Merge branch 'master' into pm-sleep

* master: (848 commits)
  SELinux: Fix RCU deref check warning in sel_netport_insert()
  binary_sysctl(): fix memory leak
  mm/vmalloc.c: remove static declaration of va from __get_vm_area_node
  ipmi_watchdog: restore settings when BMC reset
  oom: fix integer overflow of points in oom_badness
  memcg: keep root group unchanged if creation fails
  nilfs2: potential integer overflow in nilfs_ioctl_clean_segments()
  nilfs2: unbreak compat ioctl
  cpusets: stall when updating mems_allowed for mempolicy or disjoint nodemask
  evm: prevent racing during tfm allocation
  evm: key must be set once during initialization
  mmc: vub300: fix type of firmware_rom_wait_states module parameter
  Revert "mmc: enable runtime PM by default"
  mmc: sdhci: remove "state" argument from sdhci_suspend_host
  x86, dumpstack: Fix code bytes breakage due to missing KERN_CONT
  IB/qib: Correct sense on freectxts increment and decrement
  RDMA/cma: Verify private data length
  cgroups: fix a css_set not found bug in cgroup_attach_proc
  oprofile: Fix uninitialized memory access when writing to writing to oprofilefs
  Revert "xen/pv-on-hvm kexec: add xs_reset_watches to shutdown watches from old kernel"
  ...

Conflicts:
	kernel/cgroup_freezer.c

98 files changed, 2239 insertions, 1396 deletions

diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index 98ab240072e..704a2ba08ea 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -64,6 +64,8 @@ struct btrfs_worker_thread {
 	int idle;
 };
 
+static int __btrfs_start_workers(struct btrfs_workers *workers);
+
 /*
  * btrfs_start_workers uses kthread_run, which can block waiting for memory
  * for a very long time.  It will actually throttle on page writeback,
@@ -88,27 +90,10 @@ static void start_new_worker_func(struct btrfs_work *work)
 {
 	struct worker_start *start;
 	start = container_of(work, struct worker_start, work);
-	btrfs_start_workers(start->queue, 1);
+	__btrfs_start_workers(start->queue);
 	kfree(start);
 }
 
-static int start_new_worker(struct btrfs_workers *queue)
-{
-	struct worker_start *start;
-	int ret;
-
-	start = kzalloc(sizeof(*start), GFP_NOFS);
-	if (!start)
-		return -ENOMEM;
-
-	start->work.func = start_new_worker_func;
-	start->queue = queue;
-	ret = btrfs_queue_worker(queue->atomic_worker_start, &start->work);
-	if (ret)
-		kfree(start);
-	return ret;
-}
-
 /*
  * helper function to move a thread onto the idle list after it
  * has finished some requests.
@@ -153,12 +138,20 @@ static void check_busy_worker(struct btrfs_worker_thread *worker)
 static void check_pending_worker_creates(struct btrfs_worker_thread *worker)
 {
 	struct btrfs_workers *workers = worker->workers;
+	struct worker_start *start;
 	unsigned long flags;
 
 	rmb();
 	if (!workers->atomic_start_pending)
 		return;
 
+	start = kzalloc(sizeof(*start), GFP_NOFS);
+	if (!start)
+		return;
+
+	start->work.func = start_new_worker_func;
+	start->queue = workers;
+
 	spin_lock_irqsave(&workers->lock, flags);
 	if (!workers->atomic_start_pending)
 		goto out;
@@ -170,10 +163,11 @@ static void check_pending_worker_creates(struct btrfs_worker_thread *worker)
 
 	workers->num_workers_starting += 1;
 	spin_unlock_irqrestore(&workers->lock, flags);
-	start_new_worker(workers);
+	btrfs_queue_worker(workers->atomic_worker_start, &start->work);
 	return;
 
 out:
+	kfree(start);
 	spin_unlock_irqrestore(&workers->lock, flags);
 }
 
@@ -331,7 +325,7 @@ again:
 			run_ordered_completions(worker->workers, work);
 
 			check_pending_worker_creates(worker);
-
+			cond_resched();
 		}
 
 		spin_lock_irq(&worker->lock);
@@ -462,56 +456,55 @@ void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
  * starts new worker threads.  This does not enforce the max worker
  * count in case you need to temporarily go past it.
  */
-static int __btrfs_start_workers(struct btrfs_workers *workers,
-				 int num_workers)
+static int __btrfs_start_workers(struct btrfs_workers *workers)
 {
 	struct btrfs_worker_thread *worker;
 	int ret = 0;
-	int i;
 
-	for (i = 0; i < num_workers; i++) {
-		worker = kzalloc(sizeof(*worker), GFP_NOFS);
-		if (!worker) {
-			ret = -ENOMEM;
-			goto fail;
-		}
+	worker = kzalloc(sizeof(*worker), GFP_NOFS);
+	if (!worker) {
+		ret = -ENOMEM;
+		goto fail;
+	}
 
-		INIT_LIST_HEAD(&worker->pending);
-		INIT_LIST_HEAD(&worker->prio_pending);
-		INIT_LIST_HEAD(&worker->worker_list);
-		spin_lock_init(&worker->lock);
-
-		atomic_set(&worker->num_pending, 0);
-		atomic_set(&worker->refs, 1);
-		worker->workers = workers;
-		worker->task = kthread_run(worker_loop, worker,
-					   "btrfs-%s-%d", workers->name,
-					   workers->num_workers + i);
-		if (IS_ERR(worker->task)) {
-			ret = PTR_ERR(worker->task);
-			kfree(worker);
-			goto fail;
-		}
-		spin_lock_irq(&workers->lock);
-		list_add_tail(&worker->worker_list, &workers->idle_list);
-		worker->idle = 1;
-		workers->num_workers++;
-		workers->num_workers_starting--;
-		WARN_ON(workers->num_workers_starting < 0);
-		spin_unlock_irq(&workers->lock);
+	INIT_LIST_HEAD(&worker->pending);
+	INIT_LIST_HEAD(&worker->prio_pending);
+	INIT_LIST_HEAD(&worker->worker_list);
+	spin_lock_init(&worker->lock);
+
+	atomic_set(&worker->num_pending, 0);
+	atomic_set(&worker->refs, 1);
+	worker->workers = workers;
+	worker->task = kthread_run(worker_loop, worker,
+				   "btrfs-%s-%d", workers->name,
+				   workers->num_workers + 1);
+	if (IS_ERR(worker->task)) {
+		ret = PTR_ERR(worker->task);
+		kfree(worker);
+		goto fail;
 	}
+	spin_lock_irq(&workers->lock);
+	list_add_tail(&worker->worker_list, &workers->idle_list);
+	worker->idle = 1;
+	workers->num_workers++;
+	workers->num_workers_starting--;
+	WARN_ON(workers->num_workers_starting < 0);
+	spin_unlock_irq(&workers->lock);
+
 	return 0;
 fail:
-	btrfs_stop_workers(workers);
+	spin_lock_irq(&workers->lock);
+	workers->num_workers_starting--;
+	spin_unlock_irq(&workers->lock);
 	return ret;
 }
 
-int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
+int btrfs_start_workers(struct btrfs_workers *workers)
 {
 	spin_lock_irq(&workers->lock);
-	workers->num_workers_starting += num_workers;
+	workers->num_workers_starting++;
 	spin_unlock_irq(&workers->lock);
-	return __btrfs_start_workers(workers, num_workers);
+	return __btrfs_start_workers(workers);
 }
 
 /*
@@ -568,6 +561,7 @@ static struct btrfs_worker_thread *find_worker(struct btrfs_workers *workers)
 	struct btrfs_worker_thread *worker;
 	unsigned long flags;
 	struct list_head *fallback;
+	int ret;
 
 again:
 	spin_lock_irqsave(&workers->lock, flags);
@@ -584,7 +578,9 @@ again:
 			workers->num_workers_starting++;
 			spin_unlock_irqrestore(&workers->lock, flags);
 			/* we're below the limit, start another worker */
-			__btrfs_start_workers(workers, 1);
+			ret = __btrfs_start_workers(workers);
+			if (ret)
+				goto fallback;
 			goto again;
 		}
 	}
@@ -665,7 +661,7 @@ void btrfs_set_work_high_prio(struct btrfs_work *work)
 /*
  * places a struct btrfs_work into the pending queue of one of the kthreads
  */
-int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
+void btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
 {
 	struct btrfs_worker_thread *worker;
 	unsigned long flags;
@@ -673,7 +669,7 @@ int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
 
 	/* don't requeue something already on a list */
 	if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags))
-		goto out;
+		return;
 
 	worker = find_worker(workers);
 	if (workers->ordered) {
@@ -712,7 +708,4 @@ int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
 	if (wake)
 		wake_up_process(worker->task);
 	spin_unlock_irqrestore(&worker->lock, flags);
-
-out:
-	return 0;
 }
diff --git a/fs/btrfs/async-thread.h b/fs/btrfs/async-thread.h
index 5077746cf85..f34cc31fa3c 100644
--- a/fs/btrfs/async-thread.h
+++ b/fs/btrfs/async-thread.h
@@ -109,8 +109,8 @@ struct btrfs_workers {
 	char *name;
 };
 
-int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work);
-int btrfs_start_workers(struct btrfs_workers *workers, int num_workers);
+void btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work);
+int btrfs_start_workers(struct btrfs_workers *workers);
 int btrfs_stop_workers(struct btrfs_workers *workers);
 void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
 			struct btrfs_workers *async_starter);
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 04a5dfcee5a..67385033323 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2369,6 +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);
@@ -2689,7 +2692,8 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf);
 int btrfs_readpage(struct file *file, struct page *page);
 void btrfs_evict_inode(struct inode *inode);
 int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc);
-void btrfs_dirty_inode(struct inode *inode, int flags);
+int btrfs_dirty_inode(struct inode *inode);
+int btrfs_update_time(struct file *file);
 struct inode *btrfs_alloc_inode(struct super_block *sb);
 void btrfs_destroy_inode(struct inode *inode);
 int btrfs_drop_inode(struct inode *inode);
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index 5b163572e0c..9c1eccc2c50 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -640,8 +640,8 @@ static int btrfs_delayed_inode_reserve_metadata(
 	 * Now if src_rsv == delalloc_block_rsv we'll let it just steal since
 	 * we're accounted for.
 	 */
-	if (!trans->bytes_reserved &&
-	    src_rsv != &root->fs_info->delalloc_block_rsv) {
+	if (!src_rsv || (!trans->bytes_reserved &&
+	    src_rsv != &root->fs_info->delalloc_block_rsv)) {
 		ret = btrfs_block_rsv_add_noflush(root, dst_rsv, num_bytes);
 		/*
 		 * Since we're under a transaction reserve_metadata_bytes could
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index b0917590152..f99a099a774 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2190,19 +2190,27 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->endio_meta_write_workers.idle_thresh = 2;
 	fs_info->readahead_workers.idle_thresh = 2;
 
-	btrfs_start_workers(&fs_info->workers, 1);
-	btrfs_start_workers(&fs_info->generic_worker, 1);
-	btrfs_start_workers(&fs_info->submit_workers, 1);
-	btrfs_start_workers(&fs_info->delalloc_workers, 1);
-	btrfs_start_workers(&fs_info->fixup_workers, 1);
-	btrfs_start_workers(&fs_info->endio_workers, 1);
-	btrfs_start_workers(&fs_info->endio_meta_workers, 1);
-	btrfs_start_workers(&fs_info->endio_meta_write_workers, 1);
-	btrfs_start_workers(&fs_info->endio_write_workers, 1);
-	btrfs_start_workers(&fs_info->endio_freespace_worker, 1);
-	btrfs_start_workers(&fs_info->delayed_workers, 1);
-	btrfs_start_workers(&fs_info->caching_workers, 1);
-	btrfs_start_workers(&fs_info->readahead_workers, 1);
+	/*
+	 * btrfs_start_workers can really only fail because of ENOMEM so just
+	 * return -ENOMEM if any of these fail.
+	 */
+	ret = btrfs_start_workers(&fs_info->workers);
+	ret |= btrfs_start_workers(&fs_info->generic_worker);
+	ret |= btrfs_start_workers(&fs_info->submit_workers);
+	ret |= btrfs_start_workers(&fs_info->delalloc_workers);
+	ret |= btrfs_start_workers(&fs_info->fixup_workers);
+	ret |= btrfs_start_workers(&fs_info->endio_workers);
+	ret |= btrfs_start_workers(&fs_info->endio_meta_workers);
+	ret |= btrfs_start_workers(&fs_info->endio_meta_write_workers);
+	ret |= btrfs_start_workers(&fs_info->endio_write_workers);
+	ret |= btrfs_start_workers(&fs_info->endio_freespace_worker);
+	ret |= btrfs_start_workers(&fs_info->delayed_workers);
+	ret |= btrfs_start_workers(&fs_info->caching_workers);
+	ret |= btrfs_start_workers(&fs_info->readahead_workers);
+	if (ret) {
+		ret = -ENOMEM;
+		goto fail_sb_buffer;
+	}
 
 	fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
 	fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 930ae894973..f5fbe576d2b 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -2822,7 +2822,7 @@ out_free:
 	btrfs_release_path(path);
 out:
 	spin_lock(&block_group->lock);
-	if (!ret)
+	if (!ret && dcs == BTRFS_DC_SETUP)
 		block_group->cache_generation = trans->transid;
 	block_group->disk_cache_state = dcs;
 	spin_unlock(&block_group->lock);
@@ -3888,9 +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;
@@ -3909,7 +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;
@@ -3918,6 +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)
@@ -4190,12 +4204,17 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_block_rsv *block_rsv = &root->fs_info->delalloc_block_rsv;
 	u64 to_reserve = 0;
+	u64 csum_bytes;
 	unsigned nr_extents = 0;
+	int extra_reserve = 0;
 	int flush = 1;
 	int ret;
 
+	/* Need to be holding the i_mutex here if we aren't free space cache */
 	if (btrfs_is_free_space_inode(root, inode))
 		flush = 0;
+	else
+		WARN_ON(!mutex_is_locked(&inode->i_mutex));
 
 	if (flush && btrfs_transaction_in_commit(root->fs_info))
 		schedule_timeout(1);
@@ -4206,11 +4225,9 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	BTRFS_I(inode)->outstanding_extents++;
 
 	if (BTRFS_I(inode)->outstanding_extents >
-	    BTRFS_I(inode)->reserved_extents) {
+	    BTRFS_I(inode)->reserved_extents)
 		nr_extents = BTRFS_I(inode)->outstanding_extents -
 			BTRFS_I(inode)->reserved_extents;
-		BTRFS_I(inode)->reserved_extents += nr_extents;
-	}
 
 	/*
 	 * Add an item to reserve for updating the inode when we complete the
@@ -4218,11 +4235,12 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	 */
 	if (!BTRFS_I(inode)->delalloc_meta_reserved) {
 		nr_extents++;
-		BTRFS_I(inode)->delalloc_meta_reserved = 1;
+		extra_reserve = 1;
 	}
 
 	to_reserve = btrfs_calc_trans_metadata_size(root, nr_extents);
 	to_reserve += calc_csum_metadata_size(inode, num_bytes, 1);
+	csum_bytes = BTRFS_I(inode)->csum_bytes;
 	spin_unlock(&BTRFS_I(inode)->lock);
 
