1 files changed, 428 insertions, 104 deletions

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 8b353ad02f03..0ec8e228b89f 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -2538,6 +2538,12 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 		 * list before we release it.
 		 */
 		if (btrfs_delayed_ref_is_head(ref)) {
+			if (locked_ref->is_data &&
+			    locked_ref->total_ref_mod < 0) {
+				spin_lock(&delayed_refs->lock);
+				delayed_refs->pending_csums -= ref->num_bytes;
+				spin_unlock(&delayed_refs->lock);
+			}
 			btrfs_delayed_ref_unlock(locked_ref);
 			locked_ref = NULL;
 		}
@@ -2561,8 +2567,7 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 		 */
 		spin_lock(&delayed_refs->lock);
 		avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
-		avg = div64_u64(avg, 4);
-		fs_info->avg_delayed_ref_runtime = avg;
+		fs_info->avg_delayed_ref_runtime = avg >> 2;	/* div by 4 */
 		spin_unlock(&delayed_refs->lock);
 	}
 	return 0;
@@ -2624,7 +2629,26 @@ static inline u64 heads_to_leaves(struct btrfs_root *root, u64 heads)
 	 * We don't ever fill up leaves all the way so multiply by 2 just to be
 	 * closer to what we're really going to want to ouse.
 	 */
-	return div64_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(root));
+	return div_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(root));
+}
+
+/*
+ * Takes the number of bytes to be csumm'ed and figures out how many leaves it
+ * would require to store the csums for that many bytes.
+ */
+u64 btrfs_csum_bytes_to_leaves(struct btrfs_root *root, u64 csum_bytes)
+{
+	u64 csum_size;
+	u64 num_csums_per_leaf;
+	u64 num_csums;
+
+	csum_size = BTRFS_LEAF_DATA_SIZE(root) - sizeof(struct btrfs_item);
+	num_csums_per_leaf = div64_u64(csum_size,
+			(u64)btrfs_super_csum_size(root->fs_info->super_copy));
+	num_csums = div64_u64(csum_bytes, root->sectorsize);
+	num_csums += num_csums_per_leaf - 1;
+	num_csums = div64_u64(num_csums, num_csums_per_leaf);
+	return num_csums;
 }
 
 int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
@@ -2632,7 +2656,9 @@ int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_block_rsv *global_rsv;
 	u64 num_heads = trans->transaction->delayed_refs.num_heads_ready;
-	u64 num_bytes;
+	u64 csum_bytes = trans->transaction->delayed_refs.pending_csums;
+	u64 num_dirty_bgs = trans->transaction->num_dirty_bgs;
+	u64 num_bytes, num_dirty_bgs_bytes;
 	int ret = 0;
 
 	num_bytes = btrfs_calc_trans_metadata_size(root, 1);
@@ -2640,17 +2666,22 @@ int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
 	if (num_heads > 1)
 		num_bytes += (num_heads - 1) * root->nodesize;
 	num_bytes <<= 1;
+	num_bytes += btrfs_csum_bytes_to_leaves(root, csum_bytes) * root->nodesize;
+	num_dirty_bgs_bytes = btrfs_calc_trans_metadata_size(root,
+							     num_dirty_bgs);
 	global_rsv = &root->fs_info->global_block_rsv;
 
 	/*
 	 * If we can't allocate any more chunks lets make sure we have _lots_ of
 	 * wiggle room since running delayed refs can create more delayed refs.
 	 */
-	if (global_rsv->space_info->full)
+	if (global_rsv->space_info->full) {
+		num_dirty_bgs_bytes <<= 1;
 		num_bytes <<= 1;
+	}
 
 	spin_lock(&global_rsv->lock);
-	if (global_rsv->reserved <= num_bytes)
+	if (global_rsv->reserved <= num_bytes + num_dirty_bgs_bytes)
 		ret = 1;
 	spin_unlock(&global_rsv->lock);
 	return ret;
@@ -3147,8 +3178,8 @@ static int write_one_cache_group(struct btrfs_trans_handle *trans,
 	bi = btrfs_item_ptr_offset(leaf, path->slots[0]);
 	write_extent_buffer(leaf, &cache->item, bi, sizeof(cache->item));
 	btrfs_mark_buffer_dirty(leaf);
-	btrfs_release_path(path);
 fail:
+	btrfs_release_path(path);
 	if (ret)
 		btrfs_abort_transaction(trans, root, ret);
 	return ret;
@@ -3193,7 +3224,7 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group,
 	struct inode *inode = NULL;
 	u64 alloc_hint = 0;
 	int dcs = BTRFS_DC_ERROR;
-	int num_pages = 0;
+	u64 num_pages = 0;
 	int retries = 0;
 	int ret = 0;
 
