Merge tag 'v5.4.77' into toradex_5.4.y

This is the 5.4.77 stable release

45 files changed, 552 insertions, 199 deletions

diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index fe7f0bd2048e..ee9cabac1204 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -609,9 +609,9 @@ static void v9fs_mmap_vm_close(struct vm_area_struct *vma)
 	struct writeback_control wbc = {
 		.nr_to_write = LONG_MAX,
 		.sync_mode = WB_SYNC_ALL,
-		.range_start = vma->vm_pgoff * PAGE_SIZE,
+		.range_start = (loff_t)vma->vm_pgoff * PAGE_SIZE,
 		 /* absolute end, byte at end included */
-		.range_end = vma->vm_pgoff * PAGE_SIZE +
+		.range_end = (loff_t)vma->vm_pgoff * PAGE_SIZE +
 			(vma->vm_end - vma->vm_start - 1),
 	};
 
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index c05127f50637..e25133a9e9df 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -1103,6 +1103,8 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 
 	ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
 	if (ret) {
+		btrfs_tree_unlock(cow);
+		free_extent_buffer(cow);
 		btrfs_abort_transaction(trans, ret);
 		return ret;
 	}
@@ -1110,6 +1112,8 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 	if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) {
 		ret = btrfs_reloc_cow_block(trans, root, buf, cow);
 		if (ret) {
+			btrfs_tree_unlock(cow);
+			free_extent_buffer(cow);
 			btrfs_abort_transaction(trans, ret);
 			return ret;
 		}
@@ -1142,6 +1146,8 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 		if (last_ref) {
 			ret = tree_mod_log_free_eb(buf);
 			if (ret) {
+				btrfs_tree_unlock(cow);
+				free_extent_buffer(cow);
 				btrfs_abort_transaction(trans, ret);
 				return ret;
 			}
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 23b4f38e2392..27128164fac9 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -3404,6 +3404,8 @@ struct reada_control *btrfs_reada_add(struct btrfs_root *root,
 int btrfs_reada_wait(void *handle);
 void btrfs_reada_detach(void *handle);
 int btree_readahead_hook(struct extent_buffer *eb, int err);
+void btrfs_reada_remove_dev(struct btrfs_device *dev);
+void btrfs_reada_undo_remove_dev(struct btrfs_device *dev);
 
 static inline int is_fstree(u64 rootid)
 {
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index a34ee9c2f315..bef62b01824d 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -627,8 +627,7 @@ static int btrfs_delayed_inode_reserve_metadata(
 	 */
 	if (!src_rsv || (!trans->bytes_reserved &&
 			 src_rsv->type != BTRFS_BLOCK_RSV_DELALLOC)) {
-		ret = btrfs_qgroup_reserve_meta_prealloc(root,
-				fs_info->nodesize, true);
+		ret = btrfs_qgroup_reserve_meta_prealloc(root, num_bytes, true);
 		if (ret < 0)
 			return ret;
 		ret = btrfs_block_rsv_add(root, dst_rsv, num_bytes,
diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c
index 196bd241e701..96843934dcbb 100644
--- a/fs/btrfs/dev-replace.c
+++ b/fs/btrfs/dev-replace.c
@@ -190,7 +190,7 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	int ret = 0;
 
 	*device_out = NULL;
-	if (fs_info->fs_devices->seeding) {
+	if (srcdev->fs_devices->seeding) {
 		btrfs_err(fs_info, "the filesystem is a seed filesystem!");
 		return -EINVAL;
 	}
@@ -631,6 +631,9 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	}
 	btrfs_wait_ordered_roots(fs_info, U64_MAX, 0, (u64)-1);
 
+	if (!scrub_ret)
+		btrfs_reada_remove_dev(src_device);
+
 	/*
 	 * We have to use this loop approach because at this point src_device
 	 * has to be available for transaction commit to complete, yet new
@@ -639,6 +642,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	while (1) {
 		trans = btrfs_start_transaction(root, 0);
 		if (IS_ERR(trans)) {
+			btrfs_reada_undo_remove_dev(src_device);
 			mutex_unlock(&dev_replace->lock_finishing_cancel_unmount);
 			return PTR_ERR(trans);
 		}
@@ -689,6 +693,7 @@ error:
 		up_write(&dev_replace->rwsem);
 		mutex_unlock(&fs_info->chunk_mutex);
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
+		btrfs_reada_undo_remove_dev(src_device);
 		btrfs_rm_dev_replace_blocked(fs_info);
 		if (tgt_device)
 			btrfs_destroy_dev_replace_tgtdev(tgt_device);
diff --git a/fs/btrfs/reada.c b/fs/btrfs/reada.c
index 1feaeadc8cf5..2656dc8de99c 100644
--- a/fs/btrfs/reada.c
+++ b/fs/btrfs/reada.c
@@ -421,6 +421,9 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info,
 		if (!dev->bdev)
 			continue;
 
+		if (test_bit(BTRFS_DEV_STATE_NO_READA, &dev->dev_state))
+			continue;
+
 		if (dev_replace_is_ongoing &&
 		    dev == fs_info->dev_replace.tgtdev) {
 			/*
@@ -445,6 +448,8 @@ static struct reada_extent *reada_find_extent(struct btrfs_fs_info *fs_info,
 		}
 		have_zone = 1;
 	}
+	if (!have_zone)
+		radix_tree_delete(&fs_info->reada_tree, index);
 	spin_unlock(&fs_info->reada_lock);
 	up_read(&fs_info->dev_replace.rwsem);
 
@@ -1012,3 +1017,45 @@ void btrfs_reada_detach(void *handle)
 
 	kref_put(&rc->refcnt, reada_control_release);
 }
+
+/*
+ * Before removing a device (device replace or device remove ioctls), call this
+ * function to wait for all existing readahead requests on the device and to
+ * make sure no one queues more readahead requests for the device.
+ *
+ * Must be called without holding neither the device list mutex nor the device
+ * replace semaphore, otherwise it will deadlock.
+ */
+void btrfs_reada_remove_dev(struct btrfs_device *dev)
+{
+	struct btrfs_fs_info *fs_info = dev->fs_info;
+
+	/* Serialize with readahead extent creation at reada_find_extent(). */
+	spin_lock(&fs_info->reada_lock);
+	set_bit(BTRFS_DEV_STATE_NO_READA, &dev->dev_state);
+	spin_unlock(&fs_info->reada_lock);
+
+	/*
+	 * There might be readahead requests added to the radix trees which
+	 * were not yet added to the readahead work queue. We need to start
+	 * them and wait for their completion, otherwise we can end up with
+	 * use-after-free problems when dropping the last reference on the
+	 * readahead extents and their zones, as they need to access the
+	 * device structure.
+	 */
+	reada_start_machine(fs_info);
+	btrfs_flush_workqueue(fs_info->readahead_workers);
+}
+
+/*
+ * If when removing a device (device replace or device remove ioctls) an error
+ * happens after calling btrfs_reada_remove_dev(), call this to undo what that
+ * function did. This is safe to call even if btrfs_reada_remove_dev() was not
+ * called before.
+ */
+void btrfs_reada_undo_remove_dev(struct btrfs_device *dev)
+{
+	spin_lock(&dev->fs_info->reada_lock);
+	clear_bit(BTRFS_DEV_STATE_NO_READA, &dev->dev_state);
+	spin_unlock(&dev->fs_info->reada_lock);
+}
diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index b0e5dfb9be7a..88940f494428 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -3813,6 +3813,72 @@ static int update_ref_path(struct send_ctx *sctx, struct recorded_ref *ref)
 }
 