 	ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush);
@@ -4232,22 +4250,35 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 
 		spin_lock(&BTRFS_I(inode)->lock);
 		dropped = drop_outstanding_extent(inode);
-		to_free = calc_csum_metadata_size(inode, num_bytes, 0);
-		spin_unlock(&BTRFS_I(inode)->lock);
-		to_free += btrfs_calc_trans_metadata_size(root, dropped);
-
 		/*
-		 * Somebody could have come in and twiddled with the
-		 * reservation, so if we have to free more than we would have
-		 * reserved from this reservation go ahead and release those
-		 * bytes.
+		 * If the inodes csum_bytes is the same as the original
+		 * csum_bytes then we know we haven't raced with any free()ers
+		 * so we can just reduce our inodes csum bytes and carry on.
+		 * Otherwise we have to do the normal free thing to account for
+		 * the case that the free side didn't free up its reserve
+		 * because of this outstanding reservation.
 		 */
-		to_free -= to_reserve;
+		if (BTRFS_I(inode)->csum_bytes == csum_bytes)
+			calc_csum_metadata_size(inode, num_bytes, 0);
+		else
+			to_free = calc_csum_metadata_size(inode, num_bytes, 0);
+		spin_unlock(&BTRFS_I(inode)->lock);
+		if (dropped)
+			to_free += btrfs_calc_trans_metadata_size(root, dropped);
+
 		if (to_free)
 			btrfs_block_rsv_release(root, block_rsv, to_free);
 		return ret;
 	}
 
+	spin_lock(&BTRFS_I(inode)->lock);
+	if (extra_reserve) {
+		BTRFS_I(inode)->delalloc_meta_reserved = 1;
+		nr_extents--;
+	}
+	BTRFS_I(inode)->reserved_extents += nr_extents;
+	spin_unlock(&BTRFS_I(inode)->lock);
+
 	block_rsv_add_bytes(block_rsv, to_reserve, 1);
 
 	return 0;
@@ -5093,11 +5124,11 @@ 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) ?
@@ -5159,6 +5190,7 @@ 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.
@@ -5196,6 +5228,7 @@ search:
 		u64 offset;
 		int cached;
 
+		used_block_group = block_group;
 		btrfs_get_block_group(block_group);
 		search_start = block_group->key.objectid;
 
@@ -5265,84 +5298,73 @@ alloc:
 		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) {
 				/*
@@ -5378,6 +5400,7 @@ refill_cluster:
 			goto loop;
 		}
 
+unclustered_alloc:
 		offset = btrfs_find_space_for_alloc(block_group, search_start,
 						    num_bytes, empty_size);
 		/*
@@ -5404,14 +5427,14 @@ 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;
 		}
 
@@ -5419,14 +5442,14 @@ 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;
 		}
 
@@ -5435,15 +5458,19 @@ 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 9472d3de5e5..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);
@@ -2287,14 +2295,20 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 		if (!uptodate) {
 			int failed_mirror;
 			failed_mirror = (int)(unsigned long)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);
+			/*
+			 * 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) {
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index dafdfa059bf..97fbe939c05 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1167,6 +1167,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 	nrptrs = min((iov_iter_count(i) + PAGE_CACHE_SIZE - 1) /
 		     PAGE_CACHE_SIZE, PAGE_CACHE_SIZE /
 		     (sizeof(struct page *)));
+	nrptrs = min(nrptrs, current->nr_dirtied_pause - current->nr_dirtied);
+	nrptrs = max(nrptrs, 8);
 	pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL);
 	if (!pages)
 		return -ENOMEM;
@@ -1387,7 +1389,11 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
 		goto out;
 	}
 
-	file_update_time(file);
+	err = btrfs_update_time(file);
+	if (err) {
+		mutex_unlock(&inode->i_mutex);
+		goto out;
+	}
 	BTRFS_I(inode)->sequence++;
 
 	start_pos = round_down(pos, root->sectorsize);
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 6e5b7e46369..ec23d43d0c3 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -1470,6 +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++;
 
@@ -2319,6 +2320,7 @@ again:
 
 	if (!found) {
 		start = i;
+		cluster->max_size = 0;
 		found = true;
 	}
 
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 526dd51a196..0a6b928813a 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -38,6 +38,7 @@
 #include <linux/falloc.h>
 #include <linux/slab.h>
 #include <linux/ratelimit.h>
+#include <linux/mount.h>
 #include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
@@ -2031,7 +2032,7 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
 	/* insert an orphan item to track this unlinked/truncated file */
 	if (insert >= 1) {
 		ret = btrfs_insert_orphan_item(trans, root, btrfs_ino(inode));
-		BUG_ON(ret);
+		BUG_ON(ret && ret != -EEXIST);
 	}
 
 	/* insert an orphan item to track subvolume contains orphan files */
@@ -2158,6 +2159,38 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 		if (ret && ret != -ESTALE)
 			goto out;
 
+		if (ret == -ESTALE && root == root->fs_info->tree_root) {
+			struct btrfs_root *dead_root;
+			struct btrfs_fs_info *fs_info = root->fs_info;
+			int is_dead_root = 0;
+
+			/*
+			 * this is an orphan in the tree root. Currently these
+			 * could come from 2 sources:
+			 *  a) a snapshot deletion in progress
+			 *  b) a free space cache inode
+			 * We need to distinguish those two, as the snapshot
+			 * orphan must not get deleted.
+			 * find_dead_roots already ran before us, so if this
+			 * is a snapshot deletion, we should find the root
+			 * in the dead_roots list
+			 */
+			spin_lock(&fs_info->trans_lock);
+			list_for_each_entry(dead_root, &fs_info->dead_roots,
+					    root_list) {
+				if (dead_root->root_key.objectid ==
+				    found_key.objectid) {
+					is_dead_root = 1;
+					break;
+				}
+			}
+			spin_unlock(&fs_info->trans_lock);
+			if (is_dead_root) {
+				/* prevent this orphan from being found again */
+				key.offset = found_key.objectid - 1;
+				continue;
+			}
+		}
 		/*
 		 * Inode is already gone but the orphan item is still there,
 		 * kill the orphan item.
@@ -2191,7 +2224,14 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 				continue;
 			}
 			nr_truncate++;
+			/*
+			 * Need to hold the imutex for reservation purposes, not
+			 * a huge deal here but I have a WARN_ON in
+			 * btrfs_delalloc_reserve_space to catch offenders.
+			 */
+			mutex_lock(&inode->i_mutex);
 			ret = btrfs_truncate(inode);
+			mutex_unlock(&inode->i_mutex);
 		} else {
 			nr_unlink++;
 		}
@@ -3327,7 +3367,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			u64 hint_byte = 0;
 			hole_size = last_byte - cur_offset;
 
-			trans = btrfs_start_transaction(root, 2);
+			trans = btrfs_start_transaction(root, 3);
 			if (IS_ERR(trans)) {
 				err = PTR_ERR(trans);
 				break;
@@ -3337,6 +3377,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 						 cur_offset + hole_size,
 						 &hint_byte, 1);
 			if (err) {
+				btrfs_update_inode(trans, root, inode);
 				btrfs_end_transaction(trans, root);
 				break;
 			}
@@ -3346,6 +3387,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 					0, hole_size, 0, hole_size,
 					0, 0, 0);
 			if (err) {
+				btrfs_update_inode(trans, root, inode);
 				btrfs_end_transaction(trans, root);
 				break;
 			}
@@ -3353,6 +3395,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			btrfs_drop_extent_cache(inode, hole_start,
 					last_byte - 1, 0);
 
+			btrfs_update_inode(trans, root, inode);
 			btrfs_end_transaction(trans, root);
 		}
 		free_extent_map(em);
@@ -3370,6 +3413,8 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 
 static int btrfs_setsize(struct inode *inode, loff_t newsize)
 {
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct btrfs_trans_handle *trans;
 	loff_t oldsize = i_size_read(inode);
 	int ret;
 
@@ -3377,16 +3422,19 @@ static int btrfs_setsize(struct inode *inode, loff_t newsize)
 		return 0;
 
 	if (newsize > oldsize) {
-		i_size_write(inode, newsize);
-		btrfs_ordered_update_i_size(inode, i_size_read(inode), NULL);
 		truncate_pagecache(inode, oldsize, newsize);
 		ret = btrfs_cont_expand(inode, oldsize, newsize);
-		if (ret) {
-			btrfs_setsize(inode, oldsize);
+		if (ret)
 			return ret;
-		}
 
-		mark_inode_dirty(inode);
+		trans = btrfs_start_transaction(root, 1);
+		if (IS_ERR(trans))
+			return PTR_ERR(trans);
+
+		i_size_write(inode, newsize);
+		btrfs_ordered_update_i_size(inode, i_size_read(inode), NULL);
+		ret = btrfs_update_inode(trans, root, inode);
+		btrfs_end_transaction_throttle(trans, root);
 	} else {
 
 		/*
@@ -3426,9 +3474,9 @@ static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	if (attr->ia_valid) {
 		setattr_copy(inode, attr);
-		mark_inode_dirty(inode);
+		err = btrfs_dirty_inode(inode);
 
-		if (attr->ia_valid & ATTR_MODE)
+		if (!err && attr->ia_valid & ATTR_MODE)
 			err = btrfs_acl_chmod(inode);
 	}
 
@@ -3490,7 +3538,7 @@ 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
@@ -4204,42 +4252,80 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc)
  * FIXME, needs more benchmarking...there are no reasons other than performance
  * to keep or drop this code.
  */
-void btrfs_dirty_inode(struct inode *inode, int flags)
+int btrfs_dirty_inode(struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_trans_handle *trans;
 	int ret;
 
 	if (BTRFS_I(inode)->dummy_inode)
-		return;
+		return 0;
 
 	trans = btrfs_join_transaction(root);
-	BUG_ON(IS_ERR(trans));
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
 
 	ret = btrfs_update_inode(trans, root, inode);
 	if (ret && ret == -ENOSPC) {
 		/* whoops, lets try again with the full transaction */
 		btrfs_end_transaction(trans, root);
 		trans = btrfs_start_transaction(root, 1);
-		if (IS_ERR(trans)) {
-			printk_ratelimited(KERN_ERR "btrfs: fail to "
-				       "dirty  inode %llu error %ld\n",
-				       (unsigned long long)btrfs_ino(inode),
-				       PTR_ERR(trans));
-			return;
-		}
+		if (IS_ERR(trans))
+			return PTR_ERR(trans);
 
 		ret = btrfs_update_inode(trans, root, inode);
-		if (ret) {
-			printk_ratelimited(KERN_ERR "btrfs: fail to "
-				       "dirty  inode %llu error %d\n",
-				       (unsigned long long)btrfs_ino(inode),
-				       ret);
-		}
 	}
 	btrfs_end_transaction(trans, root);
 	if (BTRFS_I(inode)->delayed_node)
 		btrfs_balance_delayed_items(root);
+
+	return ret;
+}
+
+/*
+ * This is a copy of file_update_time.  We need this so we can return error on
+ * ENOSPC for updating the inode in the case of file write and mmap writes.
+ */
+int btrfs_update_time(struct file *file)
+{
+	struct inode *inode = file->f_path.dentry->d_inode;
+	struct timespec now;
+	int ret;
+	enum { S_MTIME = 1, S_CTIME = 2, S_VERSION = 4 } sync_it = 0;
+
+	/* First try to exhaust all avenues to not sync */
+	if (IS_NOCMTIME(inode))
+		return 0;
+
+	now = current_fs_time(inode->i_sb);
+	if (!timespec_equal(&inode->i_mtime, &now))
+		sync_it = S_MTIME;
+
+	if (!timespec_equal(&inode->i_ctime, &now))
+		sync_it |= S_CTIME;
+
+	if (IS_I_VERSION(inode))
+		sync_it |= S_VERSION;
+
+	if (!sync_it)
+		return 0;
+
+	/* Finally allowed to write? Takes lock. */
+	if (mnt_want_write_file(file))
+		return 0;
+
+	/* Only change inode inside the lock region */
+	if (sync_it & S_VERSION)
+		inode_inc_iversion(inode);
+	if (sync_it & S_CTIME)
+		inode->i_ctime = now;
+	if (sync_it & S_MTIME)
+		inode->i_mtime = now;
+	ret = btrfs_dirty_inode(inode);
+	if (!ret)
+		mark_inode_dirty_sync(inode);
+	mnt_drop_write(file->f_path.mnt);
+	return ret;
 }
 
 /*
@@ -4555,11 +4641,18 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
+	/*
+	* If the active LSM wants to access the inode during
+	* d_instantiate it needs these. Smack checks to see
+	* if the filesystem supports xattrs by looking at the
+	* ops vector.
+	*/
+
+	inode->i_op = &btrfs_special_inode_operations;
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 0, index);
 	if (err)
 		drop_inode = 1;
 	else {
-		inode->i_op = &btrfs_special_inode_operations;
 		init_special_inode(inode, inode->i_mode, rdev);
 		btrfs_update_inode(trans, root, inode);
 	}
@@ -4613,14 +4706,21 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
+	/*
+	* If the active LSM wants to access the inode during
+	* d_instantiate it needs these. Smack checks to see
+	* if the filesystem supports xattrs by looking at the
+	* ops vector.
+	*/
+	inode->i_fop = &btrfs_file_operations;
+	inode->i_op = &btrfs_file_inode_operations;
+
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 0, index);
 	if (err)
 		drop_inode = 1;
 	else {
 		inode->i_mapping->a_ops = &btrfs_aops;
 		inode->i_mapping->backing_dev_info = &root->fs_info->bdi;
-		inode->i_fop = &btrfs_file_operations;
-		inode->i_op = &btrfs_file_inode_operations;
 		BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
 	}
 out_unlock:
@@ -6303,7 +6403,12 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	u64 page_start;
 	u64 page_end;
 
+	/* Need this to keep space reservations serialized */
+	mutex_lock(&inode->i_mutex);
 	ret  = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE);
+	mutex_unlock(&inode->i_mutex);
+	if (!ret)
+		ret = btrfs_update_time(vma->vm_file);
 	if (ret) {
 		if (ret == -ENOMEM)
 			ret = VM_FAULT_OOM;
@@ -6515,8 +6620,9 @@ static int btrfs_truncate(struct inode *inode)
 			/* Just need the 1 for updating the inode */
 			trans = btrfs_start_transaction(root, 1);
 			if (IS_ERR(trans)) {
-				err = PTR_ERR(trans);
-				goto out;
+				ret = err = PTR_ERR(trans);
+				trans = NULL;
+				break;
 			}
 		}
 
