Merge branch 'perf/urgent' into perf/core, before applying dependent patches

Signed-off-by: Ingo Molnar <mingo@kernel.org>

27 files changed, 427 insertions, 132 deletions

diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index 9de772e..614aaa1 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -880,6 +880,8 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
  * indirect refs to their parent bytenr.
  * When roots are found, they're added to the roots list
  *
+ * NOTE: This can return values > 0
+ *
  * FIXME some caching might speed things up
  */
 static int find_parent_nodes(struct btrfs_trans_handle *trans,
@@ -1198,6 +1200,19 @@ int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+/**
+ * btrfs_check_shared - tell us whether an extent is shared
+ *
+ * @trans: optional trans handle
+ *
+ * btrfs_check_shared uses the backref walking code but will short
+ * circuit as soon as it finds a root or inode that doesn't match the
+ * one passed in. This provides a significant performance benefit for
+ * callers (such as fiemap) which want to know whether the extent is
+ * shared but do not need a ref count.
+ *
+ * Return: 0 if extent is not shared, 1 if it is shared, < 0 on error.
+ */
 int btrfs_check_shared(struct btrfs_trans_handle *trans,
 		       struct btrfs_fs_info *fs_info, u64 root_objectid,
 		       u64 inum, u64 bytenr)
@@ -1226,11 +1241,13 @@ int btrfs_check_shared(struct btrfs_trans_handle *trans,
 		ret = find_parent_nodes(trans, fs_info, bytenr, elem.seq, tmp,
 					roots, NULL, root_objectid, inum);
 		if (ret == BACKREF_FOUND_SHARED) {
+			/* this is the only condition under which we return 1 */
 			ret = 1;
 			break;
 		}
 		if (ret < 0 && ret != -ENOENT)
 			break;
+		ret = 0;
 		node = ulist_next(tmp, &uiter);
 		if (!node)
 			break;
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 0ec8e22..0ec3acd 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3180,8 +3180,6 @@ static int write_one_cache_group(struct btrfs_trans_handle *trans,
 	btrfs_mark_buffer_dirty(leaf);
 fail:
 	btrfs_release_path(path);
-	if (ret)
-		btrfs_abort_transaction(trans, root, ret);
 	return ret;
 
 }
@@ -3487,8 +3485,30 @@ again:
 				ret = 0;
 			}
 		}
-		if (!ret)
+		if (!ret) {
 			ret = write_one_cache_group(trans, root, path, cache);
+			/*
+			 * Our block group might still be attached to the list
+			 * of new block groups in the transaction handle of some
+			 * other task (struct btrfs_trans_handle->new_bgs). This
+			 * means its block group item isn't yet in the extent
+			 * tree. If this happens ignore the error, as we will
+			 * try again later in the critical section of the
+			 * transaction commit.
+			 */
+			if (ret == -ENOENT) {
+				ret = 0;
+				spin_lock(&cur_trans->dirty_bgs_lock);
+				if (list_empty(&cache->dirty_list)) {
+					list_add_tail(&cache->dirty_list,
+						      &cur_trans->dirty_bgs);
+					btrfs_get_block_group(cache);
+				}
+				spin_unlock(&cur_trans->dirty_bgs_lock);
+			} else if (ret) {
+				btrfs_abort_transaction(trans, root, ret);
+			}
+		}
 
 		/* if its not on the io list, we need to put the block group */
 		if (should_put)
@@ -3597,8 +3617,11 @@ int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
 				ret = 0;
 			}
 		}
-		if (!ret)
+		if (!ret) {
 			ret = write_one_cache_group(trans, root, path, cache);
+			if (ret)
+				btrfs_abort_transaction(trans, root, ret);
+		}
 
 		/* if its not on the io list, we need to put the block group */
 		if (should_put)
@@ -8806,6 +8829,24 @@ again:
 		goto again;
 	}
 
