Merge tag 'for-5.1-part1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "This contains usual mix of new features, core changes and fixes; full
  list below. I'm planning second pull request, with a few more fixes
  that arrived recently but too close to merge window, will send it next
  week.

  New features:

   - support zstd compression levels

   - new ioctl to unregister a device from the module (ie. reverse of
     device scan)

   - scrub prints a message to log when it's about to start or finish

  Core changes:

   - qgroups can now skip part of a tree that does not get updated
     during relocation, because this does not affect the quota
     accounting, estimated speedup in run time is about 20%

   - the compression workspace management had to be enhanced due to zstd
     requirements

   - various enospc fixes, when there's high fragmentation the
     over-reservation can cause ENOSPC that might not happen after a
     flush, in such cases try to wait if the situation improves

  Fixes:

   - various ioctls could overwrite previous return value if
     copy_to_user fails, fix this so the original error is reported

   - more reclaim vs GFP_KERNEL fixes

   - other cleanups and refactoring

   - fix a (valid) lockdep warning in a test when device replace is
     destroying worker threads

   - make qgroup async transaction commit more aggressive, this avoids
     some 'quota limit reached' errors if there are not enough data to
     trigger transaction in order to flush

   - fix deadlock between snapshot deletion and quotas when backref
     walking is called from context that already holds the same locks

   - fsync fixes:
      - fix fsync after succession of renames of different files
      - fix fsync after succession of renames and unlink/rmdir"

* tag 'for-5.1-part1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (92 commits)
  btrfs: Remove unnecessary casts in btrfs_read_root_item
  Btrfs: remove assertion when searching for a key in a node/leaf
  Btrfs: add missing error handling after doing leaf/node binary search
  btrfs: drop the lock on error in btrfs_dev_replace_cancel
  btrfs: ensure that a DUP or RAID1 block group has exactly two stripes
  btrfs: init csum_list before possible free
  Btrfs: remove no longer needed range length checks for deduplication
  Btrfs: fix fsync after succession of renames and unlink/rmdir
  Btrfs: fix fsync after succession of renames of different files
  btrfs: honor path->skip_locking in backref code
  btrfs: qgroup: Make qgroup async transaction commit more aggressive
  btrfs: qgroup: Move reserved data accounting from btrfs_delayed_ref_head to btrfs_qgroup_extent_record
  btrfs: scrub: remove unused nocow worker pointer
  btrfs: scrub: add assertions for worker pointers
  btrfs: scrub: convert scrub_workers_refcnt to refcount_t
  btrfs: scrub: add scrub_lock lockdep check in scrub_workers_get
  btrfs: scrub: fix circular locking dependency warning
  btrfs: fix comment its device list mutex not volume lock
  btrfs: extent_io: Kill the forward declaration of flush_write_bio
  btrfs: Fix grossly misleading argument names in extent io search
  ...

36 files changed, 1990 insertions, 931 deletions

diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c
index 3b66c957ea6f..5810463dc6d2 100644
--- a/fs/btrfs/acl.c
+++ b/fs/btrfs/acl.c
@@ -9,6 +9,7 @@
 #include <linux/posix_acl_xattr.h>
 #include <linux/posix_acl.h>
 #include <linux/sched.h>
+#include <linux/sched/mm.h>
 #include <linux/slab.h>
 
 #include "ctree.h"
@@ -72,8 +73,16 @@ static int __btrfs_set_acl(struct btrfs_trans_handle *trans,
 	}
 
 	if (acl) {
+		unsigned int nofs_flag;
+
 		size = posix_acl_xattr_size(acl->a_count);
+		/*
+		 * We're holding a transaction handle, so use a NOFS memory
+		 * allocation context to avoid deadlock if reclaim happens.
+		 */
+		nofs_flag = memalloc_nofs_save();
 		value = kmalloc(size, GFP_KERNEL);
+		memalloc_nofs_restore(nofs_flag);
 		if (!value) {
 			ret = -ENOMEM;
 			goto out;
diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index d522494698fa..122cb97c7909 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -139,13 +139,11 @@ __btrfs_alloc_workqueue(struct btrfs_fs_info *fs_info, const char *name,
 	}
 
 	if (flags & WQ_HIGHPRI)
-		ret->normal_wq = alloc_workqueue("%s-%s-high", flags,
-						 ret->current_active, "btrfs",
-						 name);
+		ret->normal_wq = alloc_workqueue("btrfs-%s-high", flags,
+						 ret->current_active, name);
 	else
-		ret->normal_wq = alloc_workqueue("%s-%s", flags,
-						 ret->current_active, "btrfs",
-						 name);
+		ret->normal_wq = alloc_workqueue("btrfs-%s", flags,
+						 ret->current_active, name);
 	if (!ret->normal_wq) {
 		kfree(ret);
 		return NULL;
diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index 78556447e1d5..11459fe84a29 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -712,7 +712,7 @@ out:
  * read tree blocks and add keys where required.
  */
 static int add_missing_keys(struct btrfs_fs_info *fs_info,
-			    struct preftrees *preftrees)
+			    struct preftrees *preftrees, bool lock)
 {
 	struct prelim_ref *ref;
 	struct extent_buffer *eb;
@@ -737,12 +737,14 @@ static int add_missing_keys(struct btrfs_fs_info *fs_info,
 			free_extent_buffer(eb);
 			return -EIO;
 		}
-		btrfs_tree_read_lock(eb);
+		if (lock)
+			btrfs_tree_read_lock(eb);
 		if (btrfs_header_level(eb) == 0)
 			btrfs_item_key_to_cpu(eb, &ref->key_for_search, 0);
 		else
 			btrfs_node_key_to_cpu(eb, &ref->key_for_search, 0);
-		btrfs_tree_read_unlock(eb);
+		if (lock)
+			btrfs_tree_read_unlock(eb);
 		free_extent_buffer(eb);
 		prelim_ref_insert(fs_info, &preftrees->indirect, ref, NULL);
 		cond_resched();
@@ -1227,7 +1229,7 @@ again:
 
 	btrfs_release_path(path);
 
-	ret = add_missing_keys(fs_info, &preftrees);
+	ret = add_missing_keys(fs_info, &preftrees, path->skip_locking == 0);
 	if (ret)
 		goto out;
 
@@ -1288,11 +1290,15 @@ again:
 					ret = -EIO;
 					goto out;
 				}
-				btrfs_tree_read_lock(eb);
-				btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+
+				if (!path->skip_locking) {
+					btrfs_tree_read_lock(eb);
+					btrfs_set_lock_blocking_read(eb);
+				}
 				ret = find_extent_in_eb(eb, bytenr,
 							*extent_item_pos, &eie, ignore_offset);
-				btrfs_tree_read_unlock_blocking(eb);
+				if (!path->skip_locking)
+					btrfs_tree_read_unlock_blocking(eb);
 				free_extent_buffer(eb);
 				if (ret < 0)
 					goto out;
@@ -1650,7 +1656,7 @@ char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
 		/* make sure we can use eb after releasing the path */
 		if (eb != eb_in) {
 			if (!path->skip_locking)
-				btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+				btrfs_set_lock_blocking_read(eb);
 			path->nodes[0] = NULL;
 			path->locks[0] = 0;
 		}
diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index 548057630b69..eb8e20b740d6 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -730,6 +730,28 @@ struct heuristic_ws {
 	struct list_head list;
 };
 
+static struct workspace_manager heuristic_wsm;
+
+static void heuristic_init_workspace_manager(void)
+{
+	btrfs_init_workspace_manager(&heuristic_wsm, &btrfs_heuristic_compress);
+}
+
+static void heuristic_cleanup_workspace_manager(void)
+{
+	btrfs_cleanup_workspace_manager(&heuristic_wsm);
+}
+
+static struct list_head *heuristic_get_workspace(unsigned int level)
+{
+	return btrfs_get_workspace(&heuristic_wsm, level);
+}
+
+static void heuristic_put_workspace(struct list_head *ws)
+{
+	btrfs_put_workspace(&heuristic_wsm, ws);
+}
+
 static void free_heuristic_ws(struct list_head *ws)
 {
 	struct heuristic_ws *workspace;
@@ -742,7 +764,7 @@ static void free_heuristic_ws(struct list_head *ws)
 	kfree(workspace);
 }
 
-static struct list_head *alloc_heuristic_ws(void)
+static struct list_head *alloc_heuristic_ws(unsigned int level)
 {
 	struct heuristic_ws *ws;
 
@@ -769,65 +791,59 @@ fail:
 	return ERR_PTR(-ENOMEM);
 }
 
-struct workspaces_list {
-	struct list_head idle_ws;
-	spinlock_t ws_lock;
-	/* Number of free workspaces */
-	int free_ws;
-	/* Total number of allocated workspaces */
-	atomic_t total_ws;
-	/* Waiters for a free workspace */
-	wait_queue_head_t ws_wait;
+const struct btrfs_compress_op btrfs_heuristic_compress = {
+	.init_workspace_manager = heuristic_init_workspace_manager,
+	.cleanup_workspace_manager = heuristic_cleanup_workspace_manager,
+	.get_workspace = heuristic_get_workspace,
+	.put_workspace = heuristic_put_workspace,
+	.alloc_workspace = alloc_heuristic_ws,
+	.free_workspace = free_heuristic_ws,
 };
 
-static struct workspaces_list btrfs_comp_ws[BTRFS_COMPRESS_TYPES];
-
-static struct workspaces_list btrfs_heuristic_ws;
-
 static const struct btrfs_compress_op * const btrfs_compress_op[] = {
+	/* The heuristic is represented as compression type 0 */
+	&btrfs_heuristic_compress,
 	&btrfs_zlib_compress,
 	&btrfs_lzo_compress,
 	&btrfs_zstd_compress,
 };
 
-void __init btrfs_init_compress(void)
+void btrfs_init_workspace_manager(struct workspace_manager *wsm,
+				  const struct btrfs_compress_op *ops)
 {
 	struct list_head *workspace;
-	int i;
 
-	INIT_LIST_HEAD(&btrfs_heuristic_ws.idle_ws);
-	spin_lock_init(&btrfs_heuristic_ws.ws_lock);
-	atomic_set(&btrfs_heuristic_ws.total_ws, 0);
-	init_waitqueue_head(&btrfs_heuristic_ws.ws_wait);
+	wsm->ops = ops;
 
-	workspace = alloc_heuristic_ws();
+	INIT_LIST_HEAD(&wsm->idle_ws);
+	spin_lock_init(&wsm->ws_lock);
+	atomic_set(&wsm->total_ws, 0);
+	init_waitqueue_head(&wsm->ws_wait);
+
+	/*
+	 * Preallocate one workspace for each compression type so we can
+	 * guarantee forward progress in the worst case
+	 */
+	workspace = wsm->ops->alloc_workspace(0);
 	if (IS_ERR(workspace)) {
 		pr_warn(
-	"BTRFS: cannot preallocate heuristic workspace, will try later\n");
+	"BTRFS: cannot preallocate compression workspace, will try later\n");
 	} else {
-		atomic_set(&btrfs_heuristic_ws.total_ws, 1);
-		btrfs_heuristic_ws.free_ws = 1;
-		list_add(workspace, &btrfs_heuristic_ws.idle_ws);
+		atomic_set(&wsm->total_ws, 1);
+		wsm->free_ws = 1;
+		list_add(workspace, &wsm->idle_ws);
 	}
+}
 
-	for (i = 0; i < BTRFS_COMPRESS_TYPES; i++) {
-		INIT_LIST_HEAD(&btrfs_comp_ws[i].idle_ws);
-		spin_lock_init(&btrfs_comp_ws[i].ws_lock);
-		atomic_set(&btrfs_comp_ws[i].total_ws, 0);
-		init_waitqueue_head(&btrfs_comp_ws[i].ws_wait);
+void btrfs_cleanup_workspace_manager(struct workspace_manager *wsman)
+{
+	struct list_head *ws;
 
-		/*
-		 * Preallocate one workspace for each compression type so
-		 * we can guarantee forward progress in the worst case
-		 */
-		workspace = btrfs_compress_op[i]->alloc_workspace();
-		if (IS_ERR(workspace)) {
-			pr_warn("BTRFS: cannot preallocate compression workspace, will try later\n");
-		} else {
-			atomic_set(&btrfs_comp_ws[i].total_ws, 1);
-			btrfs_comp_ws[i].free_ws = 1;
-			list_add(workspace, &btrfs_comp_ws[i].idle_ws);
-		}
+	while (!list_empty(&wsman->idle_ws)) {
+		ws = wsman->idle_ws.next;
+		list_del(ws);
+		wsman->ops->free_workspace(ws);
+		atomic_dec(&wsman->total_ws);
 	}
 }
 
@@ -837,11 +853,11 @@ void __init btrfs_init_compress(void)
  * Preallocation makes a forward progress guarantees and we do not return
  * errors.
  */
-static struct list_head *__find_workspace(int type, bool heuristic)
+struct list_head *btrfs_get_workspace(struct workspace_manager *wsm,
+				      unsigned int level)
 {
 	struct list_head *workspace;
 	int cpus = num_online_cpus();
-	int idx = type - 1;
 	unsigned nofs_flag;
 	struct list_head *idle_ws;
 	spinlock_t *ws_lock;
@@ -849,19 +865,11 @@ static struct list_head *__find_workspace(int type, bool heuristic)
 	wait_queue_head_t *ws_wait;
 	int *free_ws;
 
-	if (heuristic) {
-		idle_ws	 = &btrfs_heuristic_ws.idle_ws;
-		ws_lock	 = &btrfs_heuristic_ws.ws_lock;
-		total_ws = &btrfs_heuristic_ws.total_ws;
-		ws_wait	 = &btrfs_heuristic_ws.ws_wait;
-		free_ws	 = &btrfs_heuristic_ws.free_ws;
-	} else {
-		idle_ws	 = &btrfs_comp_ws[idx].idle_ws;
-		ws_lock	 = &btrfs_comp_ws[idx].ws_lock;
-		total_ws = &btrfs_comp_ws[idx].total_ws;
-		ws_wait	 = &btrfs_comp_ws[idx].ws_wait;
-		free_ws	 = &btrfs_comp_ws[idx].free_ws;
-	}
+	idle_ws	 = &wsm->idle_ws;
+	ws_lock	 = &wsm->ws_lock;
+	total_ws = &wsm->total_ws;
+	ws_wait	 = &wsm->ws_wait;
+	free_ws	 = &wsm->free_ws;
 
 again:
 	spin_lock(ws_lock);
@@ -892,10 +900,7 @@ again:
 	 * context of btrfs_compress_bio/btrfs_compress_pages
 	 */
 	nofs_flag = memalloc_nofs_save();
-	if (heuristic)
-		workspace = alloc_heuristic_ws();
-	else
-		workspace = btrfs_compress_op[idx]->alloc_workspace();
+	workspace = wsm->ops->alloc_workspace(level);
 	memalloc_nofs_restore(nofs_flag);
 
 	if (IS_ERR(workspace)) {
@@ -926,85 +931,47 @@ again:
 	return workspace;
 }
 
-static struct list_head *find_workspace(int type)
+static struct list_head *get_workspace(int type, int level)
 {
-	return __find_workspace(type, false);
+	return btrfs_compress_op[type]->get_workspace(level);
 }
 
 /*
  * put a workspace struct back on the list or free it if we have enough
  * idle ones sitting around
  */
-static void __free_workspace(int type, struct list_head *workspace,
-			     bool heuristic)
+void btrfs_put_workspace(struct workspace_manager *wsm, struct list_head *ws)
 {
-	int idx = type - 1;
 	struct list_head *idle_ws;
 	spinlock_t *ws_lock;
 	atomic_t *total_ws;
 	wait_queue_head_t *ws_wait;
 	int *free_ws;
 
-	if (heuristic) {
-		idle_ws	 = &btrfs_heuristic_ws.idle_ws;
-		ws_lock	 = &btrfs_heuristic_ws.ws_lock;
-		total_ws = &btrfs_heuristic_ws.total_ws;
-		ws_wait	 = &btrfs_heuristic_ws.ws_wait;
-		free_ws	 = &btrfs_heuristic_ws.free_ws;
-	} else {
-		idle_ws	 = &btrfs_comp_ws[idx].idle_ws;
-		ws_lock	 = &btrfs_comp_ws[idx].ws_lock;
-		total_ws = &btrfs_comp_ws[idx].total_ws;
-		ws_wait	 = &btrfs_comp_ws[idx].ws_wait;
-		free_ws	 = &btrfs_comp_ws[idx].free_ws;
-	}
+	idle_ws	 = &wsm->idle_ws;
+	ws_lock	 = &wsm->ws_lock;
+	total_ws = &wsm->total_ws;
+	ws_wait	 = &wsm->ws_wait;
+	free_ws	 = &wsm->free_ws;
 
 	spin_lock(ws_lock);
 	if (*free_ws <= num_online_cpus()) {
-		list_add(workspace, idle_ws);
+		list_add(ws, idle_ws);
 		(*free_ws)++;
 		spin_unlock(ws_lock);
 		goto wake;
 	}
 	spin_unlock(ws_lock);
 
-	if (heuristic)
-		free_heuristic_ws(workspace);
-	else
-		btrfs_compress_op[idx]->free_workspace(workspace);
+	wsm->ops->free_workspace(ws);
 	atomic_dec(total_ws);
 wake:
 	cond_wake_up(ws_wait);
 }
 
-static void free_workspace(int type, struct list_head *ws)
+static void put_workspace(int type, struct list_head *ws)
 {
-	return __free_workspace(type, ws, false);
-}
-
-/*
- * cleanup function for module exit
- */
-static void free_workspaces(void)
-{
-	struct list_head *workspace;
-	int i;
-
-	while (!list_empty(&btrfs_heuristic_ws.idle_ws)) {
-		workspace = btrfs_heuristic_ws.idle_ws.next;
-		list_del(workspace);
-		free_heuristic_ws(workspace);
-		atomic_dec(&btrfs_heuristic_ws.total_ws);
-	}
-
-	for (i = 0; i < BTRFS_COMPRESS_TYPES; i++) {
-		while (!list_empty(&btrfs_comp_ws[i].idle_ws)) {
-			workspace = btrfs_comp_ws[i].idle_ws.next;
-			list_del(workspace);
-			btrfs_compress_op[i]->free_workspace(workspace);
-			atomic_dec(&btrfs_comp_ws[i].total_ws);
-		}
-	}
+	return btrfs_compress_op[type]->put_workspace(ws);
 }
 
 /*
@@ -1036,18 +1003,17 @@ int btrfs_compress_pages(unsigned int type_level, struct address_space *mapping,
 			 unsigned long *total_in,
 			 unsigned long *total_out)
 {
+	int type = btrfs_compress_type(type_level);
+	int level = btrfs_compress_level(type_level);
 	struct list_head *workspace;
 	int ret;
-	int type = type_level & 0xF;
-
-	workspace = find_workspace(type);
 
-	btrfs_compress_op[type - 1]->set_level(workspace, type_level);
-	ret = btrfs_compress_op[type-1]->compress_pages(workspace, mapping,
+	workspace = get_workspace(type, level);
+	ret = btrfs_compress_op[type]->compress_pages(workspace, mapping,
 						      start, pages,
 						      out_pages,
 						      total_in, total_out);
-	free_workspace(type, workspace);
+	put_workspace(type, workspace);
 	return ret;
 }
 
@@ -1071,9 +1037,9 @@ static int btrfs_decompress_bio(struct compressed_bio *cb)
 	int ret;
 	int type = cb->compress_type;
 
-	workspace = find_workspace(type);
-	ret = btrfs_compress_op[type - 1]->decompress_bio(workspace, cb);
-	free_workspace(type, workspace);
+	workspace = get_workspace(type, 0);
+	ret = btrfs_compress_op[type]->decompress_bio(workspace, cb);
+	put_workspace(type, workspace);
 
 	return ret;
 }
@@ -1089,19 +1055,29 @@ int btrfs_decompress(int type, unsigned char *data_in, struct page *dest_page,
 	struct list_head *workspace;
 	int ret;
 
-	workspace = find_workspace(type);
-
-	ret = btrfs_compress_op[type-1]->decompress(workspace, data_in,
+	workspace = get_workspace(type, 0);
+	ret = btrfs_compress_op[type]->decompress(workspace, data_in,
 						  dest_page, start_byte,
 						  srclen, destlen);
+	put_workspace(type, workspace);
 
-	free_workspace(type, workspace);
 	return ret;
 }
 
+void __init btrfs_init_compress(void)
+{
+	int i;
+
+	for (i = 0; i < BTRFS_NR_WORKSPACE_MANAGERS; i++)
+		btrfs_compress_op[i]->init_workspace_manager();
+}
+
 void __cold btrfs_exit_compress(void)
 {
-	free_workspaces();
+	int i;
+
+	for (i = 0; i < BTRFS_NR_WORKSPACE_MANAGERS; i++)
+		btrfs_compress_op[i]->cleanup_workspace_manager();
 }
 
 /*
@@ -1512,7 +1488,7 @@ static void heuristic_collect_sample(struct inode *inode, u64 start, u64 end,
  */
 int btrfs_compress_heuristic(struct inode *inode, u64 start, u64 end)
 {
-	struct list_head *ws_list = __find_workspace(0, true);
+	struct list_head *ws_list = get_workspace(0, 0);
 	struct heuristic_ws *ws;
 	u32 i;
 	u8 byte;
@@ -1581,18 +1557,29 @@ int btrfs_compress_heuristic(struct inode *inode, u64 start, u64 end)
 	}
 
 out:
-	__free_workspace(0, ws_list, true);
+	put_workspace(0, ws_list);
 	return ret;
 }
 
-unsigned int btrfs_compress_str2level(const char *str)
+/*
+ * Convert the compression suffix (eg. after "zlib" starting with ":") to
+ * level, unrecognized string will set the default level
+ */
+unsigned int btrfs_compress_str2level(unsigned int type, const char *str)
 {
-	if (strncmp(str, "zlib", 4) != 0)
+	unsigned int level = 0;
+	int ret;
+
+	if (!type)
 		return 0;
 
-	/* Accepted form: zlib:1 up to zlib:9 and nothing left after the number */
-	if (str[4] == ':' && '1' <= str[5] && str[5] <= '9' && str[6] == 0)
-		return str[5] - '0';
+	if (str[0] == ':') {
+		ret = kstrtouint(str + 1, 10, &level);
+		if (ret)
+			level = 0;
+	}
+
+	level = btrfs_compress_op[type]->set_level(level);
 
-	return BTRFS_ZLIB_DEFAULT_LEVEL;
+	return level;
 }
diff --git a/fs/btrfs/compression.h b/fs/btrfs/compression.h
index ddda9b80bf20..9976fe0f7526 100644
--- a/fs/btrfs/compression.h
+++ b/fs/btrfs/compression.h
@@ -64,6 +64,16 @@ struct compressed_bio {
 	u32 sums;
 };
 
+static inline unsigned int btrfs_compress_type(unsigned int type_level)
+{
+	return (type_level & 0xF);
+}
+
+static inline unsigned int btrfs_compress_level(unsigned int type_level)
+{
+	return ((type_level & 0xF0) >> 4);
+}
+
 void __init btrfs_init_compress(void);
 void __cold btrfs_exit_compress(void);
 
@@ -87,7 +97,7 @@ blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start,
 blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 				 int mirror_num, unsigned long bio_flags);
 
-unsigned btrfs_compress_str2level(const char *str);
+unsigned int btrfs_compress_str2level(unsigned int type, const char *str);
 
 enum btrfs_compression_type {
 	BTRFS_COMPRESS_NONE  = 0,
@@ -97,8 +107,35 @@ enum btrfs_compression_type {
 	BTRFS_COMPRESS_TYPES = 3,
 };
 
+struct workspace_manager {
+	const struct btrfs_compress_op *ops;
+	struct list_head idle_ws;
+	spinlock_t ws_lock;
+	/* Number of free workspaces */
+	int free_ws;
+	/* Total number of allocated workspaces */
+	atomic_t total_ws;
+	/* Waiters for a free workspace */
+	wait_queue_head_t ws_wait;
+};
+
+void btrfs_init_workspace_manager(struct workspace_manager *wsm,
+				  const struct btrfs_compress_op *ops);
+struct list_head *btrfs_get_workspace(struct workspace_manager *wsm,
+				      unsigned int level);
+void btrfs_put_workspace(struct workspace_manager *wsm, struct list_head *ws);
+void btrfs_cleanup_workspace_manager(struct workspace_manager *wsm);
+
 struct btrfs_compress_op {
-	struct list_head *(*alloc_workspace)(void);
+	void (*init_workspace_manager)(void);
+
+	void (*cleanup_workspace_manager)(void);
+
+	struct list_head *(*get_workspace)(unsigned int level);
+
+	void (*put_workspace)(struct list_head *ws);
+
+	struct list_head *(*alloc_workspace)(unsigned int level);
 
 	void (*free_workspace)(struct list_head *workspace);
 
@@ -119,9 +156,18 @@ struct btrfs_compress_op {
 			  unsigned long start_byte,
 			  size_t srclen, size_t destlen);
 
-	void (*set_level)(struct list_head *ws, unsigned int type);
+	/*
+	 * This bounds the level set by the user to be within range of a
+	 * particular compression type.  It returns the level that will be used
+	 * if the level is out of bounds or the default if 0 is passed in.
+	 */
+	unsigned int (*set_level)(unsigned int level);
 };
 
+/* The heuristic workspaces are managed via the 0th workspace manager */
+#define BTRFS_NR_WORKSPACE_MANAGERS	(BTRFS_COMPRESS_TYPES + 1)
+
+extern const struct btrfs_compress_op btrfs_heuristic_compress;
 extern const struct btrfs_compress_op btrfs_zlib_compress;
 extern const struct btrfs_compress_op btrfs_lzo_compress;
 extern const struct btrfs_compress_op btrfs_zstd_compress;
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 5a6c39b44c84..324df36d28bf 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -13,6 +13,7 @@
 #include "print-tree.h"
 #include "locking.h"
 #include "volumes.h"
+#include "qgroup.h"
 
 static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
 		      *root, struct btrfs_path *path, int level);
@@ -45,11 +46,18 @@ noinline void btrfs_set_path_blocking(struct btrfs_path *p)
 	for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
 		if (!p->nodes[i] || !p->locks[i])
 			continue;
-		btrfs_set_lock_blocking_rw(p->nodes[i], p->locks[i]);
-		if (p->locks[i] == BTRFS_READ_LOCK)
+		/*
+		 * If we currently have a spinning reader or writer lock this
+		 * will bump the count of blocking holders and drop the
+		 * spinlock.
+		 */
+		if (p->locks[i] == BTRFS_READ_LOCK) {
+			btrfs_set_lock_blocking_read(p->nodes[i]);
 			p->locks[i] = BTRFS_READ_LOCK_BLOCKING;
-		else if (p->locks[i] == BTRFS_WRITE_LOCK)
+		} else if (p->locks[i] == BTRFS_WRITE_LOCK) {
+			btrfs_set_lock_blocking_write(p->nodes[i]);
 			p->locks[i] = BTRFS_WRITE_LOCK_BLOCKING;
+		}
 	}
 }
 