@@ -7076,14 +7182,21 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
+	/*
+	* If the active LSM wants to access the inode during
+	* d_instantiate it needs these. Smack checks to see
+	* if the filesystem supports xattrs by looking at the
+	* ops vector.
+	*/
+	inode->i_fop = &btrfs_file_operations;
+	inode->i_op = &btrfs_file_inode_operations;
+
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 0, index);
 	if (err)
 		drop_inode = 1;
 	else {
 		inode->i_mapping->a_ops = &btrfs_aops;
 		inode->i_mapping->backing_dev_info = &root->fs_info->bdi;
-		inode->i_fop = &btrfs_file_operations;
-		inode->i_op = &btrfs_file_inode_operations;
 		BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
 	}
 	if (drop_inode)
@@ -7353,6 +7466,7 @@ static const struct inode_operations btrfs_symlink_inode_operations = {
 	.follow_link	= page_follow_link_light,
 	.put_link	= page_put_link,
 	.getattr	= btrfs_getattr,
+	.setattr	= btrfs_setattr,
 	.permission	= btrfs_permission,
 	.setxattr	= btrfs_setxattr,
 	.getxattr	= btrfs_getxattr,
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index a90e749ed6d..c04f02c7d5b 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -252,11 +252,11 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	trans = btrfs_join_transaction(root);
 	BUG_ON(IS_ERR(trans));
 
+	btrfs_update_iflags(inode);
+	inode->i_ctime = CURRENT_TIME;
 	ret = btrfs_update_inode(trans, root, inode);
 	BUG_ON(ret);
 
-	btrfs_update_iflags(inode);
-	inode->i_ctime = CURRENT_TIME;
 	btrfs_end_transaction(trans, root);
 
 	mnt_drop_write(file->f_path.mnt);
@@ -858,8 +858,10 @@ static int cluster_pages_for_defrag(struct inode *inode,
 		return 0;
 	file_end = (isize - 1) >> PAGE_CACHE_SHIFT;
 
+	mutex_lock(&inode->i_mutex);
 	ret = btrfs_delalloc_reserve_space(inode,
 					   num_pages << PAGE_CACHE_SHIFT);
+	mutex_unlock(&inode->i_mutex);
 	if (ret)
 		return ret;
 again:
@@ -1278,7 +1280,7 @@ 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);
 	}
 
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index dff29d5e151..cfb55434a46 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2947,7 +2947,9 @@ static int relocate_file_extent_cluster(struct inode *inode,
 	index = (cluster->start - offset) >> PAGE_CACHE_SHIFT;
 	last_index = (cluster->end - offset) >> PAGE_CACHE_SHIFT;
 	while (index <= last_index) {
+		mutex_lock(&inode->i_mutex);
 		ret = btrfs_delalloc_reserve_metadata(inode, PAGE_CACHE_SIZE);
+		mutex_unlock(&inode->i_mutex);
 		if (ret)
 			goto out;
 
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index fab420db512..ddf2c90d3fc 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)
@@ -1530,18 +1535,22 @@ static noinline_for_stack int scrub_supers(struct scrub_dev *sdev)
 static noinline_for_stack int scrub_workers_get(struct btrfs_root *root)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
+	int ret = 0;
 
 	mutex_lock(&fs_info->scrub_lock);
 	if (fs_info->scrub_workers_refcnt == 0) {
 		btrfs_init_workers(&fs_info->scrub_workers, "scrub",
 			   fs_info->thread_pool_size, &fs_info->generic_worker);
 		fs_info->scrub_workers.idle_thresh = 4;
-		btrfs_start_workers(&fs_info->scrub_workers, 1);
+		ret = btrfs_start_workers(&fs_info->scrub_workers);
+		if (ret)
+			goto out;
 	}
 	++fs_info->scrub_workers_refcnt;
+out:
 	mutex_unlock(&fs_info->scrub_lock);
 
-	return 0;
+	return ret;
 }
 
 static noinline_for_stack void scrub_workers_put(struct btrfs_root *root)
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 17ee7fc5e64..200f63bc667 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -41,6 +41,7 @@
 #include <linux/slab.h>
 #include <linux/cleancache.h>
 #include <linux/mnt_namespace.h>
+#include <linux/ratelimit.h>
 #include "compat.h"
 #include "delayed-inode.h"
 #include "ctree.h"
@@ -1053,11 +1054,11 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 	u64 avail_space;
 	u64 used_space;
 	u64 min_stripe_size;
-	int min_stripes = 1;
+	int min_stripes = 1, num_stripes = 1;
 	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,
@@ -1067,20 +1068,24 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 
 	/* calc min stripe number for data space alloction */
 	type = btrfs_get_alloc_profile(root, 1);
-	if (type & BTRFS_BLOCK_GROUP_RAID0)
+	if (type & BTRFS_BLOCK_GROUP_RAID0) {
 		min_stripes = 2;
-	else if (type & BTRFS_BLOCK_GROUP_RAID1)
+		num_stripes = nr_devices;
+	} else if (type & BTRFS_BLOCK_GROUP_RAID1) {
 		min_stripes = 2;
-	else if (type & BTRFS_BLOCK_GROUP_RAID10)
+		num_stripes = 2;
+	} else if (type & BTRFS_BLOCK_GROUP_RAID10) {
 		min_stripes = 4;
+		num_stripes = 4;
+	}
 
 	if (type & BTRFS_BLOCK_GROUP_DUP)
 		min_stripe_size = 2 * BTRFS_STRIPE_LEN;
 	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;
@@ -1141,13 +1146,16 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
 	i = nr_devices - 1;
 	avail_space = 0;
 	while (nr_devices >= min_stripes) {
+		if (num_stripes > nr_devices)
+			num_stripes = nr_devices;
+
 		if (devices_info[i].max_avail >= min_stripe_size) {
 			int j;
 			u64 alloc_size;
 
-			avail_space += devices_info[i].max_avail * min_stripes;
+			avail_space += devices_info[i].max_avail * num_stripes;
 			alloc_size = devices_info[i].max_avail;
-			for (j = i + 1 - min_stripes; j <= i; j++)
+			for (j = i + 1 - num_stripes; j <= i; j++)
 				devices_info[j].max_avail -= alloc_size;
 		}
 		i--;
@@ -1264,6 +1272,16 @@ static int btrfs_unfreeze(struct super_block *sb)
 	return 0;
 }
 
+static void btrfs_fs_dirty_inode(struct inode *inode, int flags)
+{
+	int ret;
+
+	ret = btrfs_dirty_inode(inode);
+	if (ret)
+		printk_ratelimited(KERN_ERR "btrfs: fail to dirty inode %Lu "
+				   "error %d\n", btrfs_ino(inode), ret);
+}
+
 static const struct super_operations btrfs_super_ops = {
 	.drop_inode	= btrfs_drop_inode,
 	.evict_inode	= btrfs_evict_inode,
@@ -1271,7 +1289,7 @@ static const struct super_operations btrfs_super_ops = {
 	.sync_fs	= btrfs_sync_fs,
 	.show_options	= btrfs_show_options,
 	.write_inode	= btrfs_write_inode,
-	.dirty_inode	= btrfs_dirty_inode,
+	.dirty_inode	= btrfs_fs_dirty_inode,
 	.alloc_inode	= btrfs_alloc_inode,
 	.destroy_inode	= btrfs_destroy_inode,
 	.statfs		= btrfs_statfs,
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index c37433d3cd8..f4b839fd3c9 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -295,6 +295,12 @@ loop_lock:
 			btrfs_requeue_work(&device->work);
 			goto done;
 		}
+		/* unplug every 64 requests just for good measure */
+		if (batch_run % 64 == 0) {
+			blk_finish_plug(&plug);
+			blk_start_plug(&plug);
+			sync_pending = 0;
+		}
 	}
 
 	cond_resched();
@@ -1611,7 +1617,7 @@ 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);
@@ -3258,7 +3264,7 @@ static void btrfs_end_bio(struct bio *bio, int err)
 		 */
 		if (atomic_read(&bbio->error) > bbio->max_errors) {
 			err = -EIO;
-		} else if (err) {
+		} else {
 			/*
 			 * this bio is actually up to date, we didn't
 			 * go over the max number of errors
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 4144caf2f9d..173b1d22e59 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -87,7 +87,7 @@ static int ceph_set_page_dirty(struct page *page)
 	snapc = ceph_get_snap_context(ci->i_snap_realm->cached_context);
 
 	/* dirty the head */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (ci->i_head_snapc == NULL)
 		ci->i_head_snapc = ceph_get_snap_context(snapc);
 	++ci->i_wrbuffer_ref_head;
@@ -100,7 +100,7 @@ static int ceph_set_page_dirty(struct page *page)
 	     ci->i_wrbuffer_ref-1, ci->i_wrbuffer_ref_head-1,
 	     ci->i_wrbuffer_ref, ci->i_wrbuffer_ref_head,
 	     snapc, snapc->seq, snapc->num_snaps);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	/* now adjust page */
 	spin_lock_irq(&mapping->tree_lock);
@@ -391,7 +391,7 @@ static struct ceph_snap_context *get_oldest_context(struct inode *inode,
 	struct ceph_snap_context *snapc = NULL;
 	struct ceph_cap_snap *capsnap = NULL;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) {
 		dout(" cap_snap %p snapc %p has %d dirty pages\n", capsnap,
 		     capsnap->context, capsnap->dirty_pages);
@@ -407,7 +407,7 @@ static struct ceph_snap_context *get_oldest_context(struct inode *inode,
 		dout(" head snapc %p has %d dirty pages\n",
 		     snapc, ci->i_wrbuffer_ref_head);
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return snapc;
 }
 
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index 0f327c6c967..8b53193e4f7 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -309,7 +309,7 @@ void ceph_reservation_status(struct ceph_fs_client *fsc,
 /*
  * Find ceph_cap for given mds, if any.
  *
- * Called with i_lock held.
+ * Called with i_ceph_lock held.
  */
 static struct ceph_cap *__get_cap_for_mds(struct ceph_inode_info *ci, int mds)
 {
@@ -332,9 +332,9 @@ struct ceph_cap *ceph_get_cap_for_mds(struct ceph_inode_info *ci, int mds)
 {
 	struct ceph_cap *cap;
 
-	spin_lock(&ci->vfs_inode.i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	cap = __get_cap_for_mds(ci, mds);
-	spin_unlock(&ci->vfs_inode.i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return cap;
 }
 
@@ -361,15 +361,16 @@ static int __ceph_get_cap_mds(struct ceph_inode_info *ci)
 
 int ceph_get_cap_mds(struct inode *inode)
 {
+	struct ceph_inode_info *ci = ceph_inode(inode);
 	int mds;
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	mds = __ceph_get_cap_mds(ceph_inode(inode));
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return mds;
 }
 
 /*
- * Called under i_lock.
+ * Called under i_ceph_lock.
  */
 static void __insert_cap_node(struct ceph_inode_info *ci,
 			      struct ceph_cap *new)
@@ -415,7 +416,7 @@ static void __cap_set_timeouts(struct ceph_mds_client *mdsc,
  *
  * If I_FLUSH is set, leave the inode at the front of the list.
  *
- * Caller holds i_lock
+ * Caller holds i_ceph_lock
  *    -> we take mdsc->cap_delay_lock
  */
 static void __cap_delay_requeue(struct ceph_mds_client *mdsc,
@@ -457,7 +458,7 @@ static void __cap_delay_requeue_front(struct ceph_mds_client *mdsc,
 /*
  * Cancel delayed work on cap.
  *
- * Caller must hold i_lock.
+ * Caller must hold i_ceph_lock.
  */
 static void __cap_delay_cancel(struct ceph_mds_client *mdsc,
 			       struct ceph_inode_info *ci)
@@ -532,14 +533,14 @@ int ceph_add_cap(struct inode *inode,
 		wanted |= ceph_caps_for_mode(fmode);
 
 retry:
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	cap = __get_cap_for_mds(ci, mds);
 	if (!cap) {
 		if (new_cap) {
 			cap = new_cap;
 			new_cap = NULL;
 		} else {
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 			new_cap = get_cap(mdsc, caps_reservation);
 			if (new_cap == NULL)
 				return -ENOMEM;
@@ -625,7 +626,7 @@ retry:
 
 	if (fmode >= 0)
 		__ceph_get_fmode(ci, fmode);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	wake_up_all(&ci->i_cap_wq);
 	return 0;
 }
@@ -792,7 +793,7 @@ int ceph_caps_revoking(struct ceph_inode_info *ci, int mask)
 	struct rb_node *p;
 	int ret = 0;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
 		cap = rb_entry(p, struct ceph_cap, ci_node);
 		if (__cap_is_valid(cap) &&
@@ -801,7 +802,7 @@ int ceph_caps_revoking(struct ceph_inode_info *ci, int mask)
 			break;
 		}
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	dout("ceph_caps_revoking %p %s = %d\n", inode,
 	     ceph_cap_string(mask), ret);
 	return ret;
@@ -855,7 +856,7 @@ int __ceph_caps_mds_wanted(struct ceph_inode_info *ci)
 }
 
 /*
- * called under i_lock
+ * called under i_ceph_lock
  */
 static int __ceph_is_any_caps(struct ceph_inode_info *ci)
 {
@@ -865,7 +866,7 @@ static int __ceph_is_any_caps(struct ceph_inode_info *ci)
 /*
  * Remove a cap.  Take steps to deal with a racing iterate_session_caps.
  *
- * caller should hold i_lock.
+ * caller should hold i_ceph_lock.
  * caller will not hold session s_mutex if called from destroy_inode.
  */
 void __ceph_remove_cap(struct ceph_cap *cap)
@@ -1028,7 +1029,7 @@ static void __queue_cap_release(struct ceph_mds_session *session,
 
 /*
  * Queue cap releases when an inode is dropped from our cache.  Since
- * inode is about to be destroyed, there is no need for i_lock.
+ * inode is about to be destroyed, there is no need for i_ceph_lock.
  */
 void ceph_queue_caps_release(struct inode *inode)
 {
@@ -1049,7 +1050,7 @@ void ceph_queue_caps_release(struct inode *inode)
 
 /*
  * Send a cap msg on the given inode.  Update our caps state, then
- * drop i_lock and send the message.
+ * drop i_ceph_lock and send the message.
  *
  * Make note of max_size reported/requested from mds, revoked caps
  * that have now been implemented.
@@ -1061,13 +1062,13 @@ void ceph_queue_caps_release(struct inode *inode)
  * Return non-zero if delayed release, or we experienced an error
  * such that the caller should requeue + retry later.
  *
- * called with i_lock, then drops it.
+ * called with i_ceph_lock, then drops it.
  * caller should hold snap_rwsem (read), s_mutex.
  */
 static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
 		      int op, int used, int want, int retain, int flushing,
 		      unsigned *pflush_tid)
-	__releases(cap->ci->vfs_inode->i_lock)
+	__releases(cap->ci->i_ceph_lock)
 {
 	struct ceph_inode_info *ci = cap->ci;
 	struct inode *inode = &ci->vfs_inode;
@@ -1170,7 +1171,7 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
 		xattr_version = ci->i_xattrs.version;
 	}
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	ret = send_cap_msg(session, ceph_vino(inode).ino, cap_id,
 		op, keep, want, flushing, seq, flush_tid, issue_seq, mseq,
@@ -1198,13 +1199,13 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap,
  * Unless @again is true, skip cap_snaps that were already sent to
  * the MDS (i.e., during this session).
  *
- * Called under i_lock.  Takes s_mutex as needed.
+ * Called under i_ceph_lock.  Takes s_mutex as needed.
  */
 void __ceph_flush_snaps(struct ceph_inode_info *ci,
 			struct ceph_mds_session **psession,
 			int again)
-		__releases(ci->vfs_inode->i_lock)
-		__acquires(ci->vfs_inode->i_lock)
+		__releases(ci->i_ceph_lock)
+		__acquires(ci->i_ceph_lock)
 {
 	struct inode *inode = &ci->vfs_inode;
 	int mds;
@@ -1261,7 +1262,7 @@ retry:
 			session = NULL;
 		}
 		if (!session) {
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 			mutex_lock(&mdsc->mutex);
 			session = __ceph_lookup_mds_session(mdsc, mds);
 			mutex_unlock(&mdsc->mutex);
@@ -1275,7 +1276,7 @@ retry:
 			 * deletion or migration.  retry, and we'll
 			 * get a better @mds value next time.
 			 */
-			spin_lock(&inode->i_lock);
+			spin_lock(&ci->i_ceph_lock);
 			goto retry;
 		}
 