+	/*
+	 * if we are changing raid levels, try to allocate a corresponding
+	 * block group with the new raid level.
+	 */
+	alloc_flags = update_block_group_flags(root, cache->flags);
+	if (alloc_flags != cache->flags) {
+		ret = do_chunk_alloc(trans, root, alloc_flags,
+				     CHUNK_ALLOC_FORCE);
+		/*
+		 * ENOSPC is allowed here, we may have enough space
+		 * already allocated at the new raid level to
+		 * carry on
+		 */
+		if (ret == -ENOSPC)
+			ret = 0;
+		if (ret < 0)
+			goto out;
+	}
 
 	ret = set_block_group_ro(cache, 0);
 	if (!ret)
@@ -8819,7 +8860,9 @@ again:
 out:
 	if (cache->flags & BTRFS_BLOCK_GROUP_SYSTEM) {
 		alloc_flags = update_block_group_flags(root, cache->flags);
+		lock_chunks(root->fs_info->chunk_root);
 		check_system_chunk(trans, root, alloc_flags);
+		unlock_chunks(root->fs_info->chunk_root);
 	}
 	mutex_unlock(&root->fs_info->ro_block_group_mutex);
 
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 43af5a6..c32d226 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -4772,6 +4772,25 @@ struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info,
 			       start >> PAGE_CACHE_SHIFT);
 	if (eb && atomic_inc_not_zero(&eb->refs)) {
 		rcu_read_unlock();
+		/*
+		 * Lock our eb's refs_lock to avoid races with
+		 * free_extent_buffer. When we get our eb it might be flagged
+		 * with EXTENT_BUFFER_STALE and another task running
+		 * free_extent_buffer might have seen that flag set,
+		 * eb->refs == 2, that the buffer isn't under IO (dirty and
+		 * writeback flags not set) and it's still in the tree (flag
+		 * EXTENT_BUFFER_TREE_REF set), therefore being in the process
+		 * of decrementing the extent buffer's reference count twice.
+		 * So here we could race and increment the eb's reference count,
+		 * clear its stale flag, mark it as dirty and drop our reference
+		 * before the other task finishes executing free_extent_buffer,
+		 * which would later result in an attempt to free an extent
+		 * buffer that is dirty.
+		 */
+		if (test_bit(EXTENT_BUFFER_STALE, &eb->bflags)) {
+			spin_lock(&eb->refs_lock);
+			spin_unlock(&eb->refs_lock);
+		}
 		mark_extent_buffer_accessed(eb, NULL);
 		return eb;
 	}
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 41c510b..9dbe5b5 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -86,7 +86,7 @@ static struct inode *__lookup_free_space_inode(struct btrfs_root *root,
 
 	mapping_set_gfp_mask(inode->i_mapping,
 			mapping_gfp_mask(inode->i_mapping) &
-			~(GFP_NOFS & ~__GFP_HIGHMEM));
+			~(__GFP_FS | __GFP_HIGHMEM));
 
 	return inode;
 }
@@ -3466,6 +3466,7 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 	struct btrfs_free_space_ctl *ctl = root->free_ino_ctl;
 	int ret;
 	struct btrfs_io_ctl io_ctl;
+	bool release_metadata = true;
 
 	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
 		return 0;
@@ -3473,11 +3474,20 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 	memset(&io_ctl, 0, sizeof(io_ctl));
 	ret = __btrfs_write_out_cache(root, inode, ctl, NULL, &io_ctl,
 				      trans, path, 0);
-	if (!ret)
+	if (!ret) {
+		/*
+		 * At this point writepages() didn't error out, so our metadata
+		 * reservation is released when the writeback finishes, at
+		 * inode.c:btrfs_finish_ordered_io(), regardless of it finishing
+		 * with or without an error.
+		 */
+		release_metadata = false;
 		ret = btrfs_wait_cache_io(root, trans, NULL, &io_ctl, path, 0);
+	}
 
 	if (ret) {
-		btrfs_delalloc_release_metadata(inode, inode->i_size);
+		if (release_metadata)
+			btrfs_delalloc_release_metadata(inode, inode->i_size);
 #ifdef DEBUG
 		btrfs_err(root->fs_info,
 			"failed to write free ino cache for root %llu",
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index 157cc54..760c4a5 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -722,6 +722,7 @@ void btrfs_start_ordered_extent(struct inode *inode,
 int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 {
 	int ret = 0;
+	int ret_wb = 0;
 	u64 end;
 	u64 orig_end;
 	struct btrfs_ordered_extent *ordered;
@@ -741,9 +742,14 @@ int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 	if (ret)
 		return ret;
 