@@ -1288,7 +1296,7 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 		return eb;
 
 	btrfs_set_path_blocking(path);
-	btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+	btrfs_set_lock_blocking_read(eb);
 
 	if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
 		BUG_ON(tm->slot != 0);
@@ -1378,7 +1386,7 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 		free_extent_buffer(eb_root);
 		eb = alloc_dummy_extent_buffer(fs_info, logical);
 	} else {
-		btrfs_set_lock_blocking_rw(eb_root, BTRFS_READ_LOCK);
+		btrfs_set_lock_blocking_read(eb_root);
 		eb = btrfs_clone_extent_buffer(eb_root);
 		btrfs_tree_read_unlock_blocking(eb_root);
 		free_extent_buffer(eb_root);
@@ -1486,9 +1494,16 @@ noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
 	search_start = buf->start & ~((u64)SZ_1G - 1);
 
 	if (parent)
-		btrfs_set_lock_blocking(parent);
-	btrfs_set_lock_blocking(buf);
+		btrfs_set_lock_blocking_write(parent);
+	btrfs_set_lock_blocking_write(buf);
 
+	/*
+	 * Before CoWing this block for later modification, check if it's
+	 * the subtree root and do the delayed subtree trace if needed.
+	 *
+	 * Also We don't care about the error, as it's handled internally.
+	 */
+	btrfs_qgroup_trace_subtree_after_cow(trans, root, buf);
 	ret = __btrfs_cow_block(trans, root, buf, parent,
 				 parent_slot, cow_ret, search_start, 0);
 
@@ -1582,7 +1597,7 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
 	if (parent_nritems <= 1)
 		return 0;
 
-	btrfs_set_lock_blocking(parent);
+	btrfs_set_lock_blocking_write(parent);
 
 	for (i = start_slot; i <= end_slot; i++) {
 		struct btrfs_key first_key;
@@ -1641,7 +1656,7 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
 			search_start = last_block;
 
 		btrfs_tree_lock(cur);
-		btrfs_set_lock_blocking(cur);
+		btrfs_set_lock_blocking_write(cur);
 		err = __btrfs_cow_block(trans, root, cur, parent, i,
 					&cur, search_start,
 					min(16 * blocksize,
@@ -1856,7 +1871,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 		}
 
 		btrfs_tree_lock(child);
-		btrfs_set_lock_blocking(child);
+		btrfs_set_lock_blocking_write(child);
 		ret = btrfs_cow_block(trans, root, child, mid, 0, &child);
 		if (ret) {
 			btrfs_tree_unlock(child);
@@ -1894,7 +1909,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 
 	if (left) {
 		btrfs_tree_lock(left);
-		btrfs_set_lock_blocking(left);
+		btrfs_set_lock_blocking_write(left);
 		wret = btrfs_cow_block(trans, root, left,
 				       parent, pslot - 1, &left);
 		if (wret) {
@@ -1909,7 +1924,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 
 	if (right) {
 		btrfs_tree_lock(right);
-		btrfs_set_lock_blocking(right);
+		btrfs_set_lock_blocking_write(right);
 		wret = btrfs_cow_block(trans, root, right,
 				       parent, pslot + 1, &right);
 		if (wret) {
@@ -2072,7 +2087,7 @@ static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
 		u32 left_nr;
 
 		btrfs_tree_lock(left);
-		btrfs_set_lock_blocking(left);
+		btrfs_set_lock_blocking_write(left);
 
 		left_nr = btrfs_header_nritems(left);
 		if (left_nr >= BTRFS_NODEPTRS_PER_BLOCK(fs_info) - 1) {
@@ -2127,7 +2142,7 @@ static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
 		u32 right_nr;
 
 		btrfs_tree_lock(right);
-		btrfs_set_lock_blocking(right);
+		btrfs_set_lock_blocking_write(right);
 
 		right_nr = btrfs_header_nritems(right);
 		if (right_nr >= BTRFS_NODEPTRS_PER_BLOCK(fs_info) - 1) {
@@ -2529,26 +2544,6 @@ done:
 	return ret;
 }
 
-static void key_search_validate(struct extent_buffer *b,
-				const struct btrfs_key *key,
-				int level)
-{
-#ifdef CONFIG_BTRFS_ASSERT
-	struct btrfs_disk_key disk_key;
-
-	btrfs_cpu_key_to_disk(&disk_key, key);
-
-	if (level == 0)
-		ASSERT(!memcmp_extent_buffer(b, &disk_key,
-		    offsetof(struct btrfs_leaf, items[0].key),
-		    sizeof(disk_key)));
-	else
-		ASSERT(!memcmp_extent_buffer(b, &disk_key,
-		    offsetof(struct btrfs_node, ptrs[0].key),
-		    sizeof(disk_key)));
-#endif
-}
-
 static int key_search(struct extent_buffer *b, const struct btrfs_key *key,
 		      int level, int *prev_cmp, int *slot)
 {
@@ -2557,7 +2552,6 @@ static int key_search(struct extent_buffer *b, const struct btrfs_key *key,
 		return *prev_cmp;
 	}
 
-	key_search_validate(b, key, level);
 	*slot = 0;
 
 	return 0;
@@ -3005,6 +2999,8 @@ again:
 		 */
 		prev_cmp = -1;
 		ret = key_search(b, key, level, &prev_cmp, &slot);
+		if (ret < 0)
+			goto done;
 
 		if (level != 0) {
 			int dec = 0;
@@ -3771,7 +3767,7 @@ static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
 		return 1;
 
 	btrfs_tree_lock(right);
-	btrfs_set_lock_blocking(right);
+	btrfs_set_lock_blocking_write(right);
 
 	free_space = btrfs_leaf_free_space(fs_info, right);
 	if (free_space < data_size)
@@ -4005,7 +4001,7 @@ static int push_leaf_left(struct btrfs_trans_handle *trans, struct btrfs_root
 		return 1;
 
 	btrfs_tree_lock(left);
-	btrfs_set_lock_blocking(left);
+	btrfs_set_lock_blocking_write(left);
 
 	free_space = btrfs_leaf_free_space(fs_info, left);
 	if (free_space < data_size) {
@@ -5156,6 +5152,10 @@ again:
 		nritems = btrfs_header_nritems(cur);
 		level = btrfs_header_level(cur);
 		sret = btrfs_bin_search(cur, min_key, level, &slot);
+		if (sret < 0) {
+			ret = sret;
+			goto out;
+		}
 
 		/* at the lowest level, we're done, setup the path and exit */
 		if (level == path->lowest_level) {
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 7a2a2621f0d9..85140913c0f5 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -934,7 +934,8 @@ struct btrfs_fs_info {
 
 	spinlock_t delayed_iput_lock;
 	struct list_head delayed_iputs;
-	struct mutex cleaner_delayed_iput_mutex;
+	atomic_t nr_delayed_iputs;
+	wait_queue_head_t delayed_iputs_wait;
 
 	/* this protects tree_mod_seq_list */
 	spinlock_t tree_mod_seq_lock;
@@ -1074,10 +1075,13 @@ struct btrfs_fs_info {
 	atomic_t scrubs_paused;
 	atomic_t scrub_cancel_req;
 	wait_queue_head_t scrub_pause_wait;
-	int scrub_workers_refcnt;
+	/*
+	 * The worker pointers are NULL iff the refcount is 0, ie. scrub is not
+	 * running.
+	 */
+	refcount_t scrub_workers_refcnt;
 	struct btrfs_workqueue *scrub_workers;
 	struct btrfs_workqueue *scrub_wr_completion_workers;
-	struct btrfs_workqueue *scrub_nocow_workers;
 	struct btrfs_workqueue *scrub_parity_workers;
 
 #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
@@ -1199,6 +1203,24 @@ enum {
 	BTRFS_ROOT_MULTI_LOG_TASKS,
 	BTRFS_ROOT_DIRTY,
 	BTRFS_ROOT_DELETING,
+
+	/*
+	 * Reloc tree is orphan, only kept here for qgroup delayed subtree scan
+	 *
+	 * Set for the subvolume tree owning the reloc tree.
+	 */
+	BTRFS_ROOT_DEAD_RELOC_TREE,
+};
+
+/*
+ * Record swapped tree blocks of a subvolume tree for delayed subtree trace
+ * code. For detail check comment in fs/btrfs/qgroup.c.
+ */
+struct btrfs_qgroup_swapped_blocks {
+	spinlock_t lock;
+	/* RM_EMPTY_ROOT() of above blocks[] */
+	bool swapped;
+	struct rb_root blocks[BTRFS_MAX_LEVEL];
 };
 
 /*
@@ -1312,6 +1334,14 @@ struct btrfs_root {
 	u64 nr_ordered_extents;
 
 	/*
+	 * Not empty if this subvolume root has gone through tree block swap
+	 * (relocation)
+	 *
+	 * Will be used by reloc_control::dirty_subvol_roots.
+	 */
+	struct list_head reloc_dirty_list;
+
+	/*
 	 * Number of currently running SEND ioctls to prevent
 	 * manipulation with the read-only status via SUBVOL_SETFLAGS
 	 */
@@ -1328,6 +1358,9 @@ struct btrfs_root {
 	/* Number of active swapfiles */
 	atomic_t nr_swapfiles;
 
+	/* Record pairs of swapped blocks for qgroup */
+	struct btrfs_qgroup_swapped_blocks swapped_blocks;
+
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 	u64 alloc_bytenr;
 #endif
@@ -2775,7 +2808,8 @@ enum btrfs_flush_state {
 	FLUSH_DELALLOC		=	5,
 	FLUSH_DELALLOC_WAIT	=	6,
 	ALLOC_CHUNK		=	7,
-	COMMIT_TRANS		=	8,
+	ALLOC_CHUNK_FORCE	=	8,
+	COMMIT_TRANS		=	9,
 };
 
 int btrfs_alloc_data_chunk_ondemand(struct btrfs_inode *inode, u64 bytes);
@@ -3181,8 +3215,7 @@ void btrfs_extent_item_to_extent_map(struct btrfs_inode *inode,
 
 /* inode.c */
 struct extent_map *btrfs_get_extent_fiemap(struct btrfs_inode *inode,
-		struct page *page, size_t pg_offset, u64 start,
-		u64 len, int create);
+					   u64 start, u64 len);
 noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 			      u64 *orig_start, u64 *orig_block_len,
 			      u64 *ram_bytes);
@@ -3254,6 +3287,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root);
 int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size);
 void btrfs_add_delayed_iput(struct inode *inode);
 void btrfs_run_delayed_iputs(struct btrfs_fs_info *fs_info);
+int btrfs_wait_on_delayed_iputs(struct btrfs_fs_info *fs_info);
 int btrfs_prealloc_file_range(struct inode *inode, int mode,
 			      u64 start, u64 num_bytes, u64 min_size,
 			      loff_t actual_len, u64 *alloc_hint);
@@ -3261,7 +3295,7 @@ int btrfs_prealloc_file_range_trans(struct inode *inode,
 				    struct btrfs_trans_handle *trans, int mode,
 				    u64 start, u64 num_bytes, u64 min_size,
 				    loff_t actual_len, u64 *alloc_hint);
-int btrfs_run_delalloc_range(void *private_data, struct page *locked_page,
+int btrfs_run_delalloc_range(struct inode *inode, struct page *locked_page,
 		u64 start, u64 end, int *page_started, unsigned long *nr_written,
 		struct writeback_control *wbc);
 int btrfs_writepage_cow_fixup(struct page *page, u64 start, u64 end);
@@ -3490,21 +3524,18 @@ do {								\
 	rcu_read_unlock();					\
 } while (0)
 
-#ifdef CONFIG_BTRFS_ASSERT
-
 __cold
 static inline void assfail(const char *expr, const char *file, int line)
 {
-	pr_err("assertion failed: %s, file: %s, line: %d\n",
-	       expr, file, line);
-	BUG();
+	if (IS_ENABLED(CONFIG_BTRFS_ASSERT)) {
+		pr_err("assertion failed: %s, file: %s, line: %d\n",
+		       expr, file, line);
+		BUG();
+	}
 }
 
 #define ASSERT(expr)	\
 	(likely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__))
-#else
-#define ASSERT(expr)	((void)0)
-#endif
 
 /*
  * Use that for functions that are conditionally exported for sanity tests but
diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c
index cad36c99a483..7d2a413df90d 100644
--- a/fs/btrfs/delayed-ref.c
+++ b/fs/btrfs/delayed-ref.c
@@ -602,17 +602,14 @@ static void init_delayed_ref_head(struct btrfs_delayed_ref_head *head_ref,
 	RB_CLEAR_NODE(&head_ref->href_node);
 	head_ref->processing = 0;
 	head_ref->total_ref_mod = count_mod;
-	head_ref->qgroup_reserved = 0;
-	head_ref->qgroup_ref_root = 0;
 	spin_lock_init(&head_ref->lock);
 	mutex_init(&head_ref->mutex);
 
 	if (qrecord) {
 		if (ref_root && reserved) {
-			head_ref->qgroup_ref_root = ref_root;
-			head_ref->qgroup_reserved = reserved;
+			qrecord->data_rsv = reserved;
+			qrecord->data_rsv_refroot = ref_root;
 		}
-
 		qrecord->bytenr = bytenr;
 		qrecord->num_bytes = num_bytes;
 		qrecord->old_roots = NULL;
@@ -651,10 +648,6 @@ add_delayed_ref_head(struct btrfs_trans_handle *trans,
 	existing = htree_insert(&delayed_refs->href_root,
 				&head_ref->href_node);
 	if (existing) {
-		WARN_ON(qrecord && head_ref->qgroup_ref_root
-			&& head_ref->qgroup_reserved
-			&& existing->qgroup_ref_root
-			&& existing->qgroup_reserved);
 		update_existing_head_ref(trans, existing, head_ref,
 					 old_ref_mod);
 		/*
@@ -770,7 +763,7 @@ int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans,
 
 	if (test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) &&
 	    is_fstree(ref_root)) {
-		record = kmalloc(sizeof(*record), GFP_NOFS);
+		record = kzalloc(sizeof(*record), GFP_NOFS);
 		if (!record) {
 			kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref);
 			kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref);
@@ -867,7 +860,7 @@ int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans,
 
 	if (test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags) &&
 	    is_fstree(ref_root)) {
-		record = kmalloc(sizeof(*record), GFP_NOFS);
+		record = kzalloc(sizeof(*record), GFP_NOFS);
 		if (!record) {
 			kmem_cache_free(btrfs_delayed_data_ref_cachep, ref);
 			kmem_cache_free(btrfs_delayed_ref_head_cachep,
diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h
index d2af974f68a1..70606da440aa 100644
--- a/fs/btrfs/delayed-ref.h
+++ b/fs/btrfs/delayed-ref.h
@@ -103,17 +103,6 @@ struct btrfs_delayed_ref_head {
 	int ref_mod;
 
 	/*
-	 * For qgroup reserved space freeing.
-	 *
-	 * ref_root and reserved will be recorded after
-	 * BTRFS_ADD_DELAYED_EXTENT is called.
-	 * And will be used to free reserved qgroup space at
-	 * run_delayed_refs() time.
-	 */
-	u64 qgroup_ref_root;
-	u64 qgroup_reserved;
-
-	/*
 	 * when a new extent is allocated, it is just reserved in memory
 	 * The actual extent isn't inserted into the extent allocation tree
 	 * until the delayed ref is processed.  must_insert_reserved is
diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c
index 8750c835f535..ee193c5222b2 100644
--- a/fs/btrfs/dev-replace.c
+++ b/fs/btrfs/dev-replace.c
@@ -111,11 +111,11 @@ no_valid_dev_replace_entry_found:
 		break;
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED:
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED:
-		dev_replace->srcdev = btrfs_find_device(fs_info, src_devid,
-							NULL, NULL);
-		dev_replace->tgtdev = btrfs_find_device(fs_info,
+		dev_replace->srcdev = btrfs_find_device(fs_info->fs_devices,
+						src_devid, NULL, NULL, true);
+		dev_replace->tgtdev = btrfs_find_device(fs_info->fs_devices,
 							BTRFS_DEV_REPLACE_DEVID,
-							NULL, NULL);
+							NULL, NULL, true);
 		/*
 		 * allow 'btrfs dev replace_cancel' if src/tgt device is
 		 * missing
@@ -862,6 +862,7 @@ int btrfs_dev_replace_cancel(struct btrfs_fs_info *fs_info)
 			btrfs_destroy_dev_replace_tgtdev(tgt_device);
 		break;
 	default:
+		up_write(&dev_replace->rwsem);
 		result = -EINVAL;
 	}
 
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 6a2a2a951705..5216e7b3f9ad 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -17,6 +17,7 @@
 #include <linux/semaphore.h>
 #include <linux/error-injection.h>
 #include <linux/crc32c.h>
+#include <linux/sched/mm.h>
 #include <asm/unaligned.h>
 #include "ctree.h"
 #include "disk-io.h"
@@ -341,7 +342,7 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 
 	if (need_lock) {
 		btrfs_tree_read_lock(eb);
-		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+		btrfs_set_lock_blocking_read(eb);
 	}
 
 	lock_extent_bits(io_tree, eb->start, eb->start + eb->len - 1,
@@ -1120,7 +1121,7 @@ void clean_tree_block(struct btrfs_fs_info *fs_info,
 						 -buf->len,
 						 fs_info->dirty_metadata_batch);
 			/* ugh, clear_extent_buffer_dirty needs to lock the page */
-			btrfs_set_lock_blocking(buf);
+			btrfs_set_lock_blocking_write(buf);
 			clear_extent_buffer_dirty(buf);
 		}
 	}
@@ -1175,6 +1176,7 @@ static void __setup_root(struct btrfs_root *root, struct btrfs_fs_info *fs_info,
 	INIT_LIST_HEAD(&root->delalloc_root);
 	INIT_LIST_HEAD(&root->ordered_extents);
 	INIT_LIST_HEAD(&root->ordered_root);
+	INIT_LIST_HEAD(&root->reloc_dirty_list);
 	INIT_LIST_HEAD(&root->logged_list[0]);
 	INIT_LIST_HEAD(&root->logged_list[1]);
 	spin_lock_init(&root->inode_lock);
@@ -1218,6 +1220,7 @@ static void __setup_root(struct btrfs_root *root, struct btrfs_fs_info *fs_info,
 	root->anon_dev = 0;
 
 	spin_lock_init(&root->root_item_lock);
+	btrfs_qgroup_init_swapped_blocks(&root->swapped_blocks);
 }
 
 static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info,
@@ -1258,10 +1261,17 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	struct btrfs_root *tree_root = fs_info->tree_root;
 	struct btrfs_root *root;
 	struct btrfs_key key;
+	unsigned int nofs_flag;
 	int ret = 0;
 	uuid_le uuid = NULL_UUID_LE;
 
+	/*
+	 * We're holding a transaction handle, so use a NOFS memory allocation
+	 * context to avoid deadlock if reclaim happens.
+	 */
+	nofs_flag = memalloc_nofs_save();
 	root = btrfs_alloc_root(fs_info, GFP_KERNEL);
+	memalloc_nofs_restore(nofs_flag);
 	if (!root)
 		return ERR_PTR(-ENOMEM);
 
@@ -1707,9 +1717,7 @@ static int cleaner_kthread(void *arg)
 			goto sleep;
 		}
 
-		mutex_lock(&fs_info->cleaner_delayed_iput_mutex);
 		btrfs_run_delayed_iputs(fs_info);
-		mutex_unlock(&fs_info->cleaner_delayed_iput_mutex);
 
 		again = btrfs_clean_one_deleted_snapshot(root);
 		mutex_unlock(&fs_info->cleaner_mutex);
@@ -2101,7 +2109,7 @@ static void btrfs_init_scrub(struct btrfs_fs_info *fs_info)
 	atomic_set(&fs_info->scrubs_paused, 0);
 	atomic_set(&fs_info->scrub_cancel_req, 0);
 	init_waitqueue_head(&fs_info->scrub_pause_wait);
-	fs_info->scrub_workers_refcnt = 0;
+	refcount_set(&fs_info->scrub_workers_refcnt, 0);
 }
 
 static void btrfs_init_balance(struct btrfs_fs_info *fs_info)
@@ -2666,7 +2674,6 @@ int open_ctree(struct super_block *sb,
 	mutex_init(&fs_info->delete_unused_bgs_mutex);
 	mutex_init(&fs_info->reloc_mutex);
 	mutex_init(&fs_info->delalloc_root_mutex);
-	mutex_init(&fs_info->cleaner_delayed_iput_mutex);
 	seqlock_init(&fs_info->profiles_lock);
 
 	INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
@@ -2688,6 +2695,7 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->defrag_running, 0);
 	atomic_set(&fs_info->qgroup_op_seq, 0);
 	atomic_set(&fs_info->reada_works_cnt, 0);
+	atomic_set(&fs_info->nr_delayed_iputs, 0);
 	atomic64_set(&fs_info->tree_mod_seq, 0);
 	fs_info->sb = sb;
 	fs_info->max_inline = BTRFS_DEFAULT_MAX_INLINE;
@@ -2765,6 +2773,7 @@ int open_ctree(struct super_block *sb,
 	init_waitqueue_head(&fs_info->transaction_wait);
 	init_waitqueue_head(&fs_info->transaction_blocked_wait);
 	init_waitqueue_head(&fs_info->async_submit_wait);
+	init_waitqueue_head(&fs_info->delayed_iputs_wait);
 
 	INIT_LIST_HEAD(&fs_info->pinned_chunks);
 
@@ -4238,16 +4247,9 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 
 		head = rb_entry(node, struct btrfs_delayed_ref_head,
 				href_node);
-		if (!mutex_trylock(&head->mutex)) {
-			refcount_inc(&head->refs);
-			spin_unlock(&delayed_refs->lock);
-
-			mutex_lock(&head->mutex);
-			mutex_unlock(&head->mutex);
-			btrfs_put_delayed_ref_head(head);
-			spin_lock(&delayed_refs->lock);
+		if (btrfs_delayed_ref_lock(delayed_refs, head))
 			continue;
-		}
+
 		spin_lock(&head->lock);
 		while ((n = rb_first_cached(&head->ref_tree)) != NULL) {
 			ref = rb_entry(n, struct btrfs_delayed_ref_node,
@@ -4263,12 +4265,7 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 		if (head->must_insert_reserved)
 			pin_bytes = true;
 		btrfs_free_delayed_extent_op(head->extent_op);
-		delayed_refs->num_heads--;
-		if (head->processing == 0)
-			delayed_refs->num_heads_ready--;
-		atomic_dec(&delayed_refs->num_entries);
-		rb_erase_cached(&head->href_node, &delayed_refs->href_root);
-		RB_CLEAR_NODE(&head->href_node);
+		btrfs_delete_ref_head(delayed_refs, head);
 		spin_unlock(&head->lock);
 		spin_unlock(&delayed_refs->lock);
 		mutex_unlock(&head->mutex);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index d81035b7ea7d..994f0cc41799 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -2492,9 +2492,6 @@ void btrfs_cleanup_ref_head_accounting(struct btrfs_fs_info *fs_info,
 		}
 	}
 
-	/* Also free its reserved qgroup space */
-	btrfs_qgroup_free_delayed_ref(fs_info, head->qgroup_ref_root,
-				      head->qgroup_reserved);
 	btrfs_delayed_refs_rsv_release(fs_info, nr_items);
 }
 
@@ -3013,8 +3010,7 @@ again:
 	}
 
 	if (run_all) {
-		if (!list_empty(&trans->new_bgs))
-			btrfs_create_pending_block_groups(trans);
+		btrfs_create_pending_block_groups(trans);
 
 		spin_lock(&delayed_refs->lock);
 		node = rb_first_cached(&delayed_refs->href_root);
@@ -4280,10 +4276,14 @@ commit_trans:
 				/*
 				 * The cleaner kthread might still be doing iput
 				 * operations. Wait for it to finish so that
-				 * more space is released.
+				 * more space is released.  We don't need to
+				 * explicitly run the delayed iputs here because
+				 * the commit_transaction would have woken up
+				 * the cleaner.
 				 */
-				mutex_lock(&fs_info->cleaner_delayed_iput_mutex);
-				mutex_unlock(&fs_info->cleaner_delayed_iput_mutex);
+				ret = btrfs_wait_on_delayed_iputs(fs_info);
+				if (ret)
+					return ret;
 				goto again;
 			} else {
 				btrfs_end_transaction(trans);
@@ -4396,7 +4396,6 @@ static inline u64 calc_global_rsv_need_space(struct btrfs_block_rsv *global)
 static int should_alloc_chunk(struct btrfs_fs_info *fs_info,
 			      struct btrfs_space_info *sinfo, int force)
 {
-	struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
 	u64 bytes_used = btrfs_space_info_used(sinfo, false);
 	u64 thresh;
 
@@ -4404,14 +4403,6 @@ static int should_alloc_chunk(struct btrfs_fs_info *fs_info,
 		return 1;
 
