Merge branch 'for-linus' into for-2.6.33

Conflicts:
	block/cfq-iosched.c

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

65 files changed, 933 insertions, 403 deletions

diff --git a/fs/Kconfig b/fs/Kconfig
index d4bf8caad8d0..2126078a38ed 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -135,8 +135,8 @@ config TMPFS_POSIX_ACL
 
 config HUGETLBFS
 	bool "HugeTLB file system support"
-	depends on X86 || IA64 || PPC64 || SPARC64 || (SUPERH && MMU) || \
-		   (S390 && 64BIT) || SYS_SUPPORTS_HUGETLBFS || BROKEN
+	depends on X86 || IA64 || PPC_BOOK3S_64 || SPARC64 || (S390 && 64BIT) || \
+		   SYS_SUPPORTS_HUGETLBFS || BROKEN
 	help
 	  hugetlbfs is a filesystem backing for HugeTLB pages, based on
 	  ramfs. For architectures that support it, say Y here and read
diff --git a/fs/bio.c b/fs/bio.c
index 402cb84a92a1..12da5db8682c 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -325,8 +325,16 @@ static void bio_fs_destructor(struct bio *bio)
  *	@gfp_mask: allocation mask to use
  *	@nr_iovecs: number of iovecs
  *
- *	Allocate a new bio with @nr_iovecs bvecs.  If @gfp_mask
- *	contains __GFP_WAIT, the allocation is guaranteed to succeed.
+ *	bio_alloc will allocate a bio and associated bio_vec array that can hold
+ *	at least @nr_iovecs entries. Allocations will be done from the
+ *	fs_bio_set. Also see @bio_alloc_bioset and @bio_kmalloc.
+ *
+ *	If %__GFP_WAIT is set, then bio_alloc will always be able to allocate
+ *	a bio. This is due to the mempool guarantees. To make this work, callers
+ *	must never allocate more than 1 bio at a time from this pool. Callers
+ *	that need to allocate more than 1 bio must always submit the previously
+ *	allocated bio for IO before attempting to allocate a new one. Failure to
+ *	do so can cause livelocks under memory pressure.
  *
  *	RETURNS:
  *	Pointer to new bio on success, NULL on failure.
@@ -350,21 +358,13 @@ static void bio_kmalloc_destructor(struct bio *bio)
 }
 
 /**
- * bio_alloc - allocate a bio for I/O
+ * bio_kmalloc - allocate a bio for I/O using kmalloc()
  * @gfp_mask:   the GFP_ mask given to the slab allocator
  * @nr_iovecs:	number of iovecs to pre-allocate
  *
  * Description:
- *   bio_alloc will allocate a bio and associated bio_vec array that can hold
- *   at least @nr_iovecs entries. Allocations will be done from the
- *   fs_bio_set. Also see @bio_alloc_bioset.
- *
- *   If %__GFP_WAIT is set, then bio_alloc will always be able to allocate
- *   a bio. This is due to the mempool guarantees. To make this work, callers
- *   must never allocate more than 1 bio at a time from this pool. Callers
- *   that need to allocate more than 1 bio must always submit the previously
- *   allocated bio for IO before attempting to allocate a new one. Failure to
- *   do so can cause livelocks under memory pressure.
+ *   Allocate a new bio with @nr_iovecs bvecs.  If @gfp_mask contains
+ *   %__GFP_WAIT, the allocation is guaranteed to succeed.
  *
  **/
 struct bio *bio_kmalloc(gfp_t gfp_mask, int nr_iovecs)
@@ -407,7 +407,7 @@ EXPORT_SYMBOL(zero_fill_bio);
  *
  * Description:
  *   Put a reference to a &struct bio, either one you have gotten with
- *   bio_alloc or bio_get. The last put of a bio will free it.
+ *   bio_alloc, bio_get or bio_clone. The last put of a bio will free it.
  **/
 void bio_put(struct bio *bio)
 {
diff --git a/fs/block_dev.c b/fs/block_dev.c
index dde91e7e1c3a..73d6a735b8f3 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -1258,8 +1258,8 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part)
 			bd_set_size(bdev, (loff_t)bdev->bd_part->nr_sects << 9);
 		}
 	} else {
-		put_disk(disk);
 		module_put(disk->fops->owner);
+		put_disk(disk);
 		disk = NULL;
 		if (bdev->bd_contains == bdev) {
 			if (bdev->bd_disk->fops->open) {
diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c
index 69b355ae7f49..361604244271 100644
--- a/fs/btrfs/acl.c
+++ b/fs/btrfs/acl.c
@@ -27,7 +27,7 @@
 #include "btrfs_inode.h"
 #include "xattr.h"
 
-#ifdef CONFIG_BTRFS_POSIX_ACL
+#ifdef CONFIG_BTRFS_FS_POSIX_ACL
 
 static struct posix_acl *btrfs_get_acl(struct inode *inode, int type)
 {
@@ -313,7 +313,7 @@ struct xattr_handler btrfs_xattr_acl_access_handler = {
 	.set	= btrfs_xattr_acl_access_set,
 };
 
-#else /* CONFIG_BTRFS_POSIX_ACL */
+#else /* CONFIG_BTRFS_FS_POSIX_ACL */
 
 int btrfs_acl_chmod(struct inode *inode)
 {
@@ -325,4 +325,4 @@ int btrfs_init_acl(struct inode *inode, struct inode *dir)
 	return 0;
 }
 
-#endif /* CONFIG_BTRFS_POSIX_ACL */
+#endif /* CONFIG_BTRFS_FS_POSIX_ACL */
diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index 282ca085c2fb..c0861e781cdb 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -64,6 +64,51 @@ struct btrfs_worker_thread {
 };
 
 /*
+ * btrfs_start_workers uses kthread_run, which can block waiting for memory
+ * for a very long time.  It will actually throttle on page writeback,
+ * and so it may not make progress until after our btrfs worker threads
+ * process all of the pending work structs in their queue
+ *
+ * This means we can't use btrfs_start_workers from inside a btrfs worker
+ * thread that is used as part of cleaning dirty memory, which pretty much
+ * involves all of the worker threads.
+ *
+ * Instead we have a helper queue who never has more than one thread
+ * where we scheduler thread start operations.  This worker_start struct
+ * is used to contain the work and hold a pointer to the queue that needs
+ * another worker.
+ */
+struct worker_start {
+	struct btrfs_work work;
+	struct btrfs_workers *queue;
+};
+
+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);
+	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.
  */
@@ -118,11 +163,13 @@ static void check_pending_worker_creates(struct btrfs_worker_thread *worker)
 		goto out;
 
 	workers->atomic_start_pending = 0;
-	if (workers->num_workers >= workers->max_workers)
+	if (workers->num_workers + workers->num_workers_starting >=
+	    workers->max_workers)
 		goto out;
 
+	workers->num_workers_starting += 1;
 	spin_unlock_irqrestore(&workers->lock, flags);
-	btrfs_start_workers(workers, 1);
+	start_new_worker(workers);
 	return;
 
 out:
@@ -390,9 +437,11 @@ int btrfs_stop_workers(struct btrfs_workers *workers)
 /*
  * simple init on struct btrfs_workers
  */
-void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max)
+void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
+			struct btrfs_workers *async_helper)
 {
 	workers->num_workers = 0;
+	workers->num_workers_starting = 0;
 	INIT_LIST_HEAD(&workers->worker_list);
 	INIT_LIST_HEAD(&workers->idle_list);
 	INIT_LIST_HEAD(&workers->order_list);
@@ -404,14 +453,15 @@ void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max)
 	workers->name = name;
 	workers->ordered = 0;
 	workers->atomic_start_pending = 0;
-	workers->atomic_worker_start = 0;
+	workers->atomic_worker_start = async_helper;
 }
 
 /*
  * starts new worker threads.  This does not enforce the max worker
  * count in case you need to temporarily go past it.
  */
-int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
+static int __btrfs_start_workers(struct btrfs_workers *workers,
+				 int num_workers)
 {
 	struct btrfs_worker_thread *worker;
 	int ret = 0;
@@ -444,6 +494,8 @@ int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
 		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;
@@ -452,6 +504,14 @@ fail:
 	return ret;
 }
 
+int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
+{
+	spin_lock_irq(&workers->lock);
+	workers->num_workers_starting += num_workers;
+	spin_unlock_irq(&workers->lock);
+	return __btrfs_start_workers(workers, num_workers);
+}
+
 /*
  * run through the list and find a worker thread that doesn't have a lot
  * to do right now.  This can return null if we aren't yet at the thread
@@ -461,7 +521,10 @@ static struct btrfs_worker_thread *next_worker(struct btrfs_workers *workers)
 {
 	struct btrfs_worker_thread *worker;
 	struct list_head *next;
-	int enforce_min = workers->num_workers < workers->max_workers;
+	int enforce_min;
+
+	enforce_min = (workers->num_workers + workers->num_workers_starting) <
+		workers->max_workers;
 
 	/*
 	 * if we find an idle thread, don't move it to the end of the
@@ -509,15 +572,17 @@ again:
 	worker = next_worker(workers);
 
 	if (!worker) {
-		if (workers->num_workers >= workers->max_workers) {
+		if (workers->num_workers + workers->num_workers_starting >=
+		    workers->max_workers) {
 			goto fallback;
 		} else if (workers->atomic_worker_start) {
 			workers->atomic_start_pending = 1;
 			goto fallback;
 		} else {
+			workers->num_workers_starting++;
 			spin_unlock_irqrestore(&workers->lock, flags);
 			/* we're below the limit, start another worker */
-			btrfs_start_workers(workers, 1);
+			__btrfs_start_workers(workers, 1);
 			goto again;
 		}
 	}
diff --git a/fs/btrfs/async-thread.h b/fs/btrfs/async-thread.h
index fc089b95ec14..5077746cf85e 100644
--- a/fs/btrfs/async-thread.h
+++ b/fs/btrfs/async-thread.h
@@ -64,6 +64,8 @@ struct btrfs_workers {
 	/* current number of running workers */
 	int num_workers;
 
+	int num_workers_starting;
+
 	/* max number of workers allowed.  changed by btrfs_start_workers */
 	int max_workers;
 
@@ -78,9 +80,10 @@ struct btrfs_workers {
 
 	/*
 	 * are we allowed to sleep while starting workers or are we required
-	 * to start them at a later time?
+	 * to start them at a later time?  If we can't sleep, this indicates
+	 * which queue we need to use to schedule thread creation.
 	 */
-	int atomic_worker_start;
+	struct btrfs_workers *atomic_worker_start;
 
 	/* list with all the work threads.  The workers on the idle thread
 	 * may be actively servicing jobs, but they haven't yet hit the
@@ -109,7 +112,8 @@ struct btrfs_workers {
 int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work);
 int btrfs_start_workers(struct btrfs_workers *workers, int num_workers);
 int btrfs_stop_workers(struct btrfs_workers *workers);
-void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max);
+void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
+			struct btrfs_workers *async_starter);
 int btrfs_requeue_work(struct btrfs_work *work);
 void btrfs_set_work_high_prio(struct btrfs_work *work);
 #endif
diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index a54d354cefcb..f6783a42f010 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -86,6 +86,12 @@ struct btrfs_inode {
 	 * transid of the trans_handle that last modified this inode
 	 */
 	u64 last_trans;
+
+	/*
+	 * log transid when this inode was last modified
+	 */
+	u64 last_sub_trans;
+
 	/*
 	 * transid that last logged this inode
 	 */
@@ -128,12 +134,14 @@ struct btrfs_inode {
 	u64 last_unlink_trans;
 
 	/*
-	 * These two counters are for delalloc metadata reservations.  We keep
-	 * track of how many extents we've accounted for vs how many extents we
-	 * have.
+	 * Counters to keep track of the number of extent item's we may use due
+	 * to delalloc and such.  outstanding_extents is the number of extent
+	 * items we think we'll end up using, and reserved_extents is the number
+	 * of extent items we've reserved metadata for.
 	 */
-	int delalloc_reserved_extents;
-	int delalloc_extents;
+	spinlock_t accounting_lock;
+	int reserved_extents;
+	int outstanding_extents;
 
 	/*
 	 * ordered_data_close is set by truncate when a file that used
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index dd8ced9814c4..444b3e9b92a4 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -691,14 +691,17 @@ struct btrfs_space_info {
 
 	struct list_head list;
 
+	/* for controlling how we free up space for allocations */
+	wait_queue_head_t allocate_wait;
+	wait_queue_head_t flush_wait;
+	int allocating_chunk;
+	int flushing;
+
 	/* for block groups in our same type */
 	struct list_head block_groups;
 	spinlock_t lock;
 	struct rw_semaphore groups_sem;
 	atomic_t caching_threads;
-
-	int allocating_chunk;
-	wait_queue_head_t wait;
 };
 
 /*
@@ -907,6 +910,7 @@ struct btrfs_fs_info {
 	 * A third pool does submit_bio to avoid deadlocking with the other
 	 * two
 	 */
+	struct btrfs_workers generic_worker;
 	struct btrfs_workers workers;
 	struct btrfs_workers delalloc_workers;
 	struct btrfs_workers endio_workers;
@@ -914,6 +918,7 @@ struct btrfs_fs_info {
 	struct btrfs_workers endio_meta_write_workers;
 	struct btrfs_workers endio_write_workers;
 	struct btrfs_workers submit_workers;
+	struct btrfs_workers enospc_workers;
 	/*
 	 * fixup workers take dirty pages that didn't properly go through
 	 * the cow mechanism and make them safe to write.  It happens
@@ -1004,7 +1009,10 @@ struct btrfs_root {
 	atomic_t log_writers;
 	atomic_t log_commit[2];
 	unsigned long log_transid;
+	unsigned long last_log_commit;
 	unsigned long log_batch;
+	pid_t log_start_pid;
+	bool log_multiple_pids;
 
 	u64 objectid;
 	u64 last_trans;
@@ -1145,6 +1153,7 @@ struct btrfs_root {
 #define BTRFS_MOUNT_FLUSHONCOMMIT       (1 << 7)
 #define BTRFS_MOUNT_SSD_SPREAD		(1 << 8)
 #define BTRFS_MOUNT_NOSSD		(1 << 9)
+#define BTRFS_MOUNT_DISCARD		(1 << 10)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2323,7 +2332,7 @@ int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode);
 void btrfs_orphan_cleanup(struct btrfs_root *root);
 int btrfs_cont_expand(struct inode *inode, loff_t size);
 int btrfs_invalidate_inodes(struct btrfs_root *root);
-extern struct dentry_operations btrfs_dentry_operations;
+extern const struct dentry_operations btrfs_dentry_operations;
 
 /* ioctl.c */
 long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
@@ -2366,7 +2375,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options);
 int btrfs_sync_fs(struct super_block *sb, int wait);
 
 /* acl.c */
-#ifdef CONFIG_BTRFS_POSIX_ACL
+#ifdef CONFIG_BTRFS_FS_POSIX_ACL
 int btrfs_check_acl(struct inode *inode, int mask);
 #else
 #define btrfs_check_acl NULL
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index af0435f79fa6..02b6afbd7450 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -917,6 +917,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	atomic_set(&root->log_writers, 0);
 	root->log_batch = 0;
 	root->log_transid = 0;
+	root->last_log_commit = 0;
 	extent_io_tree_init(&root->dirty_log_pages,
 			     fs_info->btree_inode->i_mapping, GFP_NOFS);
 
@@ -1087,6 +1088,7 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	WARN_ON(root->log_root);
 	root->log_root = log_root;
 	root->log_transid = 0;
+	root->last_log_commit = 0;
 	return 0;
 }
 