 /*
+ * When processing the new references for an inode we may orphanize an existing
+ * directory inode because its old name conflicts with one of the new references
+ * of the current inode. Later, when processing another new reference of our
+ * inode, we might need to orphanize another inode, but the path we have in the
+ * reference reflects the pre-orphanization name of the directory we previously
+ * orphanized. For example:
+ *
+ * parent snapshot looks like:
+ *
+ * .                                     (ino 256)
+ * |----- f1                             (ino 257)
+ * |----- f2                             (ino 258)
+ * |----- d1/                            (ino 259)
+ *        |----- d2/                     (ino 260)
+ *
+ * send snapshot looks like:
+ *
+ * .                                     (ino 256)
+ * |----- d1                             (ino 258)
+ * |----- f2/                            (ino 259)
+ *        |----- f2_link/                (ino 260)
+ *        |       |----- f1              (ino 257)
+ *        |
+ *        |----- d2                      (ino 258)
+ *
+ * When processing inode 257 we compute the name for inode 259 as "d1", and we
+ * cache it in the name cache. Later when we start processing inode 258, when
+ * collecting all its new references we set a full path of "d1/d2" for its new
+ * reference with name "d2". When we start processing the new references we
+ * start by processing the new reference with name "d1", and this results in
+ * orphanizing inode 259, since its old reference causes a conflict. Then we
+ * move on the next new reference, with name "d2", and we find out we must
+ * orphanize inode 260, as its old reference conflicts with ours - but for the
+ * orphanization we use a source path corresponding to the path we stored in the
+ * new reference, which is "d1/d2" and not "o259-6-0/d2" - this makes the
+ * receiver fail since the path component "d1/" no longer exists, it was renamed
+ * to "o259-6-0/" when processing the previous new reference. So in this case we
+ * must recompute the path in the new reference and use it for the new
+ * orphanization operation.
+ */
+static int refresh_ref_path(struct send_ctx *sctx, struct recorded_ref *ref)
+{
+	char *name;
+	int ret;
+
+	name = kmemdup(ref->name, ref->name_len, GFP_KERNEL);
+	if (!name)
+		return -ENOMEM;
+
+	fs_path_reset(ref->full_path);
+	ret = get_cur_path(sctx, ref->dir, ref->dir_gen, ref->full_path);
+	if (ret < 0)
+		goto out;
+
+	ret = fs_path_add(ref->full_path, name, ref->name_len);
+	if (ret < 0)
+		goto out;
+
+	/* Update the reference's base name pointer. */
+	set_ref_path(ref, ref->full_path);
+out:
+	kfree(name);
+	return ret;
+}
+
+/*
  * This does all the move/link/unlink/rmdir magic.
  */
 static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
@@ -3880,52 +3946,56 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 			goto out;
 	}
 
+	/*
+	 * Before doing any rename and link operations, do a first pass on the
+	 * new references to orphanize any unprocessed inodes that may have a
+	 * reference that conflicts with one of the new references of the current
+	 * inode. This needs to happen first because a new reference may conflict
+	 * with the old reference of a parent directory, so we must make sure
+	 * that the path used for link and rename commands don't use an
+	 * orphanized name when an ancestor was not yet orphanized.
+	 *
+	 * Example:
+	 *
+	 * Parent snapshot:
+	 *
+	 * .                                                      (ino 256)
+	 * |----- testdir/                                        (ino 259)
+	 * |          |----- a                                    (ino 257)
+	 * |
+	 * |----- b                                               (ino 258)
+	 *
+	 * Send snapshot:
+	 *
+	 * .                                                      (ino 256)
+	 * |----- testdir_2/                                      (ino 259)
+	 * |          |----- a                                    (ino 260)
+	 * |
+	 * |----- testdir                                         (ino 257)
+	 * |----- b                                               (ino 257)
+	 * |----- b2                                              (ino 258)
+	 *
+	 * Processing the new reference for inode 257 with name "b" may happen
+	 * before processing the new reference with name "testdir". If so, we
+	 * must make sure that by the time we send a link command to create the
+	 * hard link "b", inode 259 was already orphanized, since the generated
+	 * path in "valid_path" already contains the orphanized name for 259.
+	 * We are processing inode 257, so only later when processing 259 we do
+	 * the rename operation to change its temporary (orphanized) name to
+	 * "testdir_2".
+	 */
 	list_for_each_entry(cur, &sctx->new_refs, list) {
-		/*
-		 * We may have refs where the parent directory does not exist
-		 * yet. This happens if the parent directories inum is higher
-		 * than the current inum. To handle this case, we create the
-		 * parent directory out of order. But we need to check if this
-		 * did already happen before due to other refs in the same dir.
-		 */
 		ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
 		if (ret < 0)
 			goto out;
-		if (ret == inode_state_will_create) {
-			ret = 0;
-			/*
-			 * First check if any of the current inodes refs did
-			 * already create the dir.
-			 */
-			list_for_each_entry(cur2, &sctx->new_refs, list) {
-				if (cur == cur2)
-					break;
-				if (cur2->dir == cur->dir) {
-					ret = 1;
-					break;
-				}
-			}
-
-			/*
-			 * If that did not happen, check if a previous inode
-			 * did already create the dir.
-			 */
-			if (!ret)
-				ret = did_create_dir(sctx, cur->dir);
-			if (ret < 0)
-				goto out;
-			if (!ret) {
-				ret = send_create_inode(sctx, cur->dir);
-				if (ret < 0)
-					goto out;
-			}
-		}
+		if (ret == inode_state_will_create)
+			continue;
 
 		/*
-		 * Check if this new ref would overwrite the first ref of
-		 * another unprocessed inode. If yes, orphanize the
-		 * overwritten inode. If we find an overwritten ref that is
-		 * not the first ref, simply unlink it.
+		 * Check if this new ref would overwrite the first ref of another
+		 * unprocessed inode. If yes, orphanize the overwritten inode.
+		 * If we find an overwritten ref that is not the first ref,
+		 * simply unlink it.
 		 */
 		ret = will_overwrite_ref(sctx, cur->dir, cur->dir_gen,
 				cur->name, cur->name_len,
@@ -3942,6 +4012,12 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 				struct name_cache_entry *nce;
 				struct waiting_dir_move *wdm;
 
+				if (orphanized_dir) {
+					ret = refresh_ref_path(sctx, cur);
+					if (ret < 0)
+						goto out;
+				}
+
 				ret = orphanize_inode(sctx, ow_inode, ow_gen,
 						cur->full_path);
 				if (ret < 0)
@@ -4004,6 +4080,49 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 			}
 		}
 
+	}
+
+	list_for_each_entry(cur, &sctx->new_refs, list) {
+		/*
+		 * We may have refs where the parent directory does not exist
+		 * yet. This happens if the parent directories inum is higher
+		 * than the current inum. To handle this case, we create the
+		 * parent directory out of order. But we need to check if this
+		 * did already happen before due to other refs in the same dir.
+		 */
+		ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
+		if (ret < 0)
+			goto out;
+		if (ret == inode_state_will_create) {
+			ret = 0;
+			/*
+			 * First check if any of the current inodes refs did
+			 * already create the dir.
+			 */
+			list_for_each_entry(cur2, &sctx->new_refs, list) {
+				if (cur == cur2)
+					break;
+				if (cur2->dir == cur->dir) {
+					ret = 1;
+					break;
+				}
+			}
+
+			/*
+			 * If that did not happen, check if a previous inode
+			 * did already create the dir.
+			 */
+			if (!ret)
+				ret = did_create_dir(sctx, cur->dir);
+			if (ret < 0)
+				goto out;
+			if (!ret) {
+				ret = send_create_inode(sctx, cur->dir);
+				if (ret < 0)
+					goto out;
+			}
+		}
+
 		if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root) {
 			ret = wait_for_dest_dir_move(sctx, cur, is_orphan);
 			if (ret < 0)
@@ -7233,7 +7352,7 @@ long btrfs_ioctl_send(struct file *mnt_file, struct btrfs_ioctl_send_args *arg)
 
 	alloc_size = sizeof(struct clone_root) * (arg->clone_sources_count + 1);
 
-	sctx->clone_roots = kzalloc(alloc_size, GFP_KERNEL);
+	sctx->clone_roots = kvzalloc(alloc_size, GFP_KERNEL);
 	if (!sctx->clone_roots) {
 		ret = -ENOMEM;
 		goto out;
diff --git a/fs/btrfs/tree-checker.c b/fs/btrfs/tree-checker.c
index 84b8d6ebf98f..48e46323d519 100644
--- a/fs/btrfs/tree-checker.c
+++ b/fs/btrfs/tree-checker.c
@@ -577,18 +577,36 @@ int btrfs_check_chunk_valid(struct extent_buffer *leaf,
 	u64 type;
 	u64 features;
 	bool mixed = false;
+	int raid_index;
+	int nparity;
+	int ncopies;
 
 	length = btrfs_chunk_length(leaf, chunk);
 	stripe_len = btrfs_chunk_stripe_len(leaf, chunk);
 	num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
 	sub_stripes = btrfs_chunk_sub_stripes(leaf, chunk);
 	type = btrfs_chunk_type(leaf, chunk);
+	raid_index = btrfs_bg_flags_to_raid_index(type);
+	ncopies = btrfs_raid_array[raid_index].ncopies;
+	nparity = btrfs_raid_array[raid_index].nparity;
 
 	if (!num_stripes) {
 		chunk_err(leaf, chunk, logical,
 			  "invalid chunk num_stripes, have %u", num_stripes);
 		return -EUCLEAN;
 	}
+	if (num_stripes < ncopies) {
+		chunk_err(leaf, chunk, logical,
+			  "invalid chunk num_stripes < ncopies, have %u < %d",
+			  num_stripes, ncopies);
+		return -EUCLEAN;
+	}
+	if (nparity && num_stripes == nparity) {
+		chunk_err(leaf, chunk, logical,
+			  "invalid chunk num_stripes == nparity, have %u == %d",
+			  num_stripes, nparity);
+		return -EUCLEAN;
+	}
 	if (!IS_ALIGNED(logical, fs_info->sectorsize)) {
 		chunk_err(leaf, chunk, logical,
 		"invalid chunk logical, have %llu should aligned to %u",
@@ -869,7 +887,7 @@ static int check_root_item(struct extent_buffer *leaf, struct btrfs_key *key,
 			   int slot)
 {
 	struct btrfs_fs_info *fs_info = leaf->fs_info;
-	struct btrfs_root_item ri;
+	struct btrfs_root_item ri = { 0 };
 	const u64 valid_root_flags = BTRFS_ROOT_SUBVOL_RDONLY |
 				     BTRFS_ROOT_SUBVOL_DEAD;
 