-	ret = filemap_fdatawait_range(inode->i_mapping, start, orig_end);
-	if (ret)
-		return ret;
+	/*
+	 * If we have a writeback error don't return immediately. Wait first
+	 * for any ordered extents that haven't completed yet. This is to make
+	 * sure no one can dirty the same page ranges and call writepages()
+	 * before the ordered extents complete - to avoid failures (-EEXIST)
+	 * when adding the new ordered extents to the ordered tree.
+	 */
+	ret_wb = filemap_fdatawait_range(inode->i_mapping, start, orig_end);
 
 	end = orig_end;
 	while (1) {
@@ -767,7 +773,7 @@ int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 			break;
 		end--;
 	}
-	return ret;
+	return ret_wb ? ret_wb : ret;
 }
 
 /*
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 96aebf3..174f5e1 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -4625,6 +4625,7 @@ int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 {
 	u64 chunk_offset;
 
+	ASSERT(mutex_is_locked(&extent_root->fs_info->chunk_mutex));
 	chunk_offset = find_next_chunk(extent_root->fs_info);
 	return __btrfs_alloc_chunk(trans, extent_root, chunk_offset, type);
 }
diff --git a/fs/exec.c b/fs/exec.c
index 49a1c61..1977c2a 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -659,6 +659,9 @@ int setup_arg_pages(struct linux_binprm *bprm,
 	if (stack_base > STACK_SIZE_MAX)
 		stack_base = STACK_SIZE_MAX;
 
+	/* Add space for stack randomization. */
+	stack_base += (STACK_RND_MASK << PAGE_SHIFT);
+
 	/* Make sure we didn't let the argument array grow too large. */
 	if (vma->vm_end - vma->vm_start > stack_base)
 		return -ENOMEM;
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 009a059..9a83f14 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2889,7 +2889,6 @@ extern int ext4_map_blocks(handle_t *handle, struct inode *inode,
 			   struct ext4_map_blocks *map, int flags);
 extern int ext4_ext_calc_metadata_amount(struct inode *inode,
 					 ext4_lblk_t lblocks);
-extern int ext4_extent_tree_init(handle_t *, struct inode *);
 extern int ext4_ext_calc_credits_for_single_extent(struct inode *inode,
 						   int num,
 						   struct ext4_ext_path *path);
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 3445035..d418431 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -87,6 +87,12 @@ int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle)
 		ext4_put_nojournal(handle);
 		return 0;
 	}
+
+	if (!handle->h_transaction) {
+		err = jbd2_journal_stop(handle);
+		return handle->h_err ? handle->h_err : err;
+	}
+
 	sb = handle->h_transaction->t_journal->j_private;
 	err = handle->h_err;
 	rc = jbd2_journal_stop(handle);
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index d74e0802..e003a1e 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -377,7 +377,7 @@ static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
 	ext4_lblk_t lblock = le32_to_cpu(ext->ee_block);
 	ext4_lblk_t last = lblock + len - 1;
 
-	if (lblock > last)
+	if (len == 0 || lblock > last)
 		return 0;
 	return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
 }
@@ -5396,6 +5396,14 @@ int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len)
 	loff_t new_size, ioffset;
 	int ret;
 
+	/*
+	 * We need to test this early because xfstests assumes that a
+	 * collapse range of (0, 1) will return EOPNOTSUPP if the file
+	 * system does not support collapse range.
+	 */
+	if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
+		return -EOPNOTSUPP;
+
 	/* Collapse range works only on fs block size aligned offsets. */
 	if (offset & (EXT4_CLUSTER_SIZE(sb) - 1) ||
 	    len & (EXT4_CLUSTER_SIZE(sb) - 1))
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 55b187c..0554b0b 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -4345,7 +4345,7 @@ static void ext4_update_other_inodes_time(struct super_block *sb,
 	int inode_size = EXT4_INODE_SIZE(sb);
 
 	oi.orig_ino = orig_ino;
-	ino = orig_ino & ~(inodes_per_block - 1);
+	ino = (orig_ino & ~(inodes_per_block - 1)) + 1;
 	for (i = 0; i < inodes_per_block; i++, ino++, buf += inode_size) {
 		if (ino == orig_ino)
 			continue;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index f06d058..ca9d4a2 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -294,6 +294,8 @@ static void __save_error_info(struct super_block *sb, const char *func,
 	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
 