 	/*
-	 * We need to take into account the global rsv because for all intents
-	 * and purposes it's used space.  Don't worry about locking the
-	 * global_rsv, it doesn't change except when the transaction commits.
-	 */
-	if (sinfo->flags & BTRFS_BLOCK_GROUP_METADATA)
-		bytes_used += calc_global_rsv_need_space(global_rsv);
-
-	/*
 	 * in limited mode, we want to have some free space up to
 	 * about 1% of the FS size.
 	 */
@@ -4741,7 +4732,7 @@ static void shrink_delalloc(struct btrfs_fs_info *fs_info, u64 to_reclaim,
 	struct btrfs_space_info *space_info;
 	struct btrfs_trans_handle *trans;
 	u64 delalloc_bytes;
-	u64 max_reclaim;
+	u64 async_pages;
 	u64 items;
 	long time_left;
 	unsigned long nr_pages;
@@ -4766,25 +4757,36 @@ static void shrink_delalloc(struct btrfs_fs_info *fs_info, u64 to_reclaim,
 
 	loops = 0;
 	while (delalloc_bytes && loops < 3) {
-		max_reclaim = min(delalloc_bytes, to_reclaim);
-		nr_pages = max_reclaim >> PAGE_SHIFT;
+		nr_pages = min(delalloc_bytes, to_reclaim) >> PAGE_SHIFT;
+
+		/*
+		 * Triggers inode writeback for up to nr_pages. This will invoke
+		 * ->writepages callback and trigger delalloc filling
+		 *  (btrfs_run_delalloc_range()).
+		 */
 		btrfs_writeback_inodes_sb_nr(fs_info, nr_pages, items);
+
 		/*
-		 * We need to wait for the async pages to actually start before
-		 * we do anything.
+		 * We need to wait for the compressed pages to start before
+		 * we continue.
 		 */
-		max_reclaim = atomic_read(&fs_info->async_delalloc_pages);
-		if (!max_reclaim)
+		async_pages = atomic_read(&fs_info->async_delalloc_pages);
+		if (!async_pages)
 			goto skip_async;
 
-		if (max_reclaim <= nr_pages)
-			max_reclaim = 0;
+		/*
+		 * Calculate how many compressed pages we want to be written
+		 * before we continue. I.e if there are more async pages than we
+		 * require wait_event will wait until nr_pages are written.
+		 */
+		if (async_pages <= nr_pages)
+			async_pages = 0;
 		else
-			max_reclaim -= nr_pages;
+			async_pages -= nr_pages;
 
 		wait_event(fs_info->async_submit_wait,
 			   atomic_read(&fs_info->async_delalloc_pages) <=
-			   (int)max_reclaim);
+			   (int)async_pages);
 skip_async:
 		spin_lock(&space_info->lock);
 		if (list_empty(&space_info->tickets) &&
@@ -4808,6 +4810,7 @@ skip_async:
 }
 
 struct reserve_ticket {
+	u64 orig_bytes;
 	u64 bytes;
 	int error;
 	struct list_head list;
@@ -4851,10 +4854,19 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info,
 	if (!bytes_needed)
 		return 0;
 
-	/* See if there is enough pinned space to make this reservation */
-	if (__percpu_counter_compare(&space_info->total_bytes_pinned,
-				   bytes_needed,
-				   BTRFS_TOTAL_BYTES_PINNED_BATCH) >= 0)
+	trans = btrfs_join_transaction(fs_info->extent_root);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	/*
+	 * See if there is enough pinned space to make this reservation, or if
+	 * we have block groups that are going to be freed, allowing us to
+	 * possibly do a chunk allocation the next loop through.
+	 */
+	if (test_bit(BTRFS_TRANS_HAVE_FREE_BGS, &trans->transaction->flags) ||
+	    __percpu_counter_compare(&space_info->total_bytes_pinned,
+				     bytes_needed,
+				     BTRFS_TOTAL_BYTES_PINNED_BATCH) >= 0)
 		goto commit;
 
 	/*
@@ -4862,7 +4874,7 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info,
 	 * this reservation.
 	 */
 	if (space_info != delayed_rsv->space_info)
-		return -ENOSPC;
+		goto enospc;
 
 	spin_lock(&delayed_rsv->lock);
 	reclaim_bytes += delayed_rsv->reserved;
@@ -4877,16 +4889,14 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info,
 
 	if (__percpu_counter_compare(&space_info->total_bytes_pinned,
 				   bytes_needed,
-				   BTRFS_TOTAL_BYTES_PINNED_BATCH) < 0) {
-		return -ENOSPC;
-	}
+				   BTRFS_TOTAL_BYTES_PINNED_BATCH) < 0)
+		goto enospc;
 
 commit:
-	trans = btrfs_join_transaction(fs_info->extent_root);
-	if (IS_ERR(trans))
-		return -ENOSPC;
-
 	return btrfs_commit_transaction(trans);
+enospc:
+	btrfs_end_transaction(trans);
+	return -ENOSPC;
 }
 
 /*
@@ -4939,6 +4949,7 @@ static void flush_space(struct btrfs_fs_info *fs_info,
 		btrfs_end_transaction(trans);
 		break;
 	case ALLOC_CHUNK:
+	case ALLOC_CHUNK_FORCE:
 		trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans)) {
 			ret = PTR_ERR(trans);
@@ -4946,7 +4957,8 @@ static void flush_space(struct btrfs_fs_info *fs_info,
 		}
 		ret = do_chunk_alloc(trans,
 				     btrfs_metadata_alloc_profile(fs_info),
-				     CHUNK_ALLOC_NO_FORCE);
+				     (state == ALLOC_CHUNK) ?
+				      CHUNK_ALLOC_NO_FORCE : CHUNK_ALLOC_FORCE);
 		btrfs_end_transaction(trans);
 		if (ret > 0 || ret == -ENOSPC)
 			ret = 0;
@@ -4957,9 +4969,8 @@ static void flush_space(struct btrfs_fs_info *fs_info,
 		 * bunch of pinned space, so make sure we run the iputs before
 		 * we do our pinned bytes check below.
 		 */
-		mutex_lock(&fs_info->cleaner_delayed_iput_mutex);
 		btrfs_run_delayed_iputs(fs_info);
-		mutex_unlock(&fs_info->cleaner_delayed_iput_mutex);
+		btrfs_wait_on_delayed_iputs(fs_info);
 
 		ret = may_commit_transaction(fs_info, space_info);
 		break;
@@ -5030,7 +5041,7 @@ static inline int need_do_async_reclaim(struct btrfs_fs_info *fs_info,
 		!test_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state));
 }
 
-static void wake_all_tickets(struct list_head *head)
+static bool wake_all_tickets(struct list_head *head)
 {
 	struct reserve_ticket *ticket;
 
@@ -5039,7 +5050,10 @@ static void wake_all_tickets(struct list_head *head)
 		list_del_init(&ticket->list);
 		ticket->error = -ENOSPC;
 		wake_up(&ticket->wait);
+		if (ticket->bytes != ticket->orig_bytes)
+			return true;
 	}
+	return false;
 }
 
 /*
@@ -5091,11 +5105,28 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work)
 				commit_cycles--;
 		}
 
+		/*
+		 * We don't want to force a chunk allocation until we've tried
+		 * pretty hard to reclaim space.  Think of the case where we
+		 * freed up a bunch of space and so have a lot of pinned space
+		 * to reclaim.  We would rather use that than possibly create a
+		 * underutilized metadata chunk.  So if this is our first run
+		 * through the flushing state machine skip ALLOC_CHUNK_FORCE and
+		 * commit the transaction.  If nothing has changed the next go
+		 * around then we can force a chunk allocation.
+		 */
+		if (flush_state == ALLOC_CHUNK_FORCE && !commit_cycles)
+			flush_state++;
+
 		if (flush_state > COMMIT_TRANS) {
 			commit_cycles++;
 			if (commit_cycles > 2) {
-				wake_all_tickets(&space_info->tickets);
-				space_info->flush = 0;
+				if (wake_all_tickets(&space_info->tickets)) {
+					flush_state = FLUSH_DELAYED_ITEMS_NR;
+					commit_cycles--;
+				} else {
+					space_info->flush = 0;
+				}
 			} else {
 				flush_state = FLUSH_DELAYED_ITEMS_NR;
 			}
@@ -5109,12 +5140,18 @@ void btrfs_init_async_reclaim_work(struct work_struct *work)
 	INIT_WORK(work, btrfs_async_reclaim_metadata_space);
 }
 
+static const enum btrfs_flush_state priority_flush_states[] = {
+	FLUSH_DELAYED_ITEMS_NR,
+	FLUSH_DELAYED_ITEMS,
+	ALLOC_CHUNK,
+};
+
 static void priority_reclaim_metadata_space(struct btrfs_fs_info *fs_info,
 					    struct btrfs_space_info *space_info,
 					    struct reserve_ticket *ticket)
 {
 	u64 to_reclaim;
-	int flush_state = FLUSH_DELAYED_ITEMS_NR;
+	int flush_state;
 
 	spin_lock(&space_info->lock);
 	to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info,
@@ -5125,8 +5162,10 @@ static void priority_reclaim_metadata_space(struct btrfs_fs_info *fs_info,
 	}
 	spin_unlock(&space_info->lock);
 
+	flush_state = 0;
 	do {
-		flush_space(fs_info, space_info, to_reclaim, flush_state);
+		flush_space(fs_info, space_info, to_reclaim,
+			    priority_flush_states[flush_state]);
 		flush_state++;
 		spin_lock(&space_info->lock);
 		if (ticket->bytes == 0) {
@@ -5134,23 +5173,16 @@ static void priority_reclaim_metadata_space(struct btrfs_fs_info *fs_info,
 			return;
 		}
 		spin_unlock(&space_info->lock);
-
-		/*
-		 * Priority flushers can't wait on delalloc without
-		 * deadlocking.
-		 */
-		if (flush_state == FLUSH_DELALLOC ||
-		    flush_state == FLUSH_DELALLOC_WAIT)
-			flush_state = ALLOC_CHUNK;
-	} while (flush_state < COMMIT_TRANS);
+	} while (flush_state < ARRAY_SIZE(priority_flush_states));
 }
 
 static int wait_reserve_ticket(struct btrfs_fs_info *fs_info,
 			       struct btrfs_space_info *space_info,
-			       struct reserve_ticket *ticket, u64 orig_bytes)
+			       struct reserve_ticket *ticket)
 
 {
 	DEFINE_WAIT(wait);
+	u64 reclaim_bytes = 0;
 	int ret = 0;
 
 	spin_lock(&space_info->lock);
@@ -5171,14 +5203,12 @@ static int wait_reserve_ticket(struct btrfs_fs_info *fs_info,
 		ret = ticket->error;
 	if (!list_empty(&ticket->list))
 		list_del_init(&ticket->list);
-	if (ticket->bytes && ticket->bytes < orig_bytes) {
-		u64 num_bytes = orig_bytes - ticket->bytes;
-		update_bytes_may_use(space_info, -num_bytes);
-		trace_btrfs_space_reservation(fs_info, "space_info",
-					      space_info->flags, num_bytes, 0);
-	}
+	if (ticket->bytes && ticket->bytes < ticket->orig_bytes)
+		reclaim_bytes = ticket->orig_bytes - ticket->bytes;
 	spin_unlock(&space_info->lock);
 
+	if (reclaim_bytes)
+		space_info_add_old_bytes(fs_info, space_info, reclaim_bytes);
 	return ret;
 }
 
@@ -5204,6 +5234,7 @@ static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info,
 {
 	struct reserve_ticket ticket;
 	u64 used;
+	u64 reclaim_bytes = 0;
 	int ret = 0;
 
 	ASSERT(orig_bytes);
@@ -5239,6 +5270,7 @@ static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info,
 	 * the list and we will do our own flushing further down.
 	 */
 	if (ret && flush != BTRFS_RESERVE_NO_FLUSH) {
+		ticket.orig_bytes = orig_bytes;
 		ticket.bytes = orig_bytes;
 		ticket.error = 0;
 		init_waitqueue_head(&ticket.wait);
@@ -5279,25 +5311,21 @@ static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info,
 		return ret;
 
 	if (flush == BTRFS_RESERVE_FLUSH_ALL)
-		return wait_reserve_ticket(fs_info, space_info, &ticket,
-					   orig_bytes);
+		return wait_reserve_ticket(fs_info, space_info, &ticket);
 
 	ret = 0;
 	priority_reclaim_metadata_space(fs_info, space_info, &ticket);
 	spin_lock(&space_info->lock);
 	if (ticket.bytes) {
-		if (ticket.bytes < orig_bytes) {
-			u64 num_bytes = orig_bytes - ticket.bytes;
-			update_bytes_may_use(space_info, -num_bytes);
-			trace_btrfs_space_reservation(fs_info, "space_info",
-						      space_info->flags,
-						      num_bytes, 0);
-
-		}
+		if (ticket.bytes < orig_bytes)
+			reclaim_bytes = orig_bytes - ticket.bytes;
 		list_del_init(&ticket.list);
 		ret = -ENOSPC;
 	}
 	spin_unlock(&space_info->lock);
+
+	if (reclaim_bytes)
+		space_info_add_old_bytes(fs_info, space_info, reclaim_bytes);
 	ASSERT(list_empty(&ticket.list));
 	return ret;
 }
@@ -5775,6 +5803,21 @@ int btrfs_block_rsv_refill(struct btrfs_root *root,
 	return ret;
 }
 
+static void calc_refill_bytes(struct btrfs_block_rsv *block_rsv,
+				u64 *metadata_bytes, u64 *qgroup_bytes)
+{
+	*metadata_bytes = 0;
+	*qgroup_bytes = 0;
+
+	spin_lock(&block_rsv->lock);
+	if (block_rsv->reserved < block_rsv->size)
+		*metadata_bytes = block_rsv->size - block_rsv->reserved;
+	if (block_rsv->qgroup_rsv_reserved < block_rsv->qgroup_rsv_size)
+		*qgroup_bytes = block_rsv->qgroup_rsv_size -
+			block_rsv->qgroup_rsv_reserved;
+	spin_unlock(&block_rsv->lock);
+}
+
 /**
  * btrfs_inode_rsv_refill - refill the inode block rsv.
  * @inode - the inode we are refilling.
@@ -5790,25 +5833,42 @@ static int btrfs_inode_rsv_refill(struct btrfs_inode *inode,
 {
 	struct btrfs_root *root = inode->root;
 	struct btrfs_block_rsv *block_rsv = &inode->block_rsv;
-	u64 num_bytes = 0;
-	u64 qgroup_num_bytes = 0;
+	u64 num_bytes, last = 0;
+	u64 qgroup_num_bytes;
 	int ret = -ENOSPC;
 
-	spin_lock(&block_rsv->lock);
-	if (block_rsv->reserved < block_rsv->size)
-		num_bytes = block_rsv->size - block_rsv->reserved;
-	if (block_rsv->qgroup_rsv_reserved < block_rsv->qgroup_rsv_size)
-		qgroup_num_bytes = block_rsv->qgroup_rsv_size -
-				   block_rsv->qgroup_rsv_reserved;
-	spin_unlock(&block_rsv->lock);
-
+	calc_refill_bytes(block_rsv, &num_bytes, &qgroup_num_bytes);
 	if (num_bytes == 0)
 		return 0;
 
-	ret = btrfs_qgroup_reserve_meta_prealloc(root, qgroup_num_bytes, true);
-	if (ret)
-		return ret;
-	ret = reserve_metadata_bytes(root, block_rsv, num_bytes, flush);
+	do {
+		ret = btrfs_qgroup_reserve_meta_prealloc(root, qgroup_num_bytes,
+							 true);
+		if (ret)
+			return ret;
+		ret = reserve_metadata_bytes(root, block_rsv, num_bytes, flush);
+		if (ret) {
+			btrfs_qgroup_free_meta_prealloc(root, qgroup_num_bytes);
+			last = num_bytes;
+			/*
+			 * If we are fragmented we can end up with a lot of
+			 * outstanding extents which will make our size be much
+			 * larger than our reserved amount.
+			 *
+			 * If the reservation happens here, it might be very
+			 * big though not needed in the end, if the delalloc
+			 * flushing happens.
+			 *
+			 * If this is the case try and do the reserve again.
+			 */
+			if (flush == BTRFS_RESERVE_FLUSH_ALL)
+				calc_refill_bytes(block_rsv, &num_bytes,
+						   &qgroup_num_bytes);
+			if (num_bytes == 0)
+				return 0;
+		}
+	} while (ret && last != num_bytes);
+
 	if (!ret) {
 		block_rsv_add_bytes(block_rsv, num_bytes, false);
 		trace_btrfs_space_reservation(root->fs_info, "delalloc",
@@ -5818,8 +5878,7 @@ static int btrfs_inode_rsv_refill(struct btrfs_inode *inode,
 		spin_lock(&block_rsv->lock);
 		block_rsv->qgroup_rsv_reserved += qgroup_num_bytes;
 		spin_unlock(&block_rsv->lock);
-	} else
-		btrfs_qgroup_free_meta_prealloc(root, qgroup_num_bytes);
+	}
 	return ret;
 }
 
@@ -8066,6 +8125,15 @@ loop:
 	return ret;
 }
 
+#define DUMP_BLOCK_RSV(fs_info, rsv_name)				\
+do {									\
+	struct btrfs_block_rsv *__rsv = &(fs_info)->rsv_name;		\
+	spin_lock(&__rsv->lock);					\
+	btrfs_info(fs_info, #rsv_name ": size %llu reserved %llu",	\
+		   __rsv->size, __rsv->reserved);			\
+	spin_unlock(&__rsv->lock);					\
+} while (0)
+
 static void dump_space_info(struct btrfs_fs_info *fs_info,
 			    struct btrfs_space_info *info, u64 bytes,
 			    int dump_block_groups)
@@ -8085,6 +8153,12 @@ static void dump_space_info(struct btrfs_fs_info *fs_info,
 		info->bytes_readonly);
 	spin_unlock(&info->lock);
 
+	DUMP_BLOCK_RSV(fs_info, global_block_rsv);
+	DUMP_BLOCK_RSV(fs_info, trans_block_rsv);
+	DUMP_BLOCK_RSV(fs_info, chunk_block_rsv);
+	DUMP_BLOCK_RSV(fs_info, delayed_block_rsv);
+	DUMP_BLOCK_RSV(fs_info, delayed_refs_rsv);
+
 	if (!dump_block_groups)
 		return;
 
@@ -8492,7 +8566,7 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	clean_tree_block(fs_info, buf);
 	clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);
 
-	btrfs_set_lock_blocking(buf);
+	btrfs_set_lock_blocking_write(buf);
 	set_extent_buffer_uptodate(buf);
 
 	memzero_extent_buffer(buf, 0, sizeof(struct btrfs_header));
@@ -8917,7 +8991,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 		reada = 1;
 	}
 	btrfs_tree_lock(next);
-	btrfs_set_lock_blocking(next);
+	btrfs_set_lock_blocking_write(next);
 
 	ret = btrfs_lookup_extent_info(trans, fs_info, bytenr, level - 1, 1,
 				       &wc->refs[level - 1],
@@ -8977,7 +9051,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 			return -EIO;
 		}
 		btrfs_tree_lock(next);
-		btrfs_set_lock_blocking(next);
+		btrfs_set_lock_blocking_write(next);
 	}
 
 	level--;
@@ -9089,7 +9163,7 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
 		if (!path->locks[level]) {
 			BUG_ON(level == 0);
 			btrfs_tree_lock(eb);
-			btrfs_set_lock_blocking(eb);
+			btrfs_set_lock_blocking_write(eb);
 			path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
 
 			ret = btrfs_lookup_extent_info(trans, fs_info,
@@ -9131,7 +9205,7 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
 		if (!path->locks[level] &&
 		    btrfs_header_generation(eb) == trans->transid) {
 			btrfs_tree_lock(eb);
-			btrfs_set_lock_blocking(eb);
+			btrfs_set_lock_blocking_write(eb);
 			path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
 		}
 		clean_tree_block(fs_info, eb);
@@ -9298,7 +9372,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 	if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
 		level = btrfs_header_level(root->node);
 		path->nodes[level] = btrfs_lock_root_node(root);
-		btrfs_set_lock_blocking(path->nodes[level]);
+		btrfs_set_lock_blocking_write(path->nodes[level]);
 		path->slots[level] = 0;
 		path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
 		memset(&wc->update_progress, 0,
@@ -9328,7 +9402,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 		level = btrfs_header_level(root->node);
 		while (1) {
 			btrfs_tree_lock(path->nodes[level]);
-			btrfs_set_lock_blocking(path->nodes[level]);
+			btrfs_set_lock_blocking_write(path->nodes[level]);
 			path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
 
 			ret = btrfs_lookup_extent_info(trans, fs_info,
@@ -9595,6 +9669,7 @@ static int inc_block_group_ro(struct btrfs_block_group_cache *cache, int force)
 {
 	struct btrfs_space_info *sinfo = cache->space_info;
 	u64 num_bytes;
+	u64 sinfo_used;
 	u64 min_allocable_bytes;
 	int ret = -ENOSPC;
 
@@ -9621,9 +9696,10 @@ static int inc_block_group_ro(struct btrfs_block_group_cache *cache, int force)
 
 	num_bytes = cache->key.offset - cache->reserved - cache->pinned -
 		    cache->bytes_super - btrfs_block_group_used(&cache->item);
+	sinfo_used = btrfs_space_info_used(sinfo, true);
 
-	if (btrfs_space_info_used(sinfo, true) + num_bytes +
-	    min_allocable_bytes <= sinfo->total_bytes) {
+	if (sinfo_used + num_bytes + min_allocable_bytes <=
+	    sinfo->total_bytes) {
 		sinfo->bytes_readonly += num_bytes;
 		cache->ro++;
 		list_add_tail(&cache->ro_list, &sinfo->ro_bgs);
@@ -9632,6 +9708,15 @@ static int inc_block_group_ro(struct btrfs_block_group_cache *cache, int force)
 out:
 	spin_unlock(&cache->lock);
 	spin_unlock(&sinfo->lock);
+	if (ret == -ENOSPC && btrfs_test_opt(cache->fs_info, ENOSPC_DEBUG)) {
+		btrfs_info(cache->fs_info,
+			"unable to make block group %llu ro",
+			cache->key.objectid);
+		btrfs_info(cache->fs_info,
+			"sinfo_used=%llu bg_num_bytes=%llu min_allocable=%llu",
+			sinfo_used, num_bytes, min_allocable_bytes);
+		dump_space_info(cache->fs_info, cache->space_info, 0, 0);
+	}
 	return ret;
 }
 
@@ -10781,13 +10866,10 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 	}
 
 	spin_lock(&trans->transaction->dirty_bgs_lock);
-	if (!list_empty(&block_group->dirty_list)) {
-		WARN_ON(1);
-	}
-	if (!list_empty(&block_group->io_list)) {
-		WARN_ON(1);
-	}
+	WARN_ON(!list_empty(&block_group->dirty_list));
+	WARN_ON(!list_empty(&block_group->io_list));
 	spin_unlock(&trans->transaction->dirty_bgs_lock);
+
 	btrfs_remove_free_space_cache(block_group);
 
 	spin_lock(&block_group->space_info->lock);
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 52abe4082680..ca259c75bbcd 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -147,7 +147,38 @@ static int add_extent_changeset(struct extent_state *state, unsigned bits,
 	return ret;
 }
 
-static void flush_write_bio(struct extent_page_data *epd);
+static int __must_check submit_one_bio(struct bio *bio, int mirror_num,
+				       unsigned long bio_flags)
+{
+	blk_status_t ret = 0;
+	struct bio_vec *bvec = bio_last_bvec_all(bio);
+	struct page *page = bvec->bv_page;
+	struct extent_io_tree *tree = bio->bi_private;
+	u64 start;
+
+	start = page_offset(page) + bvec->bv_offset;
+
+	bio->bi_private = NULL;
+
+	if (tree->ops)
+		ret = tree->ops->submit_bio_hook(tree->private_data, bio,
+					   mirror_num, bio_flags, start);
+	else
+		btrfsic_submit_bio(bio);
+
+	return blk_status_to_errno(ret);
+}
+
+static void flush_write_bio(struct extent_page_data *epd)
+{
+	if (epd->bio) {
+		int ret;
+
+		ret = submit_one_bio(epd->bio, 0, 0);
+		BUG_ON(ret < 0); /* -ENOMEM */
+		epd->bio = NULL;
+	}
+}
 
 int __init extent_io_init(void)
 {
@@ -281,8 +312,8 @@ do_insert:
 }
 
 static struct rb_node *__etree_search(struct extent_io_tree *tree, u64 offset,
-				      struct rb_node **prev_ret,
 				      struct rb_node **next_ret,
+				      struct rb_node **prev_ret,
 				      struct rb_node ***p_ret,
 				      struct rb_node **parent_ret)
 {
@@ -311,23 +342,23 @@ static struct rb_node *__etree_search(struct extent_io_tree *tree, u64 offset,
 	if (parent_ret)
 		*parent_ret = prev;
 
-	if (prev_ret) {
+	if (next_ret) {
 		orig_prev = prev;
 		while (prev && offset > prev_entry->end) {
 			prev = rb_next(prev);
 			prev_entry = rb_entry(prev, struct tree_entry, rb_node);
 		}
-		*prev_ret = prev;
+		*next_ret = prev;
 		prev = orig_prev;
 	}
 
-	if (next_ret) {
+	if (prev_ret) {
 		prev_entry = rb_entry(prev, struct tree_entry, rb_node);
 		while (prev && offset < prev_entry->start) {
 			prev = rb_prev(prev);
 			prev_entry = rb_entry(prev, struct tree_entry, rb_node);
 		}
-		*next_ret = prev;
+		*prev_ret = prev;
 	}
 	return NULL;
 }
@@ -338,12 +369,12 @@ tree_search_for_insert(struct extent_io_tree *tree,
 		       struct rb_node ***p_ret,
 		       struct rb_node **parent_ret)
 {
-	struct rb_node *prev = NULL;
+	struct rb_node *next= NULL;
 	struct rb_node *ret;
 
-	ret = __etree_search(tree, offset, &prev, NULL, p_ret, parent_ret);
+	ret = __etree_search(tree, offset, &next, NULL, p_ret, parent_ret);
 	if (!ret)
-		return prev;
+		return next;
 	return ret;
 }
 
@@ -585,7 +616,6 @@ int __clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 
 	if (delete)
 		bits |= ~EXTENT_CTLBITS;
-	bits |= EXTENT_FIRST_DELALLOC;
 
 	if (bits & (EXTENT_IOBITS | EXTENT_BOUNDARY))
 		clear = 1;
@@ -850,7 +880,6 @@ __set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 
 	btrfs_debug_check_extent_io_range(tree, start, end);
 
-	bits |= EXTENT_FIRST_DELALLOC;
 again:
 	if (!prealloc && gfpflags_allow_blocking(mask)) {
 		/*
@@ -2692,28 +2721,6 @@ struct bio *btrfs_bio_clone_partial(struct bio *orig, int offset, int size)
 	return bio;
 }
 