@@ -1746,21 +1748,25 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 		err = -EINVAL;
 		goto fail_iput;
 	}
-printk("thread pool is %d\n", fs_info->thread_pool_size);
-	/*
-	 * we need to start all the end_io workers up front because the
-	 * queue work function gets called at interrupt time, and so it
-	 * cannot dynamically grow.
-	 */
+
+	btrfs_init_workers(&fs_info->generic_worker,
+			   "genwork", 1, NULL);
+
 	btrfs_init_workers(&fs_info->workers, "worker",
-			   fs_info->thread_pool_size);
+			   fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 
 	btrfs_init_workers(&fs_info->delalloc_workers, "delalloc",
-			   fs_info->thread_pool_size);
+			   fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 
 	btrfs_init_workers(&fs_info->submit_workers, "submit",
 			   min_t(u64, fs_devices->num_devices,
-			   fs_info->thread_pool_size));
+			   fs_info->thread_pool_size),
+			   &fs_info->generic_worker);
+	btrfs_init_workers(&fs_info->enospc_workers, "enospc",
+			   fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 
 	/* a higher idle thresh on the submit workers makes it much more
 	 * likely that bios will be send down in a sane order to the
@@ -1774,15 +1780,20 @@ printk("thread pool is %d\n", fs_info->thread_pool_size);
 	fs_info->delalloc_workers.idle_thresh = 2;
 	fs_info->delalloc_workers.ordered = 1;
 
-	btrfs_init_workers(&fs_info->fixup_workers, "fixup", 1);
+	btrfs_init_workers(&fs_info->fixup_workers, "fixup", 1,
+			   &fs_info->generic_worker);
 	btrfs_init_workers(&fs_info->endio_workers, "endio",
-			   fs_info->thread_pool_size);
+			   fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 	btrfs_init_workers(&fs_info->endio_meta_workers, "endio-meta",
-			   fs_info->thread_pool_size);
+			   fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 	btrfs_init_workers(&fs_info->endio_meta_write_workers,
-			   "endio-meta-write", fs_info->thread_pool_size);
+			   "endio-meta-write", fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 	btrfs_init_workers(&fs_info->endio_write_workers, "endio-write",
-			   fs_info->thread_pool_size);
+			   fs_info->thread_pool_size,
+			   &fs_info->generic_worker);
 
 	/*
 	 * endios are largely parallel and should have a very
@@ -1794,12 +1805,8 @@ printk("thread pool is %d\n", fs_info->thread_pool_size);
 	fs_info->endio_write_workers.idle_thresh = 2;
 	fs_info->endio_meta_write_workers.idle_thresh = 2;
 
-	fs_info->endio_workers.atomic_worker_start = 1;
-	fs_info->endio_meta_workers.atomic_worker_start = 1;
-	fs_info->endio_write_workers.atomic_worker_start = 1;
-	fs_info->endio_meta_write_workers.atomic_worker_start = 1;
-
 	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);
@@ -1807,6 +1814,7 @@ printk("thread pool is %d\n", fs_info->thread_pool_size);
 	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->enospc_workers, 1);
 
 	fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
 	fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
@@ -2012,6 +2020,7 @@ fail_chunk_root:
 	free_extent_buffer(chunk_root->node);
 	free_extent_buffer(chunk_root->commit_root);
 fail_sb_buffer:
+	btrfs_stop_workers(&fs_info->generic_worker);
 	btrfs_stop_workers(&fs_info->fixup_workers);
 	btrfs_stop_workers(&fs_info->delalloc_workers);
 	btrfs_stop_workers(&fs_info->workers);
@@ -2020,6 +2029,7 @@ fail_sb_buffer:
 	btrfs_stop_workers(&fs_info->endio_meta_write_workers);
 	btrfs_stop_workers(&fs_info->endio_write_workers);
 	btrfs_stop_workers(&fs_info->submit_workers);
+	btrfs_stop_workers(&fs_info->enospc_workers);
 fail_iput:
 	invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
 	iput(fs_info->btree_inode);
@@ -2437,6 +2447,7 @@ int close_ctree(struct btrfs_root *root)
 
 	iput(fs_info->btree_inode);
 
+	btrfs_stop_workers(&fs_info->generic_worker);
 	btrfs_stop_workers(&fs_info->fixup_workers);
 	btrfs_stop_workers(&fs_info->delalloc_workers);
 	btrfs_stop_workers(&fs_info->workers);
@@ -2445,6 +2456,7 @@ int close_ctree(struct btrfs_root *root)
 	btrfs_stop_workers(&fs_info->endio_meta_write_workers);
 	btrfs_stop_workers(&fs_info->endio_write_workers);
 	btrfs_stop_workers(&fs_info->submit_workers);
+	btrfs_stop_workers(&fs_info->enospc_workers);
 
 	btrfs_close_devices(fs_info->fs_devices);
 	btrfs_mapping_tree_free(&fs_info->mapping_tree);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 359a754c782c..e238a0cdac67 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -1568,23 +1568,23 @@ static int remove_extent_backref(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
-#ifdef BIO_RW_DISCARD
 static void btrfs_issue_discard(struct block_device *bdev,
 				u64 start, u64 len)
 {
 	blkdev_issue_discard(bdev, start >> 9, len >> 9, GFP_KERNEL,
 			     DISCARD_FL_BARRIER);
 }
-#endif
 
 static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr,
 				u64 num_bytes)
 {
-#ifdef BIO_RW_DISCARD
 	int ret;
 	u64 map_length = num_bytes;
 	struct btrfs_multi_bio *multi = NULL;
 
+	if (!btrfs_test_opt(root, DISCARD))
+		return 0;
+
 	/* Tell the block device(s) that the sectors can be discarded */
 	ret = btrfs_map_block(&root->fs_info->mapping_tree, READ,
 			      bytenr, &map_length, &multi, 0);
@@ -1604,9 +1604,6 @@ static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr,
 	}
 
 	return ret;
-#else
-	return 0;
-#endif
 }
 
 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
@@ -2824,14 +2821,17 @@ int btrfs_unreserve_metadata_for_delalloc(struct btrfs_root *root,
 					   num_items);
 
 	spin_lock(&meta_sinfo->lock);
-	if (BTRFS_I(inode)->delalloc_reserved_extents <=
-	    BTRFS_I(inode)->delalloc_extents) {
+	spin_lock(&BTRFS_I(inode)->accounting_lock);
+	if (BTRFS_I(inode)->reserved_extents <=
+	    BTRFS_I(inode)->outstanding_extents) {
+		spin_unlock(&BTRFS_I(inode)->accounting_lock);
 		spin_unlock(&meta_sinfo->lock);
 		return 0;
 	}
+	spin_unlock(&BTRFS_I(inode)->accounting_lock);
 
-	BTRFS_I(inode)->delalloc_reserved_extents--;
-	BUG_ON(BTRFS_I(inode)->delalloc_reserved_extents < 0);
+	BTRFS_I(inode)->reserved_extents--;
+	BUG_ON(BTRFS_I(inode)->reserved_extents < 0);
 
 	if (meta_sinfo->bytes_delalloc < num_bytes) {
 		bug = true;
@@ -2864,6 +2864,107 @@ static void check_force_delalloc(struct btrfs_space_info *meta_sinfo)
 		meta_sinfo->force_delalloc = 0;
 }
 
+struct async_flush {
+	struct btrfs_root *root;
+	struct btrfs_space_info *info;
+	struct btrfs_work work;
+};
+
+static noinline void flush_delalloc_async(struct btrfs_work *work)
+{
+	struct async_flush *async;
+	struct btrfs_root *root;
+	struct btrfs_space_info *info;
+
+	async = container_of(work, struct async_flush, work);
+	root = async->root;
+	info = async->info;
+
+	btrfs_start_delalloc_inodes(root);
+	wake_up(&info->flush_wait);
+	btrfs_wait_ordered_extents(root, 0);
+
+	spin_lock(&info->lock);
+	info->flushing = 0;
+	spin_unlock(&info->lock);
+	wake_up(&info->flush_wait);
+
+	kfree(async);
+}
+
+static void wait_on_flush(struct btrfs_space_info *info)
+{
+	DEFINE_WAIT(wait);
+	u64 used;
+
+	while (1) {
+		prepare_to_wait(&info->flush_wait, &wait,
+				TASK_UNINTERRUPTIBLE);
+		spin_lock(&info->lock);
+		if (!info->flushing) {
+			spin_unlock(&info->lock);
+			break;
+		}
+
+		used = info->bytes_used + info->bytes_reserved +
+			info->bytes_pinned + info->bytes_readonly +
+			info->bytes_super + info->bytes_root +
+			info->bytes_may_use + info->bytes_delalloc;
+		if (used < info->total_bytes) {
+			spin_unlock(&info->lock);
+			break;
+		}
+		spin_unlock(&info->lock);
+		schedule();
+	}
+	finish_wait(&info->flush_wait, &wait);
+}
+
+static void flush_delalloc(struct btrfs_root *root,
+				 struct btrfs_space_info *info)
+{
+	struct async_flush *async;
+	bool wait = false;
+
+	spin_lock(&info->lock);
+
+	if (!info->flushing) {
+		info->flushing = 1;
+		init_waitqueue_head(&info->flush_wait);
+	} else {
+		wait = true;
+	}
+
+	spin_unlock(&info->lock);
+
+	if (wait) {
+		wait_on_flush(info);
+		return;
+	}
+
+	async = kzalloc(sizeof(*async), GFP_NOFS);
+	if (!async)
+		goto flush;
+
+	async->root = root;
+	async->info = info;
+	async->work.func = flush_delalloc_async;
+
+	btrfs_queue_worker(&root->fs_info->enospc_workers,
+			   &async->work);
+	wait_on_flush(info);
+	return;
+
+flush:
+	btrfs_start_delalloc_inodes(root);
+	btrfs_wait_ordered_extents(root, 0);
+
+	spin_lock(&info->lock);
+	info->flushing = 0;
+	spin_unlock(&info->lock);
+	wake_up(&info->flush_wait);
+}
+
 static int maybe_allocate_chunk(struct btrfs_root *root,
 				 struct btrfs_space_info *info)
 {
@@ -2894,7 +2995,7 @@ static int maybe_allocate_chunk(struct btrfs_root *root,
 	if (!info->allocating_chunk) {
 		info->force_alloc = 1;
 		info->allocating_chunk = 1;
-		init_waitqueue_head(&info->wait);
+		init_waitqueue_head(&info->allocate_wait);
 	} else {
 		wait = true;
 	}
@@ -2902,7 +3003,7 @@ static int maybe_allocate_chunk(struct btrfs_root *root,
 	spin_unlock(&info->lock);
 
 	if (wait) {
-		wait_event(info->wait,
+		wait_event(info->allocate_wait,
 			   !info->allocating_chunk);
 		return 1;
 	}
@@ -2923,7 +3024,7 @@ out:
 	spin_lock(&info->lock);
 	info->allocating_chunk = 0;
 	spin_unlock(&info->lock);
-	wake_up(&info->wait);
+	wake_up(&info->allocate_wait);
 
 	if (ret)
 		return 0;
@@ -2981,21 +3082,20 @@ again:
 			filemap_flush(inode->i_mapping);
 			goto again;
 		} else if (flushed == 3) {
-			btrfs_start_delalloc_inodes(root);
-			btrfs_wait_ordered_extents(root, 0);
+			flush_delalloc(root, meta_sinfo);
 			goto again;
 		}
 		spin_lock(&meta_sinfo->lock);
 		meta_sinfo->bytes_delalloc -= num_bytes;
 		spin_unlock(&meta_sinfo->lock);
 		printk(KERN_ERR "enospc, has %d, reserved %d\n",
-		       BTRFS_I(inode)->delalloc_extents,
-		       BTRFS_I(inode)->delalloc_reserved_extents);
+		       BTRFS_I(inode)->outstanding_extents,
+		       BTRFS_I(inode)->reserved_extents);
 		dump_space_info(meta_sinfo, 0, 0);
 		return -ENOSPC;
 	}
 
-	BTRFS_I(inode)->delalloc_reserved_extents++;
+	BTRFS_I(inode)->reserved_extents++;
 	check_force_delalloc(meta_sinfo);
 	spin_unlock(&meta_sinfo->lock);
 
@@ -3094,8 +3194,7 @@ again:
 		}
 
 		if (retries == 2) {
-			btrfs_start_delalloc_inodes(root);
-			btrfs_wait_ordered_extents(root, 0);
+			flush_delalloc(root, meta_sinfo);
 			goto again;
 		}
 		spin_lock(&meta_sinfo->lock);
@@ -3588,6 +3687,14 @@ static int pin_down_bytes(struct btrfs_trans_handle *trans,
 	if (is_data)
 		goto pinit;
 
+	/*
+	 * discard is sloooow, and so triggering discards on
+	 * individual btree blocks isn't a good plan.  Just
+	 * pin everything in discard mode.
+	 */
+	if (btrfs_test_opt(root, DISCARD))
+		goto pinit;
+
 	buf = btrfs_find_tree_block(root, bytenr, num_bytes);
 	if (!buf)
 		goto pinit;
@@ -4029,6 +4136,7 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 	int loop = 0;
 	bool found_uncached_bg = false;
 	bool failed_cluster_refill = false;
+	bool failed_alloc = false;
 
 	WARN_ON(num_bytes < root->sectorsize);
 	btrfs_set_key_type(ins, BTRFS_EXTENT_ITEM_KEY);
@@ -4233,14 +4341,23 @@ refill_cluster:
 
 		offset = btrfs_find_space_for_alloc(block_group, search_start,
 						    num_bytes, empty_size);
-		if (!offset && (cached || (!cached &&
-					   loop == LOOP_CACHING_NOWAIT))) {
-			goto loop;
-		} else if (!offset && (!cached &&
-				       loop > LOOP_CACHING_NOWAIT)) {
+		/*
+		 * If we didn't find a chunk, and we haven't failed on this
+		 * block group before, and this block group is in the middle of
+		 * caching and we are ok with waiting, then go ahead and wait
+		 * for progress to be made, and set failed_alloc to true.
+		 *
+		 * If failed_alloc is true then we've already waited on this
+		 * block group once and should move on to the next block group.
+		 */
+		if (!offset && !failed_alloc && !cached &&
+		    loop > LOOP_CACHING_NOWAIT) {
 			wait_block_group_cache_progress(block_group,
-					num_bytes + empty_size);
+						num_bytes + empty_size);
+			failed_alloc = true;
 			goto have_block_group;
+		} else if (!offset) {
+			goto loop;
 		}
 checks:
 		search_start = stripe_align(root, offset);
@@ -4288,6 +4405,7 @@ checks:
 		break;
 loop:
 		failed_cluster_refill = false;
+		failed_alloc = false;
 		btrfs_put_block_group(block_group);
 	}
 	up_read(&space_info->groups_sem);
@@ -4799,6 +4917,7 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
 	u64 bytenr;
 	u64 generation;
 	u64 refs;
+	u64 flags;
 	u64 last = 0;
 	u32 nritems;
 	u32 blocksize;
@@ -4836,15 +4955,19 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
 		    generation <= root->root_key.offset)
 			continue;
 
+		/* We don't lock the tree block, it's OK to be racy here */
+		ret = btrfs_lookup_extent_info(trans, root, bytenr, blocksize,
+					       &refs, &flags);
+		BUG_ON(ret);
+		BUG_ON(refs == 0);
+
 		if (wc->stage == DROP_REFERENCE) {
-			ret = btrfs_lookup_extent_info(trans, root,
-						bytenr, blocksize,
-						&refs, NULL);
-			BUG_ON(ret);
-			BUG_ON(refs == 0);
 			if (refs == 1)
 				goto reada;
 