 	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
+	if (bdev_read_only(sb->s_bdev))
+		return;
 	es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
 	es->s_last_error_time = cpu_to_le32(get_seconds());
 	strncpy(es->s_last_error_func, func, sizeof(es->s_last_error_func));
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index ef26317..07d8d8f 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -581,7 +581,7 @@ static int hostfs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
 	if (name == NULL)
 		goto out_put;
 
-	fd = file_create(name, mode & S_IFMT);
+	fd = file_create(name, mode & 0777);
 	if (fd < 0)
 		error = fd;
 	else
diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c
index b5128c6..a9079d0 100644
--- a/fs/jbd2/recovery.c
+++ b/fs/jbd2/recovery.c
@@ -842,15 +842,23 @@ static int scan_revoke_records(journal_t *journal, struct buffer_head *bh,
 {
 	jbd2_journal_revoke_header_t *header;
 	int offset, max;
+	int csum_size = 0;
+	__u32 rcount;
 	int record_len = 4;
 
 	header = (jbd2_journal_revoke_header_t *) bh->b_data;
 	offset = sizeof(jbd2_journal_revoke_header_t);
-	max = be32_to_cpu(header->r_count);
+	rcount = be32_to_cpu(header->r_count);
 
 	if (!jbd2_revoke_block_csum_verify(journal, header))
 		return -EINVAL;
 
+	if (jbd2_journal_has_csum_v2or3(journal))
+		csum_size = sizeof(struct jbd2_journal_revoke_tail);
+	if (rcount > journal->j_blocksize - csum_size)
+		return -EINVAL;
+	max = rcount;
+
 	if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT))
 		record_len = 8;
 
diff --git a/fs/jbd2/revoke.c b/fs/jbd2/revoke.c
index c6cbaef..14214da 100644
--- a/fs/jbd2/revoke.c
+++ b/fs/jbd2/revoke.c
@@ -577,7 +577,7 @@ static void write_one_revoke_record(journal_t *journal,
 {
 	int csum_size = 0;
 	struct buffer_head *descriptor;
-	int offset;
+	int sz, offset;
 	journal_header_t *header;
 
 	/* If we are already aborting, this all becomes a noop.  We
@@ -594,9 +594,14 @@ static void write_one_revoke_record(journal_t *journal,
 	if (jbd2_journal_has_csum_v2or3(journal))
 		csum_size = sizeof(struct jbd2_journal_revoke_tail);
 
+	if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT))
+		sz = 8;
+	else
+		sz = 4;
+
 	/* Make sure we have a descriptor with space left for the record */
 	if (descriptor) {
-		if (offset >= journal->j_blocksize - csum_size) {
+		if (offset + sz > journal->j_blocksize - csum_size) {
 			flush_descriptor(journal, descriptor, offset, write_op);
 			descriptor = NULL;
 		}
@@ -619,16 +624,13 @@ static void write_one_revoke_record(journal_t *journal,
 		*descriptorp = descriptor;
 	}
 
-	if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT)) {
+	if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT))
 		* ((__be64 *)(&descriptor->b_data[offset])) =
 			cpu_to_be64(record->blocknr);
-		offset += 8;
-
-	} else {
+	else
 		* ((__be32 *)(&descriptor->b_data[offset])) =
 			cpu_to_be32(record->blocknr);
-		offset += 4;
-	}
+	offset += sz;
 
 	*offsetp = offset;
 }
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index 5f09370..ff2f2e6 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -551,7 +551,6 @@ int jbd2_journal_extend(handle_t *handle, int nblocks)
 	int result;
 	int wanted;
 