@@ -3267,15 +3298,14 @@ again:
 		if (ret)
 			goto out_put;
 
-		ret = btrfs_truncate_free_space_cache(root, trans, inode);
+		ret = btrfs_truncate_free_space_cache(root, trans, NULL, inode);
 		if (ret)
 			goto out_put;
 	}
 
 	spin_lock(&block_group->lock);
 	if (block_group->cached != BTRFS_CACHE_FINISHED ||
-	    !btrfs_test_opt(root, SPACE_CACHE) ||
-	    block_group->delalloc_bytes) {
+	    !btrfs_test_opt(root, SPACE_CACHE)) {
 		/*
 		 * don't bother trying to write stuff out _if_
 		 * a) we're not cached,
@@ -3293,14 +3323,14 @@ again:
 	 * taking up quite a bit since it's not folded into the other space
 	 * cache.
 	 */
-	num_pages = (int)div64_u64(block_group->key.offset, 256 * 1024 * 1024);
+	num_pages = div_u64(block_group->key.offset, 256 * 1024 * 1024);
 	if (!num_pages)
 		num_pages = 1;
 
 	num_pages *= 16;
 	num_pages *= PAGE_CACHE_SIZE;
 
-	ret = btrfs_check_data_free_space(inode, num_pages);
+	ret = btrfs_check_data_free_space(inode, num_pages, num_pages);
 	if (ret)
 		goto out_put;
 