+			if (wc->level == 1 &&
+			    (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
+				continue;
 			if (!wc->update_ref ||
 			    generation <= root->root_key.offset)
 				continue;
@@ -4853,6 +4976,10 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
 						  &wc->update_progress);
 			if (ret < 0)
 				continue;
+		} else {
+			if (wc->level == 1 &&
+			    (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF))
+				continue;
 		}
 reada:
 		ret = readahead_tree_block(root, bytenr, blocksize,
@@ -4876,7 +5003,7 @@ reada:
 static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   struct btrfs_path *path,
-				   struct walk_control *wc)
+				   struct walk_control *wc, int lookup_info)
 {
 	int level = wc->level;
 	struct extent_buffer *eb = path->nodes[level];
@@ -4891,8 +5018,9 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
 	 * when reference count of tree block is 1, it won't increase
 	 * again. once full backref flag is set, we never clear it.
 	 */
-	if ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) ||
-	    (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag))) {
+	if (lookup_info &&
+	    ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) ||
+	     (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag)))) {
 		BUG_ON(!path->locks[level]);
 		ret = btrfs_lookup_extent_info(trans, root,
 					       eb->start, eb->len,
@@ -4953,7 +5081,7 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
 static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root,
 				 struct btrfs_path *path,
-				 struct walk_control *wc)
+				 struct walk_control *wc, int *lookup_info)
 {
 	u64 bytenr;
 	u64 generation;
@@ -4973,8 +5101,10 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	 * for the subtree
 	 */
 	if (wc->stage == UPDATE_BACKREF &&
-	    generation <= root->root_key.offset)
+	    generation <= root->root_key.offset) {
+		*lookup_info = 1;
 		return 1;
+	}
 
 	bytenr = btrfs_node_blockptr(path->nodes[level], path->slots[level]);
 	blocksize = btrfs_level_size(root, level - 1);
@@ -4987,14 +5117,19 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	btrfs_tree_lock(next);
 	btrfs_set_lock_blocking(next);
 
-	if (wc->stage == DROP_REFERENCE) {
-		ret = btrfs_lookup_extent_info(trans, root, bytenr, blocksize,
-					       &wc->refs[level - 1],
-					       &wc->flags[level - 1]);
-		BUG_ON(ret);
-		BUG_ON(wc->refs[level - 1] == 0);
+	ret = btrfs_lookup_extent_info(trans, root, bytenr, blocksize,
+				       &wc->refs[level - 1],
+				       &wc->flags[level - 1]);
+	BUG_ON(ret);
+	BUG_ON(wc->refs[level - 1] == 0);
+	*lookup_info = 0;
 
+	if (wc->stage == DROP_REFERENCE) {
 		if (wc->refs[level - 1] > 1) {
+			if (level == 1 &&
+			    (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
+				goto skip;
+
 			if (!wc->update_ref ||
 			    generation <= root->root_key.offset)
 				goto skip;
@@ -5008,12 +5143,17 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 			wc->stage = UPDATE_BACKREF;
 			wc->shared_level = level - 1;
 		}
+	} else {
+		if (level == 1 &&
+		    (wc->flags[0] & BTRFS_BLOCK_FLAG_FULL_BACKREF))
+			goto skip;
 	}
 
 	if (!btrfs_buffer_uptodate(next, generation)) {
 		btrfs_tree_unlock(next);
 		free_extent_buffer(next);
 		next = NULL;
+		*lookup_info = 1;
 	}
 
 	if (!next) {
@@ -5036,21 +5176,22 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 skip:
 	wc->refs[level - 1] = 0;
 	wc->flags[level - 1] = 0;
+	if (wc->stage == DROP_REFERENCE) {
+		if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
+			parent = path->nodes[level]->start;
+		} else {
+			BUG_ON(root->root_key.objectid !=
+			       btrfs_header_owner(path->nodes[level]));
+			parent = 0;
+		}
 
-	if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
-		parent = path->nodes[level]->start;
-	} else {
-		BUG_ON(root->root_key.objectid !=
-		       btrfs_header_owner(path->nodes[level]));
-		parent = 0;
+		ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent,
+					root->root_key.objectid, level - 1, 0);
+		BUG_ON(ret);
 	}
-
-	ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent,
-				root->root_key.objectid, level - 1, 0);
-	BUG_ON(ret);
-
 	btrfs_tree_unlock(next);
 	free_extent_buffer(next);
+	*lookup_info = 1;
 	return 1;
 }
 
@@ -5164,6 +5305,7 @@ static noinline int walk_down_tree(struct btrfs_trans_handle *trans,
 				   struct walk_control *wc)
 {
 	int level = wc->level;
+	int lookup_info = 1;
 	int ret;
 
 	while (level >= 0) {
@@ -5171,14 +5313,14 @@ static noinline int walk_down_tree(struct btrfs_trans_handle *trans,
 		    btrfs_header_nritems(path->nodes[level]))
 			break;
 
-		ret = walk_down_proc(trans, root, path, wc);
+		ret = walk_down_proc(trans, root, path, wc, lookup_info);
 		if (ret > 0)
 			break;
 
 		if (level == 0)
 			break;
 
-		ret = do_walk_down(trans, root, path, wc);
+		ret = do_walk_down(trans, root, path, wc, &lookup_info);
 		if (ret > 0) {
 			path->slots[level]++;
 			continue;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index de1793ba004a..96577e8bf9fd 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -460,7 +460,8 @@ static int clear_state_bit(struct extent_io_tree *tree,
 			    struct extent_state *state, int bits, int wake,
 			    int delete)
 {
-	int ret = state->state & bits;
+	int bits_to_clear = bits & ~EXTENT_DO_ACCOUNTING;
+	int ret = state->state & bits_to_clear;
 
 	if ((bits & EXTENT_DIRTY) && (state->state & EXTENT_DIRTY)) {
 		u64 range = state->end - state->start + 1;
@@ -468,7 +469,7 @@ static int clear_state_bit(struct extent_io_tree *tree,
 		tree->dirty_bytes -= range;
 	}
 	clear_state_cb(tree, state, bits);
-	state->state &= ~bits;
+	state->state &= ~bits_to_clear;
 	if (wake)
 		wake_up(&state->wq);
 	if (delete || state->state == 0) {
@@ -956,7 +957,8 @@ int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
 		       gfp_t mask)
 {
 	return clear_extent_bit(tree, start, end,
-				EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0,
+				EXTENT_DIRTY | EXTENT_DELALLOC |
+				EXTENT_DO_ACCOUNTING, 0, 0,
 				NULL, mask);
 }
 
@@ -1401,12 +1403,7 @@ out_failed:
 int extent_clear_unlock_delalloc(struct inode *inode,
 				struct extent_io_tree *tree,
 				u64 start, u64 end, struct page *locked_page,
-				int unlock_pages,
-				int clear_unlock,
-				int clear_delalloc, int clear_dirty,
-				int set_writeback,
-				int end_writeback,
-				int set_private2)
+				unsigned long op)
 {
 	int ret;
 	struct page *pages[16];
@@ -1416,17 +1413,21 @@ int extent_clear_unlock_delalloc(struct inode *inode,
 	int i;
 	int clear_bits = 0;
 
-	if (clear_unlock)
+	if (op & EXTENT_CLEAR_UNLOCK)
 		clear_bits |= EXTENT_LOCKED;
-	if (clear_dirty)
+	if (op & EXTENT_CLEAR_DIRTY)
 		clear_bits |= EXTENT_DIRTY;
 
-	if (clear_delalloc)
+	if (op & EXTENT_CLEAR_DELALLOC)
 		clear_bits |= EXTENT_DELALLOC;
 
+	if (op & EXTENT_CLEAR_ACCOUNTING)
+		clear_bits |= EXTENT_DO_ACCOUNTING;
+
 	clear_extent_bit(tree, start, end, clear_bits, 1, 0, NULL, GFP_NOFS);
-	if (!(unlock_pages || clear_dirty || set_writeback || end_writeback ||
-	      set_private2))
+	if (!(op & (EXTENT_CLEAR_UNLOCK_PAGE | EXTENT_CLEAR_DIRTY |
+		    EXTENT_SET_WRITEBACK | EXTENT_END_WRITEBACK |
+		    EXTENT_SET_PRIVATE2)))
 		return 0;
 
 	while (nr_pages > 0) {
@@ -1435,20 +1436,20 @@ int extent_clear_unlock_delalloc(struct inode *inode,
 				     nr_pages, ARRAY_SIZE(pages)), pages);
 		for (i = 0; i < ret; i++) {
 
-			if (set_private2)
+			if (op & EXTENT_SET_PRIVATE2)
 				SetPagePrivate2(pages[i]);
 
 			if (pages[i] == locked_page) {
 				page_cache_release(pages[i]);
 				continue;
 			}
-			if (clear_dirty)
+			if (op & EXTENT_CLEAR_DIRTY)
 				clear_page_dirty_for_io(pages[i]);
-			if (set_writeback)
+			if (op & EXTENT_SET_WRITEBACK)
 				set_page_writeback(pages[i]);
-			if (end_writeback)
+			if (op & EXTENT_END_WRITEBACK)
 				end_page_writeback(pages[i]);
-			if (unlock_pages)
+			if (op & EXTENT_CLEAR_UNLOCK_PAGE)
 				unlock_page(pages[i]);
 			page_cache_release(pages[i]);
 		}
@@ -2714,7 +2715,8 @@ int extent_invalidatepage(struct extent_io_tree *tree,
 	lock_extent(tree, start, end, GFP_NOFS);
 	wait_on_page_writeback(page);
 	clear_extent_bit(tree, start, end,
-			 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC,
+			 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
+			 EXTENT_DO_ACCOUNTING,
 			 1, 1, NULL, GFP_NOFS);
 	return 0;
 }
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 4794ec891fed..36de250a7b2b 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -15,6 +15,7 @@
 #define EXTENT_BUFFER_FILLED (1 << 8)
 #define EXTENT_BOUNDARY (1 << 9)
 #define EXTENT_NODATASUM (1 << 10)
+#define EXTENT_DO_ACCOUNTING (1 << 11)
 #define EXTENT_IOBITS (EXTENT_LOCKED | EXTENT_WRITEBACK)
 
 /* flags for bio submission */
@@ -25,6 +26,16 @@
 #define EXTENT_BUFFER_BLOCKING 1
 #define EXTENT_BUFFER_DIRTY 2
 
+/* these are flags for extent_clear_unlock_delalloc */
+#define EXTENT_CLEAR_UNLOCK_PAGE 0x1
+#define EXTENT_CLEAR_UNLOCK	 0x2
+#define EXTENT_CLEAR_DELALLOC	 0x4
+#define EXTENT_CLEAR_DIRTY	 0x8
+#define EXTENT_SET_WRITEBACK	 0x10
+#define EXTENT_END_WRITEBACK	 0x20
+#define EXTENT_SET_PRIVATE2	 0x40
+#define EXTENT_CLEAR_ACCOUNTING  0x80
+
 /*
  * page->private values.  Every page that is controlled by the extent
  * map has page->private set to one.
@@ -288,10 +299,5 @@ int extent_range_uptodate(struct extent_io_tree *tree,
 int extent_clear_unlock_delalloc(struct inode *inode,
 				struct extent_io_tree *tree,
 				u64 start, u64 end, struct page *locked_page,
-				int unlock_page,
-				int clear_unlock,
-				int clear_delalloc, int clear_dirty,
-				int set_writeback,
-				int end_writeback,
-				int set_private2);
+				unsigned long op);
 #endif
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index f19e1259a971..06550affbd27 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -878,7 +878,8 @@ again:
 			btrfs_put_ordered_extent(ordered);
 
 		clear_extent_bits(&BTRFS_I(inode)->io_tree, start_pos,
-				  last_pos - 1, EXTENT_DIRTY | EXTENT_DELALLOC,
+				  last_pos - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
+				  EXTENT_DO_ACCOUNTING,
 				  GFP_NOFS);
 		unlock_extent(&BTRFS_I(inode)->io_tree,
 			      start_pos, last_pos - 1, GFP_NOFS);
@@ -1085,8 +1086,10 @@ out_nolock:
 					btrfs_end_transaction(trans, root);
 				else
 					btrfs_commit_transaction(trans, root);
-			} else {
+			} else if (ret != BTRFS_NO_LOG_SYNC) {
 				btrfs_commit_transaction(trans, root);
+			} else {
+				btrfs_end_transaction(trans, root);
 			}
 		}
 		if (file->f_flags & O_DIRECT) {
@@ -1136,6 +1139,13 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	int ret = 0;
 	struct btrfs_trans_handle *trans;
 
+
+	/* we wait first, since the writeback may change the inode */
+	root->log_batch++;
+	/* the VFS called filemap_fdatawrite for us */
+	btrfs_wait_ordered_range(inode, 0, (u64)-1);
+	root->log_batch++;
+
 	/*
 	 * check the transaction that last modified this inode
 	 * and see if its already been committed
@@ -1143,6 +1153,11 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	if (!BTRFS_I(inode)->last_trans)
 		goto out;
 
+	/*
+	 * if the last transaction that changed this file was before
+	 * the current transaction, we can bail out now without any
+	 * syncing
+	 */
 	mutex_lock(&root->fs_info->trans_mutex);
 	if (BTRFS_I(inode)->last_trans <=
 	    root->fs_info->last_trans_committed) {
@@ -1152,13 +1167,6 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	}
 	mutex_unlock(&root->fs_info->trans_mutex);
 
-	root->log_batch++;
-	filemap_fdatawrite(inode->i_mapping);
-	btrfs_wait_ordered_range(inode, 0, (u64)-1);
-	root->log_batch++;
-
-	if (datasync && !(inode->i_state & I_DIRTY_PAGES))
-		goto out;
 	/*
 	 * ok we haven't committed the transaction yet, lets do a commit
 	 */
@@ -1187,14 +1195,18 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	 */
 	mutex_unlock(&dentry->d_inode->i_mutex);
 
-	if (ret > 0) {
-		ret = btrfs_commit_transaction(trans, root);
-	} else {
-		ret = btrfs_sync_log(trans, root);
-		if (ret == 0)
-			ret = btrfs_end_transaction(trans, root);
-		else
+	if (ret != BTRFS_NO_LOG_SYNC) {
+		if (ret > 0) {
 			ret = btrfs_commit_transaction(trans, root);
+		} else {
+			ret = btrfs_sync_log(trans, root);
+			if (ret == 0)
+				ret = btrfs_end_transaction(trans, root);
+			else
+				ret = btrfs_commit_transaction(trans, root);
+		}
+	} else {
+		ret = btrfs_end_transaction(trans, root);
 	}
 	mutex_lock(&dentry->d_inode->i_mutex);
 out:
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 112e5aa85892..dae12dc7e159 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -424,9 +424,12 @@ again:
 			 * and free up our temp pages.
 			 */
 			extent_clear_unlock_delalloc(inode,
-						     &BTRFS_I(inode)->io_tree,
-						     start, end, NULL, 1, 0,
-						     0, 1, 1, 1, 0);
+			     &BTRFS_I(inode)->io_tree,
+			     start, end, NULL,
+			     EXTENT_CLEAR_UNLOCK_PAGE | EXTENT_CLEAR_DIRTY |
+			     EXTENT_CLEAR_DELALLOC |
+			     EXTENT_CLEAR_ACCOUNTING |
+			     EXTENT_SET_WRITEBACK | EXTENT_END_WRITEBACK);
 			ret = 0;
 			goto free_pages_out;
 		}
@@ -637,11 +640,14 @@ static noinline int submit_compressed_extents(struct inode *inode,
 		 * clear dirty, set writeback and unlock the pages.
 		 */
 		extent_clear_unlock_delalloc(inode,
-					     &BTRFS_I(inode)->io_tree,
-					     async_extent->start,
-					     async_extent->start +
-					     async_extent->ram_size - 1,
-					     NULL, 1, 1, 0, 1, 1, 0, 0);
+				&BTRFS_I(inode)->io_tree,
+				async_extent->start,
+				async_extent->start +
+				async_extent->ram_size - 1,
+				NULL, EXTENT_CLEAR_UNLOCK_PAGE |
+				EXTENT_CLEAR_UNLOCK |
+				EXTENT_CLEAR_DELALLOC |
+				EXTENT_CLEAR_DIRTY | EXTENT_SET_WRITEBACK);
 
 		ret = btrfs_submit_compressed_write(inode,
 				    async_extent->start,
@@ -712,9 +718,15 @@ static noinline int cow_file_range(struct inode *inode,
 					    start, end, 0, NULL);
 		if (ret == 0) {
 			extent_clear_unlock_delalloc(inode,
-						     &BTRFS_I(inode)->io_tree,
-						     start, end, NULL, 1, 1,
-						     1, 1, 1, 1, 0);
+				     &BTRFS_I(inode)->io_tree,
+				     start, end, NULL,
+				     EXTENT_CLEAR_UNLOCK_PAGE |
+				     EXTENT_CLEAR_UNLOCK |
+				     EXTENT_CLEAR_DELALLOC |
+				     EXTENT_CLEAR_ACCOUNTING |
+				     EXTENT_CLEAR_DIRTY |
+				     EXTENT_SET_WRITEBACK |
+				     EXTENT_END_WRITEBACK);
 			*nr_written = *nr_written +
 			     (end - start + PAGE_CACHE_SIZE) / PAGE_CACHE_SIZE;
 			*page_started = 1;
@@ -738,6 +750,8 @@ static noinline int cow_file_range(struct inode *inode,
 	btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);
 
 	while (disk_num_bytes > 0) {
+		unsigned long op;
+
 		cur_alloc_size = min(disk_num_bytes, root->fs_info->max_extent);
 		ret = btrfs_reserve_extent(trans, root, cur_alloc_size,
 					   root->sectorsize, 0, alloc_hint,
@@ -789,10 +803,13 @@ static noinline int cow_file_range(struct inode *inode,
 		 * Do set the Private2 bit so we know this page was properly
 		 * setup for writepage
 		 */
+		op = unlock ? EXTENT_CLEAR_UNLOCK_PAGE : 0;
+		op |= EXTENT_CLEAR_UNLOCK | EXTENT_CLEAR_DELALLOC |
+			EXTENT_SET_PRIVATE2;
+
 		extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
 					     start, start + ram_size - 1,
-					     locked_page, unlock, 1,
-					     1, 0, 0, 0, 1);
+					     locked_page, op);
 		disk_num_bytes -= cur_alloc_size;
 		num_bytes -= cur_alloc_size;
 		alloc_hint = ins.objectid + ins.offset;
@@ -864,8 +881,8 @@ static int cow_file_range_async(struct inode *inode, struct page *locked_page,
 	u64 cur_end;
 	int limit = 10 * 1024 * 1042;
 