-	WARN_ON(!transaction);
 	if (is_handle_aborted(handle))
 		return -EROFS;
 	journal = transaction->t_journal;
@@ -627,7 +626,6 @@ int jbd2__journal_restart(handle_t *handle, int nblocks, gfp_t gfp_mask)
 	tid_t		tid;
 	int		need_to_start, ret;
 
-	WARN_ON(!transaction);
 	/* If we've had an abort of any type, don't even think about
 	 * actually doing the restart! */
 	if (is_handle_aborted(handle))
@@ -785,7 +783,6 @@ do_get_write_access(handle_t *handle, struct journal_head *jh,
 	int need_copy = 0;
 	unsigned long start_lock, time_lock;
 
-	WARN_ON(!transaction);
 	if (is_handle_aborted(handle))
 		return -EROFS;
 	journal = transaction->t_journal;
@@ -1051,7 +1048,6 @@ int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh)
 	int err;
 
 	jbd_debug(5, "journal_head %p\n", jh);
-	WARN_ON(!transaction);
 	err = -EROFS;
 	if (is_handle_aborted(handle))
 		goto out;
@@ -1266,7 +1262,6 @@ int jbd2_journal_dirty_metadata(handle_t *handle, struct buffer_head *bh)
 	struct journal_head *jh;
 	int ret = 0;
 
-	WARN_ON(!transaction);
 	if (is_handle_aborted(handle))
 		return -EROFS;
 	journal = transaction->t_journal;
@@ -1397,7 +1392,6 @@ int jbd2_journal_forget (handle_t *handle, struct buffer_head *bh)
 	int err = 0;
 	int was_modified = 0;
 
-	WARN_ON(!transaction);
 	if (is_handle_aborted(handle))
 		return -EROFS;
 	journal = transaction->t_journal;
@@ -1530,8 +1524,22 @@ int jbd2_journal_stop(handle_t *handle)
 	tid_t tid;
 	pid_t pid;
 
-	if (!transaction)
-		goto free_and_exit;
+	if (!transaction) {
+		/*
+		 * Handle is already detached from the transaction so
+		 * there is nothing to do other than decrease a refcount,
+		 * or free the handle if refcount drops to zero
+		 */
+		if (--handle->h_ref > 0) {
+			jbd_debug(4, "h_ref %d -> %d\n", handle->h_ref + 1,
+							 handle->h_ref);
+			return err;
+		} else {
+			if (handle->h_rsv_handle)
+				jbd2_free_handle(handle->h_rsv_handle);
+			goto free_and_exit;
+		}
+	}
 	journal = transaction->t_journal;
 
 	J_ASSERT(journal_current_handle() == handle);
@@ -2373,7 +2381,6 @@ int jbd2_journal_file_inode(handle_t *handle, struct jbd2_inode *jinode)
 	transaction_t *transaction = handle->h_transaction;
 	journal_t *journal;
 
-	WARN_ON(!transaction);
 	if (is_handle_aborted(handle))
 		return -EROFS;
 	journal = transaction->t_journal;
diff --git a/fs/kernfs/dir.c b/fs/kernfs/dir.c
index f131fc2..fffca95 100644
--- a/fs/kernfs/dir.c
+++ b/fs/kernfs/dir.c
@@ -518,7 +518,14 @@ static struct kernfs_node *__kernfs_new_node(struct kernfs_root *root,
 	if (!kn)
 		goto err_out1;
 