-static int __must_check submit_one_bio(struct bio *bio, int mirror_num,
-				       unsigned long bio_flags)
-{
-	blk_status_t ret = 0;
-	struct bio_vec *bvec = bio_last_bvec_all(bio);
-	struct page *page = bvec->bv_page;
-	struct extent_io_tree *tree = bio->bi_private;
-	u64 start;
-
-	start = page_offset(page) + bvec->bv_offset;
-
-	bio->bi_private = NULL;
-
-	if (tree->ops)
-		ret = tree->ops->submit_bio_hook(tree->private_data, bio,
-					   mirror_num, bio_flags, start);
-	else
-		btrfsic_submit_bio(bio);
-
-	return blk_status_to_errno(ret);
-}
-
 /*
  * @opf:	bio REQ_OP_* and REQ_* flags as one value
  * @tree:	tree so we can call our merge_bio hook
@@ -4007,17 +4014,6 @@ retry:
 	return ret;
 }
 
-static void flush_write_bio(struct extent_page_data *epd)
-{
-	if (epd->bio) {
-		int ret;
-
-		ret = submit_one_bio(epd->bio, 0, 0);
-		BUG_ON(ret < 0); /* -ENOMEM */
-		epd->bio = NULL;
-	}
-}
-
 int extent_write_full_page(struct page *page, struct writeback_control *wbc)
 {
 	int ret;
@@ -4259,8 +4255,7 @@ static struct extent_map *get_extent_skip_holes(struct inode *inode,
 		if (len == 0)
 			break;
 		len = ALIGN(len, sectorsize);
-		em = btrfs_get_extent_fiemap(BTRFS_I(inode), NULL, 0, offset,
-				len, 0);
+		em = btrfs_get_extent_fiemap(BTRFS_I(inode), offset, len);
 		if (IS_ERR_OR_NULL(em))
 			return em;
 
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 9673be3f3d1f..08749e0b9c32 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -18,17 +18,16 @@
 #define EXTENT_BOUNDARY		(1U << 9)
 #define EXTENT_NODATASUM	(1U << 10)
 #define EXTENT_CLEAR_META_RESV	(1U << 11)
-#define EXTENT_FIRST_DELALLOC	(1U << 12)
-#define EXTENT_NEED_WAIT	(1U << 13)
-#define EXTENT_DAMAGED		(1U << 14)
-#define EXTENT_NORESERVE	(1U << 15)
-#define EXTENT_QGROUP_RESERVED	(1U << 16)
-#define EXTENT_CLEAR_DATA_RESV	(1U << 17)
-#define EXTENT_DELALLOC_NEW	(1U << 18)
+#define EXTENT_NEED_WAIT	(1U << 12)
+#define EXTENT_DAMAGED		(1U << 13)
+#define EXTENT_NORESERVE	(1U << 14)
+#define EXTENT_QGROUP_RESERVED	(1U << 15)
+#define EXTENT_CLEAR_DATA_RESV	(1U << 16)
+#define EXTENT_DELALLOC_NEW	(1U << 17)
 #define EXTENT_IOBITS		(EXTENT_LOCKED | EXTENT_WRITEBACK)
 #define EXTENT_DO_ACCOUNTING    (EXTENT_CLEAR_META_RESV | \
 				 EXTENT_CLEAR_DATA_RESV)
-#define EXTENT_CTLBITS		(EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC)
+#define EXTENT_CTLBITS		(EXTENT_DO_ACCOUNTING)
 
 /*
  * flags for bio submission. The high bits indicate the compression
diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
index a042a193c120..928f729c55ba 100644
--- a/fs/btrfs/extent_map.c
+++ b/fs/btrfs/extent_map.c
@@ -210,6 +210,9 @@ static int mergable_maps(struct extent_map *prev, struct extent_map *next)
 	if (!list_empty(&prev->list) || !list_empty(&next->list))
 		return 0;
 
+	ASSERT(next->block_start != EXTENT_MAP_DELALLOC &&
+	       prev->block_start != EXTENT_MAP_DELALLOC);
+
 	if (extent_map_end(prev) == next->start &&
 	    prev->flags == next->flags &&
 	    prev->bdev == next->bdev &&
@@ -217,8 +220,6 @@ static int mergable_maps(struct extent_map *prev, struct extent_map *next)
 	      prev->block_start == EXTENT_MAP_HOLE) ||
 	     (next->block_start == EXTENT_MAP_INLINE &&
 	      prev->block_start == EXTENT_MAP_INLINE) ||
-	     (next->block_start == EXTENT_MAP_DELALLOC &&
-	      prev->block_start == EXTENT_MAP_DELALLOC) ||
 	     (next->block_start < EXTENT_MAP_LAST_BYTE - 1 &&
 	      next->block_start == extent_map_block_end(prev)))) {
 		return 1;
diff --git a/fs/btrfs/extent_map.h b/fs/btrfs/extent_map.h
index ef05a0121652..473f039fcd7c 100644
--- a/fs/btrfs/extent_map.h
+++ b/fs/btrfs/extent_map.h
@@ -9,6 +9,7 @@
 #define EXTENT_MAP_LAST_BYTE ((u64)-4)
 #define EXTENT_MAP_HOLE ((u64)-3)
 #define EXTENT_MAP_INLINE ((u64)-2)
+/* used only during fiemap calls */
 #define EXTENT_MAP_DELALLOC ((u64)-1)
 
 /* bits for the extent_map::flags field */
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index d38dc8c31533..34fe8a58b0e9 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -3218,8 +3218,7 @@ static int find_desired_extent(struct inode *inode, loff_t *offset, int whence)
 			 &cached_state);
 
 	while (start < inode->i_size) {
-		em = btrfs_get_extent_fiemap(BTRFS_I(inode), NULL, 0,
-				start, len, 0);
+		em = btrfs_get_extent_fiemap(BTRFS_I(inode), start, len);
 		if (IS_ERR(em)) {
 			ret = PTR_ERR(em);
 			em = NULL;
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 5c349667c761..3f180b857e20 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -453,7 +453,6 @@ static noinline void compress_file_range(struct inode *inode,
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	u64 blocksize = fs_info->sectorsize;
 	u64 actual_end;
-	u64 isize = i_size_read(inode);
 	int ret = 0;
 	struct page **pages = NULL;
 	unsigned long nr_pages;
@@ -467,7 +466,7 @@ static noinline void compress_file_range(struct inode *inode,
 	inode_should_defrag(BTRFS_I(inode), start, end, end - start + 1,
 			SZ_16K);
 
-	actual_end = min_t(u64, isize, end + 1);
+	actual_end = min_t(u64, i_size_read(inode), end + 1);
 again:
 	will_compress = 0;
 	nr_pages = (end >> PAGE_SHIFT) - (start >> PAGE_SHIFT) + 1;
@@ -714,9 +713,9 @@ static void free_async_extent_pages(struct async_extent *async_extent)
  * queued.  We walk all the async extents created by compress_file_range
  * and send them down to the disk.
  */
-static noinline void submit_compressed_extents(struct inode *inode,
-					      struct async_cow *async_cow)
+static noinline void submit_compressed_extents(struct async_cow *async_cow)
 {
+	struct inode *inode = async_cow->inode;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct async_extent *async_extent;
 	u64 alloc_hint = 0;
@@ -1166,8 +1165,14 @@ static noinline void async_cow_submit(struct btrfs_work *work)
 	    5 * SZ_1M)
 		cond_wake_up_nomb(&fs_info->async_submit_wait);
 
+	/*
+	 * ->inode could be NULL if async_cow_start has failed to compress,
+	 * in which case we don't have anything to submit, yet we need to
+	 * always adjust ->async_delalloc_pages as its paired with the init
+	 * happening in cow_file_range_async
+	 */
 	if (async_cow->inode)
-		submit_compressed_extents(async_cow->inode, async_cow);
+		submit_compressed_extents(async_cow);
 }
 
 static noinline void async_cow_free(struct btrfs_work *work)
@@ -1194,7 +1199,12 @@ static int cow_file_range_async(struct inode *inode, struct page *locked_page,
 	while (start < end) {
 		async_cow = kmalloc(sizeof(*async_cow), GFP_NOFS);
 		BUG_ON(!async_cow); /* -ENOMEM */
-		async_cow->inode = igrab(inode);
+		/*
+		 * igrab is called higher up in the call chain, take only the
+		 * lightweight reference for the callback lifetime
+		 */
+		ihold(inode);
+		async_cow->inode = inode;
 		async_cow->fs_info = fs_info;
 		async_cow->locked_page = locked_page;
 		async_cow->start = start;
@@ -1586,11 +1596,10 @@ static inline int need_force_cow(struct inode *inode, u64 start, u64 end)
  * Function to process delayed allocation (create CoW) for ranges which are
  * being touched for the first time.
  */
-int btrfs_run_delalloc_range(void *private_data, struct page *locked_page,
+int btrfs_run_delalloc_range(struct inode *inode, struct page *locked_page,
 		u64 start, u64 end, int *page_started, unsigned long *nr_written,
 		struct writeback_control *wbc)
 {
-	struct inode *inode = private_data;
 	int ret;
 	int force_cow = need_force_cow(inode, start, end);
 	unsigned int write_flags = wbc_to_write_flags(wbc);
@@ -3247,6 +3256,7 @@ void btrfs_add_delayed_iput(struct inode *inode)
 	if (atomic_add_unless(&inode->i_count, -1, 1))
 		return;
 
+	atomic_inc(&fs_info->nr_delayed_iputs);
 	spin_lock(&fs_info->delayed_iput_lock);
 	ASSERT(list_empty(&binode->delayed_iput));
 	list_add_tail(&binode->delayed_iput, &fs_info->delayed_iputs);
@@ -3267,11 +3277,32 @@ void btrfs_run_delayed_iputs(struct btrfs_fs_info *fs_info)
 		list_del_init(&inode->delayed_iput);
 		spin_unlock(&fs_info->delayed_iput_lock);
 		iput(&inode->vfs_inode);
+		if (atomic_dec_and_test(&fs_info->nr_delayed_iputs))
+			wake_up(&fs_info->delayed_iputs_wait);
 		spin_lock(&fs_info->delayed_iput_lock);
 	}
 	spin_unlock(&fs_info->delayed_iput_lock);
 }
 
+/**
+ * btrfs_wait_on_delayed_iputs - wait on the delayed iputs to be done running
+ * @fs_info - the fs_info for this fs
+ * @return - EINTR if we were killed, 0 if nothing's pending
+ *
+ * This will wait on any delayed iputs that are currently running with KILLABLE
+ * set.  Once they are all done running we will return, unless we are killed in
+ * which case we return EINTR. This helps in user operations like fallocate etc
+ * that might get blocked on the iputs.
+ */
+int btrfs_wait_on_delayed_iputs(struct btrfs_fs_info *fs_info)
+{
+	int ret = wait_event_killable(fs_info->delayed_iputs_wait,
+			atomic_read(&fs_info->nr_delayed_iputs) == 0);
+	if (ret)
+		return -EINTR;
+	return 0;
+}
+
 /*
  * This creates an orphan entry for the given inode in case something goes wrong
  * in the middle of an unlink.
@@ -5262,13 +5293,15 @@ static struct btrfs_trans_handle *evict_refill_and_join(struct btrfs_root *root,
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
+	u64 delayed_refs_extra = btrfs_calc_trans_metadata_size(fs_info, 1);
 	int failures = 0;
 
 	for (;;) {
 		struct btrfs_trans_handle *trans;
 		int ret;
 
-		ret = btrfs_block_rsv_refill(root, rsv, rsv->size,
+		ret = btrfs_block_rsv_refill(root, rsv,
+					     rsv->size + delayed_refs_extra,
 					     BTRFS_RESERVE_FLUSH_LIMIT);
 
 		if (ret && ++failures > 2) {
@@ -5277,9 +5310,28 @@ static struct btrfs_trans_handle *evict_refill_and_join(struct btrfs_root *root,
 			return ERR_PTR(-ENOSPC);
 		}
 
+		/*
+		 * Evict can generate a large amount of delayed refs without
+		 * having a way to add space back since we exhaust our temporary
+		 * block rsv.  We aren't allowed to do FLUSH_ALL in this case
+		 * because we could deadlock with so many things in the flushing
+		 * code, so we have to try and hold some extra space to
+		 * compensate for our delayed ref generation.  If we can't get
+		 * that space then we need see if we can steal our minimum from
+		 * the global reserve.  We will be ratelimited by the amount of
+		 * space we have for the delayed refs rsv, so we'll end up
+		 * committing and trying again.
+		 */
 		trans = btrfs_join_transaction(root);
-		if (IS_ERR(trans) || !ret)
+		if (IS_ERR(trans) || !ret) {
+			if (!IS_ERR(trans)) {
+				trans->block_rsv = &fs_info->trans_block_rsv;
+				trans->bytes_reserved = delayed_refs_extra;
+				btrfs_block_rsv_migrate(rsv, trans->block_rsv,
+							delayed_refs_extra, 1);
+			}
 			return trans;
+		}
 
 		/*
 		 * Try to steal from the global reserve if there is space for
@@ -6731,7 +6783,7 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode,
 	u64 extent_start = 0;
 	u64 extent_end = 0;
 	u64 objectid = btrfs_ino(inode);
-	u32 found_type;
+	u8 extent_type;
 	struct btrfs_path *path = NULL;
 	struct btrfs_root *root = inode->root;
 	struct btrfs_file_extent_item *item;
@@ -6786,9 +6838,7 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode,
 	if (ret < 0) {
 		err = ret;
 		goto out;
-	}
-
-	if (ret != 0) {
+	} else if (ret > 0) {
 		if (path->slots[0] == 0)
 			goto not_found;
 		path->slots[0]--;
@@ -6797,11 +6847,9 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode,
 	leaf = path->nodes[0];
 	item = btrfs_item_ptr(leaf, path->slots[0],
 			      struct btrfs_file_extent_item);
-	/* are we inside the extent that was found? */
 	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
-	found_type = found_key.type;
 	if (found_key.objectid != objectid ||
-	    found_type != BTRFS_EXTENT_DATA_KEY) {
+	    found_key.type != BTRFS_EXTENT_DATA_KEY) {
 		/*
 		 * If we backup past the first extent we want to move forward
 		 * and see if there is an extent in front of us, otherwise we'll
@@ -6812,16 +6860,16 @@ struct extent_map *btrfs_get_extent(struct btrfs_inode *inode,
 		goto next;
 	}
 
-	found_type = btrfs_file_extent_type(leaf, item);
+	extent_type = btrfs_file_extent_type(leaf, item);
 	extent_start = found_key.offset;
-	if (found_type == BTRFS_FILE_EXTENT_REG ||
-	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
+	if (extent_type == BTRFS_FILE_EXTENT_REG ||
+	    extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
 		extent_end = extent_start +
 		       btrfs_file_extent_num_bytes(leaf, item);
 
 		trace_btrfs_get_extent_show_fi_regular(inode, leaf, item,
 						       extent_start);
-	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
+	} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 		size_t size;
 
 		size = btrfs_file_extent_ram_bytes(leaf, item);
@@ -6840,9 +6888,9 @@ next:
 			if (ret < 0) {
 				err = ret;
 				goto out;
-			}
-			if (ret > 0)
+			} else if (ret > 0) {
 				goto not_found;
+			}
 			leaf = path->nodes[0];
 		}
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
@@ -6853,19 +6901,22 @@ next:
 			goto not_found;
 		if (start > found_key.offset)
 			goto next;
+
+		/* New extent overlaps with existing one */
 		em->start = start;
 		em->orig_start = start;
 		em->len = found_key.offset - start;
-		goto not_found_em;
+		em->block_start = EXTENT_MAP_HOLE;
+		goto insert;
 	}
 
 	btrfs_extent_item_to_extent_map(inode, path, item,
 			new_inline, em);
 
-	if (found_type == BTRFS_FILE_EXTENT_REG ||
-	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
+	if (extent_type == BTRFS_FILE_EXTENT_REG ||
+	    extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
 		goto insert;
-	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
+	} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 		unsigned long ptr;
 		char *map;
 		size_t size;
@@ -6916,7 +6967,6 @@ not_found:
 	em->start = start;
 	em->orig_start = start;
 	em->len = len;
-not_found_em:
 	em->block_start = EXTENT_MAP_HOLE;
 insert:
 	btrfs_release_path(path);
@@ -6946,19 +6996,17 @@ out:
 }
 
 struct extent_map *btrfs_get_extent_fiemap(struct btrfs_inode *inode,
-		struct page *page,
-		size_t pg_offset, u64 start, u64 len,
-		int create)
+					   u64 start, u64 len)
 {
 	struct extent_map *em;
 	struct extent_map *hole_em = NULL;
-	u64 range_start = start;
+	u64 delalloc_start = start;
 	u64 end;
-	u64 found;
-	u64 found_end;
+	u64 delalloc_len;
+	u64 delalloc_end;
 	int err = 0;
 
-	em = btrfs_get_extent(inode, page, pg_offset, start, len, create);
+	em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
 	if (IS_ERR(em))
 		return em;
 	/*
@@ -6983,80 +7031,84 @@ struct extent_map *btrfs_get_extent_fiemap(struct btrfs_inode *inode,
 	em = NULL;
 
 	/* ok, we didn't find anything, lets look for delalloc */
-	found = count_range_bits(&inode->io_tree, &range_start,
+	delalloc_len = count_range_bits(&inode->io_tree, &delalloc_start,
 				 end, len, EXTENT_DELALLOC, 1);
-	found_end = range_start + found;
-	if (found_end < range_start)
-		found_end = (u64)-1;
+	delalloc_end = delalloc_start + delalloc_len;
+	if (delalloc_end < delalloc_start)
+		delalloc_end = (u64)-1;
 
 	/*
-	 * we didn't find anything useful, return
-	 * the original results from get_extent()
+	 * We didn't find anything useful, return the original results from
+	 * get_extent()
 	 */
-	if (range_start > end || found_end <= start) {
+	if (delalloc_start > end || delalloc_end <= start) {
 		em = hole_em;
 		hole_em = NULL;
 		goto out;
 	}
 
-	/* adjust the range_start to make sure it doesn't
-	 * go backwards from the start they passed in
+	/*
+	 * Adjust the delalloc_start to make sure it doesn't go backwards from
+	 * the start they passed in
 	 */
-	range_start = max(start, range_start);
-	found = found_end - range_start;
+	delalloc_start = max(start, delalloc_start);
+	delalloc_len = delalloc_end - delalloc_start;
 
-	if (found > 0) {
-		u64 hole_start = start;
-		u64 hole_len = len;
+	if (delalloc_len > 0) {
+		u64 hole_start;
+		u64 hole_len;
+		const u64 hole_end = extent_map_end(hole_em);
 
 		em = alloc_extent_map();
 		if (!em) {
 			err = -ENOMEM;
 			goto out;
 		}
+		em->bdev = NULL;
+
+		ASSERT(hole_em);
 		/*
-		 * when btrfs_get_extent can't find anything it
-		 * returns one huge hole
+		 * When btrfs_get_extent can't find anything it returns one
+		 * huge hole
 		 *
-		 * make sure what it found really fits our range, and
-		 * adjust to make sure it is based on the start from
-		 * the caller
+		 * Make sure what it found really fits our range, and adjust to
+		 * make sure it is based on the start from the caller
 		 */
-		if (hole_em) {
-			u64 calc_end = extent_map_end(hole_em);
-
-			if (calc_end <= start || (hole_em->start > end)) {
-				free_extent_map(hole_em);
-				hole_em = NULL;
-			} else {
-				hole_start = max(hole_em->start, start);
-				hole_len = calc_end - hole_start;
-			}
+		if (hole_end <= start || hole_em->start > end) {
+		       free_extent_map(hole_em);
+		       hole_em = NULL;
+		} else {
+		       hole_start = max(hole_em->start, start);
+		       hole_len = hole_end - hole_start;
 		}
-		em->bdev = NULL;
-		if (hole_em && range_start > hole_start) {
-			/* our hole starts before our delalloc, so we
-			 * have to return just the parts of the hole
-			 * that go until  the delalloc starts
+
+		if (hole_em && delalloc_start > hole_start) {
+			/*
+			 * Our hole starts before our delalloc, so we have to
+			 * return just the parts of the hole that go until the
+			 * delalloc starts
 			 */
-			em->len = min(hole_len,
-				      range_start - hole_start);
+			em->len = min(hole_len, delalloc_start - hole_start);
 			em->start = hole_start;
 			em->orig_start = hole_start;
 			/*
-			 * don't adjust block start at all,
-			 * it is fixed at EXTENT_MAP_HOLE
+			 * Don't adjust block start at all, it is fixed at
+			 * EXTENT_MAP_HOLE
 			 */
 			em->block_start = hole_em->block_start;
 			em->block_len = hole_len;
 			if (test_bit(EXTENT_FLAG_PREALLOC, &hole_em->flags))
 				set_bit(EXTENT_FLAG_PREALLOC, &em->flags);
 		} else {
-			em->start = range_start;
-			em->len = found;
-			em->orig_start = range_start;
+			/*
+			 * Hole is out of passed range or it starts after
+			 * delalloc range
+			 */
+			em->start = delalloc_start;
+			em->len = delalloc_len;
+			em->orig_start = delalloc_start;
 			em->block_start = EXTENT_MAP_DELALLOC;
-			em->block_len = found;
+			em->block_len = delalloc_len;
 		}
 	} else {
 		return hole_em;
@@ -9910,7 +9962,6 @@ static struct btrfs_delalloc_work *btrfs_alloc_delalloc_work(struct inode *inode
 	init_completion(&work->completion);
 	INIT_LIST_HEAD(&work->list);
 	work->inode = inode;
-	WARN_ON_ONCE(!inode);
 	btrfs_init_work(&work->work, btrfs_flush_delalloc_helper,
 			btrfs_run_delalloc_work, NULL, NULL);
 
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 9c8e1734429c..494f0f10d70e 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1642,7 +1642,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 		btrfs_info(fs_info, "resizing devid %llu", devid);
 	}
 
-	device = btrfs_find_device(fs_info, devid, NULL, NULL);
+	device = btrfs_find_device(fs_info->fs_devices, devid, NULL, NULL, true);
 	if (!device) {
 		btrfs_info(fs_info, "resizer unable to find device %llu",
 			   devid);
@@ -3178,7 +3178,8 @@ static long btrfs_ioctl_dev_info(struct btrfs_fs_info *fs_info,
 		s_uuid = di_args->uuid;
 
 	rcu_read_lock();
-	dev = btrfs_find_device(fs_info, di_args->devid, s_uuid, NULL);
+	dev = btrfs_find_device(fs_info->fs_devices, di_args->devid, s_uuid,
+				NULL, true);
 
 	if (!dev) {
 		ret = -ENODEV;
@@ -3241,32 +3242,17 @@ static void btrfs_double_extent_lock(struct inode *inode1, u64 loff1,
 	lock_extent(&BTRFS_I(inode2)->io_tree, loff2, loff2 + len - 1);
 }
 
-static int btrfs_extent_same_range(struct inode *src, u64 loff, u64 olen,
+static int btrfs_extent_same_range(struct inode *src, u64 loff, u64 len,
 				   struct inode *dst, u64 dst_loff)
 {
-	u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize;
 	int ret;
-	u64 len = olen;
-
-	if (loff + len == src->i_size)
-		len = ALIGN(src->i_size, bs) - loff;
-	/*
-	 * For same inode case we don't want our length pushed out past i_size
-	 * as comparing that data range makes no sense.
-	 *
-	 * This effectively means we require aligned extents for the single
-	 * inode case, whereas the other cases allow an unaligned length so long
-	 * as it ends at i_size.
-	 */
-	if (dst == src && len != olen)
-		return -EINVAL;
 
 	/*
 	 * Lock destination range to serialize with concurrent readpages() and
 	 * source range to serialize with relocation.
 	 */
 	btrfs_double_extent_lock(src, loff, dst, dst_loff, len);
-	ret = btrfs_clone(src, dst, loff, olen, len, dst_loff, 1);
+	ret = btrfs_clone(src, dst, loff, len, len, dst_loff, 1);
 	btrfs_double_extent_unlock(src, loff, dst, dst_loff, len);
 
 	return ret;
@@ -3278,21 +3264,10 @@ static int btrfs_extent_same(struct inode *src, u64 loff, u64 olen,
 			     struct inode *dst, u64 dst_loff)
 {
 	int ret;
-	int num_pages = PAGE_ALIGN(BTRFS_MAX_DEDUPE_LEN) >> PAGE_SHIFT;
 	u64 i, tail_len, chunk_count;
 
-	/* don't make the dst file partly checksummed */
-	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
-	    (BTRFS_I(dst)->flags & BTRFS_INODE_NODATASUM))
-		return -EINVAL;
-
-	if (IS_SWAPFILE(src) || IS_SWAPFILE(dst))
-		return -ETXTBSY;
-
 	tail_len = olen % BTRFS_MAX_DEDUPE_LEN;
 	chunk_count = div_u64(olen, BTRFS_MAX_DEDUPE_LEN);
-	if (chunk_count == 0)
-		num_pages = PAGE_ALIGN(tail_len) >> PAGE_SHIFT;
 
 	for (i = 0; i < chunk_count; i++) {
 		ret = btrfs_extent_same_range(src, loff, BTRFS_MAX_DEDUPE_LEN,
@@ -3908,14 +3883,6 @@ static noinline int btrfs_clone_files(struct file *file, struct file *file_src,
 	 *   be either compressed or non-compressed.
 	 */
 
-	/* don't make the dst file partly checksummed */
-	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
-	    (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM))
-		return -EINVAL;
-
-	if (IS_SWAPFILE(src) || IS_SWAPFILE(inode))
-		return -ETXTBSY;
-
 	/*
 	 * VFS's generic_remap_file_range_prep() protects us from cloning the
 	 * eof block into the middle of a file, which would result in corruption
@@ -3991,6 +3958,13 @@ static int btrfs_remap_file_range_prep(struct file *file_in, loff_t pos_in,
 	else
 		btrfs_double_inode_lock(inode_in, inode_out);
 