@@ -3351,16 +3381,166 @@ int btrfs_setup_space_cache(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
+/*
+ * transaction commit does final block group cache writeback during a
+ * critical section where nothing is allowed to change the FS.  This is
+ * required in order for the cache to actually match the block group,
+ * but can introduce a lot of latency into the commit.
+ *
+ * So, btrfs_start_dirty_block_groups is here to kick off block group
+ * cache IO.  There's a chance we'll have to redo some of it if the
+ * block group changes again during the commit, but it greatly reduces
+ * the commit latency by getting rid of the easy block groups while
+ * we're still allowing others to join the commit.
+ */
+int btrfs_start_dirty_block_groups(struct btrfs_trans_handle *trans,
+				   struct btrfs_root *root)
+{
+	struct btrfs_block_group_cache *cache;
+	struct btrfs_transaction *cur_trans = trans->transaction;
+	int ret = 0;
+	int should_put;
+	struct btrfs_path *path = NULL;
+	LIST_HEAD(dirty);
+	struct list_head *io = &cur_trans->io_bgs;
+	int num_started = 0;
+	int loops = 0;
+
+	spin_lock(&cur_trans->dirty_bgs_lock);
+	if (list_empty(&cur_trans->dirty_bgs)) {
+		spin_unlock(&cur_trans->dirty_bgs_lock);
+		return 0;
+	}
+	list_splice_init(&cur_trans->dirty_bgs, &dirty);
+	spin_unlock(&cur_trans->dirty_bgs_lock);
+
+again:
+	/*
+	 * make sure all the block groups on our dirty list actually
+	 * exist
+	 */
+	btrfs_create_pending_block_groups(trans, root);
+
+	if (!path) {
+		path = btrfs_alloc_path();
+		if (!path)
+			return -ENOMEM;
+	}
+
+	/*
+	 * cache_write_mutex is here only to save us from balance or automatic
+	 * removal of empty block groups deleting this block group while we are
+	 * writing out the cache
+	 */
+	mutex_lock(&trans->transaction->cache_write_mutex);
+	while (!list_empty(&dirty)) {
+		cache = list_first_entry(&dirty,
+					 struct btrfs_block_group_cache,
+					 dirty_list);
+		/*
+		 * this can happen if something re-dirties a block
+		 * group that is already under IO.  Just wait for it to
+		 * finish and then do it all again
+		 */
+		if (!list_empty(&cache->io_list)) {
+			list_del_init(&cache->io_list);
+			btrfs_wait_cache_io(root, trans, cache,
+					    &cache->io_ctl, path,
+					    cache->key.objectid);
+			btrfs_put_block_group(cache);
+		}
+
+
+		/*
+		 * btrfs_wait_cache_io uses the cache->dirty_list to decide
+		 * if it should update the cache_state.  Don't delete
+		 * until after we wait.
+		 *
+		 * Since we're not running in the commit critical section
+		 * we need the dirty_bgs_lock to protect from update_block_group
+		 */
+		spin_lock(&cur_trans->dirty_bgs_lock);
+		list_del_init(&cache->dirty_list);
+		spin_unlock(&cur_trans->dirty_bgs_lock);
+
+		should_put = 1;
+
+		cache_save_setup(cache, trans, path);
+
+		if (cache->disk_cache_state == BTRFS_DC_SETUP) {
+			cache->io_ctl.inode = NULL;
+			ret = btrfs_write_out_cache(root, trans, cache, path);
+			if (ret == 0 && cache->io_ctl.inode) {
+				num_started++;
+				should_put = 0;
+
+				/*
+				 * the cache_write_mutex is protecting
+				 * the io_list
+				 */
+				list_add_tail(&cache->io_list, io);
+			} else {
+				/*
+				 * if we failed to write the cache, the
+				 * generation will be bad and life goes on
+				 */
+				ret = 0;
+			}
+		}
+		if (!ret)
+			ret = write_one_cache_group(trans, root, path, cache);
+
+		/* if its not on the io list, we need to put the block group */
+		if (should_put)
+			btrfs_put_block_group(cache);
+
+		if (ret)
+			break;
+
+		/*
+		 * Avoid blocking other tasks for too long. It might even save
+		 * us from writing caches for block groups that are going to be
+		 * removed.
+		 */
+		mutex_unlock(&trans->transaction->cache_write_mutex);
+		mutex_lock(&trans->transaction->cache_write_mutex);
+	}
+	mutex_unlock(&trans->transaction->cache_write_mutex);
+
+	/*
+	 * go through delayed refs for all the stuff we've just kicked off
+	 * and then loop back (just once)
+	 */
+	ret = btrfs_run_delayed_refs(trans, root, 0);
+	if (!ret && loops == 0) {
+		loops++;
+		spin_lock(&cur_trans->dirty_bgs_lock);
+		list_splice_init(&cur_trans->dirty_bgs, &dirty);
+		/*
+		 * dirty_bgs_lock protects us from concurrent block group
+		 * deletes too (not just cache_write_mutex).
+		 */
+		if (!list_empty(&dirty)) {
+			spin_unlock(&cur_trans->dirty_bgs_lock);
+			goto again;
+		}
+		spin_unlock(&cur_trans->dirty_bgs_lock);
+	}
+
+	btrfs_free_path(path);
+	return ret;
+}
+
 int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root)
 {
 	struct btrfs_block_group_cache *cache;
 	struct btrfs_transaction *cur_trans = trans->transaction;
 	int ret = 0;
+	int should_put;
 	struct btrfs_path *path;
-
-	if (list_empty(&cur_trans->dirty_bgs))
-		return 0;
+	struct list_head *io = &cur_trans->io_bgs;
+	int num_started = 0;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -3376,16 +3556,61 @@ int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
 		cache = list_first_entry(&cur_trans->dirty_bgs,
 					 struct btrfs_block_group_cache,
 					 dirty_list);
+
+		/*
+		 * this can happen if cache_save_setup re-dirties a block
+		 * group that is already under IO.  Just wait for it to
+		 * finish and then do it all again
+		 */
+		if (!list_empty(&cache->io_list)) {
+			list_del_init(&cache->io_list);
+			btrfs_wait_cache_io(root, trans, cache,
+					    &cache->io_ctl, path,
+					    cache->key.objectid);
+			btrfs_put_block_group(cache);
+		}
+
+		/*
+		 * don't remove from the dirty list until after we've waited
+		 * on any pending IO
+		 */
 		list_del_init(&cache->dirty_list);
-		if (cache->disk_cache_state == BTRFS_DC_CLEAR)
-			cache_save_setup(cache, trans, path);
+		should_put = 1;
+
+		cache_save_setup(cache, trans, path);
+
 		if (!ret)
-			ret = btrfs_run_delayed_refs(trans, root,
-						     (unsigned long) -1);
-		if (!ret && cache->disk_cache_state == BTRFS_DC_SETUP)
-			btrfs_write_out_cache(root, trans, cache, path);
+			ret = btrfs_run_delayed_refs(trans, root, (unsigned long) -1);
+
+		if (!ret && cache->disk_cache_state == BTRFS_DC_SETUP) {
+			cache->io_ctl.inode = NULL;
+			ret = btrfs_write_out_cache(root, trans, cache, path);
+			if (ret == 0 && cache->io_ctl.inode) {
+				num_started++;
+				should_put = 0;
+				list_add_tail(&cache->io_list, io);
+			} else {
+				/*
+				 * if we failed to write the cache, the
+				 * generation will be bad and life goes on
+				 */
+				ret = 0;
+			}
+		}
 		if (!ret)
 			ret = write_one_cache_group(trans, root, path, cache);
+
+		/* if its not on the io list, we need to put the block group */
+		if (should_put)
+			btrfs_put_block_group(cache);
+	}
+
+	while (!list_empty(io)) {
+		cache = list_first_entry(io, struct btrfs_block_group_cache,
+					 io_list);
+		list_del_init(&cache->io_list);
+		btrfs_wait_cache_io(root, trans, cache,
+				    &cache->io_ctl, path, cache->key.objectid);
 		btrfs_put_block_group(cache);
 	}
 