@@ -889,14 +907,21 @@ static int check_root_item(struct extent_buffer *leaf, struct btrfs_key *key,
 		return -EUCLEAN;
 	}
 
-	if (btrfs_item_size_nr(leaf, slot) != sizeof(ri)) {
+	if (btrfs_item_size_nr(leaf, slot) != sizeof(ri) &&
+	    btrfs_item_size_nr(leaf, slot) != btrfs_legacy_root_item_size()) {
 		generic_err(leaf, slot,
-			    "invalid root item size, have %u expect %zu",
-			    btrfs_item_size_nr(leaf, slot), sizeof(ri));
+			    "invalid root item size, have %u expect %zu or %u",
+			    btrfs_item_size_nr(leaf, slot), sizeof(ri),
+			    btrfs_legacy_root_item_size());
 	}
 
+	/*
+	 * For legacy root item, the members starting at generation_v2 will be
+	 * all filled with 0.
+	 * And since we allow geneartion_v2 as 0, it will still pass the check.
+	 */
 	read_extent_buffer(leaf, &ri, btrfs_item_ptr_offset(leaf, slot),
-			   sizeof(ri));
+			   btrfs_item_size_nr(leaf, slot));
 
 	/* Generation related */
 	if (btrfs_root_generation(&ri) >
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 7042b84edc89..de53e5166997 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3639,6 +3639,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 	 * search and this search we'll not find the key again and can just
 	 * bail.
 	 */
+search:
 	ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
 	if (ret != 0)
 		goto done;
@@ -3658,6 +3659,13 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans,
 
 			if (min_key.objectid != ino || min_key.type != key_type)
 				goto done;
+
+			if (need_resched()) {
+				btrfs_release_path(path);
+				cond_resched();
+				goto search;
+			}
+
 			ret = overwrite_item(trans, log, dst_path, src, i,
 					     &min_key);
 			if (ret) {
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index e798caee978e..58910a0a3e4a 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1123,16 +1123,18 @@ static noinline struct btrfs_device *device_list_add(const char *path,
 				bdput(path_bdev);
 				mutex_unlock(&fs_devices->device_list_mutex);
 				btrfs_warn_in_rcu(device->fs_info,
-			"duplicate device fsid:devid for %pU:%llu old:%s new:%s",
-					disk_super->fsid, devid,
-					rcu_str_deref(device->name), path);
+	"duplicate device %s devid %llu generation %llu scanned by %s (%d)",
+						  path, devid, found_transid,
+						  current->comm,
+						  task_pid_nr(current));
 				return ERR_PTR(-EEXIST);
 			}
 			bdput(path_bdev);
 			btrfs_info_in_rcu(device->fs_info,
-				"device fsid %pU devid %llu moved old:%s new:%s",
-				disk_super->fsid, devid,
-				rcu_str_deref(device->name), path);
+	"devid %llu device path %s changed to %s scanned by %s (%d)",
+					  devid, rcu_str_deref(device->name),
+					  path, current->comm,
+					  task_pid_nr(current));
 		}
 
 		name = rcu_string_strdup(path, GFP_NOFS);
@@ -2206,6 +2208,8 @@ int btrfs_rm_device(struct btrfs_fs_info *fs_info, const char *device_path,
 
 	mutex_unlock(&uuid_mutex);
 	ret = btrfs_shrink_device(device, 0);
+	if (!ret)
+		btrfs_reada_remove_dev(device);
 	mutex_lock(&uuid_mutex);
 	if (ret)
 		goto error_undo;
@@ -2292,6 +2296,7 @@ out:
 	return ret;
 
 error_undo:
+	btrfs_reada_undo_remove_dev(device);
 	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state)) {
 		mutex_lock(&fs_info->chunk_mutex);
 		list_add(&device->dev_alloc_list,
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index 5acf5c507ec2..aa6a6d7b2978 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -56,6 +56,7 @@ struct btrfs_io_geometry {
 #define BTRFS_DEV_STATE_MISSING		(2)
 #define BTRFS_DEV_STATE_REPLACE_TGT	(3)
 #define BTRFS_DEV_STATE_FLUSH_SENT	(4)
+#define BTRFS_DEV_STATE_NO_READA	(5)
 
 struct btrfs_device {
 	struct list_head dev_list; /* device_list_mutex */
diff --git a/fs/buffer.c b/fs/buffer.c
index 22d8ac4a8c40..0d7bd7712076 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2739,16 +2739,6 @@ int nobh_writepage(struct page *page, get_block_t *get_block,
 	/* Is the page fully outside i_size? (truncate in progress) */
 	offset = i_size & (PAGE_SIZE-1);
 	if (page->index >= end_index+1 || !offset) {
-		/*
-		 * The page may have dirty, unmapped buffers.  For example,
-		 * they may have been added in ext3_writepage().  Make them
-		 * freeable here, so the page does not leak.
-		 */
-#if 0
-		/* Not really sure about this  - do we need this ? */
-		if (page->mapping->a_ops->invalidatepage)
-			page->mapping->a_ops->invalidatepage(page, offset);
-#endif
 		unlock_page(page);
 		return 0; /* don't care */
 	}
@@ -2943,12 +2933,6 @@ int block_write_full_page(struct page *page, get_block_t *get_block,
 	/* Is the page fully outside i_size? (truncate in progress) */
 	offset = i_size & (PAGE_SIZE-1);
 	if (page->index >= end_index+1 || !offset) {
-		/*
-		 * The page may have dirty, unmapped buffers.  For example,
-		 * they may have been added in ext3_writepage().  Make them
-		 * freeable here, so the page does not leak.
-		 */
-		do_invalidatepage(page, 0, PAGE_SIZE);
 		unlock_page(page);
 		return 0; /* don't care */
 	}
diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c
index ad057ed2b30b..bd5fe8d00d00 100644
--- a/fs/cachefiles/rdwr.c
+++ b/fs/cachefiles/rdwr.c
@@ -121,7 +121,7 @@ static int cachefiles_read_reissue(struct cachefiles_object *object,
 		_debug("reissue read");
 		ret = bmapping->a_ops->readpage(NULL, backpage);
 		if (ret < 0)
-			goto unlock_discard;
+			goto discard;
 	}
 
 	/* but the page may have been read before the monitor was installed, so
@@ -138,6 +138,7 @@ static int cachefiles_read_reissue(struct cachefiles_object *object,
 
 unlock_discard:
 	unlock_page(backpage);
+discard:
 	spin_lock_irq(&object->work_lock);
 	list_del(&monitor->op_link);
 	spin_unlock_irq(&object->work_lock);
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 7ab616601141..a02e845eb0fb 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1427,7 +1427,7 @@ static vm_fault_t ceph_filemap_fault(struct vm_fault *vmf)
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_file_info *fi = vma->vm_file->private_data;
 	struct page *pinned_page = NULL;
-	loff_t off = vmf->pgoff << PAGE_SHIFT;
+	loff_t off = (loff_t)vmf->pgoff << PAGE_SHIFT;
 	int want, got, err;
 	sigset_t oldset;
 	vm_fault_t ret = VM_FAULT_SIGBUS;
diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c
index 17df90b5f57a..fd9e289f3e72 100644
--- a/fs/cifs/inode.c
+++ b/fs/cifs/inode.c
@@ -2614,13 +2614,18 @@ cifs_setattr(struct dentry *direntry, struct iattr *attrs)
 {
 	struct cifs_sb_info *cifs_sb = CIFS_SB(direntry->d_sb);
 	struct cifs_tcon *pTcon = cifs_sb_master_tcon(cifs_sb);
+	int rc, retries = 0;
 