@@ -1285,7 +1286,7 @@ retry:
 			list_del_init(&capsnap->flushing_item);
 		list_add_tail(&capsnap->flushing_item,
 			      &session->s_cap_snaps_flushing);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 
 		dout("flush_snaps %p cap_snap %p follows %lld tid %llu\n",
 		     inode, capsnap, capsnap->follows, capsnap->flush_tid);
@@ -1302,7 +1303,7 @@ retry:
 		next_follows = capsnap->follows + 1;
 		ceph_put_cap_snap(capsnap);
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		goto retry;
 	}
 
@@ -1322,11 +1323,9 @@ out:
 
 static void ceph_flush_snaps(struct ceph_inode_info *ci)
 {
-	struct inode *inode = &ci->vfs_inode;
-
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	__ceph_flush_snaps(ci, NULL, 0);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 /*
@@ -1373,7 +1372,7 @@ int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask)
  * Add dirty inode to the flushing list.  Assigned a seq number so we
  * can wait for caps to flush without starving.
  *
- * Called under i_lock.
+ * Called under i_ceph_lock.
  */
 static int __mark_caps_flushing(struct inode *inode,
 				 struct ceph_mds_session *session)
@@ -1421,9 +1420,9 @@ static int try_nonblocking_invalidate(struct inode *inode)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	u32 invalidating_gen = ci->i_rdcache_gen;
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	invalidate_mapping_pages(&inode->i_data, 0, -1);
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	if (inode->i_data.nrpages == 0 &&
 	    invalidating_gen == ci->i_rdcache_gen) {
@@ -1470,7 +1469,7 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags,
 	if (mdsc->stopping)
 		is_delayed = 1;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	if (ci->i_ceph_flags & CEPH_I_FLUSH)
 		flags |= CHECK_CAPS_FLUSH;
@@ -1480,7 +1479,7 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags,
 		__ceph_flush_snaps(ci, &session, 0);
 	goto retry_locked;
 retry:
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 retry_locked:
 	file_wanted = __ceph_caps_file_wanted(ci);
 	used = __ceph_caps_used(ci);
@@ -1634,7 +1633,7 @@ ack:
 			if (mutex_trylock(&session->s_mutex) == 0) {
 				dout("inverting session/ino locks on %p\n",
 				     session);
-				spin_unlock(&inode->i_lock);
+				spin_unlock(&ci->i_ceph_lock);
 				if (took_snap_rwsem) {
 					up_read(&mdsc->snap_rwsem);
 					took_snap_rwsem = 0;
@@ -1648,7 +1647,7 @@ ack:
 			if (down_read_trylock(&mdsc->snap_rwsem) == 0) {
 				dout("inverting snap/in locks on %p\n",
 				     inode);
-				spin_unlock(&inode->i_lock);
+				spin_unlock(&ci->i_ceph_lock);
 				down_read(&mdsc->snap_rwsem);
 				took_snap_rwsem = 1;
 				goto retry;
@@ -1664,10 +1663,10 @@ ack:
 		mds = cap->mds;  /* remember mds, so we don't repeat */
 		sent++;
 
-		/* __send_cap drops i_lock */
+		/* __send_cap drops i_ceph_lock */
 		delayed += __send_cap(mdsc, cap, CEPH_CAP_OP_UPDATE, used, want,
 				      retain, flushing, NULL);
-		goto retry; /* retake i_lock and restart our cap scan. */
+		goto retry; /* retake i_ceph_lock and restart our cap scan. */
 	}
 
 	/*
@@ -1681,7 +1680,7 @@ ack:
 	else if (!is_delayed || force_requeue)
 		__cap_delay_requeue(mdsc, ci);
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (queue_invalidate)
 		ceph_queue_invalidate(inode);
@@ -1704,7 +1703,7 @@ static int try_flush_caps(struct inode *inode, struct ceph_mds_session *session,
 	int flushing = 0;
 
 retry:
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (ci->i_ceph_flags & CEPH_I_NOFLUSH) {
 		dout("try_flush_caps skipping %p I_NOFLUSH set\n", inode);
 		goto out;
@@ -1716,7 +1715,7 @@ retry:
 		int delayed;
 
 		if (!session) {
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 			session = cap->session;
 			mutex_lock(&session->s_mutex);
 			goto retry;
@@ -1727,18 +1726,18 @@ retry:
 
 		flushing = __mark_caps_flushing(inode, session);
 
-		/* __send_cap drops i_lock */
+		/* __send_cap drops i_ceph_lock */
 		delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, used, want,
 				     cap->issued | cap->implemented, flushing,
 				     flush_tid);
 		if (!delayed)
 			goto out_unlocked;
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		__cap_delay_requeue(mdsc, ci);
 	}
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 out_unlocked:
 	if (session && unlock_session)
 		mutex_unlock(&session->s_mutex);
@@ -1753,7 +1752,7 @@ static int caps_are_flushed(struct inode *inode, unsigned tid)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int i, ret = 1;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	for (i = 0; i < CEPH_CAP_BITS; i++)
 		if ((ci->i_flushing_caps & (1 << i)) &&
 		    ci->i_cap_flush_tid[i] <= tid) {
@@ -1761,7 +1760,7 @@ static int caps_are_flushed(struct inode *inode, unsigned tid)
 			ret = 0;
 			break;
 		}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return ret;
 }
 
@@ -1868,10 +1867,10 @@ int ceph_write_inode(struct inode *inode, struct writeback_control *wbc)
 		struct ceph_mds_client *mdsc =
 			ceph_sb_to_client(inode->i_sb)->mdsc;
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		if (__ceph_caps_dirty(ci))
 			__cap_delay_requeue_front(mdsc, ci);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 	}
 	return err;
 }
@@ -1894,7 +1893,7 @@ static void kick_flushing_capsnaps(struct ceph_mds_client *mdsc,
 		struct inode *inode = &ci->vfs_inode;
 		struct ceph_cap *cap;
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		cap = ci->i_auth_cap;
 		if (cap && cap->session == session) {
 			dout("kick_flushing_caps %p cap %p capsnap %p\n", inode,
@@ -1904,7 +1903,7 @@ static void kick_flushing_capsnaps(struct ceph_mds_client *mdsc,
 			pr_err("%p auth cap %p not mds%d ???\n", inode,
 			       cap, session->s_mds);
 		}
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 	}
 }
 
@@ -1921,7 +1920,7 @@ void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc,
 		struct ceph_cap *cap;
 		int delayed = 0;
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		cap = ci->i_auth_cap;
 		if (cap && cap->session == session) {
 			dout("kick_flushing_caps %p cap %p %s\n", inode,
@@ -1932,14 +1931,14 @@ void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc,
 					     cap->issued | cap->implemented,
 					     ci->i_flushing_caps, NULL);
 			if (delayed) {
-				spin_lock(&inode->i_lock);
+				spin_lock(&ci->i_ceph_lock);
 				__cap_delay_requeue(mdsc, ci);
-				spin_unlock(&inode->i_lock);
+				spin_unlock(&ci->i_ceph_lock);
 			}
 		} else {
 			pr_err("%p auth cap %p not mds%d ???\n", inode,
 			       cap, session->s_mds);
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 		}
 	}
 }
@@ -1952,7 +1951,7 @@ static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc,
 	struct ceph_cap *cap;
 	int delayed = 0;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	cap = ci->i_auth_cap;
 	dout("kick_flushing_inode_caps %p flushing %s flush_seq %lld\n", inode,
 	     ceph_cap_string(ci->i_flushing_caps), ci->i_cap_flush_seq);
@@ -1964,12 +1963,12 @@ static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc,
 				     cap->issued | cap->implemented,
 				     ci->i_flushing_caps, NULL);
 		if (delayed) {
-			spin_lock(&inode->i_lock);
+			spin_lock(&ci->i_ceph_lock);
 			__cap_delay_requeue(mdsc, ci);
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 		}
 	} else {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 	}
 }
 
@@ -1978,7 +1977,7 @@ static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc,
  * Take references to capabilities we hold, so that we don't release
  * them to the MDS prematurely.
  *
- * Protected by i_lock.
+ * Protected by i_ceph_lock.
  */
 static void __take_cap_refs(struct ceph_inode_info *ci, int got)
 {
@@ -2016,7 +2015,7 @@ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want,
 
 	dout("get_cap_refs %p need %s want %s\n", inode,
 	     ceph_cap_string(need), ceph_cap_string(want));
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	/* make sure file is actually open */
 	file_wanted = __ceph_caps_file_wanted(ci);
@@ -2077,7 +2076,7 @@ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want,
 		     ceph_cap_string(have), ceph_cap_string(need));
 	}
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	dout("get_cap_refs %p ret %d got %s\n", inode,
 	     ret, ceph_cap_string(*got));
 	return ret;
@@ -2094,7 +2093,7 @@ static void check_max_size(struct inode *inode, loff_t endoff)
 	int check = 0;
 
 	/* do we need to explicitly request a larger max_size? */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if ((endoff >= ci->i_max_size ||
 	     endoff > (inode->i_size << 1)) &&
 	    endoff > ci->i_wanted_max_size) {
@@ -2103,7 +2102,7 @@ static void check_max_size(struct inode *inode, loff_t endoff)
 		ci->i_wanted_max_size = endoff;
 		check = 1;
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	if (check)
 		ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL);
 }
@@ -2140,9 +2139,9 @@ retry:
  */
 void ceph_get_cap_refs(struct ceph_inode_info *ci, int caps)
 {
-	spin_lock(&ci->vfs_inode.i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	__take_cap_refs(ci, caps);
-	spin_unlock(&ci->vfs_inode.i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 /*
@@ -2160,7 +2159,7 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had)
 	int last = 0, put = 0, flushsnaps = 0, wake = 0;
 	struct ceph_cap_snap *capsnap;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (had & CEPH_CAP_PIN)
 		--ci->i_pin_ref;
 	if (had & CEPH_CAP_FILE_RD)
@@ -2193,7 +2192,7 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had)
 				}
 			}
 		}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	dout("put_cap_refs %p had %s%s%s\n", inode, ceph_cap_string(had),
 	     last ? " last" : "", put ? " put" : "");
@@ -2225,7 +2224,7 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr,
 	int found = 0;
 	struct ceph_cap_snap *capsnap = NULL;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	ci->i_wrbuffer_ref -= nr;
 	last = !ci->i_wrbuffer_ref;
 
@@ -2274,7 +2273,7 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr,
 		}
 	}
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (last) {
 		ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL);
@@ -2291,7 +2290,7 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr,
  * Handle a cap GRANT message from the MDS.  (Note that a GRANT may
  * actually be a revocation if it specifies a smaller cap set.)
  *
- * caller holds s_mutex and i_lock, we drop both.
+ * caller holds s_mutex and i_ceph_lock, we drop both.
  *
  * return value:
  *  0 - ok
@@ -2302,7 +2301,7 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 			     struct ceph_mds_session *session,
 			     struct ceph_cap *cap,
 			     struct ceph_buffer *xattr_buf)
-		__releases(inode->i_lock)
+		__releases(ci->i_ceph_lock)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int mds = session->s_mds;
@@ -2453,7 +2452,7 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 	}
 	BUG_ON(cap->issued & ~cap->implemented);
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	if (writeback)
 		/*
 		 * queue inode for writeback: we can't actually call
@@ -2483,7 +2482,7 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid,
 				 struct ceph_mds_caps *m,
 				 struct ceph_mds_session *session,
 				 struct ceph_cap *cap)
-	__releases(inode->i_lock)
+	__releases(ci->i_ceph_lock)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc;
@@ -2539,7 +2538,7 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid,
 	wake_up_all(&ci->i_cap_wq);
 
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	if (drop)
 		iput(inode);
 }
@@ -2562,7 +2561,7 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid,
 	dout("handle_cap_flushsnap_ack inode %p ci %p mds%d follows %lld\n",
 	     inode, ci, session->s_mds, follows);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) {
 		if (capsnap->follows == follows) {
 			if (capsnap->flush_tid != flush_tid) {
@@ -2585,7 +2584,7 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid,
 			     capsnap, capsnap->follows);
 		}
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	if (drop)
 		iput(inode);
 }
@@ -2598,7 +2597,7 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid,
 static void handle_cap_trunc(struct inode *inode,
 			     struct ceph_mds_caps *trunc,
 			     struct ceph_mds_session *session)
-	__releases(inode->i_lock)
+	__releases(ci->i_ceph_lock)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int mds = session->s_mds;
@@ -2617,7 +2616,7 @@ static void handle_cap_trunc(struct inode *inode,
 	     inode, mds, seq, truncate_size, truncate_seq);
 	queue_trunc = ceph_fill_file_size(inode, issued,
 					  truncate_seq, truncate_size, size);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (queue_trunc)
 		ceph_queue_vmtruncate(inode);
@@ -2646,7 +2645,7 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex,
 	dout("handle_cap_export inode %p ci %p mds%d mseq %d\n",
 	     inode, ci, mds, mseq);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	/* make sure we haven't seen a higher mseq */
 	for (p = rb_first(&ci->i_caps); p; p = rb_next(p)) {
@@ -2690,7 +2689,7 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex,
 	}
 	/* else, we already released it */
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 /*
@@ -2745,9 +2744,9 @@ static void handle_cap_import(struct ceph_mds_client *mdsc,
 	up_read(&mdsc->snap_rwsem);
 
 	/* make sure we re-request max_size, if necessary */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	ci->i_requested_max_size = 0;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 /*
@@ -2762,6 +2761,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	struct ceph_mds_client *mdsc = session->s_mdsc;
 	struct super_block *sb = mdsc->fsc->sb;
 	struct inode *inode;
+	struct ceph_inode_info *ci;
 	struct ceph_cap *cap;
 	struct ceph_mds_caps *h;
 	int mds = session->s_mds;
@@ -2815,6 +2815,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 
 	/* lookup ino */
 	inode = ceph_find_inode(sb, vino);
+	ci = ceph_inode(inode);
 	dout(" op %s ino %llx.%llx inode %p\n", ceph_cap_op_name(op), vino.ino,
 	     vino.snap, inode);
 	if (!inode) {
@@ -2844,16 +2845,16 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 	}
 