-	clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, EXTENT_LOCKED |
-			 EXTENT_DELALLOC, 1, 0, NULL, GFP_NOFS);
+	clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, EXTENT_LOCKED,
+			 1, 0, NULL, GFP_NOFS);
 	while (start < end) {
 		async_cow = kmalloc(sizeof(*async_cow), GFP_NOFS);
 		async_cow->inode = inode;
@@ -1006,6 +1023,7 @@ next_slot:
 
 		if (found_key.offset > cur_offset) {
 			extent_end = found_key.offset;
+			extent_type = 0;
 			goto out_check;
 		}
 
@@ -1112,8 +1130,10 @@ out_check:
 		BUG_ON(ret);
 
 		extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
-					cur_offset, cur_offset + num_bytes - 1,
-					locked_page, 1, 1, 1, 0, 0, 0, 1);
+				cur_offset, cur_offset + num_bytes - 1,
+				locked_page, EXTENT_CLEAR_UNLOCK_PAGE |
+				EXTENT_CLEAR_UNLOCK | EXTENT_CLEAR_DELALLOC |
+				EXTENT_SET_PRIVATE2);
 		cur_offset = extent_end;
 		if (cur_offset > end)
 			break;
@@ -1178,15 +1198,17 @@ static int btrfs_split_extent_hook(struct inode *inode,
 					root->fs_info->max_extent);
 
 		/*
-		 * if we break a large extent up then leave delalloc_extents be,
-		 * since we've already accounted for the large extent.
+		 * if we break a large extent up then leave oustanding_extents
+		 * be, since we've already accounted for the large extent.
 		 */
 		if (div64_u64(new_size + root->fs_info->max_extent - 1,
 			      root->fs_info->max_extent) < num_extents)
 			return 0;
 	}
 
-	BTRFS_I(inode)->delalloc_extents++;
+	spin_lock(&BTRFS_I(inode)->accounting_lock);
+	BTRFS_I(inode)->outstanding_extents++;
+	spin_unlock(&BTRFS_I(inode)->accounting_lock);
 
 	return 0;
 }
@@ -1217,7 +1239,9 @@ static int btrfs_merge_extent_hook(struct inode *inode,
 
 	/* we're not bigger than the max, unreserve the space and go */
 	if (new_size <= root->fs_info->max_extent) {
-		BTRFS_I(inode)->delalloc_extents--;
+		spin_lock(&BTRFS_I(inode)->accounting_lock);
+		BTRFS_I(inode)->outstanding_extents--;
+		spin_unlock(&BTRFS_I(inode)->accounting_lock);
 		return 0;
 	}
 
@@ -1231,7 +1255,9 @@ static int btrfs_merge_extent_hook(struct inode *inode,
 		      root->fs_info->max_extent) > num_extents)
 		return 0;
 
-	BTRFS_I(inode)->delalloc_extents--;
+	spin_lock(&BTRFS_I(inode)->accounting_lock);
+	BTRFS_I(inode)->outstanding_extents--;
+	spin_unlock(&BTRFS_I(inode)->accounting_lock);
 
 	return 0;
 }
@@ -1253,7 +1279,9 @@ static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end,
 	if (!(old & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
 		struct btrfs_root *root = BTRFS_I(inode)->root;
 
-		BTRFS_I(inode)->delalloc_extents++;
+		spin_lock(&BTRFS_I(inode)->accounting_lock);
+		BTRFS_I(inode)->outstanding_extents++;
+		spin_unlock(&BTRFS_I(inode)->accounting_lock);
 		btrfs_delalloc_reserve_space(root, inode, end - start + 1);
 		spin_lock(&root->fs_info->delalloc_lock);
 		BTRFS_I(inode)->delalloc_bytes += end - start + 1;
@@ -1281,8 +1309,12 @@ static int btrfs_clear_bit_hook(struct inode *inode,
 	if ((state->state & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
 		struct btrfs_root *root = BTRFS_I(inode)->root;
 
-		BTRFS_I(inode)->delalloc_extents--;
-		btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
+		if (bits & EXTENT_DO_ACCOUNTING) {
+			spin_lock(&BTRFS_I(inode)->accounting_lock);
+			BTRFS_I(inode)->outstanding_extents--;
+			spin_unlock(&BTRFS_I(inode)->accounting_lock);
+			btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
+		}
 
 		spin_lock(&root->fs_info->delalloc_lock);
 		if (state->end - state->start + 1 >
@@ -3000,12 +3032,22 @@ static int btrfs_truncate_page(struct address_space *mapping, loff_t from)
 
 	if ((offset & (blocksize - 1)) == 0)
 		goto out;
+	ret = btrfs_check_data_free_space(root, inode, PAGE_CACHE_SIZE);
+	if (ret)
+		goto out;
+
+	ret = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
+	if (ret)
+		goto out;
 
 	ret = -ENOMEM;
 again:
 	page = grab_cache_page(mapping, index);
-	if (!page)
+	if (!page) {
+		btrfs_free_reserved_data_space(root, inode, PAGE_CACHE_SIZE);
+		btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
 		goto out;
+	}
 
 	page_start = page_offset(page);
 	page_end = page_start + PAGE_CACHE_SIZE - 1;
@@ -3038,6 +3080,10 @@ again:
 		goto again;
 	}
 
+	clear_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end,
+			  EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING,
+			  GFP_NOFS);
+
 	ret = btrfs_set_extent_delalloc(inode, page_start, page_end);
 	if (ret) {
 		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
@@ -3056,6 +3102,9 @@ again:
 	unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
 
 out_unlock:
+	if (ret)
+		btrfs_free_reserved_data_space(root, inode, PAGE_CACHE_SIZE);
+	btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
 	unlock_page(page);
 	page_cache_release(page);
 out:
@@ -3079,7 +3128,9 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
 	if (size <= hole_start)
 		return 0;
 
-	btrfs_truncate_page(inode->i_mapping, inode->i_size);
+	err = btrfs_truncate_page(inode->i_mapping, inode->i_size);
+	if (err)
+		return err;
 
 	while (1) {
 		struct btrfs_ordered_extent *ordered;
@@ -3448,6 +3499,7 @@ static noinline void init_btrfs_i(struct inode *inode)
 	bi->generation = 0;
 	bi->sequence = 0;
 	bi->last_trans = 0;
+	bi->last_sub_trans = 0;
 	bi->logged_trans = 0;
 	bi->delalloc_bytes = 0;
 	bi->reserved_bytes = 0;
@@ -3598,12 +3650,14 @@ static int btrfs_dentry_delete(struct dentry *dentry)
 {
 	struct btrfs_root *root;
 
-	if (!dentry->d_inode)
-		return 0;
+	if (!dentry->d_inode && !IS_ROOT(dentry))
+		dentry = dentry->d_parent;
 
-	root = BTRFS_I(dentry->d_inode)->root;
-	if (btrfs_root_refs(&root->root_item) == 0)
-		return 1;
+	if (dentry->d_inode) {
+		root = BTRFS_I(dentry->d_inode)->root;
+		if (btrfs_root_refs(&root->root_item) == 0)
+			return 1;
+	}
 	return 0;
 }
 
@@ -4808,7 +4862,8 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
 		 */
 		clear_extent_bit(tree, page_start, page_end,
 				 EXTENT_DIRTY | EXTENT_DELALLOC |
-				 EXTENT_LOCKED, 1, 0, NULL, GFP_NOFS);
+				 EXTENT_LOCKED | EXTENT_DO_ACCOUNTING, 1, 0,
+				 NULL, GFP_NOFS);
 		/*
 		 * whoever cleared the private bit is responsible
 		 * for the finish_ordered_io
@@ -4821,8 +4876,8 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
 		lock_extent(tree, page_start, page_end, GFP_NOFS);
 	}
 	clear_extent_bit(tree, page_start, page_end,
-		 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC,
-		 1, 1, NULL, GFP_NOFS);
+		 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
+		 EXTENT_DO_ACCOUNTING, 1, 1, NULL, GFP_NOFS);
 	__btrfs_releasepage(page, GFP_NOFS);
 
 	ClearPageChecked(page);
@@ -4917,7 +4972,8 @@ again:
 	 * prepare_pages in the normal write path.
 	 */
 	clear_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end,
-			  EXTENT_DIRTY | EXTENT_DELALLOC, GFP_NOFS);
+			  EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING,
+			  GFP_NOFS);
 
 	ret = btrfs_set_extent_delalloc(inode, page_start, page_end);
 	if (ret) {
@@ -4944,7 +5000,9 @@ again:
 	set_page_dirty(page);
 	SetPageUptodate(page);
 
-	BTRFS_I(inode)->last_trans = root->fs_info->generation + 1;
+	BTRFS_I(inode)->last_trans = root->fs_info->generation;
+	BTRFS_I(inode)->last_sub_trans = BTRFS_I(inode)->root->log_transid;
+
 	unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
 
 out_unlock:
@@ -4969,7 +5027,9 @@ static void btrfs_truncate(struct inode *inode)
 	if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
 		return;
 
-	btrfs_truncate_page(inode->i_mapping, inode->i_size);
+	ret = btrfs_truncate_page(inode->i_mapping, inode->i_size);
+	if (ret)
+		return;
 	btrfs_wait_ordered_range(inode, inode->i_size & (~mask), (u64)-1);
 
 	trans = btrfs_start_transaction(root, 1);
@@ -5064,9 +5124,11 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	if (!ei)
 		return NULL;
 	ei->last_trans = 0;
+	ei->last_sub_trans = 0;
 	ei->logged_trans = 0;
-	ei->delalloc_extents = 0;
-	ei->delalloc_reserved_extents = 0;
+	ei->outstanding_extents = 0;
+	ei->reserved_extents = 0;
+	spin_lock_init(&ei->accounting_lock);
 	btrfs_ordered_inode_tree_init(&ei->ordered_tree);
 	INIT_LIST_HEAD(&ei->i_orphan);
 	INIT_LIST_HEAD(&ei->ordered_operations);
@@ -5805,6 +5867,6 @@ static const struct inode_operations btrfs_symlink_inode_operations = {
 	.removexattr	= btrfs_removexattr,
 };
 
-struct dentry_operations btrfs_dentry_operations = {
+const struct dentry_operations btrfs_dentry_operations = {
 	.d_delete	= btrfs_dentry_delete,
 };
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 9a780c8d0ac8..cdbb054102b9 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -830,6 +830,7 @@ out_up_write:
 out_unlock:
 	mutex_unlock(&inode->i_mutex);
 	if (!err) {
+		shrink_dcache_sb(root->fs_info->sb);
 		btrfs_invalidate_inodes(dest);
 		d_delete(dentry);
 	}
@@ -1122,8 +1123,10 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
 					datao += off - key.offset;
 					datal -= off - key.offset;
 				}
-				if (key.offset + datao + datal > off + len)
-					datal = off + len - key.offset - datao;
+
+				if (key.offset + datal > off + len)
+					datal = off + len - key.offset;
+
 				/* disko == 0 means it's a hole */
 				if (!disko)
 					datao = 0;
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index 897fba835f89..5799bc46a309 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -306,6 +306,12 @@ int btrfs_remove_ordered_extent(struct inode *inode,
 	tree->last = NULL;
 	set_bit(BTRFS_ORDERED_COMPLETE, &entry->flags);
 
+	spin_lock(&BTRFS_I(inode)->accounting_lock);
+	BTRFS_I(inode)->outstanding_extents--;
+	spin_unlock(&BTRFS_I(inode)->accounting_lock);
+	btrfs_unreserve_metadata_for_delalloc(BTRFS_I(inode)->root,
+					      inode, 1);
+
 	spin_lock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
 	list_del_init(&entry->root_extent_list);
 
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 361ad323faac..cfcc93c93a7b 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -3518,7 +3518,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 	BUG_ON(!rc->block_group);
 
 	btrfs_init_workers(&rc->workers, "relocate",
-			   fs_info->thread_pool_size);
+			   fs_info->thread_pool_size, NULL);
 
 	rc->extent_root = extent_root;
 	btrfs_prepare_block_group_relocation(extent_root, rc->block_group);
@@ -3701,7 +3701,7 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 	mapping_tree_init(&rc->reloc_root_tree);
 	INIT_LIST_HEAD(&rc->reloc_roots);
 	btrfs_init_workers(&rc->workers, "relocate",
-			   root->fs_info->thread_pool_size);
+			   root->fs_info->thread_pool_size, NULL);
 	rc->extent_root = root->fs_info->extent_root;
 
 	set_reloc_control(rc);
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 9de9b2236419..752a5463bf53 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -66,7 +66,8 @@ enum {
 	Opt_degraded, Opt_subvol, Opt_device, Opt_nodatasum, Opt_nodatacow,
 	Opt_max_extent, Opt_max_inline, Opt_alloc_start, Opt_nobarrier,
 	Opt_ssd, Opt_nossd, Opt_ssd_spread, Opt_thread_pool, Opt_noacl,
-	Opt_compress, Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_err,
+	Opt_compress, Opt_notreelog, Opt_ratio, Opt_flushoncommit,
+	Opt_discard, Opt_err,
 };
 
 static match_table_t tokens = {
@@ -88,6 +89,7 @@ static match_table_t tokens = {
 	{Opt_notreelog, "notreelog"},
 	{Opt_flushoncommit, "flushoncommit"},
 	{Opt_ratio, "metadata_ratio=%d"},
+	{Opt_discard, "discard"},
 	{Opt_err, NULL},
 };
 
@@ -257,6 +259,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				       info->metadata_ratio);
 			}
 			break;
+		case Opt_discard:
+			btrfs_set_opt(info->mount_opt, DISCARD);
+			break;
 		default:
 			break;
 		}
@@ -344,7 +349,7 @@ static int btrfs_fill_super(struct super_block *sb,
 	sb->s_export_op = &btrfs_export_ops;
 	sb->s_xattr = btrfs_xattr_handlers;
 	sb->s_time_gran = 1;
-#ifdef CONFIG_BTRFS_POSIX_ACL
+#ifdef CONFIG_BTRFS_FS_POSIX_ACL
 	sb->s_flags |= MS_POSIXACL;
 #endif
 
