
+		 * a fair chance that the AIL push may have pushed it into
+		 * the buffer, but xfsbufd won't touch it until 30 seconds
+		 * from now, and thus we will lock up here.
+		 *
+		 * Promote the inode buffer to the front of the delwri list
+		 * and wake up xfsbufd now.
+		 */
+		xfs_promote_inode(ip);
 		xfs_iflock(ip);
 	}
 
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index f1d2802b2f0..49403579887 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -834,18 +834,14 @@ DEFINE_LOGGRANT_EVENT(xfs_log_umount_write);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_enter);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_exit);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_error);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep1);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake1);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep2);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake2);
+DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep);
+DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake_up);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_enter);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_exit);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_error);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep1);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake1);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep2);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake2);
+DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep);
+DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake_up);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_enter);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_exit);