 	/* the rest require a cap */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	cap = __get_cap_for_mds(ceph_inode(inode), mds);
 	if (!cap) {
 		dout(" no cap on %p ino %llx.%llx from mds%d\n",
 		     inode, ceph_ino(inode), ceph_snap(inode), mds);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		goto flush_cap_releases;
 	}
 
-	/* note that each of these drops i_lock for us */
+	/* note that each of these drops i_ceph_lock for us */
 	switch (op) {
 	case CEPH_CAP_OP_REVOKE:
 	case CEPH_CAP_OP_GRANT:
@@ -2869,7 +2870,7 @@ void ceph_handle_caps(struct ceph_mds_session *session,
 		break;
 
 	default:
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		pr_err("ceph_handle_caps: unknown cap op %d %s\n", op,
 		       ceph_cap_op_name(op));
 	}
@@ -2962,13 +2963,13 @@ void ceph_put_fmode(struct ceph_inode_info *ci, int fmode)
 	struct inode *inode = &ci->vfs_inode;
 	int last = 0;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	dout("put_fmode %p fmode %d %d -> %d\n", inode, fmode,
 	     ci->i_nr_by_mode[fmode], ci->i_nr_by_mode[fmode]-1);
 	BUG_ON(ci->i_nr_by_mode[fmode] == 0);
 	if (--ci->i_nr_by_mode[fmode] == 0)
 		last++;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (last && ci->i_vino.snap == CEPH_NOSNAP)
 		ceph_check_caps(ci, 0, NULL);
@@ -2991,7 +2992,7 @@ int ceph_encode_inode_release(void **p, struct inode *inode,
 	int used, dirty;
 	int ret = 0;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	used = __ceph_caps_used(ci);
 	dirty = __ceph_caps_dirty(ci);
 
@@ -3046,7 +3047,7 @@ int ceph_encode_inode_release(void **p, struct inode *inode,
 			     inode, cap, ceph_cap_string(cap->issued));
 		}
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return ret;
 }
 
@@ -3061,7 +3062,7 @@ int ceph_encode_dentry_release(void **p, struct dentry *dentry,
 
 	/*
 	 * force an record for the directory caps if we have a dentry lease.
-	 * this is racy (can't take i_lock and d_lock together), but it
+	 * this is racy (can't take i_ceph_lock and d_lock together), but it
 	 * doesn't have to be perfect; the mds will revoke anything we don't
 	 * release.
 	 */
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index bca3948e9db..3eeb9766126 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -281,18 +281,18 @@ static int ceph_readdir(struct file *filp, void *dirent, filldir_t filldir)
 	}
 
 	/* can we use the dcache? */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if ((filp->f_pos == 2 || fi->dentry) &&
 	    !ceph_test_mount_opt(fsc, NOASYNCREADDIR) &&
 	    ceph_snap(inode) != CEPH_SNAPDIR &&
 	    ceph_dir_test_complete(inode) &&
 	    __ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, 1)) {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		err = __dcache_readdir(filp, dirent, filldir);
 		if (err != -EAGAIN)
 			return err;
 	} else {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 	}
 	if (fi->dentry) {
 		err = note_last_dentry(fi, fi->dentry->d_name.name,
@@ -428,12 +428,12 @@ more:
 	 * were released during the whole readdir, and we should have
 	 * the complete dir contents in our cache.
 	 */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (ci->i_release_count == fi->dir_release_count) {
 		ceph_dir_set_complete(inode);
 		ci->i_max_offset = filp->f_pos;
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	dout("readdir %p filp %p done.\n", inode, filp);
 	return 0;
@@ -607,7 +607,7 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
 		struct ceph_inode_info *ci = ceph_inode(dir);
 		struct ceph_dentry_info *di = ceph_dentry(dentry);
 
-		spin_lock(&dir->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		dout(" dir %p flags are %d\n", dir, ci->i_ceph_flags);
 		if (strncmp(dentry->d_name.name,
 			    fsc->mount_options->snapdir_name,
@@ -615,13 +615,13 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
 		    !is_root_ceph_dentry(dir, dentry) &&
 		    ceph_dir_test_complete(dir) &&
 		    (__ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, 1))) {
-			spin_unlock(&dir->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 			dout(" dir %p complete, -ENOENT\n", dir);
 			d_add(dentry, NULL);
 			di->lease_shared_gen = ci->i_shared_gen;
 			return NULL;
 		}
-		spin_unlock(&dir->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 	}
 
 	op = ceph_snap(dir) == CEPH_SNAPDIR ?
@@ -841,12 +841,12 @@ static int drop_caps_for_unlink(struct inode *inode)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (inode->i_nlink == 1) {
 		drop |= ~(__ceph_caps_wanted(ci) | CEPH_CAP_PIN);
 		ci->i_ceph_flags |= CEPH_I_NODELAY;
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return drop;
 }
 
@@ -1015,10 +1015,10 @@ static int dir_lease_is_valid(struct inode *dir, struct dentry *dentry)
 	struct ceph_dentry_info *di = ceph_dentry(dentry);
 	int valid = 0;
 
-	spin_lock(&dir->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (ci->i_shared_gen == di->lease_shared_gen)
 		valid = __ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, 1);
-	spin_unlock(&dir->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	dout("dir_lease_is_valid dir %p v%u dentry %p v%u = %d\n",
 	     dir, (unsigned)ci->i_shared_gen, dentry,
 	     (unsigned)di->lease_shared_gen, valid);
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index ce549d31eeb..ed72428d9c7 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -147,9 +147,9 @@ int ceph_open(struct inode *inode, struct file *file)
 
 	/* trivially open snapdir */
 	if (ceph_snap(inode) == CEPH_SNAPDIR) {
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		__ceph_get_fmode(ci, fmode);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		return ceph_init_file(inode, file, fmode);
 	}
 
@@ -158,7 +158,7 @@ int ceph_open(struct inode *inode, struct file *file)
 	 * write) or any MDS (for read).  Update wanted set
 	 * asynchronously.
 	 */
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (__ceph_is_any_real_caps(ci) &&
 	    (((fmode & CEPH_FILE_MODE_WR) == 0) || ci->i_auth_cap)) {
 		int mds_wanted = __ceph_caps_mds_wanted(ci);
@@ -168,7 +168,7 @@ int ceph_open(struct inode *inode, struct file *file)
 		     inode, fmode, ceph_cap_string(wanted),
 		     ceph_cap_string(issued));
 		__ceph_get_fmode(ci, fmode);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 
 		/* adjust wanted? */
 		if ((issued & wanted) != wanted &&
@@ -180,10 +180,10 @@ int ceph_open(struct inode *inode, struct file *file)
 	} else if (ceph_snap(inode) != CEPH_NOSNAP &&
 		   (ci->i_snap_caps & wanted) == wanted) {
 		__ceph_get_fmode(ci, fmode);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		return ceph_init_file(inode, file, fmode);
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	dout("open fmode %d wants %s\n", fmode, ceph_cap_string(wanted));
 	req = prepare_open_request(inode->i_sb, flags, 0);
@@ -743,9 +743,9 @@ retry_snap:
 		 */
 		int dirty;
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		ceph_put_cap_refs(ci, got);
 
 		ret = generic_file_aio_write(iocb, iov, nr_segs, pos);
@@ -764,9 +764,9 @@ retry_snap:
 
 	if (ret >= 0) {
 		int dirty;
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		if (dirty)
 			__mark_inode_dirty(inode, dirty);
 	}
@@ -797,7 +797,8 @@ static loff_t ceph_llseek(struct file *file, loff_t offset, int origin)
 
 	mutex_lock(&inode->i_mutex);
 	__ceph_do_pending_vmtruncate(inode);
-	if (origin != SEEK_CUR || origin != SEEK_SET) {
+
+	if (origin == SEEK_END || origin == SEEK_DATA || origin == SEEK_HOLE) {
 		ret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE);
 		if (ret < 0) {
 			offset = ret;
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 116f36502f1..87fb132fb33 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -297,6 +297,8 @@ struct inode *ceph_alloc_inode(struct super_block *sb)
 
 	dout("alloc_inode %p\n", &ci->vfs_inode);
 
+	spin_lock_init(&ci->i_ceph_lock);
+
 	ci->i_version = 0;
 	ci->i_time_warp_seq = 0;
 	ci->i_ceph_flags = 0;
@@ -583,7 +585,7 @@ static int fill_inode(struct inode *inode,
 			       iinfo->xattr_len);
 	}
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	/*
 	 * provided version will be odd if inode value is projected,
@@ -680,7 +682,7 @@ static int fill_inode(struct inode *inode,
 			char *sym;
 
 			BUG_ON(symlen != inode->i_size);
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 
 			err = -ENOMEM;
 			sym = kmalloc(symlen+1, GFP_NOFS);
@@ -689,7 +691,7 @@ static int fill_inode(struct inode *inode,
 			memcpy(sym, iinfo->symlink, symlen);
 			sym[symlen] = 0;
 
-			spin_lock(&inode->i_lock);
+			spin_lock(&ci->i_ceph_lock);
 			if (!ci->i_symlink)
 				ci->i_symlink = sym;
 			else
@@ -715,7 +717,7 @@ static int fill_inode(struct inode *inode,
 	}
 
 no_change:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	/* queue truncate if we saw i_size decrease */
 	if (queue_trunc)
@@ -750,13 +752,13 @@ no_change:
 				     info->cap.flags,
 				     caps_reservation);
 		} else {
-			spin_lock(&inode->i_lock);
+			spin_lock(&ci->i_ceph_lock);
 			dout(" %p got snap_caps %s\n", inode,
 			     ceph_cap_string(le32_to_cpu(info->cap.caps)));
 			ci->i_snap_caps |= le32_to_cpu(info->cap.caps);
 			if (cap_fmode >= 0)
 				__ceph_get_fmode(ci, cap_fmode);
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 		}
 	} else if (cap_fmode >= 0) {
 		pr_warning("mds issued no caps on %llx.%llx\n",
@@ -849,19 +851,20 @@ static void ceph_set_dentry_offset(struct dentry *dn)
 {
 	struct dentry *dir = dn->d_parent;
 	struct inode *inode = dir->d_inode;
+	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_dentry_info *di;
 
 	BUG_ON(!inode);
 
 	di = ceph_dentry(dn);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (!ceph_dir_test_complete(inode)) {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		return;
 	}
 	di->offset = ceph_inode(inode)->i_max_offset++;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	spin_lock(&dir->d_lock);
 	spin_lock_nested(&dn->d_lock, DENTRY_D_LOCK_NESTED);
@@ -1308,7 +1311,7 @@ int ceph_inode_set_size(struct inode *inode, loff_t size)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	int ret = 0;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	dout("set_size %p %llu -> %llu\n", inode, inode->i_size, size);
 	inode->i_size = size;
 	inode->i_blocks = (size + (1 << 9) - 1) >> 9;
@@ -1318,7 +1321,7 @@ int ceph_inode_set_size(struct inode *inode, loff_t size)
 	    (ci->i_reported_size << 1) < ci->i_max_size)
 		ret = 1;
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return ret;
 }
 
@@ -1376,20 +1379,20 @@ static void ceph_invalidate_work(struct work_struct *work)
 	u32 orig_gen;
 	int check = 0;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	dout("invalidate_pages %p gen %d revoking %d\n", inode,
 	     ci->i_rdcache_gen, ci->i_rdcache_revoking);
 	if (ci->i_rdcache_revoking != ci->i_rdcache_gen) {
 		/* nevermind! */
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		goto out;
 	}
 	orig_gen = ci->i_rdcache_gen;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	truncate_inode_pages(&inode->i_data, 0);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (orig_gen == ci->i_rdcache_gen &&
 	    orig_gen == ci->i_rdcache_revoking) {
 		dout("invalidate_pages %p gen %d successful\n", inode,
@@ -1401,7 +1404,7 @@ static void ceph_invalidate_work(struct work_struct *work)
 		     inode, orig_gen, ci->i_rdcache_gen,
 		     ci->i_rdcache_revoking);
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (check)
 		ceph_check_caps(ci, 0, NULL);
@@ -1460,10 +1463,10 @@ void __ceph_do_pending_vmtruncate(struct inode *inode)
 	int wrbuffer_refs, wake = 0;
 
 retry:
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	if (ci->i_truncate_pending == 0) {
 		dout("__do_pending_vmtruncate %p none pending\n", inode);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		return;
 	}
 
@@ -1474,7 +1477,7 @@ retry:
 	if (ci->i_wrbuffer_ref_head < ci->i_wrbuffer_ref) {
 		dout("__do_pending_vmtruncate %p flushing snaps first\n",
 		     inode);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		filemap_write_and_wait_range(&inode->i_data, 0,
 					     inode->i_sb->s_maxbytes);
 		goto retry;
@@ -1484,15 +1487,15 @@ retry:
 	wrbuffer_refs = ci->i_wrbuffer_ref;
 	dout("__do_pending_vmtruncate %p (%d) to %lld\n", inode,
 	     ci->i_truncate_pending, to);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	truncate_inode_pages(inode->i_mapping, to);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	ci->i_truncate_pending--;
 	if (ci->i_truncate_pending == 0)
 		wake = 1;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (wrbuffer_refs == 0)
 		ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL);
@@ -1547,7 +1550,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	issued = __ceph_caps_issued(ci, NULL);
 	dout("setattr %p issued %s\n", inode, ceph_cap_string(issued));
 
@@ -1695,7 +1698,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	release &= issued;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (inode_dirty_flags)
 		__mark_inode_dirty(inode, inode_dirty_flags);
@@ -1717,7 +1720,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 	__ceph_do_pending_vmtruncate(inode);
 	return err;
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	ceph_mdsc_put_request(req);
 	return err;
 }
diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c
index 5a14c29cbba..790914a598d 100644
--- a/fs/ceph/ioctl.c
+++ b/fs/ceph/ioctl.c
@@ -241,11 +241,11 @@ static long ceph_ioctl_lazyio(struct file *file)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 
 	if ((fi->fmode & CEPH_FILE_MODE_LAZY) == 0) {
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		ci->i_nr_by_mode[fi->fmode]--;
 		fi->fmode |= CEPH_FILE_MODE_LAZY;
 		ci->i_nr_by_mode[fi->fmode]++;
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		dout("ioctl_layzio: file %p marked lazy\n", file);
 
 		ceph_check_caps(ci, 0, NULL);
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 264ab701154..6203d805eb4 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -732,21 +732,21 @@ static int __choose_mds(struct ceph_mds_client *mdsc,
 		}
 	}
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	cap = NULL;
 	if (mode == USE_AUTH_MDS)
 		cap = ci->i_auth_cap;
 	if (!cap && !RB_EMPTY_ROOT(&ci->i_caps))
 		cap = rb_entry(rb_first(&ci->i_caps), struct ceph_cap, ci_node);
 	if (!cap) {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		goto random;
 	}
 	mds = cap->session->s_mds;
 	dout("choose_mds %p %llx.%llx mds%d (%scap %p)\n",
 	     inode, ceph_vinop(inode), mds,
 	     cap == ci->i_auth_cap ? "auth " : "", cap);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return mds;
 
 random:
@@ -951,7 +951,7 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
 
 	dout("removing cap %p, ci is %p, inode is %p\n",
 	     cap, ci, &ci->vfs_inode);
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	__ceph_remove_cap(cap);
 	if (!__ceph_is_any_real_caps(ci)) {
 		struct ceph_mds_client *mdsc =
@@ -984,7 +984,7 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
 		}
 		spin_unlock(&mdsc->cap_dirty_lock);
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	while (drop--)
 		iput(inode);
 	return 0;
@@ -1015,10 +1015,10 @@ static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap,
 
 	wake_up_all(&ci->i_cap_wq);
 	if (arg) {
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		ci->i_wanted_max_size = 0;
 		ci->i_requested_max_size = 0;
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 	}
 	return 0;
 }
@@ -1151,7 +1151,7 @@ static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
 	if (session->s_trim_caps <= 0)
 		return -1;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	mine = cap->issued | cap->implemented;
 	used = __ceph_caps_used(ci);
 	oissued = __ceph_caps_issued_other(ci, cap);
@@ -1170,7 +1170,7 @@ static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
 		__ceph_remove_cap(cap);
 	} else {
 		/* try to drop referring dentries */
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		d_prune_aliases(inode);
 		dout("trim_caps_cb %p cap %p  pruned, count now %d\n",
 		     inode, cap, atomic_read(&inode->i_count));
@@ -1178,7 +1178,7 @@ static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
 	}
 