@@ -3635,19 +3860,21 @@ u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data)
  * This will check the space that the inode allocates from to make sure we have
  * enough space for bytes.
  */
-int btrfs_check_data_free_space(struct inode *inode, u64 bytes)
+int btrfs_check_data_free_space(struct inode *inode, u64 bytes, u64 write_bytes)
 {
 	struct btrfs_space_info *data_sinfo;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	u64 used;
-	int ret = 0, committed = 0, alloc_chunk = 1;
+	int ret = 0;
+	int need_commit = 2;
+	int have_pinned_space;
 
 	/* make sure bytes are sectorsize aligned */
 	bytes = ALIGN(bytes, root->sectorsize);
 
 	if (btrfs_is_free_space_inode(inode)) {
-		committed = 1;
+		need_commit = 0;
 		ASSERT(current->journal_info);
 	}
 
@@ -3669,7 +3896,7 @@ again:
 		 * if we don't have enough free bytes in this space then we need
 		 * to alloc a new chunk.
 		 */
-		if (!data_sinfo->full && alloc_chunk) {
+		if (!data_sinfo->full) {
 			u64 alloc_target;
 
 			data_sinfo->force_alloc = CHUNK_ALLOC_FORCE;
@@ -3697,8 +3924,10 @@ alloc:
 			if (ret < 0) {
 				if (ret != -ENOSPC)
 					return ret;
-				else
+				else {
+					have_pinned_space = 1;
 					goto commit_trans;
+				}
 			}
 
 			if (!data_sinfo)
@@ -3709,26 +3938,39 @@ alloc:
 
 		/*
 		 * If we don't have enough pinned space to deal with this
-		 * allocation don't bother committing the transaction.
+		 * allocation, and no removed chunk in current transaction,
+		 * don't bother committing the transaction.
 		 */
-		if (percpu_counter_compare(&data_sinfo->total_bytes_pinned,
-					   bytes) < 0)
-			committed = 1;
+		have_pinned_space = percpu_counter_compare(
+			&data_sinfo->total_bytes_pinned,
+			used + bytes - data_sinfo->total_bytes);
 		spin_unlock(&data_sinfo->lock);
 
 		/* commit the current transaction and try again */
 commit_trans:
-		if (!committed &&
+		if (need_commit &&
 		    !atomic_read(&root->fs_info->open_ioctl_trans)) {
-			committed = 1;
+			need_commit--;
 
 			trans = btrfs_join_transaction(root);
 			if (IS_ERR(trans))
 				return PTR_ERR(trans);
-			ret = btrfs_commit_transaction(trans, root);
-			if (ret)
-				return ret;
-			goto again;
+			if (have_pinned_space >= 0 ||
+			    trans->transaction->have_free_bgs ||
+			    need_commit > 0) {
+				ret = btrfs_commit_transaction(trans, root);
+				if (ret)
+					return ret;
+				/*
+				 * make sure that all running delayed iput are
+				 * done
+				 */
+				down_write(&root->fs_info->delayed_iput_sem);
+				up_write(&root->fs_info->delayed_iput_sem);
+				goto again;
+			} else {
+				btrfs_end_transaction(trans, root);
+			}
 		}
 
 		trace_btrfs_space_reservation(root->fs_info,
@@ -3736,12 +3978,16 @@ commit_trans:
 					      data_sinfo->flags, bytes, 1);
 		return -ENOSPC;
 	}
+	ret = btrfs_qgroup_reserve(root, write_bytes);
+	if (ret)
+		goto out;
 	data_sinfo->bytes_may_use += bytes;
 	trace_btrfs_space_reservation(root->fs_info, "space_info",
 				      data_sinfo->flags, bytes, 1);
+out:
 	spin_unlock(&data_sinfo->lock);
 