+	/* don't make the dst file partly checksummed */
+	if ((BTRFS_I(inode_in)->flags & BTRFS_INODE_NODATASUM) !=
+	    (BTRFS_I(inode_out)->flags & BTRFS_INODE_NODATASUM)) {
+		ret = -EINVAL;
+		goto out_unlock;
+	}
+
 	/*
 	 * Now that the inodes are locked, we need to start writeback ourselves
 	 * and can not rely on the writeback from the VFS's generic helper
@@ -4381,7 +4355,7 @@ static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
 			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
 			      0);
 
-	if (copy_to_user(arg, sa, sizeof(*sa)))
+	if (ret == 0 && copy_to_user(arg, sa, sizeof(*sa)))
 		ret = -EFAULT;
 
 	if (!(sa->flags & BTRFS_SCRUB_READONLY))
@@ -4414,7 +4388,7 @@ static long btrfs_ioctl_scrub_progress(struct btrfs_fs_info *fs_info,
 
 	ret = btrfs_scrub_progress(fs_info, sa->devid, &sa->progress);
 
-	if (copy_to_user(arg, sa, sizeof(*sa)))
+	if (ret == 0 && copy_to_user(arg, sa, sizeof(*sa)))
 		ret = -EFAULT;
 
 	kfree(sa);
@@ -4438,7 +4412,7 @@ static long btrfs_ioctl_get_dev_stats(struct btrfs_fs_info *fs_info,
 
 	ret = btrfs_get_dev_stats(fs_info, sa);
 
-	if (copy_to_user(arg, sa, sizeof(*sa)))
+	if (ret == 0 && copy_to_user(arg, sa, sizeof(*sa)))
 		ret = -EFAULT;
 
 	kfree(sa);
@@ -4484,7 +4458,7 @@ static long btrfs_ioctl_dev_replace(struct btrfs_fs_info *fs_info,
 		break;
 	}
 
-	if (copy_to_user(arg, p, sizeof(*p)))
+	if ((ret == 0 || ret == -ECANCELED) && copy_to_user(arg, p, sizeof(*p)))
 		ret = -EFAULT;
 out:
 	kfree(p);
@@ -4790,7 +4764,7 @@ do_balance:
 	ret = btrfs_balance(fs_info, bctl, bargs);
 	bctl = NULL;
 
-	if (arg) {
+	if ((ret == 0 || ret == -ECANCELED) && arg) {
 		if (copy_to_user(arg, bargs, sizeof(*bargs)))
 			ret = -EFAULT;
 	}
diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c
index 1da768e5ef75..82b84e4daad1 100644
--- a/fs/btrfs/locking.c
+++ b/fs/btrfs/locking.c
@@ -14,43 +14,58 @@
 
 static void btrfs_assert_tree_read_locked(struct extent_buffer *eb);
 
-/*
- * if we currently have a spinning reader or writer lock
- * (indicated by the rw flag) this will bump the count
- * of blocking holders and drop the spinlock.
- */
-void btrfs_set_lock_blocking_rw(struct extent_buffer *eb, int rw)
+void btrfs_set_lock_blocking_read(struct extent_buffer *eb)
 {
 	/*
-	 * no lock is required.  The lock owner may change if
-	 * we have a read lock, but it won't change to or away
-	 * from us.  If we have the write lock, we are the owner
-	 * and it'll never change.
+	 * No lock is required.  The lock owner may change if we have a read
+	 * lock, but it won't change to or away from us.  If we have the write
+	 * lock, we are the owner and it'll never change.
 	 */
 	if (eb->lock_nested && current->pid == eb->lock_owner)
 		return;
-	if (rw == BTRFS_WRITE_LOCK) {
-		if (atomic_read(&eb->blocking_writers) == 0) {
-			WARN_ON(atomic_read(&eb->spinning_writers) != 1);
-			atomic_dec(&eb->spinning_writers);
-			btrfs_assert_tree_locked(eb);
-			atomic_inc(&eb->blocking_writers);
-			write_unlock(&eb->lock);
-		}
-	} else if (rw == BTRFS_READ_LOCK) {
-		btrfs_assert_tree_read_locked(eb);
-		atomic_inc(&eb->blocking_readers);
-		WARN_ON(atomic_read(&eb->spinning_readers) == 0);
-		atomic_dec(&eb->spinning_readers);
-		read_unlock(&eb->lock);
+	btrfs_assert_tree_read_locked(eb);
+	atomic_inc(&eb->blocking_readers);
+	WARN_ON(atomic_read(&eb->spinning_readers) == 0);
+	atomic_dec(&eb->spinning_readers);
+	read_unlock(&eb->lock);
+}
+
+void btrfs_set_lock_blocking_write(struct extent_buffer *eb)
+{
+	/*
+	 * No lock is required.  The lock owner may change if we have a read
+	 * lock, but it won't change to or away from us.  If we have the write
+	 * lock, we are the owner and it'll never change.
+	 */
+	if (eb->lock_nested && current->pid == eb->lock_owner)
+		return;
+	if (atomic_read(&eb->blocking_writers) == 0) {
+		WARN_ON(atomic_read(&eb->spinning_writers) != 1);
+		atomic_dec(&eb->spinning_writers);
+		btrfs_assert_tree_locked(eb);
+		atomic_inc(&eb->blocking_writers);
+		write_unlock(&eb->lock);
 	}
 }
 
-/*
- * if we currently have a blocking lock, take the spinlock
- * and drop our blocking count
- */
-void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw)
+void btrfs_clear_lock_blocking_read(struct extent_buffer *eb)
+{
+	/*
+	 * No lock is required.  The lock owner may change if we have a read
+	 * lock, but it won't change to or away from us.  If we have the write
+	 * lock, we are the owner and it'll never change.
+	 */
+	if (eb->lock_nested && current->pid == eb->lock_owner)
+		return;
+	BUG_ON(atomic_read(&eb->blocking_readers) == 0);
+	read_lock(&eb->lock);
+	atomic_inc(&eb->spinning_readers);
+	/* atomic_dec_and_test implies a barrier */
+	if (atomic_dec_and_test(&eb->blocking_readers))
+		cond_wake_up_nomb(&eb->read_lock_wq);
+}
+
+void btrfs_clear_lock_blocking_write(struct extent_buffer *eb)
 {
 	/*
 	 * no lock is required.  The lock owner may change if
@@ -60,23 +75,13 @@ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw)
 	 */
 	if (eb->lock_nested && current->pid == eb->lock_owner)
 		return;
-
-	if (rw == BTRFS_WRITE_LOCK_BLOCKING) {
-		BUG_ON(atomic_read(&eb->blocking_writers) != 1);
-		write_lock(&eb->lock);
-		WARN_ON(atomic_read(&eb->spinning_writers));
-		atomic_inc(&eb->spinning_writers);
-		/* atomic_dec_and_test implies a barrier */
-		if (atomic_dec_and_test(&eb->blocking_writers))
-			cond_wake_up_nomb(&eb->write_lock_wq);
-	} else if (rw == BTRFS_READ_LOCK_BLOCKING) {
-		BUG_ON(atomic_read(&eb->blocking_readers) == 0);
-		read_lock(&eb->lock);
-		atomic_inc(&eb->spinning_readers);
-		/* atomic_dec_and_test implies a barrier */
-		if (atomic_dec_and_test(&eb->blocking_readers))
-			cond_wake_up_nomb(&eb->read_lock_wq);
-	}
+	BUG_ON(atomic_read(&eb->blocking_writers) != 1);
+	write_lock(&eb->lock);
+	WARN_ON(atomic_read(&eb->spinning_writers));
+	atomic_inc(&eb->spinning_writers);
+	/* atomic_dec_and_test implies a barrier */
+	if (atomic_dec_and_test(&eb->blocking_writers))
+		cond_wake_up_nomb(&eb->write_lock_wq);
 }
 
 /*
@@ -232,16 +237,9 @@ again:
 	wait_event(eb->read_lock_wq, atomic_read(&eb->blocking_readers) == 0);
 	wait_event(eb->write_lock_wq, atomic_read(&eb->blocking_writers) == 0);
 	write_lock(&eb->lock);
-	if (atomic_read(&eb->blocking_readers)) {
+	if (atomic_read(&eb->blocking_readers) ||
+	    atomic_read(&eb->blocking_writers)) {
 		write_unlock(&eb->lock);
-		wait_event(eb->read_lock_wq,
-			   atomic_read(&eb->blocking_readers) == 0);
-		goto again;
-	}
-	if (atomic_read(&eb->blocking_writers)) {
-		write_unlock(&eb->lock);
-		wait_event(eb->write_lock_wq,
-			   atomic_read(&eb->blocking_writers) == 0);
 		goto again;
 	}
 	WARN_ON(atomic_read(&eb->spinning_writers));
diff --git a/fs/btrfs/locking.h b/fs/btrfs/locking.h
index 29135def468e..595014f64830 100644
--- a/fs/btrfs/locking.h
+++ b/fs/btrfs/locking.h
@@ -17,8 +17,10 @@ void btrfs_tree_unlock(struct extent_buffer *eb);
 void btrfs_tree_read_lock(struct extent_buffer *eb);
 void btrfs_tree_read_unlock(struct extent_buffer *eb);
 void btrfs_tree_read_unlock_blocking(struct extent_buffer *eb);
-void btrfs_set_lock_blocking_rw(struct extent_buffer *eb, int rw);
-void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw);
+void btrfs_set_lock_blocking_read(struct extent_buffer *eb);
+void btrfs_set_lock_blocking_write(struct extent_buffer *eb);
+void btrfs_clear_lock_blocking_read(struct extent_buffer *eb);
+void btrfs_clear_lock_blocking_write(struct extent_buffer *eb);
 void btrfs_assert_tree_locked(struct extent_buffer *eb);
 int btrfs_try_tree_read_lock(struct extent_buffer *eb);
 int btrfs_try_tree_write_lock(struct extent_buffer *eb);
@@ -37,13 +39,4 @@ static inline void btrfs_tree_unlock_rw(struct extent_buffer *eb, int rw)
 		BUG();
 }
 
-static inline void btrfs_set_lock_blocking(struct extent_buffer *eb)
-{
-	btrfs_set_lock_blocking_rw(eb, BTRFS_WRITE_LOCK);
-}
-
-static inline void btrfs_clear_lock_blocking(struct extent_buffer *eb)
-{
-	btrfs_clear_lock_blocking_rw(eb, BTRFS_WRITE_LOCK_BLOCKING);
-}
 #endif
diff --git a/fs/btrfs/lzo.c b/fs/btrfs/lzo.c
index 90639140439f..579d53ae256f 100644
--- a/fs/btrfs/lzo.c
+++ b/fs/btrfs/lzo.c
@@ -61,6 +61,28 @@ struct workspace {
 	struct list_head list;
 };
 
+static struct workspace_manager wsm;
+
+static void lzo_init_workspace_manager(void)
+{
+	btrfs_init_workspace_manager(&wsm, &btrfs_lzo_compress);
+}
+
+static void lzo_cleanup_workspace_manager(void)
+{
+	btrfs_cleanup_workspace_manager(&wsm);
+}
+
+static struct list_head *lzo_get_workspace(unsigned int level)
+{
+	return btrfs_get_workspace(&wsm, level);
+}
+
+static void lzo_put_workspace(struct list_head *ws)
+{
+	btrfs_put_workspace(&wsm, ws);
+}
+
 static void lzo_free_workspace(struct list_head *ws)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
@@ -71,7 +93,7 @@ static void lzo_free_workspace(struct list_head *ws)
 	kfree(workspace);
 }
 
-static struct list_head *lzo_alloc_workspace(void)
+static struct list_head *lzo_alloc_workspace(unsigned int level)
 {
 	struct workspace *workspace;
 
@@ -485,11 +507,16 @@ out:
 	return ret;
 }
 
-static void lzo_set_level(struct list_head *ws, unsigned int type)
+static unsigned int lzo_set_level(unsigned int level)
 {
+	return 0;
 }
 
 const struct btrfs_compress_op btrfs_lzo_compress = {
+	.init_workspace_manager	= lzo_init_workspace_manager,
+	.cleanup_workspace_manager = lzo_cleanup_workspace_manager,
+	.get_workspace		= lzo_get_workspace,
+	.put_workspace		= lzo_put_workspace,
 	.alloc_workspace	= lzo_alloc_workspace,
 	.free_workspace		= lzo_free_workspace,
 	.compress_pages		= lzo_compress_pages,
diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index 4e473a998219..c1cd5558a646 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -1546,12 +1546,18 @@ int btrfs_qgroup_trace_extent_nolock(struct btrfs_fs_info *fs_info,
 		parent_node = *p;
 		entry = rb_entry(parent_node, struct btrfs_qgroup_extent_record,
 				 node);
-		if (bytenr < entry->bytenr)
+		if (bytenr < entry->bytenr) {
 			p = &(*p)->rb_left;
-		else if (bytenr > entry->bytenr)
+		} else if (bytenr > entry->bytenr) {
 			p = &(*p)->rb_right;
-		else
+		} else {
+			if (record->data_rsv && !entry->data_rsv) {
+				entry->data_rsv = record->data_rsv;
+				entry->data_rsv_refroot =
+					record->data_rsv_refroot;
+			}
 			return 1;
+		}
 	}
 
 	rb_link_node(&record->node, parent_node, p);
@@ -1597,7 +1603,7 @@ int btrfs_qgroup_trace_extent(struct btrfs_trans_handle *trans, u64 bytenr,
 	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)
 	    || bytenr == 0 || num_bytes == 0)
 		return 0;
-	record = kmalloc(sizeof(*record), gfp_flag);
+	record = kzalloc(sizeof(*record), gfp_flag);
 	if (!record)
 		return -ENOMEM;
 
@@ -1832,7 +1838,7 @@ static int qgroup_trace_extent_swap(struct btrfs_trans_handle* trans,
 			src_path->nodes[cur_level] = eb;
 
 			btrfs_tree_read_lock(eb);
-			btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+			btrfs_set_lock_blocking_read(eb);
 			src_path->locks[cur_level] = BTRFS_READ_LOCK_BLOCKING;
 		}
 
@@ -1973,7 +1979,7 @@ static int qgroup_trace_new_subtree_blocks(struct btrfs_trans_handle* trans,
 		dst_path->slots[cur_level] = 0;
 
 		btrfs_tree_read_lock(eb);
-		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+		btrfs_set_lock_blocking_read(eb);
 		dst_path->locks[cur_level] = BTRFS_READ_LOCK_BLOCKING;
 		need_cleanup = true;
 	}
@@ -2017,86 +2023,30 @@ out:
 	return ret;
 }
 
-/*
- * Inform qgroup to trace subtree swap used in balance.
- *
- * Unlike btrfs_qgroup_trace_subtree(), this function will only trace
- * new tree blocks whose generation is equal to (or larger than) @last_snapshot.
- *
- * Will go down the tree block pointed by @dst_eb (pointed by @dst_parent and
- * @dst_slot), and find any tree blocks whose generation is at @last_snapshot,
- * and then go down @src_eb (pointed by @src_parent and @src_slot) to find
- * the counterpart of the tree block, then mark both tree blocks as qgroup dirty,
- * and skip all tree blocks whose generation is smaller than last_snapshot.
- *
- * This would skip tons of tree blocks of original btrfs_qgroup_trace_subtree(),
- * which could be the cause of very slow balance if the file tree is large.
- *
- * @src_parent, @src_slot: pointer to src (file tree) eb.
- * @dst_parent, @dst_slot: pointer to dst (reloc tree) eb.
- */
-int btrfs_qgroup_trace_subtree_swap(struct btrfs_trans_handle *trans,
-				struct btrfs_block_group_cache *bg_cache,
-				struct extent_buffer *src_parent, int src_slot,
-				struct extent_buffer *dst_parent, int dst_slot,
-				u64 last_snapshot)
+static int qgroup_trace_subtree_swap(struct btrfs_trans_handle *trans,
+				struct extent_buffer *src_eb,
+				struct extent_buffer *dst_eb,
+				u64 last_snapshot, bool trace_leaf)
 {
 	struct btrfs_fs_info *fs_info = trans->fs_info;
 	struct btrfs_path *dst_path = NULL;
-	struct btrfs_key first_key;
-	struct extent_buffer *src_eb = NULL;
-	struct extent_buffer *dst_eb = NULL;
-	bool trace_leaf = false;
-	u64 child_gen;
-	u64 child_bytenr;
 	int level;
 	int ret;
 
 	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags))
 		return 0;
 
-	/* Check parameter order */
-	if (btrfs_node_ptr_generation(src_parent, src_slot) >
-	    btrfs_node_ptr_generation(dst_parent, dst_slot)) {
+	/* Wrong parameter order */
+	if (btrfs_header_generation(src_eb) > btrfs_header_generation(dst_eb)) {
 		btrfs_err_rl(fs_info,
 		"%s: bad parameter order, src_gen=%llu dst_gen=%llu", __func__,
-			btrfs_node_ptr_generation(src_parent, src_slot),
-			btrfs_node_ptr_generation(dst_parent, dst_slot));
+			     btrfs_header_generation(src_eb),
+			     btrfs_header_generation(dst_eb));
 		return -EUCLEAN;
 	}
 
-	/*
-	 * Only trace leaf if we're relocating data block groups, this could
-	 * reduce tons of data extents tracing for meta/sys bg relocation.
-	 */
-	if (bg_cache->flags & BTRFS_BLOCK_GROUP_DATA)
-		trace_leaf = true;
-	/* Read out real @src_eb, pointed by @src_parent and @src_slot */
-	child_bytenr = btrfs_node_blockptr(src_parent, src_slot);
-	child_gen = btrfs_node_ptr_generation(src_parent, src_slot);
-	btrfs_node_key_to_cpu(src_parent, &first_key, src_slot);
-
-	src_eb = read_tree_block(fs_info, child_bytenr, child_gen,
-			btrfs_header_level(src_parent) - 1, &first_key);
-	if (IS_ERR(src_eb)) {
-		ret = PTR_ERR(src_eb);
-		goto out;
-	}
-
-	/* Read out real @dst_eb, pointed by @src_parent and @src_slot */
-	child_bytenr = btrfs_node_blockptr(dst_parent, dst_slot);
-	child_gen = btrfs_node_ptr_generation(dst_parent, dst_slot);
-	btrfs_node_key_to_cpu(dst_parent, &first_key, dst_slot);
-
-	dst_eb = read_tree_block(fs_info, child_bytenr, child_gen,
-			btrfs_header_level(dst_parent) - 1, &first_key);
-	if (IS_ERR(dst_eb)) {
-		ret = PTR_ERR(dst_eb);
-		goto out;
-	}
-
 	if (!extent_buffer_uptodate(src_eb) || !extent_buffer_uptodate(dst_eb)) {
-		ret = -EINVAL;
+		ret = -EIO;
 		goto out;
 	}
 
@@ -2106,14 +2056,13 @@ int btrfs_qgroup_trace_subtree_swap(struct btrfs_trans_handle *trans,
 		ret = -ENOMEM;
 		goto out;
 	}
-
 	/* For dst_path */
 	extent_buffer_get(dst_eb);
 	dst_path->nodes[level] = dst_eb;
 	dst_path->slots[level] = 0;
 	dst_path->locks[level] = 0;
 
-	/* Do the generation-aware breadth-first search */
+	/* Do the generation aware breadth-first search */
 	ret = qgroup_trace_new_subtree_blocks(trans, src_eb, dst_path, level,
 					      level, last_snapshot, trace_leaf);
 	if (ret < 0)
@@ -2121,8 +2070,6 @@ int btrfs_qgroup_trace_subtree_swap(struct btrfs_trans_handle *trans,
 	ret = 0;
 
 out:
-	free_extent_buffer(src_eb);
-	free_extent_buffer(dst_eb);
 	btrfs_free_path(dst_path);
 	if (ret < 0)
 		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
@@ -2207,7 +2154,7 @@ walk_down:
 			path->slots[level] = 0;
 
 			btrfs_tree_read_lock(eb);
-			btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+			btrfs_set_lock_blocking_read(eb);
 			path->locks[level] = BTRFS_READ_LOCK_BLOCKING;
 
 			ret = btrfs_qgroup_trace_extent(trans, child_bytenr,
@@ -2576,6 +2523,11 @@ int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans)
 					goto cleanup;
 			}
 
+			/* Free the reserved data space */
+			btrfs_qgroup_free_refroot(fs_info,
+					record->data_rsv_refroot,
+					record->data_rsv,
+					BTRFS_QGROUP_RSV_DATA);
 			/*
 			 * Use SEQ_LAST as time_seq to do special search, which
 			 * doesn't lock tree or delayed_refs and search current
@@ -2842,16 +2794,15 @@ out:
 /*
  * Two limits to commit transaction in advance.
  *
- * For RATIO, it will be 1/RATIO of the remaining limit
- * (excluding data and prealloc meta) as threshold.
+ * For RATIO, it will be 1/RATIO of the remaining limit as threshold.
  * For SIZE, it will be in byte unit as threshold.
  */
-#define QGROUP_PERTRANS_RATIO		32
-#define QGROUP_PERTRANS_SIZE		SZ_32M
+#define QGROUP_FREE_RATIO		32
+#define QGROUP_FREE_SIZE		SZ_32M
 static bool qgroup_check_limits(struct btrfs_fs_info *fs_info,
 				const struct btrfs_qgroup *qg, u64 num_bytes)
 {
-	u64 limit;
+	u64 free;
 	u64 threshold;
 
 	if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_RFER) &&
@@ -2870,20 +2821,21 @@ static bool qgroup_check_limits(struct btrfs_fs_info *fs_info,
 	 */
 	if ((qg->lim_flags & (BTRFS_QGROUP_LIMIT_MAX_RFER |
 			      BTRFS_QGROUP_LIMIT_MAX_EXCL))) {
-		if (qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_EXCL)
-			limit = qg->max_excl;
-		else
-			limit = qg->max_rfer;
-		threshold = (limit - qg->rsv.values[BTRFS_QGROUP_RSV_DATA] -
-			    qg->rsv.values[BTRFS_QGROUP_RSV_META_PREALLOC]) /
-			    QGROUP_PERTRANS_RATIO;
-		threshold = min_t(u64, threshold, QGROUP_PERTRANS_SIZE);
+		if (qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_EXCL) {
+			free = qg->max_excl - qgroup_rsv_total(qg) - qg->excl;
+			threshold = min_t(u64, qg->max_excl / QGROUP_FREE_RATIO,
+					  QGROUP_FREE_SIZE);
+		} else {
+			free = qg->max_rfer - qgroup_rsv_total(qg) - qg->rfer;
+			threshold = min_t(u64, qg->max_rfer / QGROUP_FREE_RATIO,
+					  QGROUP_FREE_SIZE);
+		}
 
 		/*
 		 * Use transaction_kthread to commit transaction, so we no
 		 * longer need to bother nested transaction nor lock context.
 		 */
-		if (qg->rsv.values[BTRFS_QGROUP_RSV_META_PERTRANS] > threshold)
+		if (free < threshold)
 			btrfs_commit_transaction_locksafe(fs_info);
 	}
 
@@ -2959,7 +2911,6 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce,
 
 		qg = unode_aux_to_qgroup(unode);
 
-		trace_qgroup_update_reserve(fs_info, qg, num_bytes, type);
 		qgroup_rsv_add(fs_info, qg, num_bytes, type);
 	}
 
@@ -3026,7 +2977,6 @@ void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info,
 
 		qg = unode_aux_to_qgroup(unode);
 