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return 0;
 }
 
@@ -1296,7 +1296,7 @@ static int check_cap_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq)
 					   i_flushing_item);
 			struct inode *inode = &ci->vfs_inode;
 
-			spin_lock(&inode->i_lock);
+			spin_lock(&ci->i_ceph_lock);
 			if (ci->i_cap_flush_seq <= want_flush_seq) {
 				dout("check_cap_flush still flushing %p "
 				     "seq %lld <= %lld to mds%d\n", inode,
@@ -1304,7 +1304,7 @@ static int check_cap_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq)
 				     session->s_mds);
 				ret = 0;
 			}
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 		}
 		mutex_unlock(&session->s_mutex);
 		ceph_put_mds_session(session);
@@ -1495,6 +1495,7 @@ retry:
 			     pos, temp);
 		} else if (stop_on_nosnap && inode &&
 			   ceph_snap(inode) == CEPH_NOSNAP) {
+			spin_unlock(&temp->d_lock);
 			break;
 		} else {
 			pos -= temp->d_name.len;
@@ -2011,10 +2012,10 @@ void ceph_invalidate_dir_request(struct ceph_mds_request *req)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 
 	dout("invalidate_dir_request %p (D_COMPLETE, lease(s))\n", inode);
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	ceph_dir_clear_complete(inode);
 	ci->i_release_count++;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (req->r_dentry)
 		ceph_invalidate_dentry_lease(req->r_dentry);
@@ -2422,7 +2423,7 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
 	if (err)
 		goto out_free;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	cap->seq = 0;        /* reset cap seq */
 	cap->issue_seq = 0;  /* and issue_seq */
 
@@ -2445,7 +2446,7 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
 		rec.v1.pathbase = cpu_to_le64(pathbase);
 		reclen = sizeof(rec.v1);
 	}
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 
 	if (recon_state->flock) {
 		int num_fcntl_locks, num_flock_locks;
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index 4bb239921db..a50ca0e3947 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -20,7 +20,7 @@
  *
  *         mdsc->snap_rwsem
  *
- *         inode->i_lock
+ *         ci->i_ceph_lock
  *                 mdsc->snap_flush_lock
  *                 mdsc->cap_delay_lock
  *
diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index e2643719133..a559c80f127 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -446,7 +446,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
 		return;
 	}
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	used = __ceph_caps_used(ci);
 	dirty = __ceph_caps_dirty(ci);
 
@@ -528,7 +528,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
 		kfree(capsnap);
 	}
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 /*
@@ -537,7 +537,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
  *
  * If capsnap can now be flushed, add to snap_flush list, and return 1.
  *
- * Caller must hold i_lock.
+ * Caller must hold i_ceph_lock.
  */
 int __ceph_finish_cap_snap(struct ceph_inode_info *ci,
 			    struct ceph_cap_snap *capsnap)
@@ -739,9 +739,9 @@ static void flush_snaps(struct ceph_mds_client *mdsc)
 		inode = &ci->vfs_inode;
 		ihold(inode);
 		spin_unlock(&mdsc->snap_flush_lock);
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		__ceph_flush_snaps(ci, &session, 0);
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		iput(inode);
 		spin_lock(&mdsc->snap_flush_lock);
 	}
@@ -847,7 +847,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
 				continue;
 			ci = ceph_inode(inode);
 
-			spin_lock(&inode->i_lock);
+			spin_lock(&ci->i_ceph_lock);
 			if (!ci->i_snap_realm)
 				goto skip_inode;
 			/*
@@ -876,7 +876,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
 			oldrealm = ci->i_snap_realm;
 			ci->i_snap_realm = realm;
 			spin_unlock(&realm->inodes_with_caps_lock);
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 
 			ceph_get_snap_realm(mdsc, realm);
 			ceph_put_snap_realm(mdsc, oldrealm);
@@ -885,7 +885,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
 			continue;
 
 skip_inode:
-			spin_unlock(&inode->i_lock);
+			spin_unlock(&ci->i_ceph_lock);
 			iput(inode);
 		}
 
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index 8dc73a594a9..b48f15f101a 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -383,7 +383,7 @@ static int ceph_show_options(struct seq_file *m, struct vfsmount *mnt)
 	if (fsopt->rsize != CEPH_RSIZE_DEFAULT)
 		seq_printf(m, ",rsize=%d", fsopt->rsize);
 	if (fsopt->rasize != CEPH_RASIZE_DEFAULT)
-		seq_printf(m, ",rasize=%d", fsopt->rsize);
+		seq_printf(m, ",rasize=%d", fsopt->rasize);
 	if (fsopt->congestion_kb != default_congestion_kb())
 		seq_printf(m, ",write_congestion_kb=%d", fsopt->congestion_kb);
 	if (fsopt->caps_wanted_delay_min != CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT)
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 01bf189e08a..edcbf3774a5 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -220,7 +220,7 @@ struct ceph_dentry_info {
  * The locking for D_COMPLETE is a bit odd:
  *  - we can clear it at almost any time (see ceph_d_prune)
  *  - it is only meaningful if:
- *    - we hold dir inode i_lock
+ *    - we hold dir inode i_ceph_lock
  *    - we hold dir FILE_SHARED caps
  *    - the dentry D_COMPLETE is set
  */
@@ -250,6 +250,8 @@ struct ceph_inode_xattrs_info {
 struct ceph_inode_info {
 	struct ceph_vino i_vino;   /* ceph ino + snap */
 
+	spinlock_t i_ceph_lock;
+
 	u64 i_version;
 	u32 i_time_warp_seq;
 
@@ -271,7 +273,7 @@ struct ceph_inode_info {
 
 	struct ceph_inode_xattrs_info i_xattrs;
 
-	/* capabilities.  protected _both_ by i_lock and cap->session's
+	/* capabilities.  protected _both_ by i_ceph_lock and cap->session's
 	 * s_mutex. */
 	struct rb_root i_caps;           /* cap list */
 	struct ceph_cap *i_auth_cap;     /* authoritative cap, if any */
@@ -437,18 +439,18 @@ static inline void ceph_i_clear(struct inode *inode, unsigned mask)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	ci->i_ceph_flags &= ~mask;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 static inline void ceph_i_set(struct inode *inode, unsigned mask)
 {
 	struct ceph_inode_info *ci = ceph_inode(inode);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	ci->i_ceph_flags |= mask;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 }
 
 static inline bool ceph_i_test(struct inode *inode, unsigned mask)
@@ -456,9 +458,9 @@ static inline bool ceph_i_test(struct inode *inode, unsigned mask)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	bool r;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	r = (ci->i_ceph_flags & mask) == mask;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return r;
 }
 
@@ -508,9 +510,9 @@ extern int __ceph_caps_issued_other(struct ceph_inode_info *ci,
 static inline int ceph_caps_issued(struct ceph_inode_info *ci)
 {
 	int issued;
-	spin_lock(&ci->vfs_inode.i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	issued = __ceph_caps_issued(ci, NULL);
-	spin_unlock(&ci->vfs_inode.i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return issued;
 }
 
@@ -518,9 +520,9 @@ static inline int ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask,
 					int touch)
 {
 	int r;
-	spin_lock(&ci->vfs_inode.i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	r = __ceph_caps_issued_mask(ci, mask, touch);
-	spin_unlock(&ci->vfs_inode.i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return r;
 }
 
@@ -743,10 +745,9 @@ extern int ceph_add_cap(struct inode *inode,
 extern void __ceph_remove_cap(struct ceph_cap *cap);
 static inline void ceph_remove_cap(struct ceph_cap *cap)
 {
-	struct inode *inode = &cap->ci->vfs_inode;
-	spin_lock(&inode->i_lock);
+	spin_lock(&cap->ci->i_ceph_lock);
 	__ceph_remove_cap(cap);
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&cap->ci->i_ceph_lock);
 }
 extern void ceph_put_cap(struct ceph_mds_client *mdsc,
 			 struct ceph_cap *cap);
diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c
index 96c6739a028..a5e36e4488a 100644
--- a/fs/ceph/xattr.c
+++ b/fs/ceph/xattr.c
@@ -343,8 +343,8 @@ void __ceph_destroy_xattrs(struct ceph_inode_info *ci)
 }
 
 static int __build_xattrs(struct inode *inode)
-	__releases(inode->i_lock)
-	__acquires(inode->i_lock)
+	__releases(ci->i_ceph_lock)
+	__acquires(ci->i_ceph_lock)
 {
 	u32 namelen;
 	u32 numattr = 0;
@@ -372,7 +372,7 @@ start:
 		end = p + ci->i_xattrs.blob->vec.iov_len;
 		ceph_decode_32_safe(&p, end, numattr, bad);
 		xattr_version = ci->i_xattrs.version;
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 
 		xattrs = kcalloc(numattr, sizeof(struct ceph_xattr *),
 				 GFP_NOFS);
@@ -387,7 +387,7 @@ start:
 				goto bad_lock;
 		}
 
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		if (ci->i_xattrs.version != xattr_version) {
 			/* lost a race, retry */
 			for (i = 0; i < numattr; i++)
@@ -418,7 +418,7 @@ start:
 
 	return err;
 bad_lock:
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 bad:
 	if (xattrs) {
 		for (i = 0; i < numattr; i++)
@@ -512,7 +512,7 @@ ssize_t ceph_getxattr(struct dentry *dentry, const char *name, void *value,
 	if (vxattrs)
 		vxattr = ceph_match_vxattr(vxattrs, name);
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	dout("getxattr %p ver=%lld index_ver=%lld\n", inode,
 	     ci->i_xattrs.version, ci->i_xattrs.index_version);
 
@@ -520,14 +520,14 @@ ssize_t ceph_getxattr(struct dentry *dentry, const char *name, void *value,
 	    (ci->i_xattrs.index_version >= ci->i_xattrs.version)) {
 		goto get_xattr;
 	} else {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		/* get xattrs from mds (if we don't already have them) */
 		err = ceph_do_getattr(inode, CEPH_STAT_CAP_XATTR);
 		if (err)
 			return err;
 	}
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	if (vxattr && vxattr->readonly) {
 		err = vxattr->getxattr_cb(ci, value, size);
@@ -558,7 +558,7 @@ get_xattr:
 	memcpy(value, xattr->val, xattr->val_len);
 
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return err;
 }
 
@@ -573,7 +573,7 @@ ssize_t ceph_listxattr(struct dentry *dentry, char *names, size_t size)
 	u32 len;
 	int i;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	dout("listxattr %p ver=%lld index_ver=%lld\n", inode,
 	     ci->i_xattrs.version, ci->i_xattrs.index_version);
 
@@ -581,13 +581,13 @@ ssize_t ceph_listxattr(struct dentry *dentry, char *names, size_t size)
 	    (ci->i_xattrs.index_version >= ci->i_xattrs.version)) {
 		goto list_xattr;
 	} else {
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		err = ceph_do_getattr(inode, CEPH_STAT_CAP_XATTR);
 		if (err)
 			return err;
 	}
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 
 	err = __build_xattrs(inode);
 	if (err < 0)
@@ -619,7 +619,7 @@ list_xattr:
 		}
 
 out:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	return err;
 }
 
@@ -739,7 +739,7 @@ int ceph_setxattr(struct dentry *dentry, const char *name,
 	if (!xattr)
 		goto out;
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 retry:
 	issued = __ceph_caps_issued(ci, NULL);
 	if (!(issued & CEPH_CAP_XATTR_EXCL))
@@ -752,12 +752,12 @@ retry:
 	    required_blob_size > ci->i_xattrs.prealloc_blob->alloc_len) {
 		struct ceph_buffer *blob = NULL;
 
-		spin_unlock(&inode->i_lock);
+		spin_unlock(&ci->i_ceph_lock);
 		dout(" preaallocating new blob size=%d\n", required_blob_size);
 		blob = ceph_buffer_new(required_blob_size, GFP_NOFS);
 		if (!blob)
 			goto out;
-		spin_lock(&inode->i_lock);
+		spin_lock(&ci->i_ceph_lock);
 		if (ci->i_xattrs.prealloc_blob)
 			ceph_buffer_put(ci->i_xattrs.prealloc_blob);
 		ci->i_xattrs.prealloc_blob = blob;
@@ -770,13 +770,13 @@ retry:
 	dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL);
 	ci->i_xattrs.dirty = true;
 	inode->i_ctime = CURRENT_TIME;
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	if (dirty)
 		__mark_inode_dirty(inode, dirty);
 	return err;
 
 do_sync:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	err = ceph_sync_setxattr(dentry, name, value, size, flags);
 out:
 	kfree(newname);
@@ -833,7 +833,7 @@ int ceph_removexattr(struct dentry *dentry, const char *name)
 			return -EOPNOTSUPP;
 	}
 
-	spin_lock(&inode->i_lock);
+	spin_lock(&ci->i_ceph_lock);
 	__build_xattrs(inode);
 	issued = __ceph_caps_issued(ci, NULL);
 	dout("removexattr %p issued %s\n", inode, ceph_cap_string(issued));
@@ -846,12 +846,12 @@ int ceph_removexattr(struct dentry *dentry, const char *name)
 	ci->i_xattrs.dirty = true;
 	inode->i_ctime = CURRENT_TIME;
 
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	if (dirty)
 		__mark_inode_dirty(inode, dirty);
 	return err;
 do_sync:
-	spin_unlock(&inode->i_lock);
+	spin_unlock(&ci->i_ceph_lock);
 	err = ceph_send_removexattr(dentry, name);
 	return err;
 }
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/configfs/inode.c b/fs/configfs/inode.c
index ca418aaf635..9d8715c45f2 100644
--- a/fs/configfs/inode.c
+++ b/fs/configfs/inode.c
@@ -292,7 +292,7 @@ int __init configfs_inode_init(void)
 	return bdi_init(&configfs_backing_dev_info);
 }
 