-	if (pTcon->unix_ext)
-		return cifs_setattr_unix(direntry, attrs);
-
-	return cifs_setattr_nounix(direntry, attrs);
+	do {
+		if (pTcon->unix_ext)
+			rc = cifs_setattr_unix(direntry, attrs);
+		else
+			rc = cifs_setattr_nounix(direntry, attrs);
+		retries++;
+	} while (is_retryable_error(rc) && retries < 2);
 
 	/* BB: add cifs_setattr_legacy for really old servers */
+	return rc;
 }
 
 #if 0
diff --git a/fs/exec.c b/fs/exec.c
index de833553ae27..2441eb1a1e2d 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -1044,11 +1044,24 @@ static int exec_mmap(struct mm_struct *mm)
 	}
 
 	task_lock(tsk);
-	active_mm = tsk->active_mm;
 	membarrier_exec_mmap(mm);
-	tsk->mm = mm;
+
+	local_irq_disable();
+	active_mm = tsk->active_mm;
 	tsk->active_mm = mm;
+	tsk->mm = mm;
+	/*
+	 * This prevents preemption while active_mm is being loaded and
+	 * it and mm are being updated, which could cause problems for
+	 * lazy tlb mm refcounting when these are updated by context
+	 * switches. Not all architectures can handle irqs off over
+	 * activate_mm yet.
+	 */
+	if (!IS_ENABLED(CONFIG_ARCH_WANT_IRQS_OFF_ACTIVATE_MM))
+		local_irq_enable();
 	activate_mm(active_mm, mm);
+	if (IS_ENABLED(CONFIG_ARCH_WANT_IRQS_OFF_ACTIVATE_MM))
+		local_irq_enable();
 	tsk->mm->vmacache_seqnum = 0;
 	vmacache_flush(tsk);
 	task_unlock(tsk);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 95a8a04c77dd..cbd028a31daf 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5271,6 +5271,12 @@ static int ext4_do_update_inode(handle_t *handle,
 	if (ext4_test_inode_state(inode, EXT4_STATE_NEW))
 		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
 
+	err = ext4_inode_blocks_set(handle, raw_inode, ei);
+	if (err) {
+		spin_unlock(&ei->i_raw_lock);
+		goto out_brelse;
+	}
+
 	raw_inode->i_mode = cpu_to_le16(inode->i_mode);
 	i_uid = i_uid_read(inode);
 	i_gid = i_gid_read(inode);
@@ -5304,11 +5310,6 @@ static int ext4_do_update_inode(handle_t *handle,
 	EXT4_INODE_SET_XTIME(i_atime, inode, raw_inode);
 	EXT4_EINODE_SET_XTIME(i_crtime, ei, raw_inode);
 
-	err = ext4_inode_blocks_set(handle, raw_inode, ei);
-	if (err) {
-		spin_unlock(&ei->i_raw_lock);
-		goto out_brelse;
-	}
 	raw_inode->i_dtime = cpu_to_le32(ei->i_dtime);
 	raw_inode->i_flags = cpu_to_le32(ei->i_flags & 0xFFFFFFFF);
 	if (likely(!test_opt2(inode->i_sb, HURD_COMPAT)))
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 080e25f6ef56..ad1d4c8faf44 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -861,8 +861,10 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 
 	BUFFER_TRACE(dind, "get_write_access");
 	err = ext4_journal_get_write_access(handle, dind);
-	if (unlikely(err))
+	if (unlikely(err)) {
 		ext4_std_error(sb, err);
+		goto errout;
+	}
 
 	/* ext4_reserve_inode_write() gets a reference on the iloc */
 	err = ext4_reserve_inode_write(handle, inode, &iloc);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 4aae7e3e89a1..6a260cc8bce6 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -4684,6 +4684,7 @@ cantfind_ext4:
 
 failed_mount8:
 	ext4_unregister_sysfs(sb);
+	kobject_put(&sbi->s_kobj);
 failed_mount7:
 	ext4_unregister_li_request(sb);
 failed_mount6:
@@ -5856,6 +5857,11 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id,
 	/* Quotafile not on the same filesystem? */
 	if (path->dentry->d_sb != sb)
 		return -EXDEV;
+
+	/* Quota already enabled for this file? */
+	if (IS_NOQUOTA(d_inode(path->dentry)))
+		return -EBUSY;
+
 	/* Journaling quota? */
 	if (EXT4_SB(sb)->s_qf_names[type]) {
 		/* Quotafile not in fs root? */
diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index bbd07fe8a492..c966ccc44c15 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -108,7 +108,7 @@ struct page *f2fs_get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index)
 	return __get_meta_page(sbi, index, true);
 }
 
-struct page *f2fs_get_meta_page_nofail(struct f2fs_sb_info *sbi, pgoff_t index)
+struct page *f2fs_get_meta_page_retry(struct f2fs_sb_info *sbi, pgoff_t index)
 {
 	struct page *page;
 	int count = 0;
@@ -243,6 +243,8 @@ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages,
 					blkno * NAT_ENTRY_PER_BLOCK);
 			break;
 		case META_SIT:
+			if (unlikely(blkno >= TOTAL_SEGS(sbi)))
+				goto out;
 			/* get sit block addr */
 			fio.new_blkaddr = current_sit_addr(sbi,
 					blkno * SIT_ENTRY_PER_BLOCK);
@@ -1044,8 +1046,12 @@ int f2fs_sync_dirty_inodes(struct f2fs_sb_info *sbi, enum inode_type type)
 				get_pages(sbi, is_dir ?
 				F2FS_DIRTY_DENTS : F2FS_DIRTY_DATA));
 retry:
-	if (unlikely(f2fs_cp_error(sbi)))
+	if (unlikely(f2fs_cp_error(sbi))) {
+		trace_f2fs_sync_dirty_inodes_exit(sbi->sb, is_dir,
+				get_pages(sbi, is_dir ?
+				F2FS_DIRTY_DENTS : F2FS_DIRTY_DATA));
 		return -EIO;
+	}
 
 	spin_lock(&sbi->inode_lock[type]);
 
diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
index e9af46dc06f7..78d041f9775a 100644
--- a/fs/f2fs/dir.c
+++ b/fs/f2fs/dir.c
@@ -303,16 +303,15 @@ struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir,
 	unsigned int max_depth;
 	unsigned int level;
 
+	*res_page = NULL;
+
 	if (f2fs_has_inline_dentry(dir)) {
-		*res_page = NULL;
 		de = f2fs_find_in_inline_dir(dir, fname, res_page);
 		goto out;
 	}
 
-	if (npages == 0) {
-		*res_page = NULL;
+	if (npages == 0)
 		goto out;
-	}
 
 	max_depth = F2FS_I(dir)->i_current_depth;
 	if (unlikely(max_depth > MAX_DIR_HASH_DEPTH)) {
@@ -323,7 +322,6 @@ struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir,
 	}
 
 	for (level = 0; level < max_depth; level++) {
-		*res_page = NULL;
 		de = find_in_level(dir, level, fname, res_page);
 		if (de || IS_ERR(*res_page))
 			break;
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index b3b7e63394be..63440abe58c4 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -3149,7 +3149,7 @@ enum rw_hint f2fs_io_type_to_rw_hint(struct f2fs_sb_info *sbi,
 void f2fs_stop_checkpoint(struct f2fs_sb_info *sbi, bool end_io);
 struct page *f2fs_grab_meta_page(struct f2fs_sb_info *sbi, pgoff_t index);
 struct page *f2fs_get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index);
-struct page *f2fs_get_meta_page_nofail(struct f2fs_sb_info *sbi, pgoff_t index);
+struct page *f2fs_get_meta_page_retry(struct f2fs_sb_info *sbi, pgoff_t index);
 struct page *f2fs_get_tmp_page(struct f2fs_sb_info *sbi, pgoff_t index);
 bool f2fs_is_valid_blkaddr(struct f2fs_sb_info *sbi,
 					block_t blkaddr, int type);
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index ed12e9668184..2a4a382f28fe 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -109,7 +109,7 @@ static void clear_node_page_dirty(struct page *page)
 
 static struct page *get_current_nat_page(struct f2fs_sb_info *sbi, nid_t nid)
 {
-	return f2fs_get_meta_page_nofail(sbi, current_nat_addr(sbi, nid));
+	return f2fs_get_meta_page(sbi, current_nat_addr(sbi, nid));
 }
 