-		trace_qgroup_update_reserve(fs_info, qg, -(s64)num_bytes, type);
 		qgroup_rsv_release(fs_info, qg, num_bytes, type);
 
 		list_for_each_entry(glist, &qg->groups, next_group) {
@@ -3783,3 +3733,241 @@ void btrfs_qgroup_check_reserved_leak(struct inode *inode)
 	}
 	extent_changeset_release(&changeset);
 }
+
+void btrfs_qgroup_init_swapped_blocks(
+	struct btrfs_qgroup_swapped_blocks *swapped_blocks)
+{
+	int i;
+
+	spin_lock_init(&swapped_blocks->lock);
+	for (i = 0; i < BTRFS_MAX_LEVEL; i++)
+		swapped_blocks->blocks[i] = RB_ROOT;
+	swapped_blocks->swapped = false;
+}
+
+/*
+ * Delete all swapped blocks record of @root.
+ * Every record here means we skipped a full subtree scan for qgroup.
+ *
+ * Gets called when committing one transaction.
+ */
+void btrfs_qgroup_clean_swapped_blocks(struct btrfs_root *root)
+{
+	struct btrfs_qgroup_swapped_blocks *swapped_blocks;
+	int i;
+
+	swapped_blocks = &root->swapped_blocks;
+
+	spin_lock(&swapped_blocks->lock);
+	if (!swapped_blocks->swapped)
+		goto out;
+	for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
+		struct rb_root *cur_root = &swapped_blocks->blocks[i];
+		struct btrfs_qgroup_swapped_block *entry;
+		struct btrfs_qgroup_swapped_block *next;
+
+		rbtree_postorder_for_each_entry_safe(entry, next, cur_root,
+						     node)
+			kfree(entry);
+		swapped_blocks->blocks[i] = RB_ROOT;
+	}
+	swapped_blocks->swapped = false;
+out:
+	spin_unlock(&swapped_blocks->lock);
+}
+
+/*
+ * Add subtree roots record into @subvol_root.
+ *
+ * @subvol_root:	tree root of the subvolume tree get swapped
+ * @bg:			block group under balance
+ * @subvol_parent/slot:	pointer to the subtree root in subvolume tree
+ * @reloc_parent/slot:	pointer to the subtree root in reloc tree
+ *			BOTH POINTERS ARE BEFORE TREE SWAP
+ * @last_snapshot:	last snapshot generation of the subvolume tree
+ */
+int btrfs_qgroup_add_swapped_blocks(struct btrfs_trans_handle *trans,
+		struct btrfs_root *subvol_root,
+		struct btrfs_block_group_cache *bg,
+		struct extent_buffer *subvol_parent, int subvol_slot,
+		struct extent_buffer *reloc_parent, int reloc_slot,
+		u64 last_snapshot)
+{
+	struct btrfs_fs_info *fs_info = subvol_root->fs_info;
+	struct btrfs_qgroup_swapped_blocks *blocks = &subvol_root->swapped_blocks;
+	struct btrfs_qgroup_swapped_block *block;
+	struct rb_node **cur;
+	struct rb_node *parent = NULL;
+	int level = btrfs_header_level(subvol_parent) - 1;
+	int ret = 0;
+
+	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags))
+		return 0;
+
+	if (btrfs_node_ptr_generation(subvol_parent, subvol_slot) >
+	    btrfs_node_ptr_generation(reloc_parent, reloc_slot)) {
+		btrfs_err_rl(fs_info,
+		"%s: bad parameter order, subvol_gen=%llu reloc_gen=%llu",
+			__func__,
+			btrfs_node_ptr_generation(subvol_parent, subvol_slot),
+			btrfs_node_ptr_generation(reloc_parent, reloc_slot));
+		return -EUCLEAN;
+	}
+
+	block = kmalloc(sizeof(*block), GFP_NOFS);
+	if (!block) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	/*
+	 * @reloc_parent/slot is still before swap, while @block is going to
+	 * record the bytenr after swap, so we do the swap here.
+	 */
+	block->subvol_bytenr = btrfs_node_blockptr(reloc_parent, reloc_slot);
+	block->subvol_generation = btrfs_node_ptr_generation(reloc_parent,
+							     reloc_slot);
+	block->reloc_bytenr = btrfs_node_blockptr(subvol_parent, subvol_slot);
+	block->reloc_generation = btrfs_node_ptr_generation(subvol_parent,
+							    subvol_slot);
+	block->last_snapshot = last_snapshot;
+	block->level = level;
+	if (bg->flags & BTRFS_BLOCK_GROUP_DATA)
+		block->trace_leaf = true;
+	else
+		block->trace_leaf = false;
+	btrfs_node_key_to_cpu(reloc_parent, &block->first_key, reloc_slot);
+
+	/* Insert @block into @blocks */
+	spin_lock(&blocks->lock);
+	cur = &blocks->blocks[level].rb_node;
+	while (*cur) {
+		struct btrfs_qgroup_swapped_block *entry;
+
+		parent = *cur;
+		entry = rb_entry(parent, struct btrfs_qgroup_swapped_block,
+				 node);
+
+		if (entry->subvol_bytenr < block->subvol_bytenr) {
+			cur = &(*cur)->rb_left;
+		} else if (entry->subvol_bytenr > block->subvol_bytenr) {
+			cur = &(*cur)->rb_right;
+		} else {
+			if (entry->subvol_generation !=
+					block->subvol_generation ||
+			    entry->reloc_bytenr != block->reloc_bytenr ||
+			    entry->reloc_generation !=
+					block->reloc_generation) {
+				/*
+				 * Duplicated but mismatch entry found.
+				 * Shouldn't happen.
+				 *
+				 * Marking qgroup inconsistent should be enough
+				 * for end users.
+				 */
+				WARN_ON(IS_ENABLED(CONFIG_BTRFS_DEBUG));
+				ret = -EEXIST;
+			}
+			kfree(block);
+			goto out_unlock;
+		}
+	}
+	rb_link_node(&block->node, parent, cur);
+	rb_insert_color(&block->node, &blocks->blocks[level]);
+	blocks->swapped = true;
+out_unlock:
+	spin_unlock(&blocks->lock);
+out:
+	if (ret < 0)
+		fs_info->qgroup_flags |=
+			BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
+	return ret;
+}
+
+/*
+ * Check if the tree block is a subtree root, and if so do the needed
+ * delayed subtree trace for qgroup.
+ *
+ * This is called during btrfs_cow_block().
+ */
+int btrfs_qgroup_trace_subtree_after_cow(struct btrfs_trans_handle *trans,
+					 struct btrfs_root *root,
+					 struct extent_buffer *subvol_eb)
+{
+	struct btrfs_fs_info *fs_info = root->fs_info;
+	struct btrfs_qgroup_swapped_blocks *blocks = &root->swapped_blocks;
+	struct btrfs_qgroup_swapped_block *block;
+	struct extent_buffer *reloc_eb = NULL;
+	struct rb_node *node;
+	bool found = false;
+	bool swapped = false;
+	int level = btrfs_header_level(subvol_eb);
+	int ret = 0;
+	int i;
+
+	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags))
+		return 0;
+	if (!is_fstree(root->root_key.objectid) || !root->reloc_root)
+		return 0;
+
+	spin_lock(&blocks->lock);
+	if (!blocks->swapped) {
+		spin_unlock(&blocks->lock);
+		return 0;
+	}
+	node = blocks->blocks[level].rb_node;
+
+	while (node) {
+		block = rb_entry(node, struct btrfs_qgroup_swapped_block, node);
+		if (block->subvol_bytenr < subvol_eb->start) {
+			node = node->rb_left;
+		} else if (block->subvol_bytenr > subvol_eb->start) {
+			node = node->rb_right;
+		} else {
+			found = true;
+			break;
+		}
+	}
+	if (!found) {
+		spin_unlock(&blocks->lock);
+		goto out;
+	}
+	/* Found one, remove it from @blocks first and update blocks->swapped */
+	rb_erase(&block->node, &blocks->blocks[level]);
+	for (i = 0; i < BTRFS_MAX_LEVEL; i++) {
+		if (RB_EMPTY_ROOT(&blocks->blocks[i])) {
+			swapped = true;
+			break;
+		}
+	}
+	blocks->swapped = swapped;
+	spin_unlock(&blocks->lock);
+
+	/* Read out reloc subtree root */
+	reloc_eb = read_tree_block(fs_info, block->reloc_bytenr,
+				   block->reloc_generation, block->level,
+				   &block->first_key);
+	if (IS_ERR(reloc_eb)) {
+		ret = PTR_ERR(reloc_eb);
+		reloc_eb = NULL;
+		goto free_out;
+	}
+	if (!extent_buffer_uptodate(reloc_eb)) {
+		ret = -EIO;
+		goto free_out;
+	}
+
+	ret = qgroup_trace_subtree_swap(trans, reloc_eb, subvol_eb,
+			block->last_snapshot, block->trace_leaf);
+free_out:
+	kfree(block);
+	free_extent_buffer(reloc_eb);
+out:
+	if (ret < 0) {
+		btrfs_err_rl(fs_info,
+			     "failed to account subtree at bytenr %llu: %d",
+			     subvol_eb->start, ret);
+		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
+	}
+	return ret;
+}
diff --git a/fs/btrfs/qgroup.h b/fs/btrfs/qgroup.h
index 20c6bd5fa701..46ba7bd2961c 100644
--- a/fs/btrfs/qgroup.h
+++ b/fs/btrfs/qgroup.h
@@ -6,6 +6,8 @@
 #ifndef BTRFS_QGROUP_H
 #define BTRFS_QGROUP_H
 
+#include <linux/spinlock.h>
+#include <linux/rbtree.h>
 #include "ulist.h"
 #include "delayed-ref.h"
 
@@ -38,6 +40,66 @@
  */
 
 /*
+ * Special performance optimization for balance.
+ *
+ * For balance, we need to swap subtree of subvolume and reloc trees.
+ * In theory, we need to trace all subtree blocks of both subvolume and reloc
+ * trees, since their owner has changed during such swap.
+ *
+ * However since balance has ensured that both subtrees are containing the
+ * same contents and have the same tree structures, such swap won't cause
+ * qgroup number change.
+ *
+ * But there is a race window between subtree swap and transaction commit,
+ * during that window, if we increase/decrease tree level or merge/split tree
+ * blocks, we still need to trace the original subtrees.
+ *
+ * So for balance, we use a delayed subtree tracing, whose workflow is:
+ *
+ * 1) Record the subtree root block get swapped.
+ *
+ *    During subtree swap:
+ *    O = Old tree blocks
+ *    N = New tree blocks
+ *          reloc tree                     subvolume tree X
+ *             Root                               Root
+ *            /    \                             /    \
+ *          NA     OB                          OA      OB
+ *        /  |     |  \                      /  |      |  \
+ *      NC  ND     OE  OF                   OC  OD     OE  OF
+ *
+ *   In this case, NA and OA are going to be swapped, record (NA, OA) into
+ *   subvolume tree X.
+ *
+ * 2) After subtree swap.
+ *          reloc tree                     subvolume tree X
+ *             Root                               Root
+ *            /    \                             /    \
+ *          OA     OB                          NA      OB
+ *        /  |     |  \                      /  |      |  \
+ *      OC  OD     OE  OF                   NC  ND     OE  OF
+ *
+ * 3a) COW happens for OB
+ *     If we are going to COW tree block OB, we check OB's bytenr against
+ *     tree X's swapped_blocks structure.
+ *     If it doesn't fit any, nothing will happen.
+ *
+ * 3b) COW happens for NA
+ *     Check NA's bytenr against tree X's swapped_blocks, and get a hit.
+ *     Then we do subtree scan on both subtrees OA and NA.
+ *     Resulting 6 tree blocks to be scanned (OA, OC, OD, NA, NC, ND).
+ *
+ *     Then no matter what we do to subvolume tree X, qgroup numbers will
+ *     still be correct.
+ *     Then NA's record gets removed from X's swapped_blocks.
+ *
+ * 4)  Transaction commit
+ *     Any record in X's swapped_blocks gets removed, since there is no
+ *     modification to the swapped subtrees, no need to trigger heavy qgroup
+ *     subtree rescan for them.
+ */
+
+/*
  * Record a dirty extent, and info qgroup to update quota on it
  * TODO: Use kmem cache to alloc it.
  */
@@ -45,9 +107,38 @@ struct btrfs_qgroup_extent_record {
 	struct rb_node node;
 	u64 bytenr;
 	u64 num_bytes;
+
+	/*
+	 * For qgroup reserved data space freeing.
+	 *
+	 * @data_rsv_refroot and @data_rsv will be recorded after
+	 * BTRFS_ADD_DELAYED_EXTENT is called.
+	 * And will be used to free reserved qgroup space at
+	 * transaction commit time.
+	 */
+	u32 data_rsv;		/* reserved data space needs to be freed */
+	u64 data_rsv_refroot;	/* which root the reserved data belongs to */
 	struct ulist *old_roots;
 };
 
+struct btrfs_qgroup_swapped_block {
+	struct rb_node node;
+
+	int level;
+	bool trace_leaf;
+
+	/* bytenr/generation of the tree block in subvolume tree after swap */
+	u64 subvol_bytenr;
+	u64 subvol_generation;
+
+	/* bytenr/generation of the tree block in reloc tree after swap */
+	u64 reloc_bytenr;
+	u64 reloc_generation;
+
+	u64 last_snapshot;
+	struct btrfs_key first_key;
+};
+
 /*
  * Qgroup reservation types:
  *
@@ -236,12 +327,6 @@ int btrfs_qgroup_trace_leaf_items(struct btrfs_trans_handle *trans,
 int btrfs_qgroup_trace_subtree(struct btrfs_trans_handle *trans,
 			       struct extent_buffer *root_eb,
 			       u64 root_gen, int root_level);
-
-int btrfs_qgroup_trace_subtree_swap(struct btrfs_trans_handle *trans,
-				struct btrfs_block_group_cache *bg_cache,
-				struct extent_buffer *src_parent, int src_slot,
-				struct extent_buffer *dst_parent, int dst_slot,
-				u64 last_snapshot);
 int btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans, u64 bytenr,
 				u64 num_bytes, struct ulist *old_roots,
 				struct ulist *new_roots);
@@ -252,15 +337,6 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, u64 srcid,
 void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info,
 			       u64 ref_root, u64 num_bytes,
 			       enum btrfs_qgroup_rsv_type type);
-static inline void btrfs_qgroup_free_delayed_ref(struct btrfs_fs_info *fs_info,
-						 u64 ref_root, u64 num_bytes)
-{
-	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags))
-		return;
-	trace_btrfs_qgroup_free_delayed_ref(fs_info, ref_root, num_bytes);
-	btrfs_qgroup_free_refroot(fs_info, ref_root, num_bytes,
-				  BTRFS_QGROUP_RSV_DATA);
-}
 
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 int btrfs_verify_qgroup_counts(struct btrfs_fs_info *fs_info, u64 qgroupid,
@@ -325,4 +401,18 @@ void btrfs_qgroup_convert_reserved_meta(struct btrfs_root *root, int num_bytes);
 
 void btrfs_qgroup_check_reserved_leak(struct inode *inode);
 
+/* btrfs_qgroup_swapped_blocks related functions */
+void btrfs_qgroup_init_swapped_blocks(
+	struct btrfs_qgroup_swapped_blocks *swapped_blocks);
+
+void btrfs_qgroup_clean_swapped_blocks(struct btrfs_root *root);
+int btrfs_qgroup_add_swapped_blocks(struct btrfs_trans_handle *trans,
+		struct btrfs_root *subvol_root,
+		struct btrfs_block_group_cache *bg,
+		struct extent_buffer *subvol_parent, int subvol_slot,
+		struct extent_buffer *reloc_parent, int reloc_slot,
+		u64 last_snapshot);
+int btrfs_qgroup_trace_subtree_after_cow(struct btrfs_trans_handle *trans,
+		struct btrfs_root *root, struct extent_buffer *eb);
+
 #endif
diff --git a/fs/btrfs/ref-verify.c b/fs/btrfs/ref-verify.c
index c3557c12656b..d09b6cdb785a 100644
--- a/fs/btrfs/ref-verify.c
+++ b/fs/btrfs/ref-verify.c
@@ -583,7 +583,7 @@ static int walk_down_tree(struct btrfs_root *root, struct btrfs_path *path,
 				return -EIO;
 			}
 			btrfs_tree_read_lock(eb);
-			btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+			btrfs_set_lock_blocking_read(eb);
 			path->nodes[level-1] = eb;
 			path->slots[level-1] = 0;
 			path->locks[level-1] = BTRFS_READ_LOCK_BLOCKING;
@@ -987,7 +987,7 @@ int btrfs_build_ref_tree(struct btrfs_fs_info *fs_info)
 		return -ENOMEM;
 
 	eb = btrfs_read_lock_root_node(fs_info->extent_root);
-	btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+	btrfs_set_lock_blocking_read(eb);
 	level = btrfs_header_level(eb);
 	path->nodes[level] = eb;
 	path->slots[level] = 0;
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 272b287f8cf0..ddf028509931 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -162,6 +162,8 @@ struct reloc_control {
 	struct mapping_tree reloc_root_tree;
 	/* list of reloc trees */
 	struct list_head reloc_roots;
+	/* list of subvolume trees that get relocated */
+	struct list_head dirty_subvol_roots;
 	/* size of metadata reservation for merging reloc trees */
 	u64 merging_rsv_size;
 	/* size of relocated tree nodes */
@@ -1467,15 +1469,17 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans,
 	struct btrfs_root_item *root_item;
 	int ret;
 
-	if (!root->reloc_root)
+	if (test_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state) ||
+	    !root->reloc_root)
 		goto out;
 
 	reloc_root = root->reloc_root;
 	root_item = &reloc_root->root_item;
 
+	/* root->reloc_root will stay until current relocation finished */
 	if (fs_info->reloc_ctl->merge_reloc_tree &&
 	    btrfs_root_refs(root_item) == 0) {
-		root->reloc_root = NULL;
+		set_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state);
 		__del_reloc_root(reloc_root);
 	}
 
@@ -1773,7 +1777,7 @@ again:
 	btrfs_node_key_to_cpu(path->nodes[lowest_level], &key, slot);
 
 	eb = btrfs_lock_root_node(dest);
-	btrfs_set_lock_blocking(eb);
+	btrfs_set_lock_blocking_write(eb);
 	level = btrfs_header_level(eb);
 
 	if (level < lowest_level) {
@@ -1786,7 +1790,7 @@ again:
 		ret = btrfs_cow_block(trans, dest, eb, NULL, 0, &eb);
 		BUG_ON(ret);
 	}
-	btrfs_set_lock_blocking(eb);
+	btrfs_set_lock_blocking_write(eb);
 
 	if (next_key) {
 		next_key->objectid = (u64)-1;
@@ -1802,6 +1806,8 @@ again:
 		BUG_ON(level < lowest_level);
 
 		ret = btrfs_bin_search(parent, &key, level, &slot);
+		if (ret < 0)
+			break;
 		if (ret && slot > 0)
 			slot--;
 
@@ -1852,7 +1858,7 @@ again:
 						      slot, &eb);
 				BUG_ON(ret);
 			}
-			btrfs_set_lock_blocking(eb);
+			btrfs_set_lock_blocking_write(eb);
 
 			btrfs_tree_unlock(parent);
 			free_extent_buffer(parent);
@@ -1885,15 +1891,18 @@ again:
 		 *    If not traced, we will leak data numbers
 		 * 2) Fs subtree
 		 *    If not traced, we will double count old data
-		 *    and tree block numbers, if current trans doesn't free
-		 *    data reloc tree inode.
+		 *
+		 * We don't scan the subtree right now, but only record
+		 * the swapped tree blocks.
+		 * The real subtree rescan is delayed until we have new
+		 * CoW on the subtree root node before transaction commit.
 		 */
-		ret = btrfs_qgroup_trace_subtree_swap(trans, rc->block_group,
-				parent, slot, path->nodes[level],
-				path->slots[level], last_snapshot);
+		ret = btrfs_qgroup_add_swapped_blocks(trans, dest,
+				rc->block_group, parent, slot,
+				path->nodes[level], path->slots[level],
+				last_snapshot);
 		if (ret < 0)
 			break;
-
 		/*
 		 * swap blocks in fs tree and reloc tree.
 		 */
@@ -2121,6 +2130,58 @@ static int find_next_key(struct btrfs_path *path, int level,
 }
 
 /*
+ * Insert current subvolume into reloc_control::dirty_subvol_roots
+ */
+static void insert_dirty_subvol(struct btrfs_trans_handle *trans,
+				struct reloc_control *rc,
+				struct btrfs_root *root)
+{
+	struct btrfs_root *reloc_root = root->reloc_root;
+	struct btrfs_root_item *reloc_root_item;
+
+	/* @root must be a subvolume tree root with a valid reloc tree */
+	ASSERT(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID);
+	ASSERT(reloc_root);
+
+	reloc_root_item = &reloc_root->root_item;
+	memset(&reloc_root_item->drop_progress, 0,
+		sizeof(reloc_root_item->drop_progress));
+	reloc_root_item->drop_level = 0;
+	btrfs_set_root_refs(reloc_root_item, 0);
+	btrfs_update_reloc_root(trans, root);
+
+	if (list_empty(&root->reloc_dirty_list)) {
+		btrfs_grab_fs_root(root);
+		list_add_tail(&root->reloc_dirty_list, &rc->dirty_subvol_roots);
+	}
+}
+
+static int clean_dirty_subvols(struct reloc_control *rc)
+{
+	struct btrfs_root *root;
+	struct btrfs_root *next;
+	int ret = 0;
+
+	list_for_each_entry_safe(root, next, &rc->dirty_subvol_roots,
+				 reloc_dirty_list) {
+		struct btrfs_root *reloc_root = root->reloc_root;
+
+		clear_bit(BTRFS_ROOT_DEAD_RELOC_TREE, &root->state);
+		list_del_init(&root->reloc_dirty_list);
+		root->reloc_root = NULL;
+		if (reloc_root) {
+			int ret2;
+
+			ret2 = btrfs_drop_snapshot(reloc_root, NULL, 0, 1);
+			if (ret2 < 0 && !ret)
+				ret = ret2;
+		}
+		btrfs_put_fs_root(root);
+	}
+	return ret;
+}
+
+/*
  * merge the relocated tree blocks in reloc tree with corresponding
  * fs tree.
  */
@@ -2128,7 +2189,6 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 					       struct btrfs_root *root)
 {
 	struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
-	LIST_HEAD(inode_list);
 	struct btrfs_key key;
 	struct btrfs_key next_key;
 	struct btrfs_trans_handle *trans = NULL;
@@ -2259,13 +2319,8 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 out:
 	btrfs_free_path(path);
 
-	if (err == 0) {
-		memset(&root_item->drop_progress, 0,
-		       sizeof(root_item->drop_progress));
-		root_item->drop_level = 0;
-		btrfs_set_root_refs(root_item, 0);
-		btrfs_update_reloc_root(trans, root);
-	}
+	if (err == 0)
+		insert_dirty_subvol(trans, rc, root);
 
 	if (trans)
 		btrfs_end_transaction_throttle(trans);
@@ -2410,14 +2465,6 @@ again:
 		} else {
 			list_del_init(&reloc_root->root_list);
 		}
-
-		ret = btrfs_drop_snapshot(reloc_root, rc->block_rsv, 0, 1);
-		if (ret < 0) {
-			if (list_empty(&reloc_root->root_list))
-				list_add_tail(&reloc_root->root_list,
-					      &reloc_roots);
-			goto out;
-		}
 	}
 
 	if (found) {
@@ -2685,6 +2732,10 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 			if (!lowest) {
 				ret = btrfs_bin_search(upper->eb, key,
 						       upper->level, &slot);
+				if (ret < 0) {
+					err = ret;
+					goto next;
+				}
 				BUG_ON(ret);
 				bytenr = btrfs_node_blockptr(upper->eb, slot);
 				if (node->eb->start == bytenr)
@@ -2720,6 +2771,10 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 		} else {
 			ret = btrfs_bin_search(upper->eb, key, upper->level,
 					       &slot);
+			if (ret < 0) {
+				err = ret;
+				goto next;
+			}
 			BUG_ON(ret);
 		}
 
@@ -2752,7 +2807,7 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 			goto next;
 		}
 		btrfs_tree_lock(eb);
-		btrfs_set_lock_blocking(eb);
+		btrfs_set_lock_blocking_write(eb);
 
 		if (!node->eb) {
 			ret = btrfs_cow_block(trans, root, eb, upper->eb,
@@ -4079,6 +4134,9 @@ restart:
 		goto out_free;
 	}
 	btrfs_commit_transaction(trans);
+	ret = clean_dirty_subvols(rc);
+	if (ret < 0 && !err)
+		err = ret;
 out_free:
 	btrfs_free_block_rsv(fs_info, rc->block_rsv);
 	btrfs_free_path(path);
@@ -4173,6 +4231,7 @@ static struct reloc_control *alloc_reloc_control(void)
 		return NULL;
 
 	INIT_LIST_HEAD(&rc->reloc_roots);
+	INIT_LIST_HEAD(&rc->dirty_subvol_roots);
 	backref_cache_init(&rc->backref_cache);
 	mapping_tree_init(&rc->reloc_root_tree);
 	extent_io_tree_init(&rc->processed_blocks, NULL);
@@ -4468,6 +4527,10 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 		goto out_free;
 	}
 	err = btrfs_commit_transaction(trans);
+
+	ret = clean_dirty_subvols(rc);
+	if (ret < 0 && !err)
+		err = ret;
 out_free:
 	kfree(rc);
 out:
diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c
index 65bda0682928..0d2b957ca3a3 100644
--- a/fs/btrfs/root-tree.c
+++ b/fs/btrfs/root-tree.c
@@ -21,12 +21,12 @@ static void btrfs_read_root_item(struct extent_buffer *eb, int slot,
 				struct btrfs_root_item *item)
 {
 	uuid_le uuid;
-	int len;
+	u32 len;
 	int need_reset = 0;
 
 	len = btrfs_item_size_nr(eb, slot);
 	read_extent_buffer(eb, item, btrfs_item_ptr_offset(eb, slot),
-			min_t(int, len, (int)sizeof(*item)));
+			   min_t(u32, len, sizeof(*item)));
 	if (len < sizeof(*item))
 		need_reset = 1;
 	if (!need_reset && btrfs_root_generation(item)
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index 6dcd36d7b849..a99588536c79 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -584,6 +584,7 @@ static noinline_for_stack struct scrub_ctx *scrub_setup_ctx(
 	sctx->pages_per_rd_bio = SCRUB_PAGES_PER_RD_BIO;
 	sctx->curr = -1;
 	sctx->fs_info = fs_info;
+	INIT_LIST_HEAD(&sctx->csum_list);
 	for (i = 0; i < SCRUB_BIOS_PER_SCTX; ++i) {
 		struct scrub_bio *sbio;
 
@@ -608,7 +609,6 @@ static noinline_for_stack struct scrub_ctx *scrub_setup_ctx(
 	atomic_set(&sctx->workers_pending, 0);
 	atomic_set(&sctx->cancel_req, 0);
 	sctx->csum_size = btrfs_super_csum_size(fs_info->super_copy);
-	INIT_LIST_HEAD(&sctx->csum_list);
 
 	spin_lock_init(&sctx->list_lock);
 	spin_lock_init(&sctx->stat_lock);
@@ -3741,25 +3741,33 @@ static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info,
 	unsigned int flags = WQ_FREEZABLE | WQ_UNBOUND;
 	int max_active = fs_info->thread_pool_size;
 
-	if (fs_info->scrub_workers_refcnt == 0) {
+	lockdep_assert_held(&fs_info->scrub_lock);
+
+	if (refcount_read(&fs_info->scrub_workers_refcnt) == 0) {
+		ASSERT(fs_info->scrub_workers == NULL);
 		fs_info->scrub_workers = btrfs_alloc_workqueue(fs_info, "scrub",
 				flags, is_dev_replace ? 1 : max_active, 4);
 		if (!fs_info->scrub_workers)
 			goto fail_scrub_workers;
 
+		ASSERT(fs_info->scrub_wr_completion_workers == NULL);
 		fs_info->scrub_wr_completion_workers =
 			btrfs_alloc_workqueue(fs_info, "scrubwrc", flags,
 					      max_active, 2);
 		if (!fs_info->scrub_wr_completion_workers)
 			goto fail_scrub_wr_completion_workers;
 
+		ASSERT(fs_info->scrub_parity_workers == NULL);
 		fs_info->scrub_parity_workers =
 			btrfs_alloc_workqueue(fs_info, "scrubparity", flags,
 					      max_active, 2);
 		if (!fs_info->scrub_parity_workers)
 			goto fail_scrub_parity_workers;
+
+		refcount_set(&fs_info->scrub_workers_refcnt, 1);
+	} else {
+		refcount_inc(&fs_info->scrub_workers_refcnt);
 	}
-	++fs_info->scrub_workers_refcnt;
 	return 0;
 
 fail_scrub_parity_workers:
@@ -3770,16 +3778,6 @@ fail_scrub_workers:
 	return -ENOMEM;
 }
 