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 0b8f36d4400a..bca82a4ca8e6 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -344,10 +344,10 @@ int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans,
 /*
  * when btree blocks are allocated, they have some corresponding bits set for
  * them in one of two extent_io trees.  This is used to make sure all of
- * those extents are on disk for transaction or log commit
+ * those extents are sent to disk but does not wait on them
  */
-int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
-					struct extent_io_tree *dirty_pages)
+int btrfs_write_marked_extents(struct btrfs_root *root,
+			       struct extent_io_tree *dirty_pages)
 {
 	int ret;
 	int err = 0;
@@ -394,6 +394,29 @@ int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
 			page_cache_release(page);
 		}
 	}
+	if (err)
+		werr = err;
+	return werr;
+}
+
+/*
+ * when btree blocks are allocated, they have some corresponding bits set for
+ * them in one of two extent_io trees.  This is used to make sure all of
+ * those extents are on disk for transaction or log commit.  We wait
+ * on all the pages and clear them from the dirty pages state tree
+ */
+int btrfs_wait_marked_extents(struct btrfs_root *root,
+			      struct extent_io_tree *dirty_pages)
+{
+	int ret;
+	int err = 0;
+	int werr = 0;
+	struct page *page;
+	struct inode *btree_inode = root->fs_info->btree_inode;
+	u64 start = 0;
+	u64 end;
+	unsigned long index;
+
 	while (1) {
 		ret = find_first_extent_bit(dirty_pages, 0, &start, &end,
 					    EXTENT_DIRTY);
@@ -424,6 +447,22 @@ int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
 	return werr;
 }
 
+/*
+ * when btree blocks are allocated, they have some corresponding bits set for
+ * them in one of two extent_io trees.  This is used to make sure all of
+ * those extents are on disk for transaction or log commit
+ */
+int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
+					struct extent_io_tree *dirty_pages)
+{
+	int ret;
+	int ret2;
+
+	ret = btrfs_write_marked_extents(root, dirty_pages);
+	ret2 = btrfs_wait_marked_extents(root, dirty_pages);
+	return ret || ret2;
+}
+
 int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root)
 {
diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h
index 663c67404918..d4e3e7a6938c 100644
--- a/fs/btrfs/transaction.h
+++ b/fs/btrfs/transaction.h
@@ -79,6 +79,7 @@ static inline void btrfs_set_inode_last_trans(struct btrfs_trans_handle *trans,
 					      struct inode *inode)
 {
 	BTRFS_I(inode)->last_trans = trans->transaction->transid;
+	BTRFS_I(inode)->last_sub_trans = BTRFS_I(inode)->root->log_transid;
 }
 
 int btrfs_end_transaction(struct btrfs_trans_handle *trans,
@@ -107,5 +108,9 @@ int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root);
 int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
 					struct extent_io_tree *dirty_pages);
+int btrfs_write_marked_extents(struct btrfs_root *root,
+					struct extent_io_tree *dirty_pages);
+int btrfs_wait_marked_extents(struct btrfs_root *root,
+					struct extent_io_tree *dirty_pages);
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info);
 #endif
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 7827841b55cb..741666a7676a 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -137,11 +137,20 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
 
 	mutex_lock(&root->log_mutex);
 	if (root->log_root) {
+		if (!root->log_start_pid) {
+			root->log_start_pid = current->pid;
+			root->log_multiple_pids = false;
+		} else if (root->log_start_pid != current->pid) {
+			root->log_multiple_pids = true;
+		}
+
 		root->log_batch++;
 		atomic_inc(&root->log_writers);
 		mutex_unlock(&root->log_mutex);
 		return 0;
 	}
+	root->log_multiple_pids = false;
+	root->log_start_pid = current->pid;
 	mutex_lock(&root->fs_info->tree_log_mutex);
 	if (!root->fs_info->log_root_tree) {
 		ret = btrfs_init_log_root_tree(trans, root->fs_info);
@@ -1971,6 +1980,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	int ret;
 	struct btrfs_root *log = root->log_root;
 	struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
+	u64 log_transid = 0;
 
 	mutex_lock(&root->log_mutex);
 	index1 = root->log_transid % 2;
@@ -1987,10 +1997,11 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 
 	while (1) {
 		unsigned long batch = root->log_batch;
-		mutex_unlock(&root->log_mutex);
-		schedule_timeout_uninterruptible(1);
-		mutex_lock(&root->log_mutex);
-
+		if (root->log_multiple_pids) {
+			mutex_unlock(&root->log_mutex);
+			schedule_timeout_uninterruptible(1);
+			mutex_lock(&root->log_mutex);
+		}
 		wait_for_writer(trans, root);
 		if (batch == root->log_batch)
 			break;
@@ -2003,14 +2014,19 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 		goto out;
 	}
 
-	ret = btrfs_write_and_wait_marked_extents(log, &log->dirty_log_pages);
+	/* we start IO on  all the marked extents here, but we don't actually
+	 * wait for them until later.
+	 */
+	ret = btrfs_write_marked_extents(log, &log->dirty_log_pages);
 	BUG_ON(ret);
 
 	btrfs_set_root_node(&log->root_item, log->node);
 
 	root->log_batch = 0;
+	log_transid = root->log_transid;
 	root->log_transid++;
 	log->log_transid = root->log_transid;
+	root->log_start_pid = 0;
 	smp_mb();
 	/*
 	 * log tree has been flushed to disk, new modifications of
@@ -2036,6 +2052,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 
 	index2 = log_root_tree->log_transid % 2;
 	if (atomic_read(&log_root_tree->log_commit[index2])) {
+		btrfs_wait_marked_extents(log, &log->dirty_log_pages);
 		wait_log_commit(trans, log_root_tree,
 				log_root_tree->log_transid);
 		mutex_unlock(&log_root_tree->log_mutex);
@@ -2055,6 +2072,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	 * check the full commit flag again
 	 */
 	if (root->fs_info->last_trans_log_full_commit == trans->transid) {
+		btrfs_wait_marked_extents(log, &log->dirty_log_pages);
 		mutex_unlock(&log_root_tree->log_mutex);
 		ret = -EAGAIN;
 		goto out_wake_log_root;
@@ -2063,6 +2081,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	ret = btrfs_write_and_wait_marked_extents(log_root_tree,
 				&log_root_tree->dirty_log_pages);
 	BUG_ON(ret);
+	btrfs_wait_marked_extents(log, &log->dirty_log_pages);
 
 	btrfs_set_super_log_root(&root->fs_info->super_for_commit,
 				log_root_tree->node->start);
@@ -2082,9 +2101,14 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	 * the running transaction open, so a full commit can't hop
 	 * in and cause problems either.
 	 */
-	write_ctree_super(trans, root->fs_info->tree_root, 2);
+	write_ctree_super(trans, root->fs_info->tree_root, 1);
 	ret = 0;
 
+	mutex_lock(&root->log_mutex);
+	if (root->last_log_commit < log_transid)
+		root->last_log_commit = log_transid;
+	mutex_unlock(&root->log_mutex);
+
 out_wake_log_root:
 	atomic_set(&log_root_tree->log_commit[index2], 0);
 	smp_mb();
@@ -2852,6 +2876,21 @@ out:
 	return ret;
 }
 
+static int inode_in_log(struct btrfs_trans_handle *trans,
+		 struct inode *inode)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	int ret = 0;
+
+	mutex_lock(&root->log_mutex);
+	if (BTRFS_I(inode)->logged_trans == trans->transid &&
+	    BTRFS_I(inode)->last_sub_trans <= root->last_log_commit)
+		ret = 1;
+	mutex_unlock(&root->log_mutex);
+	return ret;
+}
+
+
 /*
  * helper function around btrfs_log_inode to make sure newly created
  * parent directories also end up in the log.  A minimal inode and backref
@@ -2891,6 +2930,11 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 	if (ret)
 		goto end_no_trans;
 
+	if (inode_in_log(trans, inode)) {
+		ret = BTRFS_NO_LOG_SYNC;
+		goto end_no_trans;
+	}
+
 	start_log_trans(trans, root);
 
 	ret = btrfs_log_inode(trans, root, inode, inode_only);
diff --git a/fs/btrfs/tree-log.h b/fs/btrfs/tree-log.h
index d09c7609e16b..0776eacb5083 100644
--- a/fs/btrfs/tree-log.h
+++ b/fs/btrfs/tree-log.h
@@ -19,6 +19,9 @@
 #ifndef __TREE_LOG_
 #define __TREE_LOG_
 
+/* return value for btrfs_log_dentry_safe that means we don't need to log it at all */
+#define BTRFS_NO_LOG_SYNC 256
+
 int btrfs_sync_log(struct btrfs_trans_handle *trans,
 		   struct btrfs_root *root);
 int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root);
diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c
index b0fc93f95fd0..b6dd5967c48a 100644
--- a/fs/btrfs/xattr.c
+++ b/fs/btrfs/xattr.c
@@ -260,7 +260,7 @@ err:
  * attributes are handled directly.
  */
 struct xattr_handler *btrfs_xattr_handlers[] = {
-#ifdef CONFIG_BTRFS_POSIX_ACL
+#ifdef CONFIG_BTRFS_FS_POSIX_ACL
 	&btrfs_xattr_acl_access_handler,
 	&btrfs_xattr_acl_default_handler,
 #endif
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 43003e0bef18..b09098079916 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -1577,7 +1577,8 @@ cifs_get_tcp_session(struct smb_vol *volume_info)
 
 out_err:
 	if (tcp_ses) {
-		kfree(tcp_ses->hostname);
+		if (!IS_ERR(tcp_ses->hostname))
+			kfree(tcp_ses->hostname);
 		if (tcp_ses->ssocket)
 			sock_release(tcp_ses->ssocket);
 		kfree(tcp_ses);
diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 240cef14fe58..70736eb4b516 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -316,6 +316,10 @@ int dlm_lowcomms_connect_node(int nodeid)
 {
 	struct connection *con;
 
+	/* with sctp there's no connecting without sending */
+	if (dlm_config.ci_protocol != 0)
+		return 0;
+
 	if (nodeid == dlm_our_nodeid())
 		return 0;
 
@@ -455,9 +459,9 @@ static void process_sctp_notification(struct connection *con,
 			int prim_len, ret;
 			int addr_len;
 			struct connection *new_con;
-			struct file *file;
 			sctp_peeloff_arg_t parg;
 			int parglen = sizeof(parg);
+			int err;
 
 			/*
 			 * We get this before any data for an association.
@@ -512,19 +516,22 @@ static void process_sctp_notification(struct connection *con,
 			ret = kernel_getsockopt(con->sock, IPPROTO_SCTP,
 						SCTP_SOCKOPT_PEELOFF,
 						(void *)&parg, &parglen);
-			if (ret) {
+			if (ret < 0) {
 				log_print("Can't peel off a socket for "
-					  "connection %d to node %d: err=%d\n",
+					  "connection %d to node %d: err=%d",
 					  parg.associd, nodeid, ret);
+				return;
+			}
+			new_con->sock = sockfd_lookup(parg.sd, &err);
+			if (!new_con->sock) {
+				log_print("sockfd_lookup error %d", err);
+				return;
 			}
-			file = fget(parg.sd);
-			new_con->sock = SOCKET_I(file->f_dentry->d_inode);
 			add_sock(new_con->sock, new_con);
-			fput(file);
-			put_unused_fd(parg.sd);
+			sockfd_put(new_con->sock);
 
-			log_print("got new/restarted association %d nodeid %d",
-				 (int)sn->sn_assoc_change.sac_assoc_id, nodeid);
+			log_print("connecting to %d sctp association %d",
+				 nodeid, (int)sn->sn_assoc_change.sac_assoc_id);
 
 			/* Send any pending writes */
 			clear_bit(CF_CONNECT_PENDING, &new_con->flags);
@@ -837,8 +844,6 @@ static void sctp_init_assoc(struct connection *con)
 	if (con->retries++ > MAX_CONNECT_RETRIES)
 		return;
 
-	log_print("Initiating association with node %d", con->nodeid);
-
 	if (nodeid_to_addr(con->nodeid, (struct sockaddr *)&rem_addr)) {
 		log_print("no address for nodeid %d", con->nodeid);
 		return;
@@ -855,11 +860,14 @@ static void sctp_init_assoc(struct connection *con)
 	outmessage.msg_flags = MSG_EOR;
 
 	spin_lock(&con->writequeue_lock);
-	e = list_entry(con->writequeue.next, struct writequeue_entry,
-		       list);
 
-	BUG_ON((struct list_head *) e == &con->writequeue);
+	if (list_empty(&con->writequeue)) {
+		spin_unlock(&con->writequeue_lock);
+		log_print("writequeue empty for nodeid %d", con->nodeid);
+		return;
+	}
 
+	e = list_first_entry(&con->writequeue, struct writequeue_entry, list);
 	len = e->len;
 	offset = e->offset;
 	spin_unlock(&con->writequeue_lock);
diff --git a/fs/ecryptfs/Kconfig b/fs/ecryptfs/Kconfig
index 8aadb99b7634..1cd6d9d3e29a 100644
--- a/fs/ecryptfs/Kconfig
+++ b/fs/ecryptfs/Kconfig
@@ -1,8 +1,9 @@
 config ECRYPT_FS
 	tristate "eCrypt filesystem layer support (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && KEYS && NET
+	depends on EXPERIMENTAL && KEYS && CRYPTO
 	select CRYPTO_ECB
 	select CRYPTO_CBC
+	select CRYPTO_MD5
 	help
 	  Encrypted filesystem that operates on the VFS layer.  See
 	  <file:Documentation/filesystems/ecryptfs.txt> to learn more about
diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c
index 101fe4c7b1ee..c6ac85d6c701 100644
--- a/fs/ecryptfs/main.c
+++ b/fs/ecryptfs/main.c
@@ -35,6 +35,7 @@
 #include <linux/key.h>
 #include <linux/parser.h>
 #include <linux/fs_stack.h>
+#include <linux/ima.h>
 #include "ecryptfs_kernel.h"
 
 /**
@@ -118,6 +119,7 @@ int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry)
 	const struct cred *cred = current_cred();
 	struct ecryptfs_inode_info *inode_info =
 		ecryptfs_inode_to_private(ecryptfs_dentry->d_inode);
+	int opened_lower_file = 0;
 	int rc = 0;
 
 	mutex_lock(&inode_info->lower_file_mutex);
@@ -134,9 +136,12 @@ int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry)
 			       "for lower_dentry [0x%p] and lower_mnt [0x%p]; "
 			       "rc = [%d]\n", lower_dentry, lower_mnt, rc);
 			inode_info->lower_file = NULL;
-		}
+		} else
+			opened_lower_file = 1;
 	}
 	mutex_unlock(&inode_info->lower_file_mutex);
+	if (opened_lower_file)
+		ima_counts_get(inode_info->lower_file);
 	return rc;
 }
 