 static struct page *get_next_nat_page(struct f2fs_sb_info *sbi, nid_t nid)
diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 7d8578401267..5ba677f85533 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -2310,7 +2310,9 @@ int f2fs_npages_for_summary_flush(struct f2fs_sb_info *sbi, bool for_ra)
  */
 struct page *f2fs_get_sum_page(struct f2fs_sb_info *sbi, unsigned int segno)
 {
-	return f2fs_get_meta_page_nofail(sbi, GET_SUM_BLOCK(sbi, segno));
+	if (unlikely(f2fs_cp_error(sbi)))
+		return ERR_PTR(-EIO);
+	return f2fs_get_meta_page_retry(sbi, GET_SUM_BLOCK(sbi, segno));
 }
 
 void f2fs_update_meta_page(struct f2fs_sb_info *sbi,
@@ -2582,7 +2584,11 @@ static void change_curseg(struct f2fs_sb_info *sbi, int type)
 	__next_free_blkoff(sbi, curseg, 0);
 
 	sum_page = f2fs_get_sum_page(sbi, new_segno);
-	f2fs_bug_on(sbi, IS_ERR(sum_page));
+	if (IS_ERR(sum_page)) {
+		/* GC won't be able to use stale summary pages by cp_error */
+		memset(curseg->sum_blk, 0, SUM_ENTRY_SIZE);
+		return;
+	}
 	sum_node = (struct f2fs_summary_block *)page_address(sum_page);
 	memcpy(curseg->sum_blk, sum_node, SUM_ENTRY_SIZE);
 	f2fs_put_page(sum_page, 1);
@@ -3713,7 +3719,7 @@ int f2fs_lookup_journal_in_cursum(struct f2fs_journal *journal, int type,
 static struct page *get_current_sit_page(struct f2fs_sb_info *sbi,
 					unsigned int segno)
 {
-	return f2fs_get_meta_page_nofail(sbi, current_sit_addr(sbi, segno));
+	return f2fs_get_meta_page(sbi, current_sit_addr(sbi, segno));
 }
 
 static struct page *get_next_sit_page(struct f2fs_sb_info *sbi,
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 0290a22ebccf..9e1685a30bf8 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -873,7 +873,8 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 out_free:
 	kfree(gl->gl_lksb.sb_lvbptr);
 	kmem_cache_free(cachep, gl);
-	atomic_dec(&sdp->sd_glock_disposal);
+	if (atomic_dec_and_test(&sdp->sd_glock_disposal))
+		wake_up(&sdp->sd_glock_wait);
 
 out:
 	return ret;
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 5f89c515f5bb..33a6b074209d 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -694,6 +694,7 @@ struct gfs2_sbd {
 	struct super_block *sd_vfs;
 	struct gfs2_pcpu_lkstats __percpu *sd_lkstats;
 	struct kobject sd_kobj;
+	struct completion sd_kobj_unregister;
 	unsigned long sd_flags;	/* SDF_... */
 	struct gfs2_sb_host sd_sb;
 
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index e0c55765b06d..29b27d769860 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -169,15 +169,19 @@ static int gfs2_check_sb(struct gfs2_sbd *sdp, int silent)
 		return -EINVAL;
 	}
 
-	/*  If format numbers match exactly, we're done.  */
-
-	if (sb->sb_fs_format == GFS2_FORMAT_FS &&
-	    sb->sb_multihost_format == GFS2_FORMAT_MULTI)
-		return 0;
+	if (sb->sb_fs_format != GFS2_FORMAT_FS ||
+	    sb->sb_multihost_format != GFS2_FORMAT_MULTI) {
+		fs_warn(sdp, "Unknown on-disk format, unable to mount\n");
+		return -EINVAL;
+	}
 
-	fs_warn(sdp, "Unknown on-disk format, unable to mount\n");
+	if (sb->sb_bsize < 512 || sb->sb_bsize > PAGE_SIZE ||
+	    (sb->sb_bsize & (sb->sb_bsize - 1))) {
+		pr_warn("Invalid superblock size\n");
+		return -EINVAL;
+	}
 
-	return -EINVAL;
+	return 0;
 }
 
 static void end_bio_io_page(struct bio *bio)
@@ -1094,26 +1098,14 @@ static int gfs2_fill_super(struct super_block *sb, struct fs_context *fc)
 	}
 
 	error = init_names(sdp, silent);
-	if (error) {
-		/* In this case, we haven't initialized sysfs, so we have to
-		   manually free the sdp. */
-		free_sbd(sdp);
-		sb->s_fs_info = NULL;
-		return error;
-	}
+	if (error)
+		goto fail_free;
 
 	snprintf(sdp->sd_fsname, sizeof(sdp->sd_fsname), "%s", sdp->sd_table_name);
 
 	error = gfs2_sys_fs_add(sdp);
-	/*
-	 * If we hit an error here, gfs2_sys_fs_add will have called function
-	 * kobject_put which causes the sysfs usage count to go to zero, which
-	 * causes sysfs to call function gfs2_sbd_release, which frees sdp.
-	 * Subsequent error paths here will call gfs2_sys_fs_del, which also
-	 * kobject_put to free sdp.
-	 */
 	if (error)
-		return error;
+		goto fail_free;
 
 	gfs2_create_debugfs_file(sdp);
 
@@ -1210,9 +1202,9 @@ fail_lm:
 	gfs2_lm_unmount(sdp);
 fail_debug:
 	gfs2_delete_debugfs_file(sdp);
-	/* gfs2_sys_fs_del must be the last thing we do, since it causes
-	 * sysfs to call function gfs2_sbd_release, which frees sdp. */
 	gfs2_sys_fs_del(sdp);
+fail_free:
+	free_sbd(sdp);
 	sb->s_fs_info = NULL;
 	return error;
 }
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index 5fa1eec4fb4f..5935ce5ae563 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -695,6 +695,7 @@ restart:
 
 	/*  At this point, we're through participating in the lockspace  */
 	gfs2_sys_fs_del(sdp);
+	free_sbd(sdp);
 }
 
 /**
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index dd15b8e4af2c..1c6e52dc878e 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -302,7 +302,7 @@ static void gfs2_sbd_release(struct kobject *kobj)
 {
 	struct gfs2_sbd *sdp = container_of(kobj, struct gfs2_sbd, sd_kobj);
 
-	free_sbd(sdp);
+	complete(&sdp->sd_kobj_unregister);
 }
 
 static struct kobj_type gfs2_ktype = {
@@ -652,6 +652,7 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 	sprintf(ro, "RDONLY=%d", sb_rdonly(sb));
 	sprintf(spectator, "SPECTATOR=%d", sdp->sd_args.ar_spectator ? 1 : 0);
 
+	init_completion(&sdp->sd_kobj_unregister);
 	sdp->sd_kobj.kset = gfs2_kset;
 	error = kobject_init_and_add(&sdp->sd_kobj, &gfs2_ktype, NULL,
 				     "%s", sdp->sd_table_name);
@@ -682,6 +683,7 @@ fail_tune:
 fail_reg:
 	fs_err(sdp, "error %d adding sysfs files\n", error);
 	kobject_put(&sdp->sd_kobj);
+	wait_for_completion(&sdp->sd_kobj_unregister);
 	sb->s_fs_info = NULL;
 	return error;
 }
@@ -692,6 +694,7 @@ void gfs2_sys_fs_del(struct gfs2_sbd *sdp)
 	sysfs_remove_group(&sdp->sd_kobj, &tune_group);
 	sysfs_remove_group(&sdp->sd_kobj, &lock_module_group);
 	kobject_put(&sdp->sd_kobj);
+	wait_for_completion(&sdp->sd_kobj_unregister);
 }
 
 static int gfs2_uevent(struct kset *kset, struct kobject *kobj,
diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index 9287eb666322..2db17fdf516b 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -31,9 +31,9 @@ int nfs_mountpoint_expiry_timeout = 500 * HZ;
 /*
  * nfs_path - reconstruct the path given an arbitrary dentry
  * @base - used to return pointer to the end of devname part of path
- * @dentry - pointer to dentry
+ * @dentry_in - pointer to dentry
  * @buffer - result buffer
- * @buflen - length of buffer
+ * @buflen_in - length of buffer
  * @flags - options (see below)
  *
  * Helper function for constructing the server pathname
@@ -48,15 +48,19 @@ int nfs_mountpoint_expiry_timeout = 500 * HZ;
  *		       the original device (export) name
  *		       (if unset, the original name is returned verbatim)
  */
-char *nfs_path(char **p, struct dentry *dentry, char *buffer, ssize_t buflen,
-	       unsigned flags)
+char *nfs_path(char **p, struct dentry *dentry_in, char *buffer,
+	       ssize_t buflen_in, unsigned flags)
 {
 	char *end;
 	int namelen;
 	unsigned seq;
 	const char *base;
+	struct dentry *dentry;
+	ssize_t buflen;
 
 rename_retry:
+	buflen = buflen_in;
+	dentry = dentry_in;
 	end = buffer+buflen;
 	*--end = '\0';
 	buflen--;
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index bb322d9de313..c4a98cbda6dd 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -570,6 +570,14 @@ static inline bool nfs4_stateid_is_newer(const nfs4_stateid *s1, const nfs4_stat
 	return (s32)(be32_to_cpu(s1->seqid) - be32_to_cpu(s2->seqid)) > 0;
 }
 