-void __exit configfs_inode_exit(void)
+void configfs_inode_exit(void)
 {
 	bdi_destroy(&configfs_backing_dev_info);
 }
diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c
index ecc62178bed..276e15cafd5 100644
--- a/fs/configfs/mount.c
+++ b/fs/configfs/mount.c
@@ -143,28 +143,26 @@ static int __init configfs_init(void)
 		goto out;
 
 	config_kobj = kobject_create_and_add("config", kernel_kobj);
-	if (!config_kobj) {
-		kmem_cache_destroy(configfs_dir_cachep);
-		configfs_dir_cachep = NULL;
-		goto out;
-	}
+	if (!config_kobj)
+		goto out2;
+
+	err = configfs_inode_init();
+	if (err)
+		goto out3;
 
 	err = register_filesystem(&configfs_fs_type);
-	if (err) {
-		printk(KERN_ERR "configfs: Unable to register filesystem!\n");
-		kobject_put(config_kobj);
-		kmem_cache_destroy(configfs_dir_cachep);
-		configfs_dir_cachep = NULL;
-		goto out;
-	}
+	if (err)
+		goto out4;
 
-	err = configfs_inode_init();
-	if (err) {
-		unregister_filesystem(&configfs_fs_type);
-		kobject_put(config_kobj);
-		kmem_cache_destroy(configfs_dir_cachep);
-		configfs_dir_cachep = NULL;
-	}
+	return 0;
+out4:
+	printk(KERN_ERR "configfs: Unable to register filesystem!\n");
+	configfs_inode_exit();
+out3:
+	kobject_put(config_kobj);
+out2:
+	kmem_cache_destroy(configfs_dir_cachep);
+	configfs_dir_cachep = NULL;
 out:
 	return err;
 }
diff --git a/fs/dcache.c b/fs/dcache.c
index 10ba92def3f..89509b5a090 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -2439,16 +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;
@@ -2483,10 +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;
 
@@ -2500,15 +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
  *
@@ -2519,10 +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;
@@ -2533,7 +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;
 }
@@ -2541,8 +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)) {
@@ -2579,7 +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;
 
 	/*
@@ -2594,9 +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);
@@ -2617,7 +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)
@@ -2625,9 +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);
@@ -2758,19 +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/extents.c b/fs/ext4/extents.c
index 61fa9e1614a..607b1557d29 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -1095,7 +1095,7 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
 		  le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
 		  ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
 
-	neh->eh_depth = cpu_to_le16(neh->eh_depth + 1);
+	neh->eh_depth = cpu_to_le16(le16_to_cpu(neh->eh_depth) + 1);
 	ext4_mark_inode_dirty(handle, inode);
 out:
 	brelse(bh);
@@ -2955,7 +2955,6 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 	/* Pre-conditions */
 	BUG_ON(!ext4_ext_is_uninitialized(ex));
 	BUG_ON(!in_range(map->m_lblk, ee_block, ee_len));
-	BUG_ON(map->m_lblk + map->m_len > ee_block + ee_len);
 
 	/*
 	 * Attempt to transfer newly initialized blocks from the currently
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index fffec40d599..92655fd8965 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1339,8 +1339,11 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd,
 					clear_buffer_unwritten(bh);
 				}
 
-				/* skip page if block allocation undone */
-				if (buffer_delay(bh) || buffer_unwritten(bh))
+				/*
+				 * skip page if block allocation undone and
+				 * block is dirty
+				 */
+				if (ext4_bh_delay_or_unwritten(NULL, bh))
 					skip_page = 1;
 				bh = bh->b_this_page;
 				block_start += bh->b_size;
@@ -2387,7 +2390,6 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 	pgoff_t index;
 	struct inode *inode = mapping->host;
 	handle_t *handle;
-	loff_t page_len;
 
 	index = pos >> PAGE_CACHE_SHIFT;
 
@@ -2434,13 +2436,6 @@ retry:
 		 */
 		if (pos + len > inode->i_size)
 			ext4_truncate_failed_write(inode);
-	} else {
-		page_len = pos & (PAGE_CACHE_SIZE - 1);
-		if (page_len > 0) {
-			ret = ext4_discard_partial_page_buffers_no_lock(handle,
-				inode, page, pos - page_len, page_len,
-				EXT4_DISCARD_PARTIAL_PG_ZERO_UNMAPPED);
-		}
 	}
 
 	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
@@ -2483,7 +2478,6 @@ static int ext4_da_write_end(struct file *file,
 	loff_t new_i_size;
 	unsigned long start, end;
 	int write_mode = (int)(unsigned long)fsdata;
-	loff_t page_len;
 
 	if (write_mode == FALL_BACK_TO_NONDELALLOC) {
 		if (ext4_should_order_data(inode)) {
@@ -2508,7 +2502,7 @@ static int ext4_da_write_end(struct file *file,
 	 */
 
 	new_i_size = pos + copied;
-	if (new_i_size > EXT4_I(inode)->i_disksize) {
+	if (copied && new_i_size > EXT4_I(inode)->i_disksize) {
 		if (ext4_da_should_update_i_disksize(page, end)) {
 			down_write(&EXT4_I(inode)->i_data_sem);
 			if (new_i_size > EXT4_I(inode)->i_disksize) {
@@ -2532,16 +2526,6 @@ static int ext4_da_write_end(struct file *file,
 	}
 	ret2 = generic_write_end(file, mapping, pos, len, copied,
 							page, fsdata);
-
-	page_len = PAGE_CACHE_SIZE -
-			((pos + copied - 1) & (PAGE_CACHE_SIZE - 1));
-
-	if (page_len > 0) {
-		ret = ext4_discard_partial_page_buffers_no_lock(handle,
-			inode, page, pos + copied - 1, page_len,
-			EXT4_DISCARD_PARTIAL_PG_ZERO_UNMAPPED);
-	}
-
 	copied = ret2;
 	if (ret2 < 0)
 		ret = ret2;
@@ -2781,10 +2765,11 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
  		  iocb->private, io_end->inode->i_ino, iocb, offset,
 		  size);
 
+	iocb->private = NULL;
+
 	/* if not aio dio with unwritten extents, just free io and return */
 	if (!(io_end->flag & EXT4_IO_END_UNWRITTEN)) {
 		ext4_free_io_end(io_end);
-		iocb->private = NULL;
 out:
 		if (is_async)
 			aio_complete(iocb, ret, 0);
@@ -2808,7 +2793,6 @@ out:
 
 	/* queue the work to convert unwritten extents to written */
 	queue_work(wq, &io_end->work);
-	iocb->private = NULL;
 
 	/* XXX: probably should move into the real I/O completion handler */
 	inode_dio_done(inode);
@@ -3203,26 +3187,8 @@ int ext4_discard_partial_page_buffers_no_lock(handle_t *handle,
 
 	iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);
 
-	if (!page_has_buffers(page)) {
-		/*
-		 * If the range to be discarded covers a partial block
-		 * we need to get the page buffers.  This is because
-		 * partial blocks cannot be released and the page needs
-		 * to be updated with the contents of the block before
-		 * we write the zeros on top of it.
-		 */
-		if ((from & (blocksize - 1)) ||
-		    ((from + length) & (blocksize - 1))) {
-			create_empty_buffers(page, blocksize, 0);
-		} else {
-			/*
-			 * If there are no partial blocks,
-			 * there is nothing to update,
-			 * so we can return now
-			 */
-			return 0;
-		}
-	}
+	if (!page_has_buffers(page))
+		create_empty_buffers(page, blocksize, 0);
 
 	/* Find the buffer that contains "offset" */
 	bh = page_buffers(page);
diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c
index 7ce1d0b19c9..7e106c810c6 100644
--- a/fs/ext4/page-io.c
+++ b/fs/ext4/page-io.c
@@ -385,6 +385,18 @@ int ext4_bio_write_page(struct ext4_io_submit *io,
 
 		block_end = block_start + blocksize;
 		if (block_start >= len) {
+			/*
+			 * Comments copied from block_write_full_page_endio:
+			 *
+			 * The page straddles i_size.  It must be zeroed out on
+			 * each and every writepage invocation because it may
+			 * be mmapped.  "A file is mapped in multiples of the
+			 * page size.  For a file that is not a multiple of
+			 * the  page size, the remaining memory is zeroed when
+			 * mapped, and writes to that region are not written
+			 * out to the file."
+			 */
+			zero_user_segment(page, block_start, block_end);
 			clear_buffer_dirty(bh);
 			set_buffer_uptodate(bh);
 			continue;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 1c7bbd00e7e..d0666c8d15f 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1155,9 +1155,9 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
 		seq_puts(seq, ",block_validity");
 
 	if (!test_opt(sb, INIT_INODE_TABLE))
-		seq_puts(seq, ",noinit_inode_table");
+		seq_puts(seq, ",noinit_itable");
 	else if (sbi->s_li_wait_mult != EXT4_DEF_LI_WAIT_MULT)
-		seq_printf(seq, ",init_inode_table=%u",
+		seq_printf(seq, ",init_itable=%u",
 			   (unsigned) sbi->s_li_wait_mult);
 
 	ext4_show_quota_options(seq, sb);
@@ -1333,8 +1333,7 @@ enum {
 	Opt_nomblk_io_submit, Opt_block_validity, Opt_noblock_validity,
 	Opt_inode_readahead_blks, Opt_journal_ioprio,
 	Opt_dioread_nolock, Opt_dioread_lock,
-	Opt_discard, Opt_nodiscard,
-	Opt_init_inode_table, Opt_noinit_inode_table,
+	Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
 };
 
 static const match_table_t tokens = {
@@ -1407,9 +1406,9 @@ static const match_table_t tokens = {
 	{Opt_dioread_lock, "dioread_lock"},
 	{Opt_discard, "discard"},
 	{Opt_nodiscard, "nodiscard"},
-	{Opt_init_inode_table, "init_itable=%u"},
-	{Opt_init_inode_table, "init_itable"},
-	{Opt_noinit_inode_table, "noinit_itable"},
+	{Opt_init_itable, "init_itable=%u"},
+	{Opt_init_itable, "init_itable"},
+	{Opt_noinit_itable, "noinit_itable"},
 	{Opt_err, NULL},
 };
 
@@ -1892,7 +1891,7 @@ set_qf_format:
 		case Opt_dioread_lock:
 			clear_opt(sb, DIOREAD_NOLOCK);
 			break;
-		case Opt_init_inode_table:
+		case Opt_init_itable:
 			set_opt(sb, INIT_INODE_TABLE);
 			if (args[0].from) {
 				if (match_int(&args[0], &option))
@@ -1903,7 +1902,7 @@ set_qf_format:
 				return 0;
 			sbi->s_li_wait_mult = option;
 			break;
-		case Opt_noinit_inode_table:
+		case Opt_noinit_itable:
 			clear_opt(sb, INIT_INODE_TABLE);
 			break;
 		default:
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 271fde50f0e..e4c7af393c2 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -156,6 +156,7 @@ __bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages,
  * bdi_start_writeback - start writeback
  * @bdi: the backing device to write from
  * @nr_pages: the number of pages to write
+ * @reason: reason why some writeback work was initiated
  *
  * Description:
  *   This does WB_SYNC_NONE opportunistic writeback. The IO is only
@@ -1221,6 +1222,7 @@ static void wait_sb_inodes(struct super_block *sb)
  * writeback_inodes_sb_nr -	writeback dirty inodes from given super_block
  * @sb: the superblock
  * @nr: the number of pages to write
+ * @reason: reason why some writeback work initiated
  *
  * Start writeback on some inodes on this super_block. No guarantees are made
  * on how many (if any) will be written, and this function does not wait
@@ -1249,6 +1251,7 @@ EXPORT_SYMBOL(writeback_inodes_sb_nr);
 /**
  * writeback_inodes_sb	-	writeback dirty inodes from given super_block
  * @sb: the superblock
+ * @reason: reason why some writeback work was initiated
  *
  * Start writeback on some inodes on this super_block. No guarantees are made
  * on how many (if any) will be written, and this function does not wait
@@ -1263,6 +1266,7 @@ EXPORT_SYMBOL(writeback_inodes_sb);
 /**
  * writeback_inodes_sb_if_idle	-	start writeback if none underway
  * @sb: the superblock
+ * @reason: reason why some writeback work was initiated
  *
  * Invoke writeback_inodes_sb if no writeback is currently underway.
  * Returns 1 if writeback was started, 0 if not.
@@ -1283,6 +1287,7 @@ EXPORT_SYMBOL(writeback_inodes_sb_if_idle);
  * writeback_inodes_sb_if_idle	-	start writeback if none underway
  * @sb: the superblock
  * @nr: the number of pages to write
+ * @reason: reason why some writeback work was initiated
  *
  * Invoke writeback_inodes_sb if no writeback is currently underway.
  * Returns 1 if writeback was started, 0 if not.
diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index 5cb8614508c..2aaf3eaaf13 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -1512,7 +1512,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode,
 	else if (outarg->offset + num > file_size)
 		num = file_size - outarg->offset;
 
-	while (num) {
+	while (num && req->num_pages < FUSE_MAX_PAGES_PER_REQ) {
 		struct page *page;
 		unsigned int this_num;
 
@@ -1526,6 +1526,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode,
 
 		num -= this_num;
 		total_len += this_num;
+		index++;
 	}
 	req->misc.retrieve_in.offset = outarg->offset;
 	req->misc.retrieve_in.size = total_len;
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 594f07a81c2..0c84100acd4 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -1556,7 +1556,7 @@ static loff_t fuse_file_llseek(struct file *file, loff_t offset, int origin)
 	struct inode *inode = file->f_path.dentry->d_inode;
 
 	mutex_lock(&inode->i_mutex);
-	if (origin != SEEK_CUR || origin != SEEK_SET) {
+	if (origin != SEEK_CUR && origin != SEEK_SET) {
 		retval = fuse_update_attributes(inode, NULL, file, NULL);
 		if (retval)
 			goto exit;
@@ -1567,6 +1567,10 @@ static loff_t fuse_file_llseek(struct file *file, loff_t offset, int origin)
 		offset += i_size_read(inode);
 		break;
 	case SEEK_CUR:
+		if (offset == 0) {
+			retval = file->f_pos;
+			goto exit;
+		}
 		offset += file->f_pos;
 		break;
 	case SEEK_DATA:
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index 3e6d7275647..aa83109b943 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -1138,28 +1138,28 @@ static int __init fuse_fs_init(void)
 {
 	int err;
 
-	err = register_filesystem(&fuse_fs_type);
-	if (err)
-		goto out;
-
-	err = register_fuseblk();
-	if (err)
-		goto out_unreg;
-
 	fuse_inode_cachep = kmem_cache_create("fuse_inode",
 					      sizeof(struct fuse_inode),
 					      0, SLAB_HWCACHE_ALIGN,
 					      fuse_inode_init_once);
 	err = -ENOMEM;
 	if (!fuse_inode_cachep)
-		goto out_unreg2;
+		goto out;
+
+	err = register_fuseblk();
+	if (err)
+		goto out2;
+
+	err = register_filesystem(&fuse_fs_type);
+	if (err)
+		goto out3;
 
 	return 0;
 