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index 72743d360509..7a520a862f49 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -2321,7 +2321,18 @@ static int ext3_commit_super(struct super_block *sb,
 
 	if (!sbh)
 		return error;
-	es->s_wtime = cpu_to_le32(get_seconds());
+	/*
+	 * If the file system is mounted read-only, don't update the
+	 * superblock write time.  This avoids updating the superblock
+	 * write time when we are mounting the root file system
+	 * read/only but we need to replay the journal; at that point,
+	 * for people who are east of GMT and who make their clock
+	 * tick in localtime for Windows bug-for-bug compatibility,
+	 * the clock is set in the future, and this will cause e2fsck
+	 * to complain and force a full file system check.
+	 */
+	if (!(sb->s_flags & MS_RDONLY))
+		es->s_wtime = cpu_to_le32(get_seconds());
 	es->s_free_blocks_count = cpu_to_le32(ext3_count_free_blocks(sb));
 	es->s_free_inodes_count = cpu_to_le32(ext3_count_free_inodes(sb));
 	BUFFER_TRACE(sbh, "marking dirty");
diff --git a/fs/file.c b/fs/file.c
index f313314f996f..87e129030ab1 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -10,6 +10,7 @@
 #include <linux/fs.h>
 #include <linux/mm.h>
 #include <linux/time.h>
+#include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include <linux/file.h>
diff --git a/fs/hfs/btree.c b/fs/hfs/btree.c
index 9b9d6395bad3..052f214ea6f0 100644
--- a/fs/hfs/btree.c
+++ b/fs/hfs/btree.c
@@ -58,6 +58,11 @@ struct hfs_btree *hfs_btree_open(struct super_block *sb, u32 id, btree_keycmp ke
 	}
 	unlock_new_inode(tree->inode);
 
+	if (!HFS_I(tree->inode)->first_blocks) {
+		printk(KERN_ERR "hfs: invalid btree extent records (0 size).\n");
+		goto free_inode;
+	}
+
 	mapping = tree->inode->i_mapping;
 	page = read_mapping_page(mapping, 0, NULL);
 	if (IS_ERR(page))
diff --git a/fs/hfsplus/wrapper.c b/fs/hfsplus/wrapper.c
index 175d08eacc86..bed78ac8f6d1 100644
--- a/fs/hfsplus/wrapper.c
+++ b/fs/hfsplus/wrapper.c
@@ -99,6 +99,10 @@ int hfsplus_read_wrapper(struct super_block *sb)
 
 	if (hfsplus_get_last_session(sb, &part_start, &part_size))
 		return -EINVAL;
+	if ((u64)part_start + part_size > 0x100000000ULL) {
+		pr_err("hfs: volumes larger than 2TB are not supported yet\n");
+		return -EINVAL;
+	}
 	while (1) {
 		bh = sb_bread512(sb, part_start + HFSPLUS_VOLHEAD_SECTOR, vhdr);
 		if (!bh)
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 63976c0ccc25..99ea196f071f 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -1180,7 +1180,7 @@ static int nfs4_init_client(struct nfs_client *clp,
 				      1, flags & NFS_MOUNT_NORESVPORT);
 	if (error < 0)
 		goto error;
-	memcpy(clp->cl_ipaddr, ip_addr, sizeof(clp->cl_ipaddr));
+	strlcpy(clp->cl_ipaddr, ip_addr, sizeof(clp->cl_ipaddr));
 
 	error = nfs_idmap_new(clp);
 	if (error < 0) {
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 32062c33c859..7cb298525eef 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1536,6 +1536,8 @@ nfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
 		old_dentry->d_parent->d_name.name, old_dentry->d_name.name,
 		dentry->d_parent->d_name.name, dentry->d_name.name);
 
+	nfs_inode_return_delegation(inode);
+
 	d_drop(dentry);
 	error = NFS_PROTO(dir)->link(inode, dir, &dentry->d_name);
 	if (error == 0) {
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 6c3210099d51..e1d415e97849 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -457,6 +457,7 @@ static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq)
 	};
 	struct rpc_task_setup task_setup_data = {
 		.rpc_client = NFS_CLIENT(inode),
+		.rpc_message = &msg,
 		.callback_ops = &nfs_write_direct_ops,
 		.workqueue = nfsiod_workqueue,
 		.flags = RPC_TASK_ASYNC,
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index 2636c26d56fa..fa3408f20112 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -121,7 +121,7 @@ static struct vfsmount *try_location(struct nfs_clone_mount *mountdata,
 
 	mnt_path = nfs4_pathname_string(&location->rootpath, page2, PAGE_SIZE);
 	if (IS_ERR(mnt_path))
-		return mnt;
+		return ERR_CAST(mnt_path);
 	mountdata->mnt_path = mnt_path;
 	maxbuflen = mnt_path - 1 - page2;
 
@@ -132,15 +132,15 @@ static struct vfsmount *try_location(struct nfs_clone_mount *mountdata,
 		if (buf->len <= 0 || buf->len >= maxbuflen)
 			continue;
 
-		mountdata->addr = (struct sockaddr *)&addr;
-
 		if (memchr(buf->data, IPV6_SCOPE_DELIMITER, buf->len))
 			continue;
-		mountdata->addrlen = nfs_parse_server_name(buf->data,
-				buf->len,
-				mountdata->addr, mountdata->addrlen);
+
+		mountdata->addrlen = nfs_parse_server_name(buf->data, buf->len,
+				(struct sockaddr *)&addr, sizeof(addr));
 		if (mountdata->addrlen == 0)
 			continue;
+
+		mountdata->addr = (struct sockaddr *)&addr;
 		rpc_set_port(mountdata->addr, NFS_PORT);
 
 		memcpy(page2, buf->data, buf->len);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index ed7c269e2514..ff37454fa783 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -72,12 +72,17 @@ static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle,
 /* Prevent leaks of NFSv4 errors into userland */
 static int nfs4_map_errors(int err)
 {
-	if (err < -1000) {
+	if (err >= -1000)
+		return err;
+	switch (err) {
+	case -NFS4ERR_RESOURCE:
+		return -EREMOTEIO;
+	default:
 		dprintk("%s could not handle NFSv4 error %d\n",
 				__func__, -err);
-		return -EIO;
+		break;
 	}
-	return err;
+	return -EIO;
 }
 
 /*
@@ -3060,9 +3065,6 @@ static void nfs4_renew_done(struct rpc_task *task, void *data)
 	if (time_before(clp->cl_last_renewal,timestamp))
 		clp->cl_last_renewal = timestamp;
 	spin_unlock(&clp->cl_lock);
-	dprintk("%s calling put_rpccred on rpc_cred %p\n", __func__,
-				task->tk_msg.rpc_cred);
-	put_rpccred(task->tk_msg.rpc_cred);
 }
 
 static const struct rpc_call_ops nfs4_renew_ops = {
@@ -4877,7 +4879,6 @@ void nfs41_sequence_call_done(struct rpc_task *task, void *data)
 	nfs41_sequence_free_slot(clp, task->tk_msg.rpc_resp);
 	dprintk("%s rpc_cred %p\n", __func__, task->tk_msg.rpc_cred);
 
-	put_rpccred(task->tk_msg.rpc_cred);
 	kfree(task->tk_msg.rpc_argp);
 	kfree(task->tk_msg.rpc_resp);
 
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index e27c6cef18f2..0156c01c212c 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -127,12 +127,6 @@ nfs4_schedule_state_renewal(struct nfs_client *clp)
 }
 
 void
-nfs4_renewd_prepare_shutdown(struct nfs_server *server)
-{
-	cancel_delayed_work(&server->nfs_client->cl_renewd);
-}
-
-void
 nfs4_kill_renewd(struct nfs_client *clp)
 {
 	cancel_delayed_work_sync(&clp->cl_renewd);
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 83ad47cbdd8a..20b4e30e6c82 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -5681,7 +5681,6 @@ static struct {
 	{ NFS4ERR_SERVERFAULT,	-ESERVERFAULT	},
 	{ NFS4ERR_BADTYPE,	-EBADTYPE	},
 	{ NFS4ERR_LOCKED,	-EAGAIN		},
-	{ NFS4ERR_RESOURCE,	-EREMOTEIO	},
 	{ NFS4ERR_SYMLINK,	-ELOOP		},
 	{ NFS4ERR_OP_ILLEGAL,	-EOPNOTSUPP	},
 	{ NFS4ERR_DEADLOCK,	-EDEADLK	},
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 29786d3b9326..90be551b80c1 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -728,22 +728,24 @@ static void nfs_umount_begin(struct super_block *sb)
 	unlock_kernel();
 }
 
-static struct nfs_parsed_mount_data *nfs_alloc_parsed_mount_data(int flags)
+static struct nfs_parsed_mount_data *nfs_alloc_parsed_mount_data(unsigned int version)
 {
 	struct nfs_parsed_mount_data *data;
 
 	data = kzalloc(sizeof(*data), GFP_KERNEL);
 	if (data) {
-		data->flags		= flags;
 		data->rsize		= NFS_MAX_FILE_IO_SIZE;
 		data->wsize		= NFS_MAX_FILE_IO_SIZE;
 		data->acregmin		= NFS_DEF_ACREGMIN;
 		data->acregmax		= NFS_DEF_ACREGMAX;
 		data->acdirmin		= NFS_DEF_ACDIRMIN;
 		data->acdirmax		= NFS_DEF_ACDIRMAX;
+		data->mount_server.port	= NFS_UNSPEC_PORT;
 		data->nfs_server.port	= NFS_UNSPEC_PORT;
+		data->nfs_server.protocol = XPRT_TRANSPORT_TCP;
 		data->auth_flavors[0]	= RPC_AUTH_UNIX;
 		data->auth_flavor_len	= 1;
+		data->version		= version;
 		data->minorversion	= 0;
 	}
 	return data;
@@ -776,15 +778,13 @@ static int nfs_verify_server_address(struct sockaddr *addr)
  * Select between a default port value and a user-specified port value.
  * If a zero value is set, then autobind will be used.
  */
-static void nfs_set_default_port(struct sockaddr *sap, const int parsed_port,
+static void nfs_set_port(struct sockaddr *sap, int *port,
 				 const unsigned short default_port)
 {
-	unsigned short port = default_port;
+	if (*port == NFS_UNSPEC_PORT)
+		*port = default_port;
 
-	if (parsed_port != NFS_UNSPEC_PORT)
-		port = parsed_port;
-
-	rpc_set_port(sap, port);
+	rpc_set_port(sap, *port);
 }
 
 /*
@@ -1253,6 +1253,7 @@ static int nfs_parse_mount_options(char *raw,
 			default:
 				dfprintk(MOUNT, "NFS:   unrecognized "
 						"transport protocol\n");
+				kfree(string);
 				return 0;
 			}
 			break;
@@ -1475,7 +1476,7 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args,
 		args->mount_server.addrlen = args->nfs_server.addrlen;
 	}
 	request.salen = args->mount_server.addrlen;
-	nfs_set_default_port(request.sap, args->mount_server.port, 0);
+	nfs_set_port(request.sap, &args->mount_server.port, 0);
 
 	/*
 	 * Now ask the mount server to map our export path
@@ -1711,8 +1712,6 @@ static int nfs_validate_mount_data(void *options,
 
 		if (!(data->flags & NFS_MOUNT_TCP))
 			args->nfs_server.protocol = XPRT_TRANSPORT_UDP;
-		else
-			args->nfs_server.protocol = XPRT_TRANSPORT_TCP;
 		/* N.B. caller will free nfs_server.hostname in all cases */
 		args->nfs_server.hostname = kstrdup(data->hostname, GFP_KERNEL);
 		args->namlen		= data->namlen;
@@ -1767,7 +1766,7 @@ static int nfs_validate_mount_data(void *options,
 			goto out_v4_not_compiled;
 #endif
 
-		nfs_set_default_port(sap, args->nfs_server.port, 0);
+		nfs_set_port(sap, &args->nfs_server.port, 0);
 
 		nfs_set_mount_transport_protocol(args);
 
@@ -1848,9 +1847,10 @@ nfs_compare_remount_data(struct nfs_server *nfss,
 	    data->acdirmin != nfss->acdirmin / HZ ||
 	    data->acdirmax != nfss->acdirmax / HZ ||
 	    data->timeo != (10U * nfss->client->cl_timeout->to_initval / HZ) ||
+	    data->nfs_server.port != nfss->port ||
 	    data->nfs_server.addrlen != nfss->nfs_client->cl_addrlen ||
-	    memcmp(&data->nfs_server.address, &nfss->nfs_client->cl_addr,
-		   data->nfs_server.addrlen) != 0)
+	    !rpc_cmp_addr((struct sockaddr *)&data->nfs_server.address,
+			  (struct sockaddr *)&nfss->nfs_client->cl_addr))
 		return -EINVAL;
 
 	return 0;
@@ -1893,6 +1893,7 @@ nfs_remount(struct super_block *sb, int *flags, char *raw_data)
 	data->acdirmin = nfss->acdirmin / HZ;
 	data->acdirmax = nfss->acdirmax / HZ;
 	data->timeo = 10U * nfss->client->cl_timeout->to_initval / HZ;
+	data->nfs_server.port = nfss->port;
 	data->nfs_server.addrlen = nfss->nfs_client->cl_addrlen;
 	memcpy(&data->nfs_server.address, &nfss->nfs_client->cl_addr,
 		data->nfs_server.addrlen);
@@ -2106,7 +2107,7 @@ static int nfs_get_sb(struct file_system_type *fs_type,
 	};
 	int error = -ENOMEM;
 
-	data = nfs_alloc_parsed_mount_data(NFS_MOUNT_VER3 | NFS_MOUNT_TCP);
+	data = nfs_alloc_parsed_mount_data(3);
 	mntfh = kzalloc(sizeof(*mntfh), GFP_KERNEL);
 	if (data == NULL || mntfh == NULL)
 		goto out_free_fh;
@@ -2331,7 +2332,7 @@ static int nfs4_validate_text_mount_data(void *options,
 {
 	struct sockaddr *sap = (struct sockaddr *)&args->nfs_server.address;
 
-	nfs_set_default_port(sap, args->nfs_server.port, NFS_PORT);
+	nfs_set_port(sap, &args->nfs_server.port, NFS_PORT);
 
 	nfs_validate_transport_protocol(args);
 
@@ -2376,7 +2377,6 @@ static int nfs4_validate_mount_data(void *options,
 	if (data == NULL)
 		goto out_no_data;
 
-	args->version		= 4;
 	switch (data->version) {
 	case 1:
 		if (data->host_addrlen > sizeof(args->nfs_server.address))
@@ -2660,7 +2660,7 @@ static int nfs4_get_sb(struct file_system_type *fs_type,
 	struct nfs_parsed_mount_data *data;
 	int error = -ENOMEM;
 
-	data = nfs_alloc_parsed_mount_data(0);
+	data = nfs_alloc_parsed_mount_data(4);
 	if (data == NULL)
 		goto out_free_data;
 
@@ -2690,7 +2690,6 @@ static void nfs4_kill_super(struct super_block *sb)
 	dprintk("--> %s\n", __func__);
 	nfs_super_return_all_delegations(sb);
 	kill_anon_super(sb);
-	nfs4_renewd_prepare_shutdown(server);
 	nfs_fscache_release_super_cookie(sb);
 	nfs_free_server(server);
 	dprintk("<-- %s\n", __func__);
diff --git a/fs/notify/dnotify/dnotify.c b/fs/notify/dnotify/dnotify.c
index 828a889be909..7e54e52964dd 100644
--- a/fs/notify/dnotify/dnotify.c
+++ b/fs/notify/dnotify/dnotify.c
@@ -91,6 +91,7 @@ static int dnotify_handle_event(struct fsnotify_group *group,
 	struct dnotify_struct *dn;
 	struct dnotify_struct **prev;
 	struct fown_struct *fown;
+	__u32 test_mask = event->mask & ~FS_EVENT_ON_CHILD;
 
 	to_tell = event->to_tell;
 
@@ -106,7 +107,7 @@ static int dnotify_handle_event(struct fsnotify_group *group,
 	spin_lock(&entry->lock);
 	prev = &dnentry->dn;
 	while ((dn = *prev) != NULL) {
-		if ((dn->dn_mask & event->mask) == 0) {
+		if ((dn->dn_mask & test_mask) == 0) {
 			prev = &dn->dn_next;
 			continue;
 		}
diff --git a/fs/notify/inode_mark.c b/fs/notify/inode_mark.c
index c8a07c65482b..3165d85aada2 100644
--- a/fs/notify/inode_mark.c
+++ b/fs/notify/inode_mark.c
@@ -324,11 +324,11 @@ int fsnotify_add_mark(struct fsnotify_mark_entry *entry,
 	spin_lock(&group->mark_lock);
 	spin_lock(&inode->i_lock);
 