-	return 0;
+	return ret;
 }
 
 /*
@@ -4298,8 +4544,13 @@ out:
 static inline int need_do_async_reclaim(struct btrfs_space_info *space_info,
 					struct btrfs_fs_info *fs_info, u64 used)
 {
-	return (used >= div_factor_fine(space_info->total_bytes, 98) &&
-		!btrfs_fs_closing(fs_info) &&
+	u64 thresh = div_factor_fine(space_info->total_bytes, 98);
+
+	/* If we're just plain full then async reclaim just slows us down. */
+	if (space_info->bytes_used >= thresh)
+		return 0;
+
+	return (used >= thresh && !btrfs_fs_closing(fs_info) &&
 		!test_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state));
 }
 
@@ -4354,10 +4605,7 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work)
 		if (!btrfs_need_do_async_reclaim(space_info, fs_info,
 						 flush_state))
 			return;
-	} while (flush_state <= COMMIT_TRANS);
-
-	if (btrfs_need_do_async_reclaim(space_info, fs_info, flush_state))
-		queue_work(system_unbound_wq, work);
+	} while (flush_state < COMMIT_TRANS);
 }
 
 void btrfs_init_async_reclaim_work(struct work_struct *work)
@@ -4700,6 +4948,11 @@ void btrfs_free_block_rsv(struct btrfs_root *root,
 	kfree(rsv);
 }
 
+void __btrfs_free_block_rsv(struct btrfs_block_rsv *rsv)
+{
+	kfree(rsv);
+}
+
 int btrfs_block_rsv_add(struct btrfs_root *root,
 			struct btrfs_block_rsv *block_rsv, u64 num_bytes,
 			enum btrfs_reserve_flush_enum flush)
@@ -4812,10 +5065,10 @@ static u64 calc_global_metadata_size(struct btrfs_fs_info *fs_info)
 
 	num_bytes = (data_used >> fs_info->sb->s_blocksize_bits) *
 		    csum_size * 2;
-	num_bytes += div64_u64(data_used + meta_used, 50);
+	num_bytes += div_u64(data_used + meta_used, 50);
 
 	if (num_bytes * 3 > meta_used)
-		num_bytes = div64_u64(meta_used, 3);
+		num_bytes = div_u64(meta_used, 3);
 
 	return ALIGN(num_bytes, fs_info->extent_root->nodesize << 10);
 }
@@ -4998,8 +5251,6 @@ void btrfs_subvolume_release_metadata(struct btrfs_root *root,
 				      u64 qgroup_reserved)
 {
 	btrfs_block_rsv_release(root, rsv, (u64)-1);
-	if (qgroup_reserved)
-		btrfs_qgroup_free(root, qgroup_reserved);
 }
 
 /**
@@ -5066,30 +5317,18 @@ static u64 calc_csum_metadata_size(struct inode *inode, u64 num_bytes,
 				   int reserve)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	u64 csum_size;
-	int num_csums_per_leaf;
-	int num_csums;
-	int old_csums;
+	u64 old_csums, num_csums;
 
 	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM &&
 	    BTRFS_I(inode)->csum_bytes == 0)
 		return 0;
 
-	old_csums = (int)div64_u64(BTRFS_I(inode)->csum_bytes, root->sectorsize);
+	old_csums = btrfs_csum_bytes_to_leaves(root, BTRFS_I(inode)->csum_bytes);
 	if (reserve)
 		BTRFS_I(inode)->csum_bytes += num_bytes;
 	else
 		BTRFS_I(inode)->csum_bytes -= num_bytes;
-	csum_size = BTRFS_LEAF_DATA_SIZE(root) - sizeof(struct btrfs_item);
-	num_csums_per_leaf = (int)div64_u64(csum_size,
-					    sizeof(struct btrfs_csum_item) +
-					    sizeof(struct btrfs_disk_key));
-	num_csums = (int)div64_u64(BTRFS_I(inode)->csum_bytes, root->sectorsize);
-	num_csums = num_csums + num_csums_per_leaf - 1;
-	num_csums = num_csums / num_csums_per_leaf;
-
-	old_csums = old_csums + num_csums_per_leaf - 1;
-	old_csums = old_csums / num_csums_per_leaf;
+	num_csums = btrfs_csum_bytes_to_leaves(root, BTRFS_I(inode)->csum_bytes);
 