+static inline bool nfs4_stateid_is_next(const nfs4_stateid *s1, const nfs4_stateid *s2)
+{
+	u32 seq1 = be32_to_cpu(s1->seqid);
+	u32 seq2 = be32_to_cpu(s2->seqid);
+
+	return seq2 == seq1 + 1U || (seq2 == 1U && seq1 == 0xffffffffU);
+}
+
 static inline void nfs4_stateid_seqid_inc(nfs4_stateid *s1)
 {
 	u32 seqid = be32_to_cpu(s1->seqid);
diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c
index 534b6fd70ffd..6b31cb5f9c9d 100644
--- a/fs/nfs/nfs4file.c
+++ b/fs/nfs/nfs4file.c
@@ -138,7 +138,8 @@ static ssize_t __nfs4_copy_file_range(struct file *file_in, loff_t pos_in,
 	/* Only offload copy if superblock is the same */
 	if (file_inode(file_in)->i_sb != file_inode(file_out)->i_sb)
 		return -EXDEV;
-	if (!nfs_server_capable(file_inode(file_out), NFS_CAP_COPY))
+	if (!nfs_server_capable(file_inode(file_out), NFS_CAP_COPY) ||
+	    !nfs_server_capable(file_inode(file_in), NFS_CAP_COPY))
 		return -EOPNOTSUPP;
 	if (file_inode(file_in) == file_inode(file_out))
 		return -EOPNOTSUPP;
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 00435556db0c..ddc900df461c 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1515,19 +1515,6 @@ static void nfs_state_log_update_open_stateid(struct nfs4_state *state)
 		wake_up_all(&state->waitq);
 }
 
-static void nfs_state_log_out_of_order_open_stateid(struct nfs4_state *state,
-		const nfs4_stateid *stateid)
-{
-	u32 state_seqid = be32_to_cpu(state->open_stateid.seqid);
-	u32 stateid_seqid = be32_to_cpu(stateid->seqid);
-
-	if (stateid_seqid == state_seqid + 1U ||
-	    (stateid_seqid == 1U && state_seqid == 0xffffffffU))
-		nfs_state_log_update_open_stateid(state);
-	else
-		set_bit(NFS_STATE_CHANGE_WAIT, &state->flags);
-}
-
 static void nfs_test_and_clear_all_open_stateid(struct nfs4_state *state)
 {
 	struct nfs_client *clp = state->owner->so_server->nfs_client;
@@ -1553,21 +1540,19 @@ static void nfs_test_and_clear_all_open_stateid(struct nfs4_state *state)
  * i.e. The stateid seqids have to be initialised to 1, and
  * are then incremented on every state transition.
  */
-static bool nfs_need_update_open_stateid(struct nfs4_state *state,
+static bool nfs_stateid_is_sequential(struct nfs4_state *state,
 		const nfs4_stateid *stateid)
 {
-	if (test_bit(NFS_OPEN_STATE, &state->flags) == 0 ||
-	    !nfs4_stateid_match_other(stateid, &state->open_stateid)) {
+	if (test_bit(NFS_OPEN_STATE, &state->flags)) {
+		/* The common case - we're updating to a new sequence number */
+		if (nfs4_stateid_match_other(stateid, &state->open_stateid) &&
+			nfs4_stateid_is_next(&state->open_stateid, stateid)) {
+			return true;
+		}
+	} else {
+		/* This is the first OPEN in this generation */
 		if (stateid->seqid == cpu_to_be32(1))
-			nfs_state_log_update_open_stateid(state);
-		else
-			set_bit(NFS_STATE_CHANGE_WAIT, &state->flags);
-		return true;
-	}
-
-	if (nfs4_stateid_is_newer(stateid, &state->open_stateid)) {
-		nfs_state_log_out_of_order_open_stateid(state, stateid);
-		return true;
+			return true;
 	}
 	return false;
 }
@@ -1641,16 +1626,16 @@ static void nfs_set_open_stateid_locked(struct nfs4_state *state,
 	int status = 0;
 	for (;;) {
 
-		if (!nfs_need_update_open_stateid(state, stateid))
-			return;
-		if (!test_bit(NFS_STATE_CHANGE_WAIT, &state->flags))
+		if (nfs_stateid_is_sequential(state, stateid))
 			break;
+
 		if (status)
 			break;
 		/* Rely on seqids for serialisation with NFSv4.0 */
 		if (!nfs4_has_session(NFS_SERVER(state->inode)->nfs_client))
 			break;
 
+		set_bit(NFS_STATE_CHANGE_WAIT, &state->flags);
 		prepare_to_wait(&state->waitq, &wait, TASK_KILLABLE);
 		/*
 		 * Ensure we process the state changes in the same order
@@ -1661,6 +1646,7 @@ static void nfs_set_open_stateid_locked(struct nfs4_state *state,
 		spin_unlock(&state->owner->so_lock);
 		rcu_read_unlock();
 		trace_nfs4_open_stateid_update_wait(state->inode, stateid, 0);
+
 		if (!signal_pending(current)) {
 			if (schedule_timeout(5*HZ) == 0)
 				status = -EAGAIN;
@@ -3397,7 +3383,8 @@ static bool nfs4_refresh_open_old_stateid(nfs4_stateid *dst,
 	__be32 seqid_open;
 	u32 dst_seqid;
 	bool ret;
-	int seq;
+	int seq, status = -EAGAIN;
+	DEFINE_WAIT(wait);
 
 	for (;;) {
 		ret = false;
@@ -3409,15 +3396,41 @@ static bool nfs4_refresh_open_old_stateid(nfs4_stateid *dst,
 				continue;
 			break;
 		}
+
+		write_seqlock(&state->seqlock);
 		seqid_open = state->open_stateid.seqid;
-		if (read_seqretry(&state->seqlock, seq))
-			continue;
 
 		dst_seqid = be32_to_cpu(dst->seqid);
-		if ((s32)(dst_seqid - be32_to_cpu(seqid_open)) >= 0)
-			dst->seqid = cpu_to_be32(dst_seqid + 1);
-		else
+
+		/* Did another OPEN bump the state's seqid?  try again: */
+		if ((s32)(be32_to_cpu(seqid_open) - dst_seqid) > 0) {
 			dst->seqid = seqid_open;
+			write_sequnlock(&state->seqlock);
+			ret = true;
+			break;
+		}
+
+		/* server says we're behind but we haven't seen the update yet */
+		set_bit(NFS_STATE_CHANGE_WAIT, &state->flags);
+		prepare_to_wait(&state->waitq, &wait, TASK_KILLABLE);
+		write_sequnlock(&state->seqlock);
+		trace_nfs4_close_stateid_update_wait(state->inode, dst, 0);
+
+		if (signal_pending(current))
+			status = -EINTR;
+		else
+			if (schedule_timeout(5*HZ) != 0)
+				status = 0;
+
+		finish_wait(&state->waitq, &wait);
+
+		if (!status)
+			continue;
+		if (status == -EINTR)
+			break;
+
+		/* we slept the whole 5 seconds, we must have lost a seqid */
+		dst->seqid = cpu_to_be32(dst_seqid + 1);
 		ret = true;
 		break;
 	}
@@ -7846,9 +7859,11 @@ int nfs4_proc_secinfo(struct inode *dir, const struct qstr *name,
  * both PNFS and NON_PNFS flags set, and not having one of NON_PNFS, PNFS, or
  * DS flags set.
  */
-static int nfs4_check_cl_exchange_flags(u32 flags)
+static int nfs4_check_cl_exchange_flags(u32 flags, u32 version)
 {
-	if (flags & ~EXCHGID4_FLAG_MASK_R)
+	if (version >= 2 && (flags & ~EXCHGID4_2_FLAG_MASK_R))
+		goto out_inval;
+	else if (version < 2 && (flags & ~EXCHGID4_FLAG_MASK_R))
 		goto out_inval;
 	if ((flags & EXCHGID4_FLAG_USE_PNFS_MDS) &&
 	    (flags & EXCHGID4_FLAG_USE_NON_PNFS))
@@ -8261,7 +8276,8 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, const struct cred *cre
 	if (status  != 0)
 		goto out;
 