-static noinline_for_stack void scrub_workers_put(struct btrfs_fs_info *fs_info)
-{
-	if (--fs_info->scrub_workers_refcnt == 0) {
-		btrfs_destroy_workqueue(fs_info->scrub_workers);
-		btrfs_destroy_workqueue(fs_info->scrub_wr_completion_workers);
-		btrfs_destroy_workqueue(fs_info->scrub_parity_workers);
-	}
-	WARN_ON(fs_info->scrub_workers_refcnt < 0);
-}
-
 int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		    u64 end, struct btrfs_scrub_progress *progress,
 		    int readonly, int is_dev_replace)
@@ -3788,6 +3786,9 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	int ret;
 	struct btrfs_device *dev;
 	unsigned int nofs_flag;
+	struct btrfs_workqueue *scrub_workers = NULL;
+	struct btrfs_workqueue *scrub_wr_comp = NULL;
+	struct btrfs_workqueue *scrub_parity = NULL;
 
 	if (btrfs_fs_closing(fs_info))
 		return -EINVAL;
@@ -3835,7 +3836,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		return PTR_ERR(sctx);
 
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
-	dev = btrfs_find_device(fs_info, devid, NULL, NULL);
+	dev = btrfs_find_device(fs_info->fs_devices, devid, NULL, NULL, true);
 	if (!dev || (test_bit(BTRFS_DEV_STATE_MISSING, &dev->dev_state) &&
 		     !is_dev_replace)) {
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
@@ -3903,6 +3904,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	 */
 	nofs_flag = memalloc_nofs_save();
 	if (!is_dev_replace) {
+		btrfs_info(fs_info, "scrub: started on devid %llu", devid);
 		/*
 		 * by holding device list mutex, we can
 		 * kick off writing super in log tree sync.
@@ -3925,11 +3927,26 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	if (progress)
 		memcpy(progress, &sctx->stat, sizeof(*progress));
 
+	if (!is_dev_replace)
+		btrfs_info(fs_info, "scrub: %s on devid %llu with status: %d",
+			ret ? "not finished" : "finished", devid, ret);
+
 	mutex_lock(&fs_info->scrub_lock);
 	dev->scrub_ctx = NULL;
-	scrub_workers_put(fs_info);
+	if (refcount_dec_and_test(&fs_info->scrub_workers_refcnt)) {
+		scrub_workers = fs_info->scrub_workers;
+		scrub_wr_comp = fs_info->scrub_wr_completion_workers;
+		scrub_parity = fs_info->scrub_parity_workers;
+
+		fs_info->scrub_workers = NULL;
+		fs_info->scrub_wr_completion_workers = NULL;
+		fs_info->scrub_parity_workers = NULL;
+	}
 	mutex_unlock(&fs_info->scrub_lock);
 
+	btrfs_destroy_workqueue(scrub_workers);
+	btrfs_destroy_workqueue(scrub_wr_comp);
+	btrfs_destroy_workqueue(scrub_parity);
 	scrub_put_ctx(sctx);
 
 	return ret;
@@ -4012,7 +4029,7 @@ int btrfs_scrub_progress(struct btrfs_fs_info *fs_info, u64 devid,
 	struct scrub_ctx *sctx = NULL;
 
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
-	dev = btrfs_find_device(fs_info, devid, NULL, NULL);
+	dev = btrfs_find_device(fs_info->fs_devices, devid, NULL, NULL, true);
 	if (dev)
 		sctx = dev->scrub_ctx;
 	if (sctx)
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 0a3f122dd61f..120e4340792a 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -529,7 +529,9 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
 				if (token != Opt_compress &&
 				    token != Opt_compress_force)
 					info->compress_level =
-					  btrfs_compress_str2level(args[0].from);
+					  btrfs_compress_str2level(
+							BTRFS_COMPRESS_ZLIB,
+							args[0].from + 4);
 				btrfs_set_opt(info->mount_opt, COMPRESS);
 				btrfs_clear_opt(info->mount_opt, NODATACOW);
 				btrfs_clear_opt(info->mount_opt, NODATASUM);
@@ -542,9 +544,13 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
 				btrfs_clear_opt(info->mount_opt, NODATASUM);
 				btrfs_set_fs_incompat(info, COMPRESS_LZO);
 				no_compress = 0;
-			} else if (strcmp(args[0].from, "zstd") == 0) {
+			} else if (strncmp(args[0].from, "zstd", 4) == 0) {
 				compress_type = "zstd";
 				info->compress_type = BTRFS_COMPRESS_ZSTD;
+				info->compress_level =
+					btrfs_compress_str2level(
+							 BTRFS_COMPRESS_ZSTD,
+							 args[0].from + 4);
 				btrfs_set_opt(info->mount_opt, COMPRESS);
 				btrfs_clear_opt(info->mount_opt, NODATACOW);
 				btrfs_clear_opt(info->mount_opt, NODATASUM);
@@ -2190,6 +2196,9 @@ static long btrfs_control_ioctl(struct file *file, unsigned int cmd,
 		ret = PTR_ERR_OR_ZERO(device);
 		mutex_unlock(&uuid_mutex);
 		break;
+	case BTRFS_IOC_FORGET_DEV:
+		ret = btrfs_forget_devices(vol->name);
+		break;
 	case BTRFS_IOC_DEVICES_READY:
 		mutex_lock(&uuid_mutex);
 		device = btrfs_scan_one_device(vol->name, FMODE_READ,
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 4ec2b660d014..acdad6d658f5 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -122,6 +122,7 @@ static noinline void switch_commit_roots(struct btrfs_transaction *trans)
 		if (is_fstree(root->root_key.objectid))
 			btrfs_unpin_free_ino(root);
 		clear_btree_io_tree(&root->dirty_log_pages);
+		btrfs_qgroup_clean_swapped_blocks(root);
 	}
 
 	/* We can free old roots now. */
@@ -845,8 +846,7 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	btrfs_trans_release_metadata(trans);
 	trans->block_rsv = NULL;
 
-	if (!list_empty(&trans->new_bgs))
-		btrfs_create_pending_block_groups(trans);
+	btrfs_create_pending_block_groups(trans);
 
 	btrfs_trans_release_chunk_metadata(trans);
 
@@ -1532,7 +1532,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 		goto fail;
 	}
 
-	btrfs_set_lock_blocking(old);
+	btrfs_set_lock_blocking_write(old);
 
 	ret = btrfs_copy_root(trans, root, old, &tmp, objectid);
 	/* clean up in any case */
@@ -1943,8 +1943,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
 	cur_trans->delayed_refs.flushing = 1;
 	smp_wmb();
 
-	if (!list_empty(&trans->new_bgs))
-		btrfs_create_pending_block_groups(trans);
+	btrfs_create_pending_block_groups(trans);
 
 	ret = btrfs_run_delayed_refs(trans, 0);
 	if (ret) {
diff --git a/fs/btrfs/tree-defrag.c b/fs/btrfs/tree-defrag.c
index 3c0987ab587d..5f9e2dd413af 100644
--- a/fs/btrfs/tree-defrag.c
+++ b/fs/btrfs/tree-defrag.c
@@ -52,7 +52,7 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
 		u32 nritems;
 
 		root_node = btrfs_lock_root_node(root);
-		btrfs_set_lock_blocking(root_node);
+		btrfs_set_lock_blocking_write(root_node);
 		nritems = btrfs_header_nritems(root_node);
 		root->defrag_max.objectid = 0;
 		/* from above we know this is not a leaf */
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index ac232b3d6d7e..f06454a55e00 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -27,6 +27,7 @@
 #define LOG_INODE_ALL 0
 #define LOG_INODE_EXISTS 1
 #define LOG_OTHER_INODE 2
+#define LOG_OTHER_INODE_ALL 3
 
 /*
  * directory trouble cases
@@ -1330,6 +1331,67 @@ out:
 	return ret;
 }
 
+static int add_link(struct btrfs_trans_handle *trans, struct btrfs_root *root,
+		    struct inode *dir, struct inode *inode, const char *name,
+		    int namelen, u64 ref_index)
+{
+	struct btrfs_dir_item *dir_item;
+	struct btrfs_key key;
+	struct btrfs_path *path;
+	struct inode *other_inode = NULL;
+	int ret;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
+	dir_item = btrfs_lookup_dir_item(NULL, root, path,
+					 btrfs_ino(BTRFS_I(dir)),
+					 name, namelen, 0);
+	if (!dir_item) {
+		btrfs_release_path(path);
+		goto add_link;
+	} else if (IS_ERR(dir_item)) {
+		ret = PTR_ERR(dir_item);
+		goto out;
+	}
+
+	/*
+	 * Our inode's dentry collides with the dentry of another inode which is
+	 * in the log but not yet processed since it has a higher inode number.
+	 * So delete that other dentry.
+	 */
+	btrfs_dir_item_key_to_cpu(path->nodes[0], dir_item, &key);
+	btrfs_release_path(path);
+	other_inode = read_one_inode(root, key.objectid);
+	if (!other_inode) {
+		ret = -ENOENT;
+		goto out;
+	}
+	ret = btrfs_unlink_inode(trans, root, BTRFS_I(dir), BTRFS_I(other_inode),
+				 name, namelen);
+	if (ret)
+		goto out;
+	/*
+	 * If we dropped the link count to 0, bump it so that later the iput()
+	 * on the inode will not free it. We will fixup the link count later.
+	 */
+	if (other_inode->i_nlink == 0)
+		inc_nlink(other_inode);
+
+	ret = btrfs_run_delayed_items(trans);
+	if (ret)
+		goto out;
+add_link:
+	ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode),
+			     name, namelen, 0, ref_index);
+out:
+	iput(other_inode);
+	btrfs_free_path(path);
+
+	return ret;
+}
+
 /*
  * replay one inode back reference item found in the log tree.
  * eb, slot and key refer to the buffer and key found in the log tree.
@@ -1466,9 +1528,8 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 				goto out;
 
 			/* insert our name */
-			ret = btrfs_add_link(trans, BTRFS_I(dir),
-					BTRFS_I(inode),
-					name, namelen, 0, ref_index);
+			ret = add_link(trans, root, dir, inode, name, namelen,
+				       ref_index);
 			if (ret)
 				goto out;
 
@@ -2663,7 +2724,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 
 				if (trans) {
 					btrfs_tree_lock(next);
-					btrfs_set_lock_blocking(next);
+					btrfs_set_lock_blocking_write(next);
 					clean_tree_block(fs_info, next);
 					btrfs_wait_tree_block_writeback(next);
 					btrfs_tree_unlock(next);
@@ -2747,7 +2808,7 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
 
 				if (trans) {
 					btrfs_tree_lock(next);
-					btrfs_set_lock_blocking(next);
+					btrfs_set_lock_blocking_write(next);
 					clean_tree_block(fs_info, next);
 					btrfs_wait_tree_block_writeback(next);
 					btrfs_tree_unlock(next);
@@ -2829,7 +2890,7 @@ static int walk_log_tree(struct btrfs_trans_handle *trans,
 
 			if (trans) {
 				btrfs_tree_lock(next);
-				btrfs_set_lock_blocking(next);
+				btrfs_set_lock_blocking_write(next);
 				clean_tree_block(fs_info, next);
 				btrfs_wait_tree_block_writeback(next);
 				btrfs_tree_unlock(next);
@@ -3706,6 +3767,8 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans,
 		found_key.type = 0;
 		ret = btrfs_bin_search(path->nodes[0], &found_key, 0,
 				       &start_slot);
+		if (ret < 0)
+			break;
 
 		ret = btrfs_del_items(trans, log, path, start_slot,
 				      path->slots[0] - start_slot + 1);
@@ -4717,7 +4780,7 @@ static int btrfs_check_ref_name_override(struct extent_buffer *eb,
 					 const int slot,
 					 const struct btrfs_key *key,
 					 struct btrfs_inode *inode,
-					 u64 *other_ino)
+					 u64 *other_ino, u64 *other_parent)
 {
 	int ret;
 	struct btrfs_path *search_path;
@@ -4780,8 +4843,13 @@ static int btrfs_check_ref_name_override(struct extent_buffer *eb,
 			btrfs_dir_item_key_to_cpu(search_path->nodes[0],
 						  di, &di_key);
 			if (di_key.type == BTRFS_INODE_ITEM_KEY) {
-				ret = 1;
-				*other_ino = di_key.objectid;
+				if (di_key.objectid != key->objectid) {
+					ret = 1;
+					*other_ino = di_key.objectid;
+					*other_parent = parent;
+				} else {
+					ret = 0;
+				}
 			} else {
 				ret = -EAGAIN;
 			}
@@ -4801,6 +4869,144 @@ out:
 	return ret;
 }
 
+struct btrfs_ino_list {
+	u64 ino;
+	u64 parent;
+	struct list_head list;
+};
+
+static int log_conflicting_inodes(struct btrfs_trans_handle *trans,
+				  struct btrfs_root *root,
+				  struct btrfs_path *path,
+				  struct btrfs_log_ctx *ctx,
+				  u64 ino, u64 parent)
+{
+	struct btrfs_ino_list *ino_elem;
+	LIST_HEAD(inode_list);
+	int ret = 0;
+
+	ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS);
+	if (!ino_elem)
+		return -ENOMEM;
+	ino_elem->ino = ino;
+	ino_elem->parent = parent;
+	list_add_tail(&ino_elem->list, &inode_list);
+
+	while (!list_empty(&inode_list)) {
+		struct btrfs_fs_info *fs_info = root->fs_info;
+		struct btrfs_key key;
+		struct inode *inode;
+
+		ino_elem = list_first_entry(&inode_list, struct btrfs_ino_list,
+					    list);
+		ino = ino_elem->ino;
+		parent = ino_elem->parent;
+		list_del(&ino_elem->list);
+		kfree(ino_elem);
+		if (ret)
+			continue;
+
+		btrfs_release_path(path);
+
+		key.objectid = ino;
+		key.type = BTRFS_INODE_ITEM_KEY;
+		key.offset = 0;
+		inode = btrfs_iget(fs_info->sb, &key, root, NULL);
+		/*
+		 * If the other inode that had a conflicting dir entry was
+		 * deleted in the current transaction, we need to log its parent
+		 * directory.
+		 */
+		if (IS_ERR(inode)) {
+			ret = PTR_ERR(inode);
+			if (ret == -ENOENT) {
+				key.objectid = parent;
+				inode = btrfs_iget(fs_info->sb, &key, root,
+						   NULL);
+				if (IS_ERR(inode)) {
+					ret = PTR_ERR(inode);
+				} else {
+					ret = btrfs_log_inode(trans, root,
+						      BTRFS_I(inode),
+						      LOG_OTHER_INODE_ALL,
+						      0, LLONG_MAX, ctx);
+					iput(inode);
+				}
+			}
+			continue;
+		}
+		/*
+		 * We are safe logging the other inode without acquiring its
+		 * lock as long as we log with the LOG_INODE_EXISTS mode. We
+		 * are safe against concurrent renames of the other inode as
+		 * well because during a rename we pin the log and update the
+		 * log with the new name before we unpin it.
+		 */
+		ret = btrfs_log_inode(trans, root, BTRFS_I(inode),
+				      LOG_OTHER_INODE, 0, LLONG_MAX, ctx);
+		if (ret) {
+			iput(inode);
+			continue;
+		}
+
+		key.objectid = ino;
+		key.type = BTRFS_INODE_REF_KEY;
+		key.offset = 0;
+		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+		if (ret < 0) {
+			iput(inode);
+			continue;
+		}
+
+		while (true) {
+			struct extent_buffer *leaf = path->nodes[0];
+			int slot = path->slots[0];
+			u64 other_ino = 0;
+			u64 other_parent = 0;
+
+			if (slot >= btrfs_header_nritems(leaf)) {
+				ret = btrfs_next_leaf(root, path);
+				if (ret < 0) {
+					break;
+				} else if (ret > 0) {
+					ret = 0;
+					break;
+				}
+				continue;
+			}
+
+			btrfs_item_key_to_cpu(leaf, &key, slot);
+			if (key.objectid != ino ||
+			    (key.type != BTRFS_INODE_REF_KEY &&
+			     key.type != BTRFS_INODE_EXTREF_KEY)) {
+				ret = 0;
+				break;
+			}
+
+			ret = btrfs_check_ref_name_override(leaf, slot, &key,
+					BTRFS_I(inode), &other_ino,
+					&other_parent);
+			if (ret < 0)
+				break;
+			if (ret > 0) {
+				ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS);
+				if (!ino_elem) {
+					ret = -ENOMEM;
+					break;
+				}
+				ino_elem->ino = other_ino;
+				ino_elem->parent = other_parent;
+				list_add_tail(&ino_elem->list, &inode_list);
+				ret = 0;
+			}
+			path->slots[0]++;
+		}
+		iput(inode);
+	}
+
+	return ret;
+}
+
 /* log a single inode in the tree log.
  * At least one parent directory for this inode must exist in the tree
  * or be logged already.
@@ -4840,6 +5046,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	u64 logged_isize = 0;
 	bool need_log_inode_item = true;
 	bool xattrs_logged = false;
+	bool recursive_logging = false;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -4885,8 +5092,12 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 		return ret;
 	}
 
-	if (inode_only == LOG_OTHER_INODE) {
-		inode_only = LOG_INODE_EXISTS;
+	if (inode_only == LOG_OTHER_INODE || inode_only == LOG_OTHER_INODE_ALL) {
+		recursive_logging = true;
+		if (inode_only == LOG_OTHER_INODE)
+			inode_only = LOG_INODE_EXISTS;
+		else
+			inode_only = LOG_INODE_ALL;
 		mutex_lock_nested(&inode->log_mutex, SINGLE_DEPTH_NESTING);
 	} else {
 		mutex_lock(&inode->log_mutex);
@@ -4981,20 +5192,19 @@ again:
 
 		if ((min_key.type == BTRFS_INODE_REF_KEY ||
 		     min_key.type == BTRFS_INODE_EXTREF_KEY) &&
-		    inode->generation == trans->transid) {
+		    inode->generation == trans->transid &&
+		    !recursive_logging) {
 			u64 other_ino = 0;
+			u64 other_parent = 0;
 
 			ret = btrfs_check_ref_name_override(path->nodes[0],
 					path->slots[0], &min_key, inode,
-					&other_ino);
+					&other_ino, &other_parent);
 			if (ret < 0) {
 				err = ret;
 				goto out_unlock;
 			} else if (ret > 0 && ctx &&
 				   other_ino != btrfs_ino(BTRFS_I(ctx->inode))) {
-				struct btrfs_key inode_key;
-				struct inode *other_inode;
-
 				if (ins_nr > 0) {
 					ins_nr++;
 				} else {
@@ -5010,43 +5220,13 @@ again:
 					goto out_unlock;
 				}
 				ins_nr = 0;
-				btrfs_release_path(path);
-				inode_key.objectid = other_ino;
-				inode_key.type = BTRFS_INODE_ITEM_KEY;
-				inode_key.offset = 0;
-				other_inode = btrfs_iget(fs_info->sb,
-							 &inode_key, root,
-							 NULL);
-				/*
-				 * If the other inode that had a conflicting dir
-				 * entry was deleted in the current transaction,
-				 * we don't need to do more work nor fallback to
-				 * a transaction commit.
-				 */
-				if (other_inode == ERR_PTR(-ENOENT)) {
-					goto next_key;
-				} else if (IS_ERR(other_inode)) {
-					err = PTR_ERR(other_inode);
-					goto out_unlock;
-				}
-				/*
-				 * We are safe logging the other inode without
-				 * acquiring its i_mutex as long as we log with
-				 * the LOG_INODE_EXISTS mode. We're safe against
-				 * concurrent renames of the other inode as well
-				 * because during a rename we pin the log and
-				 * update the log with the new name before we
-				 * unpin it.
-				 */
-				err = btrfs_log_inode(trans, root,
-						BTRFS_I(other_inode),
-						LOG_OTHER_INODE, 0, LLONG_MAX,
-						ctx);
-				iput(other_inode);
+
+				err = log_conflicting_inodes(trans, root, path,
+						ctx, other_ino, other_parent);
 				if (err)
 					goto out_unlock;
-				else
-					goto next_key;
+				btrfs_release_path(path);
+				goto next_key;
 			}
 		}
 
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 15561926ab32..9024eee889b9 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -415,27 +415,6 @@ static struct btrfs_device *__alloc_device(void)
 	return dev;
 }
 
-/*
- * Find a device specified by @devid or @uuid in the list of @fs_devices, or
- * return NULL.
- *
- * If devid and uuid are both specified, the match must be exact, otherwise
- * only devid is used.
- */
-static struct btrfs_device *find_device(struct btrfs_fs_devices *fs_devices,
-		u64 devid, const u8 *uuid)
-{
-	struct btrfs_device *dev;
-
-	list_for_each_entry(dev, &fs_devices->devices, dev_list) {
-		if (dev->devid == devid &&
-		    (!uuid || !memcmp(dev->uuid, uuid, BTRFS_UUID_SIZE))) {
-			return dev;
-		}
-	}
-	return NULL;
-}
-
 static noinline struct btrfs_fs_devices *find_fsid(
 		const u8 *fsid, const u8 *metadata_fsid)
 {
@@ -734,6 +713,17 @@ static void pending_bios_fn(struct btrfs_work *work)
 	run_scheduled_bios(device);
 }
 
+static bool device_path_matched(const char *path, struct btrfs_device *device)
+{
+	int found;
+
+	rcu_read_lock();
+	found = strcmp(rcu_str_deref(device->name), path);
+	rcu_read_unlock();
+
+	return found == 0;
+}
+
 /*
  *  Search and remove all stale (devices which are not mounted) devices.
  *  When both inputs are NULL, it will search and release all stale devices.
@@ -741,52 +731,57 @@ static void pending_bios_fn(struct btrfs_work *work)
  *		matching this path only.
  *  skip_dev:	Optional. Will skip this device when searching for the stale
  *		devices.
+ *  Return:	0 for success or if @path is NULL.
+ * 		-EBUSY if @path is a mounted device.
+ * 		-ENOENT if @path does not match any device in the list.
  */
-static void btrfs_free_stale_devices(const char *path,
+static int btrfs_free_stale_devices(const char *path,
 				     struct btrfs_device *skip_device)
 {
 	struct btrfs_fs_devices *fs_devices, *tmp_fs_devices;
 	struct btrfs_device *device, *tmp_device;
+	int ret = 0;
+
+	if (path)
+		ret = -ENOENT;
 
 	list_for_each_entry_safe(fs_devices, tmp_fs_devices, &fs_uuids, fs_list) {
-		mutex_lock(&fs_devices->device_list_mutex);
-		if (fs_devices->opened) {
-			mutex_unlock(&fs_devices->device_list_mutex);
-			continue;
-		}
 
+		mutex_lock(&fs_devices->device_list_mutex);
 		list_for_each_entry_safe(device, tmp_device,
 					 &fs_devices->devices, dev_list) {
-			int not_found = 0;
-
 			if (skip_device && skip_device == device)
 				continue;
 			if (path && !device->name)
 				continue;
-
-			rcu_read_lock();
-			if (path)
-				not_found = strcmp(rcu_str_deref(device->name),
-						   path);
-			rcu_read_unlock();
-			if (not_found)
+			if (path && !device_path_matched(path, device))
 				continue;
+			if (fs_devices->opened) {
+				/* for an already deleted device return 0 */
+				if (path && ret != 0)
+					ret = -EBUSY;
+				break;
+			}
 
 			/* delete the stale device */
 			fs_devices->num_devices--;
 			list_del(&device->dev_list);
 			btrfs_free_device(device);
 
+			ret = 0;
 			if (fs_devices->num_devices == 0)
 				break;
 		}
 		mutex_unlock(&fs_devices->device_list_mutex);
+
 		if (fs_devices->num_devices == 0) {
 			btrfs_sysfs_remove_fsid(fs_devices);
 			list_del(&fs_devices->fs_list);
 			free_fs_devices(fs_devices);
 		}
 	}
+
+	return ret;
 }
 
 static int btrfs_open_one_device(struct btrfs_fs_devices *fs_devices,
@@ -968,8 +963,8 @@ static noinline struct btrfs_device *device_list_add(const char *path,
 		device = NULL;
 	} else {
 		mutex_lock(&fs_devices->device_list_mutex);
-		device = find_device(fs_devices, devid,
-				disk_super->dev_item.uuid);
+		device = btrfs_find_device(fs_devices, devid,
+				disk_super->dev_item.uuid, NULL, false);
 
 		/*
 		 * If this disk has been pulled into an fs devices created by
@@ -1134,7 +1129,6 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 	mutex_lock(&orig->device_list_mutex);
 	fs_devices->total_devices = orig->total_devices;
 
-	/* We have held the volume lock, it is safe to get the devices. */
 	list_for_each_entry(orig_dev, &orig->devices, dev_list) {
 		struct rcu_string *name;
 
@@ -1451,6 +1445,17 @@ static int btrfs_read_disk_super(struct block_device *bdev, u64 bytenr,
 	return 0;
 }
 
+int btrfs_forget_devices(const char *path)
+{
+	int ret;
+
+	mutex_lock(&uuid_mutex);
+	ret = btrfs_free_stale_devices(strlen(path) ? path : NULL, NULL);
+	mutex_unlock(&uuid_mutex);
+
+	return ret;
+}
+
 /*
  * Look for a btrfs signature on a device. This may be called out of the mount path
  * and we are not allowed to call set_blocksize during the scan. The superblock
@@ -2385,11 +2390,11 @@ static struct btrfs_device *btrfs_find_device_by_path(
 	devid = btrfs_stack_device_id(&disk_super->dev_item);
 	dev_uuid = disk_super->dev_item.uuid;
 	if (btrfs_fs_incompat(fs_info, METADATA_UUID))
-		device = btrfs_find_device(fs_info, devid, dev_uuid,
-				disk_super->metadata_uuid);
+		device = btrfs_find_device(fs_info->fs_devices, devid, dev_uuid,
+					   disk_super->metadata_uuid, true);
 	else
-		device = btrfs_find_device(fs_info, devid,
-				dev_uuid, disk_super->fsid);
+		device = btrfs_find_device(fs_info->fs_devices, devid, dev_uuid,
+					   disk_super->fsid, true);
 
 	brelse(bh);
 	if (!device)
@@ -2398,50 +2403,38 @@ static struct btrfs_device *btrfs_find_device_by_path(
 	return device;
 }
 