 	/* No change, no need to reserve more */
 	if (old_csums == num_csums)
@@ -5163,8 +5402,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	spin_unlock(&BTRFS_I(inode)->lock);
 
 	if (root->fs_info->quota_enabled) {
-		ret = btrfs_qgroup_reserve(root, num_bytes +
-					   nr_extents * root->nodesize);
+		ret = btrfs_qgroup_reserve(root, nr_extents * root->nodesize);
 		if (ret)
 			goto out_fail;
 	}
@@ -5172,8 +5410,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush);
 	if (unlikely(ret)) {
 		if (root->fs_info->quota_enabled)
-			btrfs_qgroup_free(root, num_bytes +
-						nr_extents * root->nodesize);
+			btrfs_qgroup_free(root, nr_extents * root->nodesize);
 		goto out_fail;
 	}
 
@@ -5290,10 +5527,6 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes)
 
 	trace_btrfs_space_reservation(root->fs_info, "delalloc",
 				      btrfs_ino(inode), to_free, 0);
-	if (root->fs_info->quota_enabled) {
-		btrfs_qgroup_free(root, num_bytes +
-					dropped * root->nodesize);
-	}
 
 	btrfs_block_rsv_release(root, &root->fs_info->delalloc_block_rsv,
 				to_free);
@@ -5318,7 +5551,7 @@ int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes)
 {
 	int ret;
 
-	ret = btrfs_check_data_free_space(inode, num_bytes);
+	ret = btrfs_check_data_free_space(inode, num_bytes, num_bytes);
 	if (ret)
 		return ret;
 
@@ -5390,14 +5623,6 @@ static int update_block_group(struct btrfs_trans_handle *trans,
 		if (!alloc && cache->cached == BTRFS_CACHE_NO)
 			cache_block_group(cache, 1);
 
-		spin_lock(&trans->transaction->dirty_bgs_lock);
-		if (list_empty(&cache->dirty_list)) {
-			list_add_tail(&cache->dirty_list,
-				      &trans->transaction->dirty_bgs);
-			btrfs_get_block_group(cache);
-		}
-		spin_unlock(&trans->transaction->dirty_bgs_lock);
-
 		byte_in_group = bytenr - cache->key.objectid;
 		WARN_ON(byte_in_group > cache->key.offset);
 
@@ -5446,6 +5671,16 @@ static int update_block_group(struct btrfs_trans_handle *trans,
 				spin_unlock(&info->unused_bgs_lock);
 			}
 		}
+
+		spin_lock(&trans->transaction->dirty_bgs_lock);
+		if (list_empty(&cache->dirty_list)) {
+			list_add_tail(&cache->dirty_list,
+				      &trans->transaction->dirty_bgs);
+				trans->transaction->num_dirty_bgs++;
+			btrfs_get_block_group(cache);
+		}
+		spin_unlock(&trans->transaction->dirty_bgs_lock);
+
 		btrfs_put_block_group(cache);
 		total -= num_bytes;
 		bytenr += num_bytes;
@@ -6956,15 +7191,15 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root,
 		return -ENOSPC;
 	}
 
-	if (btrfs_test_opt(root, DISCARD))
-		ret = btrfs_discard_extent(root, start, len, NULL);
-
 	if (pin)
 		pin_down_extent(root, cache, start, len, 1);
 	else {
+		if (btrfs_test_opt(root, DISCARD))
+			ret = btrfs_discard_extent(root, start, len, NULL);
 		btrfs_add_free_space(cache, start, len);
 		btrfs_update_reserved_bytes(cache, len, RESERVE_FREE, delalloc);
 	}
+
 	btrfs_put_block_group(cache);
 
 	trace_btrfs_reserved_extent_free(root, start, len);
@@ -7095,9 +7330,9 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 	ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
 				      ins, size);
 	if (ret) {
+		btrfs_free_path(path);
 		btrfs_free_and_pin_reserved_extent(root, ins->objectid,
 						   root->nodesize);
-		btrfs_free_path(path);
 		return ret;
 	}
 
@@ -7217,7 +7452,7 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	btrfs_set_header_generation(buf, trans->transid);
 	btrfs_set_buffer_lockdep_class(root->root_key.objectid, buf, level);
 	btrfs_tree_lock(buf);
-	clean_tree_block(trans, root, buf);
+	clean_tree_block(trans, root->fs_info, buf);
 	clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);
 