-	status = nfs4_check_cl_exchange_flags(resp->flags);
+	status = nfs4_check_cl_exchange_flags(resp->flags,
+			clp->cl_mvops->minor_version);
 	if (status  != 0)
 		goto out;
 
diff --git a/fs/nfs/nfs4trace.h b/fs/nfs/nfs4trace.h
index 9398c0b6e0a3..2295a934a154 100644
--- a/fs/nfs/nfs4trace.h
+++ b/fs/nfs/nfs4trace.h
@@ -1291,6 +1291,7 @@ DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_setattr);
 DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_delegreturn);
 DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_open_stateid_update);
 DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_open_stateid_update_wait);
+DEFINE_NFS4_INODE_STATEID_EVENT(nfs4_close_stateid_update_wait);
 
 DECLARE_EVENT_CLASS(nfs4_getattr_event,
 		TP_PROTO(
diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c
index c83ddac22f38..754c763374dd 100644
--- a/fs/nfsd/nfsproc.c
+++ b/fs/nfsd/nfsproc.c
@@ -118,6 +118,13 @@ done:
 	return nfsd_return_attrs(nfserr, resp);
 }
 
+/* Obsolete, replaced by MNTPROC_MNT. */
+static __be32
+nfsd_proc_root(struct svc_rqst *rqstp)
+{
+	return nfs_ok;
+}
+
 /*
  * Look up a path name component
  * Note: the dentry in the resp->fh may be negative if the file
@@ -203,6 +210,13 @@ nfsd_proc_read(struct svc_rqst *rqstp)
 	return fh_getattr(&resp->fh, &resp->stat);
 }
 
+/* Reserved */
+static __be32
+nfsd_proc_writecache(struct svc_rqst *rqstp)
+{
+	return nfs_ok;
+}
+
 /*
  * Write data to a file
  * N.B. After this call resp->fh needs an fh_put
@@ -617,6 +631,7 @@ static const struct svc_procedure nfsd_procedures2[18] = {
 		.pc_xdrressize = ST+AT,
 	},
 	[NFSPROC_ROOT] = {
+		.pc_func = nfsd_proc_root,
 		.pc_decode = nfssvc_decode_void,
 		.pc_encode = nfssvc_encode_void,
 		.pc_argsize = sizeof(struct nfsd_void),
@@ -654,6 +669,7 @@ static const struct svc_procedure nfsd_procedures2[18] = {
 		.pc_xdrressize = ST+AT+1+NFSSVC_MAXBLKSIZE_V2/4,
 	},
 	[NFSPROC_WRITECACHE] = {
+		.pc_func = nfsd_proc_writecache,
 		.pc_decode = nfssvc_decode_void,
 		.pc_encode = nfssvc_encode_void,
 		.pc_argsize = sizeof(struct nfsd_void),
diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c
index e4b52783819d..992b74f9c941 100644
--- a/fs/ubifs/debug.c
+++ b/fs/ubifs/debug.c
@@ -1123,6 +1123,7 @@ int dbg_check_dir(struct ubifs_info *c, const struct inode *dir)
 			err = PTR_ERR(dent);
 			if (err == -ENOENT)
 				break;
+			kfree(pdent);
 			return err;
 		}
 
diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c
index 5f2ac5ef0891..f78c3e3ef931 100644
--- a/fs/ubifs/journal.c
+++ b/fs/ubifs/journal.c
@@ -894,6 +894,7 @@ int ubifs_jnl_write_inode(struct ubifs_info *c, const struct inode *inode)
 				if (err == -ENOENT)
 					break;
 
+				kfree(pxent);
 				goto out_release;
 			}
 
@@ -906,6 +907,7 @@ int ubifs_jnl_write_inode(struct ubifs_info *c, const struct inode *inode)
 				ubifs_err(c, "dead directory entry '%s', error %d",
 					  xent->name, err);
 				ubifs_ro_mode(c, err);
+				kfree(pxent);
 				kfree(xent);
 				goto out_release;
 			}
@@ -936,8 +938,6 @@ int ubifs_jnl_write_inode(struct ubifs_info *c, const struct inode *inode)
 					  inode->i_ino);
 	release_head(c, BASEHD);
 
-	ubifs_add_auth_dirt(c, lnum);
-
 	if (last_reference) {
 		err = ubifs_tnc_remove_ino(c, inode->i_ino);
 		if (err)
@@ -947,6 +947,8 @@ int ubifs_jnl_write_inode(struct ubifs_info *c, const struct inode *inode)
 	} else {
 		union ubifs_key key;
 
+		ubifs_add_auth_dirt(c, lnum);
+
 		ino_key_init(c, &key, inode->i_ino);
 		err = ubifs_tnc_add(c, &key, lnum, offs, ilen, hash);
 	}
diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c
index 283f9eb48410..b0117878b3a0 100644
--- a/fs/ubifs/orphan.c
+++ b/fs/ubifs/orphan.c
@@ -173,6 +173,7 @@ int ubifs_add_orphan(struct ubifs_info *c, ino_t inum)
 			err = PTR_ERR(xent);
 			if (err == -ENOENT)
 				break;
+			kfree(pxent);
 			return err;
 		}
 
@@ -182,6 +183,7 @@ int ubifs_add_orphan(struct ubifs_info *c, ino_t inum)
 
 		xattr_orphan = orphan_add(c, xattr_inum, orphan);
 		if (IS_ERR(xattr_orphan)) {
+			kfree(pxent);
 			kfree(xent);
 			return PTR_ERR(xattr_orphan);
 		}
diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index 7fc2f3f07c16..e49bd69dfc1c 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -1092,14 +1092,20 @@ static int ubifs_parse_options(struct ubifs_info *c, char *options,
 			break;
 		}
 		case Opt_auth_key:
-			c->auth_key_name = kstrdup(args[0].from, GFP_KERNEL);
-			if (!c->auth_key_name)
-				return -ENOMEM;
+			if (!is_remount) {
+				c->auth_key_name = kstrdup(args[0].from,
+								GFP_KERNEL);
+				if (!c->auth_key_name)
+					return -ENOMEM;
+			}
 			break;
 		case Opt_auth_hash_name:
-			c->auth_hash_name = kstrdup(args[0].from, GFP_KERNEL);
-			if (!c->auth_hash_name)
-				return -ENOMEM;
+			if (!is_remount) {
+				c->auth_hash_name = kstrdup(args[0].from,
+								GFP_KERNEL);
+				if (!c->auth_hash_name)
+					return -ENOMEM;
+			}
 			break;
 		case Opt_ignore:
 			break;
@@ -1123,6 +1129,18 @@ static int ubifs_parse_options(struct ubifs_info *c, char *options,
 	return 0;
 }
 
+/*
+ * ubifs_release_options - release mount parameters which have been dumped.
+ * @c: UBIFS file-system description object
+ */
+static void ubifs_release_options(struct ubifs_info *c)
+{
+	kfree(c->auth_key_name);
+	c->auth_key_name = NULL;
+	kfree(c->auth_hash_name);
+	c->auth_hash_name = NULL;
+}
+
 /**
  * destroy_journal - destroy journal data structures.
  * @c: UBIFS file-system description object
@@ -1295,7 +1313,7 @@ static int mount_ubifs(struct ubifs_info *c)
 
 	err = ubifs_read_superblock(c);
 	if (err)
-		goto out_free;
+		goto out_auth;
 
 	c->probing = 0;
 
@@ -1307,18 +1325,18 @@ static int mount_ubifs(struct ubifs_info *c)
 		ubifs_err(c, "'compressor \"%s\" is not compiled in",
 			  ubifs_compr_name(c, c->default_compr));
 		err = -ENOTSUPP;
-		goto out_free;
+		goto out_auth;
 	}
 
 	err = init_constants_sb(c);
 	if (err)
-		goto out_free;
+		goto out_auth;
 
 	sz = ALIGN(c->max_idx_node_sz, c->min_io_size) * 2;
 	c->cbuf = kmalloc(sz, GFP_NOFS);
 	if (!c->cbuf) {
 		err = -ENOMEM;
-		goto out_free;
+		goto out_auth;
 	}
 
 	err = alloc_wbufs(c);
@@ -1593,6 +1611,8 @@ out_wbufs:
 	free_wbufs(c);
 out_cbuf:
 	kfree(c->cbuf);
+out_auth:
+	ubifs_exit_authentication(c);
 out_free:
 	kfree(c->write_reserve_buf);
 	kfree(c->bu.buf);
@@ -1632,8 +1652,7 @@ static void ubifs_umount(struct ubifs_info *c)
 	ubifs_lpt_free(c, 0);
 	ubifs_exit_authentication(c);
 