-	entry->group = group;
-	entry->inode = inode;
-
 	lentry = fsnotify_find_mark_entry(group, inode);
 	if (!lentry) {
+		entry->group = group;
+		entry->inode = inode;
+
 		hlist_add_head(&entry->i_list, &inode->i_fsnotify_mark_entries);
 		list_add(&entry->g_list, &group->mark_entries);
 
diff --git a/fs/notify/notification.c b/fs/notify/notification.c
index 3816d5750dd5..b8bf53b4c108 100644
--- a/fs/notify/notification.c
+++ b/fs/notify/notification.c
@@ -143,7 +143,7 @@ static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new
 			/* remember, after old was put on the wait_q we aren't
 			 * allowed to look at the inode any more, only thing
 			 * left to check was if the file_name is the same */
-			if (old->name_len &&
+			if (!old->name_len ||
 			    !strcmp(old->file_name, new->file_name))
 				return true;
 			break;
diff --git a/fs/pipe.c b/fs/pipe.c
index 52c415114838..ae17d026aaa3 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -777,36 +777,55 @@ pipe_rdwr_release(struct inode *inode, struct file *filp)
 static int
 pipe_read_open(struct inode *inode, struct file *filp)
 {
-	/* We could have perhaps used atomic_t, but this and friends
-	   below are the only places.  So it doesn't seem worthwhile.  */
+	int ret = -ENOENT;
+
 	mutex_lock(&inode->i_mutex);
-	inode->i_pipe->readers++;
+
+	if (inode->i_pipe) {
+		ret = 0;
+		inode->i_pipe->readers++;
+	}
+
 	mutex_unlock(&inode->i_mutex);
 
-	return 0;
+	return ret;
 }
 
 static int
 pipe_write_open(struct inode *inode, struct file *filp)
 {
+	int ret = -ENOENT;
+
 	mutex_lock(&inode->i_mutex);
-	inode->i_pipe->writers++;
+
+	if (inode->i_pipe) {
+		ret = 0;
+		inode->i_pipe->writers++;
+	}
+
 	mutex_unlock(&inode->i_mutex);
 
-	return 0;
+	return ret;
 }
 
 static int
 pipe_rdwr_open(struct inode *inode, struct file *filp)
 {
+	int ret = -ENOENT;
+
 	mutex_lock(&inode->i_mutex);
-	if (filp->f_mode & FMODE_READ)
-		inode->i_pipe->readers++;
-	if (filp->f_mode & FMODE_WRITE)
-		inode->i_pipe->writers++;
+
+	if (inode->i_pipe) {
+		ret = 0;
+		if (filp->f_mode & FMODE_READ)
+			inode->i_pipe->readers++;
+		if (filp->f_mode & FMODE_WRITE)
+			inode->i_pipe->writers++;
+	}
+
 	mutex_unlock(&inode->i_mutex);
 
-	return 0;
+	return ret;
 }
 
 /*
diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c
index 56013371f9f3..a44a7897fd4d 100644
--- a/fs/proc/kcore.c
+++ b/fs/proc/kcore.c
@@ -23,7 +23,6 @@
 #include <asm/io.h>
 #include <linux/list.h>
 #include <linux/ioport.h>
-#include <linux/mm.h>
 #include <linux/memory.h>
 #include <asm/sections.h>
 
diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c
index c7bff4f603ff..a65239cfd97e 100644
--- a/fs/proc/meminfo.c
+++ b/fs/proc/meminfo.c
@@ -99,7 +99,7 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
 		"VmallocUsed:    %8lu kB\n"
 		"VmallocChunk:   %8lu kB\n"
 #ifdef CONFIG_MEMORY_FAILURE
-		"HardwareCorrupted: %8lu kB\n"
+		"HardwareCorrupted: %5lu kB\n"
 #endif
 		,
 		K(i.totalram),
diff --git a/fs/proc/page.c b/fs/proc/page.c
index 2281c2cbfe2b..5033ce0d254b 100644
--- a/fs/proc/page.c
+++ b/fs/proc/page.c
@@ -94,6 +94,7 @@ static const struct file_operations proc_kpagecount_operations = {
 #define KPF_COMPOUND_TAIL	16
 #define KPF_HUGE		17
 #define KPF_UNEVICTABLE		18
+#define KPF_HWPOISON		19
 #define KPF_NOPAGE		20
 
 #define KPF_KSM			21
@@ -180,6 +181,10 @@ static u64 get_uflags(struct page *page)
 	u |= kpf_copy_bit(k, KPF_UNEVICTABLE,	PG_unevictable);
 	u |= kpf_copy_bit(k, KPF_MLOCKED,	PG_mlocked);
 
+#ifdef CONFIG_MEMORY_FAILURE
+	u |= kpf_copy_bit(k, KPF_HWPOISON,	PG_hwpoison);
+#endif
+
 #ifdef CONFIG_IA64_UNCACHED_ALLOCATOR
 	u |= kpf_copy_bit(k, KPF_UNCACHED,	PG_uncached);
 #endif
diff --git a/fs/romfs/storage.c b/fs/romfs/storage.c
index b3208adf8e71..71e2b4d50a0a 100644
--- a/fs/romfs/storage.c
+++ b/fs/romfs/storage.c
@@ -253,11 +253,11 @@ ssize_t romfs_dev_strnlen(struct super_block *sb,
 
 #ifdef CONFIG_ROMFS_ON_MTD
 	if (sb->s_mtd)
-		return romfs_mtd_strnlen(sb, pos, limit);
+		return romfs_mtd_strnlen(sb, pos, maxlen);
 #endif
 #ifdef CONFIG_ROMFS_ON_BLOCK
 	if (sb->s_bdev)
-		return romfs_blk_strnlen(sb, pos, limit);
+		return romfs_blk_strnlen(sb, pos, maxlen);
 #endif
 	return -EIO;
 }
diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c
index 0050fc40e8c9..5fad489ce5bc 100644
--- a/fs/sysfs/dir.c
+++ b/fs/sysfs/dir.c
@@ -894,7 +894,8 @@ int sysfs_move_dir(struct kobject *kobj, struct kobject *new_parent_kobj)
 
 	mutex_lock(&sysfs_rename_mutex);
 	BUG_ON(!sd->s_parent);
-	new_parent_sd = new_parent_kobj->sd ? new_parent_kobj->sd : &sysfs_root;
+	new_parent_sd = (new_parent_kobj && new_parent_kobj->sd) ?
+		new_parent_kobj->sd : &sysfs_root;
 
 	error = 0;
 	if (sd->s_parent == new_parent_sd)
diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c
index 561a9c050cef..f5ea4680f15f 100644
--- a/fs/sysfs/file.c
+++ b/fs/sysfs/file.c
@@ -268,7 +268,7 @@ static int sysfs_get_open_dirent(struct sysfs_dirent *sd,
 	struct sysfs_open_dirent *od, *new_od = NULL;
 
  retry:
-	spin_lock(&sysfs_open_dirent_lock);
+	spin_lock_irq(&sysfs_open_dirent_lock);
 
 	if (!sd->s_attr.open && new_od) {
 		sd->s_attr.open = new_od;
@@ -281,7 +281,7 @@ static int sysfs_get_open_dirent(struct sysfs_dirent *sd,
 		list_add_tail(&buffer->list, &od->buffers);
 	}
 
-	spin_unlock(&sysfs_open_dirent_lock);
+	spin_unlock_irq(&sysfs_open_dirent_lock);
 
 	if (od) {
 		kfree(new_od);
@@ -315,8 +315,9 @@ static void sysfs_put_open_dirent(struct sysfs_dirent *sd,
 				  struct sysfs_buffer *buffer)
 {
 	struct sysfs_open_dirent *od = sd->s_attr.open;
+	unsigned long flags;
 
-	spin_lock(&sysfs_open_dirent_lock);
+	spin_lock_irqsave(&sysfs_open_dirent_lock, flags);
 
 	list_del(&buffer->list);
 	if (atomic_dec_and_test(&od->refcnt))
@@ -324,7 +325,7 @@ static void sysfs_put_open_dirent(struct sysfs_dirent *sd,
 	else
 		od = NULL;
 
-	spin_unlock(&sysfs_open_dirent_lock);
+	spin_unlock_irqrestore(&sysfs_open_dirent_lock, flags);
 
 	kfree(od);
 }
@@ -456,8 +457,9 @@ static unsigned int sysfs_poll(struct file *filp, poll_table *wait)
 void sysfs_notify_dirent(struct sysfs_dirent *sd)
 {
 	struct sysfs_open_dirent *od;
+	unsigned long flags;
 
-	spin_lock(&sysfs_open_dirent_lock);
+	spin_lock_irqsave(&sysfs_open_dirent_lock, flags);
 
 	od = sd->s_attr.open;
 	if (od) {
@@ -465,7 +467,7 @@ void sysfs_notify_dirent(struct sysfs_dirent *sd)
 		wake_up_interruptible(&od->poll);
 	}
 
-	spin_unlock(&sysfs_open_dirent_lock);
+	spin_unlock_irqrestore(&sysfs_open_dirent_lock, flags);
 }
 EXPORT_SYMBOL_GPL(sysfs_notify_dirent);
 
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 381854461b28..c2e30eea74dc 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -186,19 +186,37 @@ xfs_destroy_ioend(
 }
 
 /*
+ * If the end of the current ioend is beyond the current EOF,
+ * return the new EOF value, otherwise zero.
+ */
+STATIC xfs_fsize_t
+xfs_ioend_new_eof(
+	xfs_ioend_t		*ioend)
+{
+	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
+	xfs_fsize_t		isize;
+	xfs_fsize_t		bsize;
+
+	bsize = ioend->io_offset + ioend->io_size;
+	isize = MAX(ip->i_size, ip->i_new_size);
+	isize = MIN(isize, bsize);
+	return isize > ip->i_d.di_size ? isize : 0;
+}
+
+/*
  * Update on-disk file size now that data has been written to disk.
  * The current in-memory file size is i_size.  If a write is beyond
  * eof i_new_size will be the intended file size until i_size is
  * updated.  If this write does not extend all the way to the valid
  * file size then restrict this update to the end of the write.
  */
+
 STATIC void
 xfs_setfilesize(
 	xfs_ioend_t		*ioend)
 {
 	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
 	xfs_fsize_t		isize;
-	xfs_fsize_t		bsize;
 
 	ASSERT((ip->i_d.di_mode & S_IFMT) == S_IFREG);
 	ASSERT(ioend->io_type != IOMAP_READ);
@@ -206,16 +224,10 @@ xfs_setfilesize(
 	if (unlikely(ioend->io_error))
 		return;
 
-	bsize = ioend->io_offset + ioend->io_size;
-
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-
-	isize = MAX(ip->i_size, ip->i_new_size);
-	isize = MIN(isize, bsize);
-
-	if (ip->i_d.di_size < isize) {
+	isize = xfs_ioend_new_eof(ioend);
+	if (isize) {
 		ip->i_d.di_size = isize;
-		ip->i_update_core = 1;
 		xfs_mark_inode_dirty_sync(ip);
 	}
 
@@ -404,10 +416,16 @@ xfs_submit_ioend_bio(
 	struct bio	*bio)
 {
 	atomic_inc(&ioend->io_remaining);
-
 	bio->bi_private = ioend;
 	bio->bi_end_io = xfs_end_bio;
 
+	/*
+	 * If the I/O is beyond EOF we mark the inode dirty immediately
+	 * but don't update the inode size until I/O completion.
+	 */
+	if (xfs_ioend_new_eof(ioend))
+		xfs_mark_inode_dirty_sync(XFS_I(ioend->io_inode));
+
 	submit_bio(WRITE, bio);
 	ASSERT(!bio_flagged(bio, BIO_EOPNOTSUPP));
 	bio_put(bio);
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index 629370974e57..eff61e2732af 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -176,14 +176,7 @@ xfs_file_fsync(
 	struct dentry		*dentry,
 	int			datasync)
 {
-	struct inode		*inode = dentry->d_inode;
-	struct xfs_inode	*ip = XFS_I(inode);
-	int			error;
-
-	/* capture size updates in I/O completion before writing the inode. */
-	error = filemap_fdatawait(inode->i_mapping);
-	if (error)
-		return error;
+	struct xfs_inode	*ip = XFS_I(dentry->d_inode);
 
 	xfs_iflags_clear(ip, XFS_ITRUNCATED);
 	return -xfs_fsync(ip);
diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c
index da0159d99f82..cd42ef78f6b5 100644
--- a/fs/xfs/linux-2.6/xfs_iops.c
+++ b/fs/xfs/linux-2.6/xfs_iops.c
@@ -57,19 +57,22 @@
 #include <linux/fiemap.h>
 
 /*
- * Bring the atime in the XFS inode uptodate.
- * Used before logging the inode to disk or when the Linux inode goes away.
+ * Bring the timestamps in the XFS inode uptodate.
+ *
+ * Used before writing the inode to disk.
  */
 void
-xfs_synchronize_atime(
+xfs_synchronize_times(
 	xfs_inode_t	*ip)
 {
 	struct inode	*inode = VFS_I(ip);
 
-	if (!(inode->i_state & I_CLEAR)) {
-		ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec;
-		ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec;
-	}
+	ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec;
+	ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec;
+	ip->i_d.di_ctime.t_sec = (__int32_t)inode->i_ctime.tv_sec;
+	ip->i_d.di_ctime.t_nsec = (__int32_t)inode->i_ctime.tv_nsec;
+	ip->i_d.di_mtime.t_sec = (__int32_t)inode->i_mtime.tv_sec;
+	ip->i_d.di_mtime.t_nsec = (__int32_t)inode->i_mtime.tv_nsec;
 }
 
 /*
@@ -106,32 +109,20 @@ xfs_ichgtime(
 	if ((flags & XFS_ICHGTIME_MOD) &&
 	    !timespec_equal(&inode->i_mtime, &tv)) {
 		inode->i_mtime = tv;
-		ip->i_d.di_mtime.t_sec = (__int32_t)tv.tv_sec;
-		ip->i_d.di_mtime.t_nsec = (__int32_t)tv.tv_nsec;
 		sync_it = 1;
 	}
 	if ((flags & XFS_ICHGTIME_CHG) &&
 	    !timespec_equal(&inode->i_ctime, &tv)) {
 		inode->i_ctime = tv;
-		ip->i_d.di_ctime.t_sec = (__int32_t)tv.tv_sec;
-		ip->i_d.di_ctime.t_nsec = (__int32_t)tv.tv_nsec;
 		sync_it = 1;
 	}
 
 	/*
-	 * We update the i_update_core field _after_ changing
-	 * the timestamps in order to coordinate properly with
-	 * xfs_iflush() so that we don't lose timestamp updates.
-	 * This keeps us from having to hold the inode lock
-	 * while doing this.  We use the SYNCHRONIZE macro to
-	 * ensure that the compiler does not reorder the update
-	 * of i_update_core above the timestamp updates above.
+	 * Update complete - now make sure everyone knows that the inode
+	 * is dirty.
 	 */
-	if (sync_it) {
-		SYNCHRONIZE();
-		ip->i_update_core = 1;
+	if (sync_it)
 		xfs_mark_inode_dirty_sync(ip);
-	}
 }
 
 /*
@@ -506,10 +497,8 @@ xfs_vn_getattr(
 	stat->gid = ip->i_d.di_gid;
 	stat->ino = ip->i_ino;
 	stat->atime = inode->i_atime;
-	stat->mtime.tv_sec = ip->i_d.di_mtime.t_sec;
-	stat->mtime.tv_nsec = ip->i_d.di_mtime.t_nsec;
-	stat->ctime.tv_sec = ip->i_d.di_ctime.t_sec;
-	stat->ctime.tv_nsec = ip->i_d.di_ctime.t_nsec;
+	stat->mtime = inode->i_mtime;
+	stat->ctime = inode->i_ctime;
 	stat->blocks =
 		XFS_FSB_TO_BB(mp, ip->i_d.di_nblocks + ip->i_delayed_blks);
 
diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c
index 49e4a6aea73c..072050f8d346 100644
--- a/fs/xfs/linux-2.6/xfs_lrw.c
+++ b/fs/xfs/linux-2.6/xfs_lrw.c
@@ -667,7 +667,7 @@ start:
 		xip->i_new_size = new_size;
 
 	if (likely(!(ioflags & IO_INVIS)))
-		xfs_ichgtime(xip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
+		file_update_time(file);
 