 	btrfs_set_lock_blocking(buf);
@@ -7311,7 +7546,7 @@ static void unuse_block_rsv(struct btrfs_fs_info *fs_info,
  * returns the key for the extent through ins, and a tree buffer for
  * the first block of the extent through buf.
  *
- * returns the tree buffer or NULL.
+ * returns the tree buffer or an ERR_PTR on error.
  */
 struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 					struct btrfs_root *root,
@@ -7322,6 +7557,7 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 	struct btrfs_key ins;
 	struct btrfs_block_rsv *block_rsv;
 	struct extent_buffer *buf;
+	struct btrfs_delayed_extent_op *extent_op;
 	u64 flags = 0;
 	int ret;
 	u32 blocksize = root->nodesize;
@@ -7342,13 +7578,14 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_reserve_extent(root, blocksize, blocksize,
 				   empty_size, hint, &ins, 0, 0);
-	if (ret) {
-		unuse_block_rsv(root->fs_info, block_rsv, blocksize);
-		return ERR_PTR(ret);
-	}
+	if (ret)
+		goto out_unuse;
 
 	buf = btrfs_init_new_buffer(trans, root, ins.objectid, level);
-	BUG_ON(IS_ERR(buf)); /* -ENOMEM */
+	if (IS_ERR(buf)) {
+		ret = PTR_ERR(buf);
+		goto out_free_reserved;
+	}
 
 	if (root_objectid == BTRFS_TREE_RELOC_OBJECTID) {
 		if (parent == 0)
@@ -7358,9 +7595,11 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 		BUG_ON(parent > 0);
 
 	if (root_objectid != BTRFS_TREE_LOG_OBJECTID) {
-		struct btrfs_delayed_extent_op *extent_op;
 		extent_op = btrfs_alloc_delayed_extent_op();
-		BUG_ON(!extent_op); /* -ENOMEM */
+		if (!extent_op) {
+			ret = -ENOMEM;
+			goto out_free_buf;
+		}
 		if (key)
 			memcpy(&extent_op->key, key, sizeof(extent_op->key));
 		else
@@ -7375,13 +7614,24 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 		extent_op->level = level;
 
 		ret = btrfs_add_delayed_tree_ref(root->fs_info, trans,
-					ins.objectid,
-					ins.offset, parent, root_objectid,
-					level, BTRFS_ADD_DELAYED_EXTENT,
-					extent_op, 0);
-		BUG_ON(ret); /* -ENOMEM */
+						 ins.objectid, ins.offset,
+						 parent, root_objectid, level,
+						 BTRFS_ADD_DELAYED_EXTENT,
+						 extent_op, 0);
+		if (ret)
+			goto out_free_delayed;
 	}
 	return buf;
+
+out_free_delayed:
+	btrfs_free_delayed_extent_op(extent_op);
+out_free_buf:
+	free_extent_buffer(buf);
+out_free_reserved:
+	btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 0);
+out_unuse:
+	unuse_block_rsv(root->fs_info, block_rsv, blocksize);
+	return ERR_PTR(ret);
 }
 
 struct walk_control {
@@ -7815,7 +8065,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	bytenr = btrfs_node_blockptr(path->nodes[level], path->slots[level]);
 	blocksize = root->nodesize;
 
-	next = btrfs_find_tree_block(root, bytenr);
+	next = btrfs_find_tree_block(root->fs_info, bytenr);
 	if (!next) {
 		next = btrfs_find_create_tree_block(root, bytenr);
 		if (!next)
@@ -8016,7 +8266,7 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
 			btrfs_set_lock_blocking(eb);
 			path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
 		}
-		clean_tree_block(trans, root, eb);
+		clean_tree_block(trans, root->fs_info, eb);
 	}
 
 	if (eb == root->node) {
@@ -8533,10 +8783,30 @@ int btrfs_set_block_group_ro(struct btrfs_root *root,
 
 	BUG_ON(cache->ro);
 
+again:
 	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
+	/*
+	 * we're not allowed to set block groups readonly after the dirty
+	 * block groups cache has started writing.  If it already started,
+	 * back off and let this transaction commit
+	 */
+	mutex_lock(&root->fs_info->ro_block_group_mutex);
+	if (trans->transaction->dirty_bg_run) {
+		u64 transid = trans->transid;
+
+		mutex_unlock(&root->fs_info->ro_block_group_mutex);
+		btrfs_end_transaction(trans, root);
+
+		ret = btrfs_wait_for_commit(root, transid);
+		if (ret)
+			return ret;
+		goto again;
+	}
+
+
 	ret = set_block_group_ro(cache, 0);
 	if (!ret)
 		goto out;
@@ -8551,6 +8821,7 @@ out:
 		alloc_flags = update_block_group_flags(root, cache->flags);
 		check_system_chunk(trans, root, alloc_flags);
 	}
+	mutex_unlock(&root->fs_info->ro_block_group_mutex);
 
 	btrfs_end_transaction(trans, root);
 	return ret;
@@ -8720,7 +8991,7 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr)
 		min_free <<= 1;
 	} else if (index == BTRFS_RAID_RAID0) {
 		dev_min = fs_devices->rw_devices;
-		do_div(min_free, dev_min);
+		min_free = div64_u64(min_free, dev_min);
 	}
 