- out_unreg2:
+ out3:
 	unregister_fuseblk();
- out_unreg:
-	unregister_filesystem(&fuse_fs_type);
+ out2:
+	kmem_cache_destroy(fuse_inode_cachep);
  out:
 	return err;
 }
diff --git a/fs/namespace.c b/fs/namespace.c
index 6d3a1963879..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);
 
@@ -2776,3 +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/ncpfs/inode.c b/fs/ncpfs/inode.c
index 5b5fa33b6b9..cbd1a61c110 100644
--- a/fs/ncpfs/inode.c
+++ b/fs/ncpfs/inode.c
@@ -548,7 +548,7 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent)
 
 	error = bdi_setup_and_register(&server->bdi, "ncpfs", BDI_CAP_MAP_COPY);
 	if (error)
-		goto out_bdi;
+		goto out_fput;
 
 	server->ncp_filp = ncp_filp;
 	server->ncp_sock = sock;
@@ -559,7 +559,7 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent)
 		error = -EBADF;
 		server->info_filp = fget(data.info_fd);
 		if (!server->info_filp)
-			goto out_fput;
+			goto out_bdi;
 		error = -ENOTSOCK;
 		sock_inode = server->info_filp->f_path.dentry->d_inode;
 		if (!S_ISSOCK(sock_inode->i_mode))
@@ -746,9 +746,9 @@ out_nls:
 out_fput2:
 	if (server->info_filp)
 		fput(server->info_filp);
-out_fput:
-	bdi_destroy(&server->bdi);
 out_bdi:
+	bdi_destroy(&server->bdi);
+out_fput:
 	/* 23/12/1998 Marcin Dalecki <dalecki@cs.net.pl>:
 	 * 
 	 * The previously used put_filp(ncp_filp); was bogus, since
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index eca56d4b39c..606ef0f20ae 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -147,7 +147,7 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 	 * origin == SEEK_END || SEEK_DATA || SEEK_HOLE => we must revalidate
 	 * the cached file length
 	 */
-	if (origin != SEEK_SET || origin != SEEK_CUR) {
+	if (origin != SEEK_SET && origin != SEEK_CUR) {
 		struct inode *inode = filp->f_mapping->host;
 
 		int retval = nfs_revalidate_file_size(inode, filp);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b28bb19b04f..dcda0ba7af6 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -39,6 +39,8 @@
 #include <linux/delay.h>
 #include <linux/errno.h>
 #include <linux/string.h>
+#include <linux/ratelimit.h>
+#include <linux/printk.h>
 #include <linux/slab.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/gss_api.h>
@@ -895,6 +897,8 @@ out:
 
 static int can_open_delegated(struct nfs_delegation *delegation, fmode_t fmode)
 {
+	if (delegation == NULL)
+		return 0;
 	if ((delegation->type & fmode) != fmode)
 		return 0;
 	if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags))
@@ -1037,8 +1041,7 @@ static struct nfs4_state *nfs4_try_open_cached(struct nfs4_opendata *opendata)
 		}
 		rcu_read_lock();
 		delegation = rcu_dereference(nfsi->delegation);
-		if (delegation == NULL ||
-		    !can_open_delegated(delegation, fmode)) {
+		if (!can_open_delegated(delegation, fmode)) {
 			rcu_read_unlock();
 			break;
 		}
@@ -1092,7 +1095,12 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data
 		if (delegation)
 			delegation_flags = delegation->flags;
 		rcu_read_unlock();
-		if ((delegation_flags & 1UL<<NFS_DELEGATION_NEED_RECLAIM) == 0)
+		if (data->o_arg.claim == NFS4_OPEN_CLAIM_DELEGATE_CUR) {
+			pr_err_ratelimited("NFS: Broken NFSv4 server %s is "
+					"returning a delegation for "
+					"OPEN(CLAIM_DELEGATE_CUR)\n",
+					NFS_CLIENT(inode)->cl_server);
+		} else if ((delegation_flags & 1UL<<NFS_DELEGATION_NEED_RECLAIM) == 0)
 			nfs_inode_set_delegation(state->inode,
 					data->owner->so_cred,
 					&data->o_res);
@@ -1424,11 +1432,9 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata)
 			goto out_no_action;
 		rcu_read_lock();
 		delegation = rcu_dereference(NFS_I(data->state->inode)->delegation);
-		if (delegation != NULL &&
-		    test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) == 0) {
-			rcu_read_unlock();
-			goto out_no_action;
-		}
+		if (data->o_arg.claim != NFS4_OPEN_CLAIM_DELEGATE_CUR &&
+		    can_open_delegated(delegation, data->o_arg.fmode))
+			goto unlock_no_action;
 		rcu_read_unlock();
 	}
 	/* Update sequence id. */
@@ -1445,6 +1451,8 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata)
 		return;
 	rpc_call_start(task);
 	return;
+unlock_no_action:
+	rcu_read_unlock();
 out_no_action:
 	task->tk_action = NULL;
 
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 39914be40b0..6a7107ae6b7 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1156,11 +1156,13 @@ restart:
 		if (status >= 0) {
 			status = nfs4_reclaim_locks(state, ops);
 			if (status >= 0) {
+				spin_lock(&state->state_lock);
 				list_for_each_entry(lock, &state->lock_states, ls_locks) {
 					if (!(lock->ls_flags & NFS_LOCK_INITIALIZED))
 						printk("%s: Lock reclaim failed!\n",
 							__func__);
 				}
+				spin_unlock(&state->state_lock);
 				nfs4_put_open_state(state);
 				goto restart;
 			}
@@ -1224,10 +1226,12 @@ static void nfs4_clear_open_state(struct nfs4_state *state)
 	clear_bit(NFS_O_RDONLY_STATE, &state->flags);
 	clear_bit(NFS_O_WRONLY_STATE, &state->flags);
 	clear_bit(NFS_O_RDWR_STATE, &state->flags);
+	spin_lock(&state->state_lock);
 	list_for_each_entry(lock, &state->lock_states, ls_locks) {
 		lock->ls_seqid.flags = 0;
 		lock->ls_flags &= ~NFS_LOCK_INITIALIZED;
 	}
+	spin_unlock(&state->state_lock);
 }
 
 static void nfs4_reset_seqids(struct nfs_server *server,
@@ -1350,12 +1354,14 @@ static void nfs4_warn_keyexpired(const char *s)
 static int nfs4_recovery_handle_error(struct nfs_client *clp, int error)
 {
 	switch (error) {
+		case 0:
+			break;
 		case -NFS4ERR_CB_PATH_DOWN:
 			nfs_handle_cb_pathdown(clp);
-			return 0;
+			break;
 		case -NFS4ERR_NO_GRACE:
 			nfs4_state_end_reclaim_reboot(clp);
-			return 0;
+			break;
 		case -NFS4ERR_STALE_CLIENTID:
 		case -NFS4ERR_LEASE_MOVED:
 			set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
@@ -1375,13 +1381,15 @@ static int nfs4_recovery_handle_error(struct nfs_client *clp, int error)
 		case -NFS4ERR_SEQ_MISORDERED:
 			set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state);
 			/* Zero session reset errors */
-			return 0;
+			break;
 		case -EKEYEXPIRED:
 			/* Nothing we can do */
 			nfs4_warn_keyexpired(clp->cl_hostname);
-			return 0;
+			break;
+		default:
+			return error;
 	}
-	return error;
+	return 0;
 }
 
 static int nfs4_do_reclaim(struct nfs_client *clp, const struct nfs4_state_recovery_ops *ops)
@@ -1428,7 +1436,7 @@ static int nfs4_check_lease(struct nfs_client *clp)
 	struct rpc_cred *cred;
 	const struct nfs4_state_maintenance_ops *ops =
 		clp->cl_mvops->state_renewal_ops;
-	int status = -NFS4ERR_EXPIRED;
+	int status;
 
 	/* Is the client already known to have an expired lease? */
 	if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
@@ -1438,6 +1446,7 @@ static int nfs4_check_lease(struct nfs_client *clp)
 	spin_unlock(&clp->cl_lock);
 	if (cred == NULL) {
 		cred = nfs4_get_setclientid_cred(clp);
+		status = -ENOKEY;
 		if (cred == NULL)
 			goto out;
 	}
@@ -1525,16 +1534,16 @@ void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags)
 {
 	if (!flags)
 		return;
-	else if (flags & SEQ4_STATUS_RESTART_RECLAIM_NEEDED)
+	if (flags & SEQ4_STATUS_RESTART_RECLAIM_NEEDED)
 		nfs41_handle_server_reboot(clp);
-	else if (flags & (SEQ4_STATUS_EXPIRED_ALL_STATE_REVOKED |
+	if (flags & (SEQ4_STATUS_EXPIRED_ALL_STATE_REVOKED |
 			    SEQ4_STATUS_EXPIRED_SOME_STATE_REVOKED |
 			    SEQ4_STATUS_ADMIN_STATE_REVOKED |
 			    SEQ4_STATUS_LEASE_MOVED))
 		nfs41_handle_state_revoked(clp);
-	else if (flags & SEQ4_STATUS_RECALLABLE_STATE_REVOKED)
+	if (flags & SEQ4_STATUS_RECALLABLE_STATE_REVOKED)
 		nfs41_handle_recallable_state_revoked(clp);
-	else if (flags & (SEQ4_STATUS_CB_PATH_DOWN |
+	if (flags & (SEQ4_STATUS_CB_PATH_DOWN |
 			    SEQ4_STATUS_BACKCHANNEL_FAULT |
 			    SEQ4_STATUS_CB_PATH_DOWN_SESSION))
 		nfs41_handle_cb_path_down(clp);
@@ -1662,10 +1671,10 @@ static void nfs4_state_manager(struct nfs_client *clp)
 
 		if (test_and_clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state)) {
 			status = nfs4_check_lease(clp);
+			if (status < 0)
+				goto out_error;
 			if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
 				continue;
-			if (status < 0 && status != -NFS4ERR_CB_PATH_DOWN)
-				goto out_error;
 		}
 
 		/* Initialize or reset the session */
diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c
index 41d6743d303..ac258beeda3 100644
--- a/fs/nilfs2/ioctl.c
+++ b/fs/nilfs2/ioctl.c
@@ -625,6 +625,9 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp,
 		if (argv[n].v_nmembs > nsegs * nilfs->ns_blocks_per_segment)
 			goto out_free;
 
+		if (argv[n].v_nmembs >= UINT_MAX / argv[n].v_size)
+			goto out_free;
+
 		len = argv[n].v_size * argv[n].v_nmembs;
 		base = (void __user *)(unsigned long)argv[n].v_base;
 		if (len == 0) {
@@ -842,6 +845,19 @@ long nilfs_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	case FS_IOC32_GETVERSION:
 		cmd = FS_IOC_GETVERSION;
 		break;
+	case NILFS_IOCTL_CHANGE_CPMODE:
+	case NILFS_IOCTL_DELETE_CHECKPOINT:
+	case NILFS_IOCTL_GET_CPINFO:
+	case NILFS_IOCTL_GET_CPSTAT:
+	case NILFS_IOCTL_GET_SUINFO:
+	case NILFS_IOCTL_GET_SUSTAT:
+	case NILFS_IOCTL_GET_VINFO:
+	case NILFS_IOCTL_GET_BDESCS:
+	case NILFS_IOCTL_CLEAN_SEGMENTS:
+	case NILFS_IOCTL_SYNC:
+	case NILFS_IOCTL_RESIZE:
+	case NILFS_IOCTL_SET_ALLOC_RANGE:
+		break;
 	default:
 		return -ENOIOCTLCMD;
 	}
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/root.c b/fs/proc/root.c
index 9a8a2b77b87..03102d97818 100644
--- a/fs/proc/root.c
+++ b/fs/proc/root.c
@@ -91,20 +91,18 @@ static struct file_system_type proc_fs_type = {
 
 void __init proc_root_init(void)
 {
-	struct vfsmount *mnt;
 	int err;
 
 	proc_init_inodecache();
 	err = register_filesystem(&proc_fs_type);
 	if (err)
 		return;
-	mnt = kern_mount_data(&proc_fs_type, &init_pid_ns);
-	if (IS_ERR(mnt)) {
+	err = pid_ns_prepare_proc(&init_pid_ns);
+	if (err) {
 		unregister_filesystem(&proc_fs_type);
 		return;
 	}
 
-	init_pid_ns.proc_mnt = mnt;
 	proc_symlink("mounts", NULL, "self/mounts");
 
 	proc_net_init();
@@ -209,5 +207,5 @@ int pid_ns_prepare_proc(struct pid_namespace *ns)
 
 void pid_ns_release_proc(struct pid_namespace *ns)
 {
-	mntput(ns->proc_mnt);
+	kern_unmount(ns->proc_mnt);
 }
diff --git a/fs/proc/stat.c b/fs/proc/stat.c
index 42b274da92c..2a30d67dd6b 100644
--- a/fs/proc/stat.c
+++ b/fs/proc/stat.c
@@ -32,7 +32,7 @@ static cputime64_t get_idle_time(int cpu)
 		idle = kstat_cpu(cpu).cpustat.idle;
 		idle = cputime64_add(idle, arch_idle_time(cpu));
 	} else
-		idle = usecs_to_cputime(idle_time);
+		idle = nsecs_to_jiffies64(1000 * idle_time);
 
 	return idle;
 }
@@ -46,7 +46,7 @@ static cputime64_t get_iowait_time(int cpu)
 		/* !NO_HZ so we can rely on cpustat.iowait */
 		iowait = kstat_cpu(cpu).cpustat.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/ubifs/super.c b/fs/ubifs/super.c
index 20403dc5d43..ae0e76bb6eb 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -2264,19 +2264,12 @@ static int __init ubifs_init(void)
 		return -EINVAL;
 	}
 
-	err = register_filesystem(&ubifs_fs_type);
-	if (err) {
-		ubifs_err("cannot register file system, error %d", err);
-		return err;
-	}
-
-	err = -ENOMEM;
 	ubifs_inode_slab = kmem_cache_create("ubifs_inode_slab",
 				sizeof(struct ubifs_inode), 0,
 				SLAB_MEM_SPREAD | SLAB_RECLAIM_ACCOUNT,
 				&inode_slab_ctor);
 	if (!ubifs_inode_slab)
-		goto out_reg;
+		return -ENOMEM;
 
 	register_shrinker(&ubifs_shrinker_info);
 
@@ -2288,15 +2281,20 @@ static int __init ubifs_init(void)
 	if (err)
 		goto out_compr;
 
+	err = register_filesystem(&ubifs_fs_type);
+	if (err) {
+		ubifs_err("cannot register file system, error %d", err);
+		goto out_dbg;
+	}
 	return 0;
 
+out_dbg:
+	dbg_debugfs_exit();
 out_compr:
 	ubifs_compressors_exit();
 out_shrinker:
 	unregister_shrinker(&ubifs_shrinker_info);
 	kmem_cache_destroy(ubifs_inode_slab);
-out_reg:
-	unregister_filesystem(&ubifs_fs_type);
 	return err;
 }
 /* late_initcall to let compressors initialize first */
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_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);