 	/*
 	 * If the offset is beyond the size of the file, we have a couple
diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c
index 9e41f91aa269..3d4a0c84d634 100644
--- a/fs/xfs/linux-2.6/xfs_quotaops.c
+++ b/fs/xfs/linux-2.6/xfs_quotaops.c
@@ -80,7 +80,7 @@ xfs_fs_set_xstate(
 
 	if (sb->s_flags & MS_RDONLY)
 		return -EROFS;
-	if (!XFS_IS_QUOTA_RUNNING(mp))
+	if (op != Q_XQUOTARM && !XFS_IS_QUOTA_RUNNING(mp))
 		return -ENOSYS;
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index bdd41c8c342f..18a4b8e11df2 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -977,6 +977,28 @@ xfs_fs_inode_init_once(
 }
 
 /*
+ * Dirty the XFS inode when mark_inode_dirty_sync() is called so that
+ * we catch unlogged VFS level updates to the inode. Care must be taken
+ * here - the transaction code calls mark_inode_dirty_sync() to mark the
+ * VFS inode dirty in a transaction and clears the i_update_core field;
+ * it must clear the field after calling mark_inode_dirty_sync() to
+ * correctly indicate that the dirty state has been propagated into the
+ * inode log item.
+ *
+ * We need the barrier() to maintain correct ordering between unlogged
+ * updates and the transaction commit code that clears the i_update_core
+ * field. This requires all updates to be completed before marking the
+ * inode dirty.
+ */
+STATIC void
+xfs_fs_dirty_inode(
+	struct inode	*inode)
+{
+	barrier();
+	XFS_I(inode)->i_update_core = 1;
+}
+
+/*
  * Attempt to flush the inode, this will actually fail
  * if the inode is pinned, but we dirty the inode again
  * at the point when it is unpinned after a log write,
@@ -1126,7 +1148,7 @@ xfs_fs_put_super(
 }
 
 STATIC int
-xfs_fs_sync_super(
+xfs_fs_sync_fs(
 	struct super_block	*sb,
 	int			wait)
 {
@@ -1134,23 +1156,23 @@ xfs_fs_sync_super(
 	int			error;
 
 	/*
-	 * Treat a sync operation like a freeze.  This is to work
-	 * around a race in sync_inodes() which works in two phases
-	 * - an asynchronous flush, which can write out an inode
-	 * without waiting for file size updates to complete, and a
-	 * synchronous flush, which wont do anything because the
-	 * async flush removed the inode's dirty flag.  Also
-	 * sync_inodes() will not see any files that just have
-	 * outstanding transactions to be flushed because we don't
-	 * dirty the Linux inode until after the transaction I/O
-	 * completes.
+	 * Not much we can do for the first async pass.  Writing out the
+	 * superblock would be counter-productive as we are going to redirty
+	 * when writing out other data and metadata (and writing out a single
+	 * block is quite fast anyway).
+	 *
+	 * Try to asynchronously kick off quota syncing at least.
 	 */
-	if (wait || unlikely(sb->s_frozen == SB_FREEZE_WRITE))
-		error = xfs_quiesce_data(mp);
-	else
-		error = xfs_sync_fsdata(mp, 0);
+	if (!wait) {
+		xfs_qm_sync(mp, SYNC_TRYLOCK);
+		return 0;
+	}
+
+	error = xfs_quiesce_data(mp);
+	if (error)
+		return -error;
 
-	if (unlikely(laptop_mode)) {
+	if (laptop_mode) {
 		int	prev_sync_seq = mp->m_sync_seq;
 
 		/*
@@ -1169,7 +1191,7 @@ xfs_fs_sync_super(
 				mp->m_sync_seq != prev_sync_seq);
 	}
 
-	return -error;
+	return 0;
 }
 
 STATIC int
@@ -1539,10 +1561,11 @@ xfs_fs_get_sb(
 static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
+	.dirty_inode		= xfs_fs_dirty_inode,
 	.write_inode		= xfs_fs_write_inode,
 	.clear_inode		= xfs_fs_clear_inode,
 	.put_super		= xfs_fs_put_super,
-	.sync_fs		= xfs_fs_sync_super,
+	.sync_fs		= xfs_fs_sync_fs,
 	.freeze_fs		= xfs_fs_freeze,
 	.statfs			= xfs_fs_statfs,
 	.remount_fs		= xfs_fs_remount,
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index 320be6aea492..961df0a22c78 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -309,11 +309,15 @@ xfs_sync_attr(
 STATIC int
 xfs_commit_dummy_trans(
 	struct xfs_mount	*mp,
-	uint			log_flags)
+	uint			flags)
 {
 	struct xfs_inode	*ip = mp->m_rootip;
 	struct xfs_trans	*tp;
 	int			error;
+	int			log_flags = XFS_LOG_FORCE;
+
+	if (flags & SYNC_WAIT)
+		log_flags |= XFS_LOG_SYNC;
 
 	/*
 	 * Put a dummy transaction in the log to tell recovery
@@ -331,13 +335,12 @@ xfs_commit_dummy_trans(
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
 	xfs_trans_ihold(tp, ip);
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-	/* XXX(hch): ignoring the error here.. */
 	error = xfs_trans_commit(tp, 0);
-
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 
+	/* the log force ensures this transaction is pushed to disk */
 	xfs_log_force(mp, 0, log_flags);
-	return 0;
+	return error;
 }
 
 int
@@ -385,7 +388,20 @@ xfs_sync_fsdata(
 	else
 		XFS_BUF_ASYNC(bp);
 
-	return xfs_bwrite(mp, bp);
+	error = xfs_bwrite(mp, bp);
+	if (error)
+		return error;
+
+	/*
+	 * If this is a data integrity sync make sure all pending buffers
+	 * are flushed out for the log coverage check below.
+	 */
+	if (flags & SYNC_WAIT)
+		xfs_flush_buftarg(mp->m_ddev_targp, 1);
+
+	if (xfs_log_need_covered(mp))
+		error = xfs_commit_dummy_trans(mp, flags);
+	return error;
 
  out_brelse:
 	xfs_buf_relse(bp);
@@ -419,14 +435,16 @@ xfs_quiesce_data(
 	/* push non-blocking */
 	xfs_sync_data(mp, 0);
 	xfs_qm_sync(mp, SYNC_TRYLOCK);
-	xfs_filestream_flush(mp);
 
-	/* push and block */
+	/* push and block till complete */
 	xfs_sync_data(mp, SYNC_WAIT);
 	xfs_qm_sync(mp, SYNC_WAIT);
 
+	/* drop inode references pinned by filestreams */
+	xfs_filestream_flush(mp);
+
 	/* write superblock and hoover up shutdown errors */
-	error = xfs_sync_fsdata(mp, 0);
+	error = xfs_sync_fsdata(mp, SYNC_WAIT);
 
 	/* flush data-only devices */
 	if (mp->m_rtdev_targp)
@@ -570,8 +588,6 @@ xfs_sync_worker(
 		/* dgc: errors ignored here */
 		error = xfs_qm_sync(mp, SYNC_TRYLOCK);
 		error = xfs_sync_fsdata(mp, SYNC_TRYLOCK);
-		if (xfs_log_need_covered(mp))
-			error = xfs_commit_dummy_trans(mp, XFS_LOG_FORCE);
 	}
 	mp->m_sync_seq++;
 	wake_up(&mp->m_wait_single_sync_task);
diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c
index 4e4276b956e8..5d1a3b98a6e6 100644
--- a/fs/xfs/quota/xfs_qm_syscalls.c
+++ b/fs/xfs/quota/xfs_qm_syscalls.c
@@ -876,7 +876,6 @@ xfs_dqrele_inode(
 		ip->i_gdquot = NULL;
 	}
 	xfs_iput(ip, XFS_ILOCK_EXCL);
-	IRELE(ip);
 
 	return 0;
 }
diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c
index 7465f9ee125f..ab89a7e94a0f 100644
--- a/fs/xfs/xfs_dfrag.c
+++ b/fs/xfs/xfs_dfrag.c
@@ -206,10 +206,10 @@ xfs_swap_extents(
 	 * process that the file was not changed out from
 	 * under it.
 	 */
-	if ((sbp->bs_ctime.tv_sec != ip->i_d.di_ctime.t_sec) ||
-	    (sbp->bs_ctime.tv_nsec != ip->i_d.di_ctime.t_nsec) ||
-	    (sbp->bs_mtime.tv_sec != ip->i_d.di_mtime.t_sec) ||
-	    (sbp->bs_mtime.tv_nsec != ip->i_d.di_mtime.t_nsec)) {
+	if ((sbp->bs_ctime.tv_sec != VFS_I(ip)->i_ctime.tv_sec) ||
+	    (sbp->bs_ctime.tv_nsec != VFS_I(ip)->i_ctime.tv_nsec) ||
+	    (sbp->bs_mtime.tv_sec != VFS_I(ip)->i_mtime.tv_sec) ||
+	    (sbp->bs_mtime.tv_nsec != VFS_I(ip)->i_mtime.tv_nsec)) {
 		error = XFS_ERROR(EBUSY);
 		goto out_unlock;
 	}
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index fa913e459442..41ad537c49e9 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -854,6 +854,7 @@ xfs_dir2_leaf_getdents(
 			 */
 			ra_want = howmany(bufsize + mp->m_dirblksize,
 					  mp->m_sb.sb_blocksize) - 1;
+			ASSERT(ra_want >= 0);
 
 			/*
 			 * If we don't have as many as we want, and we haven't
@@ -1088,7 +1089,8 @@ xfs_dir2_leaf_getdents(
 		 */
 		ptr += length;
 		curoff += length;
-		bufsize -= length;
+		/* bufsize may have just been a guess; don't go negative */
+		bufsize = bufsize > length ? bufsize - length : 0;
 	}
 
 	/*
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index ab64f3efb43b..0785797db828 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -880,6 +880,7 @@ nextag:
 				 * Not in range - save last search
 				 * location and allocate a new inode
 				 */
+				xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
 				pag->pagl_leftrec = trec.ir_startino;
 				pag->pagl_rightrec = rec.ir_startino;
 				pag->pagl_pagino = pagino;
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index c1dc7ef5a1d8..b92a4fa2a0a1 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -3068,9 +3068,9 @@ xfs_iflush_int(
 	SYNCHRONIZE();
 
 	/*
-	 * Make sure to get the latest atime from the Linux inode.
+	 * Make sure to get the latest timestamps from the Linux inode.
 	 */
-	xfs_synchronize_atime(ip);
+	xfs_synchronize_times(ip);
 
 	if (XFS_TEST_ERROR(be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC,
 			       mp, XFS_ERRTAG_IFLUSH_1, XFS_RANDOM_IFLUSH_1)) {
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 0b38b9a869ec..41555de1d1db 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -504,7 +504,7 @@ void		xfs_ichgtime(xfs_inode_t *, int);
 void		xfs_lock_inodes(xfs_inode_t **, int, uint);
 void		xfs_lock_two_inodes(xfs_inode_t *, xfs_inode_t *, uint);
 
-void		xfs_synchronize_atime(xfs_inode_t *);
+void		xfs_synchronize_times(xfs_inode_t *);
 void		xfs_mark_inode_dirty_sync(xfs_inode_t *);
 
 #if defined(XFS_INODE_TRACE)
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index 47d5b663c37e..9794b876d6ff 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -232,6 +232,15 @@ xfs_inode_item_format(
 	nvecs	     = 1;
 
 	/*
+	 * Make sure the linux inode is dirty. We do this before
+	 * clearing i_update_core as the VFS will call back into
+	 * XFS here and set i_update_core, so we need to dirty the
+	 * inode first so that the ordering of i_update_core and
+	 * unlogged modifications still works as described below.
+	 */
+	xfs_mark_inode_dirty_sync(ip);
+
+	/*
 	 * Clear i_update_core if the timestamps (or any other
 	 * non-transactional modification) need flushing/logging
 	 * and we're about to log them with the rest of the core.
@@ -263,14 +272,9 @@ xfs_inode_item_format(
 	}
 
 	/*
-	 * Make sure to get the latest atime from the Linux inode.
+	 * Make sure to get the latest timestamps from the Linux inode.
 	 */
-	xfs_synchronize_atime(ip);
-
-	/*
-	 * make sure the linux inode is dirty
-	 */
-	xfs_mark_inode_dirty_sync(ip);
+	xfs_synchronize_times(ip);
 
 	vecp->i_addr = (xfs_caddr_t)&ip->i_d;
 	vecp->i_len  = sizeof(struct xfs_icdinode);
diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c
index b68f9107e26c..62efab2f3839 100644
--- a/fs/xfs/xfs_itable.c
+++ b/fs/xfs/xfs_itable.c
@@ -59,6 +59,7 @@ xfs_bulkstat_one_iget(
 {
 	xfs_icdinode_t	*dic;	/* dinode core info pointer */
 	xfs_inode_t	*ip;		/* incore inode pointer */
+	struct inode	*inode;
 	int		error;
 
 	error = xfs_iget(mp, NULL, ino,
@@ -72,6 +73,7 @@ xfs_bulkstat_one_iget(
 	ASSERT(ip->i_imap.im_blkno != 0);
 
 	dic = &ip->i_d;
+	inode = VFS_I(ip);
 
 	/* xfs_iget returns the following without needing
 	 * further change.
@@ -83,16 +85,19 @@ xfs_bulkstat_one_iget(
 	buf->bs_uid = dic->di_uid;
 	buf->bs_gid = dic->di_gid;
 	buf->bs_size = dic->di_size;
+
 	/*
-	 * We are reading the atime from the Linux inode because the
-	 * dinode might not be uptodate.
+	 * We need to read the timestamps from the Linux inode because
+	 * the VFS keeps writing directly into the inode structure instead
+	 * of telling us about the updates.
 	 */
-	buf->bs_atime.tv_sec = VFS_I(ip)->i_atime.tv_sec;
-	buf->bs_atime.tv_nsec = VFS_I(ip)->i_atime.tv_nsec;
-	buf->bs_mtime.tv_sec = dic->di_mtime.t_sec;
-	buf->bs_mtime.tv_nsec = dic->di_mtime.t_nsec;
-	buf->bs_ctime.tv_sec = dic->di_ctime.t_sec;
-	buf->bs_ctime.tv_nsec = dic->di_ctime.t_nsec;
+	buf->bs_atime.tv_sec = inode->i_atime.tv_sec;
+	buf->bs_atime.tv_nsec = inode->i_atime.tv_nsec;
+	buf->bs_mtime.tv_sec = inode->i_mtime.tv_sec;
+	buf->bs_mtime.tv_nsec = inode->i_mtime.tv_nsec;
+	buf->bs_ctime.tv_sec = inode->i_ctime.tv_sec;
+	buf->bs_ctime.tv_nsec = inode->i_ctime.tv_nsec;
+
 	buf->bs_xflags = xfs_ip2xflags(ip);
 	buf->bs_extsize = dic->di_extsize << mp->m_sb.sb_blocklog;
 	buf->bs_extents = dic->di_nextents;
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index a434f287962d..b572f7e840e0 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -2476,12 +2476,6 @@ xfs_reclaim(
 	ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0);
 
 	/*
-	 * Make sure the atime in the XFS inode is correct before freeing the
-	 * Linux inode.
-	 */
-	xfs_synchronize_atime(ip);
-
-	/*
 	 * If we have nothing to flush with this inode then complete the
 	 * teardown now, otherwise break the link between the xfs inode and the
 	 * linux inode and clean up the xfs inode later. This avoids flushing