 	/* We need to do this so that we can look at pending chunks */
@@ -8992,6 +9263,7 @@ btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size)
 	INIT_LIST_HEAD(&cache->bg_list);
 	INIT_LIST_HEAD(&cache->ro_list);
 	INIT_LIST_HEAD(&cache->dirty_list);
+	INIT_LIST_HEAD(&cache->io_list);
 	btrfs_init_free_space_ctl(cache);
 	atomic_set(&cache->trimming, 0);
 
@@ -9355,7 +9627,38 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 		goto out;
 	}
 
+	/*
+	 * get the inode first so any iput calls done for the io_list
+	 * aren't the final iput (no unlinks allowed now)
+	 */
 	inode = lookup_free_space_inode(tree_root, block_group, path);
+
+	mutex_lock(&trans->transaction->cache_write_mutex);
+	/*
+	 * make sure our free spache cache IO is done before remove the
+	 * free space inode
+	 */
+	spin_lock(&trans->transaction->dirty_bgs_lock);
+	if (!list_empty(&block_group->io_list)) {
+		list_del_init(&block_group->io_list);
+
+		WARN_ON(!IS_ERR(inode) && inode != block_group->io_ctl.inode);
+
+		spin_unlock(&trans->transaction->dirty_bgs_lock);
+		btrfs_wait_cache_io(root, trans, block_group,
+				    &block_group->io_ctl, path,
+				    block_group->key.objectid);
+		btrfs_put_block_group(block_group);
+		spin_lock(&trans->transaction->dirty_bgs_lock);
+	}
+
+	if (!list_empty(&block_group->dirty_list)) {
+		list_del_init(&block_group->dirty_list);
+		btrfs_put_block_group(block_group);
+	}
+	spin_unlock(&trans->transaction->dirty_bgs_lock);
+	mutex_unlock(&trans->transaction->cache_write_mutex);
+
 	if (!IS_ERR(inode)) {
 		ret = btrfs_orphan_add(trans, inode);
 		if (ret) {
@@ -9448,18 +9751,29 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 
 	spin_lock(&trans->transaction->dirty_bgs_lock);
 	if (!list_empty(&block_group->dirty_list)) {
-		list_del_init(&block_group->dirty_list);
-		btrfs_put_block_group(block_group);
+		WARN_ON(1);
+	}
+	if (!list_empty(&block_group->io_list)) {
+		WARN_ON(1);
 	}
 	spin_unlock(&trans->transaction->dirty_bgs_lock);
-
 	btrfs_remove_free_space_cache(block_group);
 
 	spin_lock(&block_group->space_info->lock);
 	list_del_init(&block_group->ro_list);
+
+	if (btrfs_test_opt(root, ENOSPC_DEBUG)) {
+		WARN_ON(block_group->space_info->total_bytes
+			< block_group->key.offset);
+		WARN_ON(block_group->space_info->bytes_readonly
+			< block_group->key.offset);
+		WARN_ON(block_group->space_info->disk_total
+			< block_group->key.offset * factor);
+	}
 	block_group->space_info->total_bytes -= block_group->key.offset;
 	block_group->space_info->bytes_readonly -= block_group->key.offset;
 	block_group->space_info->disk_total -= block_group->key.offset * factor;
+
 	spin_unlock(&block_group->space_info->lock);
 
 	memcpy(&key, &block_group->key, sizeof(key));
@@ -9647,8 +9961,18 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info)
 		mutex_unlock(&fs_info->unused_bg_unpin_mutex);
 
 		/* Reset pinned so btrfs_put_block_group doesn't complain */
+		spin_lock(&space_info->lock);
+		spin_lock(&block_group->lock);
+
+		space_info->bytes_pinned -= block_group->pinned;
+		space_info->bytes_readonly += block_group->pinned;
+		percpu_counter_add(&space_info->total_bytes_pinned,
+				   -block_group->pinned);
 		block_group->pinned = 0;
 
+		spin_unlock(&block_group->lock);
+		spin_unlock(&space_info->lock);
+
 		/*
 		 * Btrfs_remove_chunk will abort the transaction if things go
 		 * horribly wrong.