-static struct btrfs_device *btrfs_find_device_missing_or_by_path(
-		struct btrfs_fs_info *fs_info, const char *device_path)
-{
-	struct btrfs_device *device = NULL;
-	if (strcmp(device_path, "missing") == 0) {
-		struct list_head *devices;
-		struct btrfs_device *tmp;
-
-		devices = &fs_info->fs_devices->devices;
-		list_for_each_entry(tmp, devices, dev_list) {
-			if (test_bit(BTRFS_DEV_STATE_IN_FS_METADATA,
-					&tmp->dev_state) && !tmp->bdev) {
-				device = tmp;
-				break;
-			}
-		}
-
-		if (!device)
-			return ERR_PTR(-ENOENT);
-	} else {
-		device = btrfs_find_device_by_path(fs_info, device_path);
-	}
-
-	return device;
-}
-
 /*
  * Lookup a device given by device id, or the path if the id is 0.
  */
 struct btrfs_device *btrfs_find_device_by_devspec(
-		struct btrfs_fs_info *fs_info, u64 devid, const char *devpath)
+		struct btrfs_fs_info *fs_info, u64 devid,
+		const char *device_path)
 {
 	struct btrfs_device *device;
 
 	if (devid) {
-		device = btrfs_find_device(fs_info, devid, NULL, NULL);
+		device = btrfs_find_device(fs_info->fs_devices, devid, NULL,
+					   NULL, true);
 		if (!device)
 			return ERR_PTR(-ENOENT);
-	} else {
-		if (!devpath || !devpath[0])
-			return ERR_PTR(-EINVAL);
-		device = btrfs_find_device_missing_or_by_path(fs_info, devpath);
+		return device;
 	}
-	return device;
+
+	if (!device_path || !device_path[0])
+		return ERR_PTR(-EINVAL);
+
+	if (strcmp(device_path, "missing") == 0) {
+		/* Find first missing device */
+		list_for_each_entry(device, &fs_info->fs_devices->devices,
+				    dev_list) {
+			if (test_bit(BTRFS_DEV_STATE_IN_FS_METADATA,
+				     &device->dev_state) && !device->bdev)
+				return device;
+		}
+		return ERR_PTR(-ENOENT);
+	}
+
+	return btrfs_find_device_by_path(fs_info, device_path);
 }
 
 /*
@@ -2563,7 +2556,8 @@ next_slot:
 				   BTRFS_UUID_SIZE);
 		read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
 				   BTRFS_FSID_SIZE);
-		device = btrfs_find_device(fs_info, devid, dev_uuid, fs_uuid);
+		device = btrfs_find_device(fs_info->fs_devices, devid, dev_uuid,
+					   fs_uuid, true);
 		BUG_ON(!device); /* Logic error */
 
 		if (device->fs_devices->seeding) {
@@ -6616,21 +6610,36 @@ blk_status_t btrfs_map_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
 	return BLK_STS_OK;
 }
 
-struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
-				       u8 *uuid, u8 *fsid)
+/*
+ * Find a device specified by @devid or @uuid in the list of @fs_devices, or
+ * return NULL.
+ *
+ * If devid and uuid are both specified, the match must be exact, otherwise
+ * only devid is used.
+ *
+ * If @seed is true, traverse through the seed devices.
+ */
+struct btrfs_device *btrfs_find_device(struct btrfs_fs_devices *fs_devices,
+				       u64 devid, u8 *uuid, u8 *fsid,
+				       bool seed)
 {
 	struct btrfs_device *device;
-	struct btrfs_fs_devices *cur_devices;
 
-	cur_devices = fs_info->fs_devices;
-	while (cur_devices) {
+	while (fs_devices) {
 		if (!fsid ||
-		    !memcmp(cur_devices->metadata_uuid, fsid, BTRFS_FSID_SIZE)) {
-			device = find_device(cur_devices, devid, uuid);
-			if (device)
-				return device;
+		    !memcmp(fs_devices->metadata_uuid, fsid, BTRFS_FSID_SIZE)) {
+			list_for_each_entry(device, &fs_devices->devices,
+					    dev_list) {
+				if (device->devid == devid &&
+				    (!uuid || memcmp(device->uuid, uuid,
+						     BTRFS_UUID_SIZE) == 0))
+					return device;
+			}
 		}
-		cur_devices = cur_devices->seed;
+		if (seed)
+			fs_devices = fs_devices->seed;
+		else
+			return NULL;
 	}
 	return NULL;
 }
@@ -6782,10 +6791,10 @@ static int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info,
 	}
 
 	if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes != 2) ||
-	    (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) ||
+	    (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes != 2) ||
 	    (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) ||
 	    (type & BTRFS_BLOCK_GROUP_RAID6 && num_stripes < 3) ||
-	    (type & BTRFS_BLOCK_GROUP_DUP && num_stripes > 2) ||
+	    (type & BTRFS_BLOCK_GROUP_DUP && num_stripes != 2) ||
 	    ((type & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0 &&
 	     num_stripes != 1)) {
 		btrfs_err(fs_info,
@@ -6875,8 +6884,8 @@ static int read_one_chunk(struct btrfs_fs_info *fs_info, struct btrfs_key *key,
 		read_extent_buffer(leaf, uuid, (unsigned long)
 				   btrfs_stripe_dev_uuid_nr(chunk, i),
 				   BTRFS_UUID_SIZE);
-		map->stripes[i].dev = btrfs_find_device(fs_info, devid,
-							uuid, NULL);
+		map->stripes[i].dev = btrfs_find_device(fs_info->fs_devices,
+							devid, uuid, NULL, true);
 		if (!map->stripes[i].dev &&
 		    !btrfs_test_opt(fs_info, DEGRADED)) {
 			free_extent_map(em);
@@ -7015,7 +7024,8 @@ static int read_one_dev(struct btrfs_fs_info *fs_info,
 			return PTR_ERR(fs_devices);
 	}
 
-	device = btrfs_find_device(fs_info, devid, dev_uuid, fs_uuid);
+	device = btrfs_find_device(fs_info->fs_devices, devid, dev_uuid,
+				   fs_uuid, true);
 	if (!device) {
 		if (!btrfs_test_opt(fs_info, DEGRADED)) {
 			btrfs_report_missing_device(fs_info, devid,
@@ -7605,7 +7615,8 @@ int btrfs_get_dev_stats(struct btrfs_fs_info *fs_info,
 	int i;
 
 	mutex_lock(&fs_devices->device_list_mutex);
-	dev = btrfs_find_device(fs_info, stats->devid, NULL, NULL);
+	dev = btrfs_find_device(fs_info->fs_devices, stats->devid, NULL, NULL,
+				true);
 	mutex_unlock(&fs_devices->device_list_mutex);
 
 	if (!dev) {
@@ -7819,7 +7830,7 @@ static int verify_one_dev_extent(struct btrfs_fs_info *fs_info,
 	}
 
 	/* Make sure no dev extent is beyond device bondary */
-	dev = btrfs_find_device(fs_info, devid, NULL, NULL);
+	dev = btrfs_find_device(fs_info->fs_devices, devid, NULL, NULL, true);
 	if (!dev) {
 		btrfs_err(fs_info, "failed to find devid %llu", devid);
 		ret = -EUCLEAN;
@@ -7828,7 +7839,8 @@ static int verify_one_dev_extent(struct btrfs_fs_info *fs_info,
 
 	/* It's possible this device is a dummy for seed device */
 	if (dev->disk_total_bytes == 0) {
-		dev = find_device(fs_info->fs_devices->seed, devid, NULL);
+		dev = btrfs_find_device(fs_info->fs_devices->seed, devid, NULL,
+					NULL, false);
 		if (!dev) {
 			btrfs_err(fs_info, "failed to find seed devid %llu",
 				  devid);
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index ed806649a473..3ad9d58d1b66 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -416,6 +416,7 @@ int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 		       fmode_t flags, void *holder);
 struct btrfs_device *btrfs_scan_one_device(const char *path,
 					   fmode_t flags, void *holder);
+int btrfs_forget_devices(const char *path);
 int btrfs_close_devices(struct btrfs_fs_devices *fs_devices);
 void btrfs_free_extra_devids(struct btrfs_fs_devices *fs_devices, int step);
 void btrfs_assign_next_active_device(struct btrfs_device *device,
@@ -433,8 +434,8 @@ void __exit btrfs_cleanup_fs_uuids(void);
 int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len);
 int btrfs_grow_device(struct btrfs_trans_handle *trans,
 		      struct btrfs_device *device, u64 new_size);
-struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
-				       u8 *uuid, u8 *fsid);
+struct btrfs_device *btrfs_find_device(struct btrfs_fs_devices *fs_devices,
+				       u64 devid, u8 *uuid, u8 *fsid, bool seed);
 int btrfs_shrink_device(struct btrfs_device *device, u64 new_size);
 int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *path);
 int btrfs_balance(struct btrfs_fs_info *fs_info,
diff --git a/fs/btrfs/zlib.c b/fs/btrfs/zlib.c
index 970ff3e35bb3..b86b7ad6b900 100644
--- a/fs/btrfs/zlib.c
+++ b/fs/btrfs/zlib.c
@@ -27,6 +27,33 @@ struct workspace {
 	int level;
 };
 
+static struct workspace_manager wsm;
+
+static void zlib_init_workspace_manager(void)
+{
+	btrfs_init_workspace_manager(&wsm, &btrfs_zlib_compress);
+}
+
+static void zlib_cleanup_workspace_manager(void)
+{
+	btrfs_cleanup_workspace_manager(&wsm);
+}
+
+static struct list_head *zlib_get_workspace(unsigned int level)
+{
+	struct list_head *ws = btrfs_get_workspace(&wsm, level);
+	struct workspace *workspace = list_entry(ws, struct workspace, list);
+
+	workspace->level = level;
+
+	return ws;
+}
+
+static void zlib_put_workspace(struct list_head *ws)
+{
+	btrfs_put_workspace(&wsm, ws);
+}
+
 static void zlib_free_workspace(struct list_head *ws)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
@@ -36,7 +63,7 @@ static void zlib_free_workspace(struct list_head *ws)
 	kfree(workspace);
 }
 
-static struct list_head *zlib_alloc_workspace(void)
+static struct list_head *zlib_alloc_workspace(unsigned int level)
 {
 	struct workspace *workspace;
 	int workspacesize;
@@ -48,6 +75,7 @@ static struct list_head *zlib_alloc_workspace(void)
 	workspacesize = max(zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL),
 			zlib_inflate_workspacesize());
 	workspace->strm.workspace = kvmalloc(workspacesize, GFP_KERNEL);
+	workspace->level = level;
 	workspace->buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
 	if (!workspace->strm.workspace || !workspace->buf)
 		goto fail;
@@ -390,18 +418,19 @@ next:
 	return ret;
 }
 
-static void zlib_set_level(struct list_head *ws, unsigned int type)
+static unsigned int zlib_set_level(unsigned int level)
 {
-	struct workspace *workspace = list_entry(ws, struct workspace, list);
-	unsigned level = (type & 0xF0) >> 4;
-
-	if (level > 9)
-		level = 9;
+	if (!level)
+		return BTRFS_ZLIB_DEFAULT_LEVEL;
 
-	workspace->level = level > 0 ? level : 3;
+	return min_t(unsigned int, level, 9);
 }
 
 const struct btrfs_compress_op btrfs_zlib_compress = {
+	.init_workspace_manager	= zlib_init_workspace_manager,
+	.cleanup_workspace_manager = zlib_cleanup_workspace_manager,
+	.get_workspace		= zlib_get_workspace,
+	.put_workspace		= zlib_put_workspace,
 	.alloc_workspace	= zlib_alloc_workspace,
 	.free_workspace		= zlib_free_workspace,
 	.compress_pages		= zlib_compress_pages,
diff --git a/fs/btrfs/zstd.c b/fs/btrfs/zstd.c
index af6ec59972f5..3e418a3aeb11 100644
--- a/fs/btrfs/zstd.c
+++ b/fs/btrfs/zstd.c
@@ -6,25 +6,31 @@
  */
 
 #include <linux/bio.h>
+#include <linux/bitmap.h>
 #include <linux/err.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
+#include <linux/sched/mm.h>
 #include <linux/pagemap.h>
 #include <linux/refcount.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/zstd.h>
 #include "compression.h"
+#include "ctree.h"
 
 #define ZSTD_BTRFS_MAX_WINDOWLOG 17
 #define ZSTD_BTRFS_MAX_INPUT (1 << ZSTD_BTRFS_MAX_WINDOWLOG)
 #define ZSTD_BTRFS_DEFAULT_LEVEL 3
+#define ZSTD_BTRFS_MAX_LEVEL 15
+/* 307s to avoid pathologically clashing with transaction commit */
+#define ZSTD_BTRFS_RECLAIM_JIFFIES (307 * HZ)
 
-static ZSTD_parameters zstd_get_btrfs_parameters(size_t src_len)
+static ZSTD_parameters zstd_get_btrfs_parameters(unsigned int level,
+						 size_t src_len)
 {
-	ZSTD_parameters params = ZSTD_getParams(ZSTD_BTRFS_DEFAULT_LEVEL,
-						src_len, 0);
+	ZSTD_parameters params = ZSTD_getParams(level, src_len, 0);
 
 	if (params.cParams.windowLog > ZSTD_BTRFS_MAX_WINDOWLOG)
 		params.cParams.windowLog = ZSTD_BTRFS_MAX_WINDOWLOG;
@@ -36,11 +42,290 @@ struct workspace {
 	void *mem;
 	size_t size;
 	char *buf;
+	unsigned int level;
+	unsigned int req_level;
+	unsigned long last_used; /* jiffies */
 	struct list_head list;
+	struct list_head lru_list;
 	ZSTD_inBuffer in_buf;
 	ZSTD_outBuffer out_buf;
 };
 
+/*
+ * Zstd Workspace Management
+ *
+ * Zstd workspaces have different memory requirements depending on the level.
+ * The zstd workspaces are managed by having individual lists for each level
+ * and a global lru.  Forward progress is maintained by protecting a max level
+ * workspace.
+ *
+ * Getting a workspace is done by using the bitmap to identify the levels that
+ * have available workspaces and scans up.  This lets us recycle higher level
+ * workspaces because of the monotonic memory guarantee.  A workspace's
+ * last_used is only updated if it is being used by the corresponding memory
+ * level.  Putting a workspace involves adding it back to the appropriate places
+ * and adding it back to the lru if necessary.
+ *
+ * A timer is used to reclaim workspaces if they have not been used for
+ * ZSTD_BTRFS_RECLAIM_JIFFIES.  This helps keep only active workspaces around.
+ * The upper bound is provided by the workqueue limit which is 2 (percpu limit).
+ */
+
+struct zstd_workspace_manager {
+	const struct btrfs_compress_op *ops;
+	spinlock_t lock;
+	struct list_head lru_list;
+	struct list_head idle_ws[ZSTD_BTRFS_MAX_LEVEL];
+	unsigned long active_map;
+	wait_queue_head_t wait;
+	struct timer_list timer;
+};
+
+static struct zstd_workspace_manager wsm;
+
+static size_t zstd_ws_mem_sizes[ZSTD_BTRFS_MAX_LEVEL];
+
+static inline struct workspace *list_to_workspace(struct list_head *list)
+{
+	return container_of(list, struct workspace, list);
+}
+
+/*
+ * zstd_reclaim_timer_fn - reclaim timer
+ * @t: timer
+ *
+ * This scans the lru_list and attempts to reclaim any workspace that hasn't
+ * been used for ZSTD_BTRFS_RECLAIM_JIFFIES.
+ */
+static void zstd_reclaim_timer_fn(struct timer_list *timer)
+{
+	unsigned long reclaim_threshold = jiffies - ZSTD_BTRFS_RECLAIM_JIFFIES;
+	struct list_head *pos, *next;
+
+	spin_lock(&wsm.lock);
+
+	if (list_empty(&wsm.lru_list)) {
+		spin_unlock(&wsm.lock);
+		return;
+	}
+
+	list_for_each_prev_safe(pos, next, &wsm.lru_list) {
+		struct workspace *victim = container_of(pos, struct workspace,
+							lru_list);
+		unsigned int level;
+
+		if (time_after(victim->last_used, reclaim_threshold))
+			break;
+
+		/* workspace is in use */
+		if (victim->req_level)
+			continue;
+
+		level = victim->level;
+		list_del(&victim->lru_list);
+		list_del(&victim->list);
+		wsm.ops->free_workspace(&victim->list);
+
+		if (list_empty(&wsm.idle_ws[level - 1]))
+			clear_bit(level - 1, &wsm.active_map);
+
+	}
+
+	if (!list_empty(&wsm.lru_list))
+		mod_timer(&wsm.timer, jiffies + ZSTD_BTRFS_RECLAIM_JIFFIES);
+
+	spin_unlock(&wsm.lock);
+}
+
+/*
+ * zstd_calc_ws_mem_sizes - calculate monotonic memory bounds
+ *
+ * It is possible based on the level configurations that a higher level
+ * workspace uses less memory than a lower level workspace.  In order to reuse
+ * workspaces, this must be made a monotonic relationship.  This precomputes
+ * the required memory for each level and enforces the monotonicity between
+ * level and memory required.
+ */
+static void zstd_calc_ws_mem_sizes(void)
+{
+	size_t max_size = 0;
+	unsigned int level;
+
+	for (level = 1; level <= ZSTD_BTRFS_MAX_LEVEL; level++) {
+		ZSTD_parameters params =
+			zstd_get_btrfs_parameters(level, ZSTD_BTRFS_MAX_INPUT);
+		size_t level_size =
+			max_t(size_t,
+			      ZSTD_CStreamWorkspaceBound(params.cParams),
+			      ZSTD_DStreamWorkspaceBound(ZSTD_BTRFS_MAX_INPUT));
+
+		max_size = max_t(size_t, max_size, level_size);
+		zstd_ws_mem_sizes[level - 1] = max_size;
+	}
+}
+
+static void zstd_init_workspace_manager(void)
+{
+	struct list_head *ws;
+	int i;
+
+	zstd_calc_ws_mem_sizes();
+
+	wsm.ops = &btrfs_zstd_compress;
+	spin_lock_init(&wsm.lock);
+	init_waitqueue_head(&wsm.wait);
+	timer_setup(&wsm.timer, zstd_reclaim_timer_fn, 0);
+
+	INIT_LIST_HEAD(&wsm.lru_list);
+	for (i = 0; i < ZSTD_BTRFS_MAX_LEVEL; i++)
+		INIT_LIST_HEAD(&wsm.idle_ws[i]);
+
+	ws = wsm.ops->alloc_workspace(ZSTD_BTRFS_MAX_LEVEL);
+	if (IS_ERR(ws)) {
+		pr_warn(
+		"BTRFS: cannot preallocate zstd compression workspace\n");
+	} else {
+		set_bit(ZSTD_BTRFS_MAX_LEVEL - 1, &wsm.active_map);
+		list_add(ws, &wsm.idle_ws[ZSTD_BTRFS_MAX_LEVEL - 1]);
+	}
+}
+
+static void zstd_cleanup_workspace_manager(void)
+{
+	struct workspace *workspace;
+	int i;
+
+	del_timer(&wsm.timer);
+
+	for (i = 0; i < ZSTD_BTRFS_MAX_LEVEL; i++) {
+		while (!list_empty(&wsm.idle_ws[i])) {
+			workspace = container_of(wsm.idle_ws[i].next,
+						 struct workspace, list);
+			list_del(&workspace->list);
+			list_del(&workspace->lru_list);
+			wsm.ops->free_workspace(&workspace->list);
+		}
+	}
+}
+
+/*
+ * zstd_find_workspace - find workspace
+ * @level: compression level
+ *
+ * This iterates over the set bits in the active_map beginning at the requested
+ * compression level.  This lets us utilize already allocated workspaces before
+ * allocating a new one.  If the workspace is of a larger size, it is used, but
+ * the place in the lru_list and last_used times are not updated.  This is to
+ * offer the opportunity to reclaim the workspace in favor of allocating an
+ * appropriately sized one in the future.
+ */
+static struct list_head *zstd_find_workspace(unsigned int level)
+{
+	struct list_head *ws;
+	struct workspace *workspace;
+	int i = level - 1;
+
+	spin_lock(&wsm.lock);
+	for_each_set_bit_from(i, &wsm.active_map, ZSTD_BTRFS_MAX_LEVEL) {
+		if (!list_empty(&wsm.idle_ws[i])) {
+			ws = wsm.idle_ws[i].next;
+			workspace = list_to_workspace(ws);
+			list_del_init(ws);
+			/* keep its place if it's a lower level using this */
+			workspace->req_level = level;
+			if (level == workspace->level)
+				list_del(&workspace->lru_list);
+			if (list_empty(&wsm.idle_ws[i]))
+				clear_bit(i, &wsm.active_map);
+			spin_unlock(&wsm.lock);
+			return ws;
+		}
+	}
+	spin_unlock(&wsm.lock);
+
+	return NULL;
+}
+
+/*
+ * zstd_get_workspace - zstd's get_workspace
+ * @level: compression level
+ *
+ * If @level is 0, then any compression level can be used.  Therefore, we begin
+ * scanning from 1.  We first scan through possible workspaces and then after
+ * attempt to allocate a new workspace.  If we fail to allocate one due to
+ * memory pressure, go to sleep waiting for the max level workspace to free up.
+ */
+static struct list_head *zstd_get_workspace(unsigned int level)
+{
+	struct list_head *ws;
+	unsigned int nofs_flag;
+
+	/* level == 0 means we can use any workspace */
+	if (!level)
+		level = 1;
+
+again:
+	ws = zstd_find_workspace(level);
+	if (ws)
+		return ws;
+
+	nofs_flag = memalloc_nofs_save();
+	ws = wsm.ops->alloc_workspace(level);
+	memalloc_nofs_restore(nofs_flag);
+
+	if (IS_ERR(ws)) {
+		DEFINE_WAIT(wait);
+
+		prepare_to_wait(&wsm.wait, &wait, TASK_UNINTERRUPTIBLE);
+		schedule();
+		finish_wait(&wsm.wait, &wait);
+
+		goto again;
+	}
+
+	return ws;
+}
+
+/*
+ * zstd_put_workspace - zstd put_workspace
+ * @ws: list_head for the workspace
+ *
+ * When putting back a workspace, we only need to update the LRU if we are of
+ * the requested compression level.  Here is where we continue to protect the
+ * max level workspace or update last_used accordingly.  If the reclaim timer
+ * isn't set, it is also set here.  Only the max level workspace tries and wakes
+ * up waiting workspaces.
+ */
+static void zstd_put_workspace(struct list_head *ws)
+{
+	struct workspace *workspace = list_to_workspace(ws);
+
+	spin_lock(&wsm.lock);
+
+	/* A node is only taken off the lru if we are the corresponding level */
+	if (workspace->req_level == workspace->level) {
+		/* Hide a max level workspace from reclaim */
+		if (list_empty(&wsm.idle_ws[ZSTD_BTRFS_MAX_LEVEL - 1])) {
+			INIT_LIST_HEAD(&workspace->lru_list);
+		} else {
+			workspace->last_used = jiffies;
+			list_add(&workspace->lru_list, &wsm.lru_list);
+			if (!timer_pending(&wsm.timer))
+				mod_timer(&wsm.timer,
+					  jiffies + ZSTD_BTRFS_RECLAIM_JIFFIES);
+		}
+	}
+
+	set_bit(workspace->level - 1, &wsm.active_map);
+	list_add(&workspace->list, &wsm.idle_ws[workspace->level - 1]);
+	workspace->req_level = 0;
+
+	spin_unlock(&wsm.lock);
+
+	if (workspace->level == ZSTD_BTRFS_MAX_LEVEL)
+		cond_wake_up(&wsm.wait);
+}
+
 static void zstd_free_workspace(struct list_head *ws)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
@@ -50,25 +335,25 @@ static void zstd_free_workspace(struct list_head *ws)
 	kfree(workspace);
 }
 
-static struct list_head *zstd_alloc_workspace(void)
+static struct list_head *zstd_alloc_workspace(unsigned int level)
 {
-	ZSTD_parameters params =
-			zstd_get_btrfs_parameters(ZSTD_BTRFS_MAX_INPUT);
 	struct workspace *workspace;
 
 	workspace = kzalloc(sizeof(*workspace), GFP_KERNEL);
 	if (!workspace)
 		return ERR_PTR(-ENOMEM);
 
-	workspace->size = max_t(size_t,
-			ZSTD_CStreamWorkspaceBound(params.cParams),
-			ZSTD_DStreamWorkspaceBound(ZSTD_BTRFS_MAX_INPUT));
+	workspace->size = zstd_ws_mem_sizes[level - 1];
+	workspace->level = level;
+	workspace->req_level = level;
+	workspace->last_used = jiffies;
 	workspace->mem = kvmalloc(workspace->size, GFP_KERNEL);
 	workspace->buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
 	if (!workspace->mem || !workspace->buf)
 		goto fail;
 
 	INIT_LIST_HEAD(&workspace->list);
+	INIT_LIST_HEAD(&workspace->lru_list);
 
 	return &workspace->list;
 fail:
@@ -95,7 +380,8 @@ static int zstd_compress_pages(struct list_head *ws,
 	unsigned long len = *total_out;
 	const unsigned long nr_dest_pages = *out_pages;
 	unsigned long max_out = nr_dest_pages * PAGE_SIZE;
-	ZSTD_parameters params = zstd_get_btrfs_parameters(len);
+	ZSTD_parameters params = zstd_get_btrfs_parameters(workspace->req_level,
+							   len);
 
 	*out_pages = 0;
 	*total_out = 0;
@@ -419,11 +705,19 @@ finish:
 	return ret;
 }
 
-static void zstd_set_level(struct list_head *ws, unsigned int type)
+static unsigned int zstd_set_level(unsigned int level)
 {
+	if (!level)
+		return ZSTD_BTRFS_DEFAULT_LEVEL;
+
+	return min_t(unsigned int, level, ZSTD_BTRFS_MAX_LEVEL);
 }
 
 const struct btrfs_compress_op btrfs_zstd_compress = {
+	.init_workspace_manager = zstd_init_workspace_manager,
+	.cleanup_workspace_manager = zstd_cleanup_workspace_manager,
+	.get_workspace = zstd_get_workspace,
+	.put_workspace = zstd_put_workspace,
 	.alloc_workspace = zstd_alloc_workspace,
 	.free_workspace = zstd_free_workspace,
 	.compress_pages = zstd_compress_pages,