-	kfree(c->auth_key_name);
-	kfree(c->auth_hash_name);
+	ubifs_release_options(c);
 	kfree(c->cbuf);
 	kfree(c->rcvrd_mst_node);
 	kfree(c->mst_node);
@@ -2201,6 +2220,7 @@ out_umount:
 out_unlock:
 	mutex_unlock(&c->umount_mutex);
 out_close:
+	ubifs_release_options(c);
 	ubi_close_volume(c->ubi);
 out:
 	return err;
diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c
index e8e7b0e9532e..33742ee3945b 100644
--- a/fs/ubifs/tnc.c
+++ b/fs/ubifs/tnc.c
@@ -2885,6 +2885,7 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum)
 			err = PTR_ERR(xent);
 			if (err == -ENOENT)
 				break;
+			kfree(pxent);
 			return err;
 		}
 
@@ -2898,6 +2899,7 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum)
 		fname_len(&nm) = le16_to_cpu(xent->nlen);
 		err = ubifs_tnc_remove_nm(c, &key1, &nm);
 		if (err) {
+			kfree(pxent);
 			kfree(xent);
 			return err;
 		}
@@ -2906,6 +2908,7 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum)
 		highest_ino_key(c, &key2, xattr_inum);
 		err = ubifs_tnc_remove_range(c, &key1, &key2);
 		if (err) {
+			kfree(pxent);
 			kfree(xent);
 			return err;
 		}
diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c
index 9aefbb60074f..a0b9b349efe6 100644
--- a/fs/ubifs/xattr.c
+++ b/fs/ubifs/xattr.c
@@ -522,6 +522,7 @@ int ubifs_purge_xattrs(struct inode *host)
 				  xent->name, err);
 			ubifs_ro_mode(c, err);
 			kfree(pxent);
+			kfree(xent);
 			return err;
 		}
 
@@ -531,6 +532,7 @@ int ubifs_purge_xattrs(struct inode *host)
 		err = remove_xattr(c, host, xino, &nm);
 		if (err) {
 			kfree(pxent);
+			kfree(xent);
 			iput(xino);
 			ubifs_err(c, "cannot remove xattr, error %d", err);
 			return err;
diff --git a/fs/udf/super.c b/fs/udf/super.c
index a0cd766b41cd..4aba4878ed96 100644
--- a/fs/udf/super.c
+++ b/fs/udf/super.c
@@ -1703,7 +1703,8 @@ static noinline int udf_process_sequence(
 					"Pointers (max %u supported)\n",
 					UDF_MAX_TD_NESTING);
 				brelse(bh);
-				return -EIO;
+				ret = -EIO;
+				goto out;
 			}
 
 			vdp = (struct volDescPtr *)bh->b_data;
@@ -1723,7 +1724,8 @@ static noinline int udf_process_sequence(
 			curr = get_volume_descriptor_record(ident, bh, &data);
 			if (IS_ERR(curr)) {
 				brelse(bh);
-				return PTR_ERR(curr);
+				ret = PTR_ERR(curr);
+				goto out;
 			}
 			/* Descriptor we don't care about? */
 			if (!curr)
@@ -1745,28 +1747,31 @@ static noinline int udf_process_sequence(
 	 */
 	if (!data.vds[VDS_POS_PRIMARY_VOL_DESC].block) {
 		udf_err(sb, "Primary Volume Descriptor not found!\n");
-		return -EAGAIN;
+		ret = -EAGAIN;
+		goto out;
 	}
 	ret = udf_load_pvoldesc(sb, data.vds[VDS_POS_PRIMARY_VOL_DESC].block);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	if (data.vds[VDS_POS_LOGICAL_VOL_DESC].block) {
 		ret = udf_load_logicalvol(sb,
 				data.vds[VDS_POS_LOGICAL_VOL_DESC].block,
 				fileset);
 		if (ret < 0)
-			return ret;
+			goto out;
 	}
 
 	/* Now handle prevailing Partition Descriptors */
 	for (i = 0; i < data.num_part_descs; i++) {
 		ret = udf_load_partdesc(sb, data.part_descs_loc[i].rec.block);
 		if (ret < 0)
-			return ret;
+			goto out;
 	}
-
-	return 0;
+	ret = 0;
+out:
+	kfree(data.part_descs_loc);
+	return ret;
 }
 
 /*
diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index f8db3fe616df..c114d24be619 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -4985,20 +4985,25 @@ xfs_bmap_del_extent_real(
 
 	flags = XFS_ILOG_CORE;
 	if (whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip)) {
-		xfs_fsblock_t	bno;
 		xfs_filblks_t	len;
 		xfs_extlen_t	mod;
 
-		bno = div_u64_rem(del->br_startblock, mp->m_sb.sb_rextsize,
-				  &mod);
-		ASSERT(mod == 0);
 		len = div_u64_rem(del->br_blockcount, mp->m_sb.sb_rextsize,
 				  &mod);
 		ASSERT(mod == 0);
 
-		error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
-		if (error)
-			goto done;
+		if (!(bflags & XFS_BMAPI_REMAP)) {
+			xfs_fsblock_t	bno;
+
+			bno = div_u64_rem(del->br_startblock,
+					mp->m_sb.sb_rextsize, &mod);
+			ASSERT(mod == 0);
+
+			error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
+			if (error)
+				goto done;
+		}
+
 		do_fx = 0;
 		nblks = len * mp->m_sb.sb_rextsize;
 		qfield = XFS_TRANS_DQ_RTBCOUNT;
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index bf0435dbec43..b3021d9b34a5 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -622,7 +622,6 @@ xfs_ioc_space(
 	error = xfs_break_layouts(inode, &iolock, BREAK_UNMAP);
 	if (error)
 		goto out_unlock;
-	inode_dio_wait(inode);
 
 	switch (bf->l_whence) {
 	case 0: /*SEEK_SET*/
@@ -668,6 +667,31 @@ xfs_ioc_space(
 		goto out_unlock;
 	}
 
+	/*
+	 * Must wait for all AIO to complete before we continue as AIO can
+	 * change the file size on completion without holding any locks we
+	 * currently hold. We must do this first because AIO can update both
+	 * the on disk and in memory inode sizes, and the operations that follow
+	 * require the in-memory size to be fully up-to-date.
+	 */
+	inode_dio_wait(inode);
+
+	/*
+	 * Now that AIO and DIO has drained we can flush and (if necessary)
+	 * invalidate the cached range over the first operation we are about to
+	 * run. We include zero range here because it starts with a hole punch
+	 * over the target range.
+	 */
+	switch (cmd) {
+	case XFS_IOC_ZERO_RANGE:
+	case XFS_IOC_UNRESVSP:
+	case XFS_IOC_UNRESVSP64:
+		error = xfs_flush_unmap_range(ip, bf->l_start, bf->l_len);
+		if (error)
+			goto out_unlock;
+		break;
+	}
+
 	switch (cmd) {
 	case XFS_IOC_ZERO_RANGE:
 		flags |= XFS_PREALLOC_SET;
diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c
index b58366937082..6d5ddc4e5135 100644
--- a/fs/xfs/xfs_rtalloc.c
+++ b/fs/xfs/xfs_rtalloc.c
@@ -1021,10 +1021,13 @@ xfs_growfs_rt(
 		xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
 		xfs_trans_ijoin(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
 		/*
-		 * Update the bitmap inode's size.
+		 * Update the bitmap inode's size ondisk and incore.  We need
+		 * to update the incore size so that inode inactivation won't
+		 * punch what it thinks are "posteof" blocks.
 		 */
 		mp->m_rbmip->i_d.di_size =
 			nsbp->sb_rbmblocks * nsbp->sb_blocksize;
+		i_size_write(VFS_I(mp->m_rbmip), mp->m_rbmip->i_d.di_size);
 		xfs_trans_log_inode(tp, mp->m_rbmip, XFS_ILOG_CORE);
 		/*
 		 * Get the summary inode into the transaction.
@@ -1032,9 +1035,12 @@ xfs_growfs_rt(
 		xfs_ilock(mp->m_rsumip, XFS_ILOCK_EXCL);
 		xfs_trans_ijoin(tp, mp->m_rsumip, XFS_ILOCK_EXCL);
 		/*
-		 * Update the summary inode's size.
+		 * Update the summary inode's size.  We need to update the
+		 * incore size so that inode inactivation won't punch what it
+		 * thinks are "posteof" blocks.
 		 */
 		mp->m_rsumip->i_d.di_size = nmp->m_rsumsize;
+		i_size_write(VFS_I(mp->m_rsumip), mp->m_rsumip->i_d.di_size);
 		xfs_trans_log_inode(tp, mp->m_rsumip, XFS_ILOG_CORE);
 		/*
 		 * Copy summary data from old to new sizes.