-	ret = ida_simple_get(&root->ino_ida, 1, 0, GFP_KERNEL);
+	/*
+	 * If the ino of the sysfs entry created for a kmem cache gets
+	 * allocated from an ida layer, which is accounted to the memcg that
+	 * owns the cache, the memcg will get pinned forever. So do not account
+	 * ino ida allocations.
+	 */
+	ret = ida_simple_get(&root->ino_ida, 1, 0,
+			     GFP_KERNEL | __GFP_NOACCOUNT);
 	if (ret < 0)
 		goto err_out2;
 	kn->ino = ret;
diff --git a/fs/namei.c b/fs/namei.c
index 4a8d998b..fe30d3b 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1415,6 +1415,7 @@ static int lookup_fast(struct nameidata *nd,
 	 */
 	if (nd->flags & LOOKUP_RCU) {
 		unsigned seq;
+		bool negative;
 		dentry = __d_lookup_rcu(parent, &nd->last, &seq);
 		if (!dentry)
 			goto unlazy;
@@ -1424,8 +1425,11 @@ static int lookup_fast(struct nameidata *nd,
 		 * the dentry name information from lookup.
 		 */
 		*inode = dentry->d_inode;
+		negative = d_is_negative(dentry);
 		if (read_seqcount_retry(&dentry->d_seq, seq))
 			return -ECHILD;
+		if (negative)
+			return -ENOENT;
 
 		/*
 		 * This sequence count validates that the parent had no
@@ -1472,6 +1476,10 @@ unlazy:
 		goto need_lookup;
 	}
 
+	if (unlikely(d_is_negative(dentry))) {
+		dput(dentry);
+		return -ENOENT;
+	}
 	path->mnt = mnt;
 	path->dentry = dentry;
 	err = follow_managed(path, nd->flags);
@@ -1583,10 +1591,10 @@ static inline int walk_component(struct nameidata *nd, struct path *path,
 			goto out_err;
 
 		inode = path->dentry->d_inode;
+		err = -ENOENT;
+		if (d_is_negative(path->dentry))
+			goto out_path_put;
 	}
-	err = -ENOENT;
-	if (d_is_negative(path->dentry))
-		goto out_path_put;
 
 	if (should_follow_link(path->dentry, follow)) {
 		if (nd->flags & LOOKUP_RCU) {
@@ -3036,14 +3044,13 @@ retry_lookup:
 
 	BUG_ON(nd->flags & LOOKUP_RCU);
 	inode = path->dentry->d_inode;
-finish_lookup:
-	/* we _can_ be in RCU mode here */
 	error = -ENOENT;
 	if (d_is_negative(path->dentry)) {
 		path_to_nameidata(path, nd);
 		goto out;
 	}
-
+finish_lookup:
+	/* we _can_ be in RCU mode here */
 	if (should_follow_link(path->dentry, !symlink_ok)) {
 		if (nd->flags & LOOKUP_RCU) {
 			if (unlikely(nd->path.mnt != path->mnt ||
@@ -3226,7 +3233,7 @@ static struct file *path_openat(int dfd, struct filename *pathname,
 
 	if (unlikely(file->f_flags & __O_TMPFILE)) {
 		error = do_tmpfile(dfd, pathname, nd, flags, op, file, &opened);
-		goto out;
+		goto out2;
 	}
 
 	error = path_init(dfd, pathname, flags, nd);
@@ -3256,6 +3263,7 @@ static struct file *path_openat(int dfd, struct filename *pathname,
 	}
 out:
 	path_cleanup(nd);
+out2:
 	if (!(opened & FILE_OPENED)) {
 		BUG_ON(!error);
 		put_filp(file);
diff --git a/fs/namespace.c b/fs/namespace.c
index 1f4f9da..1b9e111 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -3179,6 +3179,12 @@ bool fs_fully_visible(struct file_system_type *type)
 		if (mnt->mnt.mnt_sb->s_type != type)
 			continue;
 
+		/* This mount is not fully visible if it's root directory
+		 * is not the root directory of the filesystem.
+		 */
+		if (mnt->mnt.mnt_root != mnt->mnt.mnt_sb->s_root)
+			continue;
+
 		/* This mount is not fully visible if there are any child mounts
 		 * that cover anything except for empty directories.
 		 */
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 45b35b9..55e1e3a 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -38,6 +38,7 @@
 #include <linux/mm.h>
 #include <linux/delay.h>
 #include <linux/errno.h>
+#include <linux/file.h>
 #include <linux/string.h>
 #include <linux/ratelimit.h>
 #include <linux/printk.h>
@@ -5604,6 +5605,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl,
 	p->server = server;
 	atomic_inc(&lsp->ls_count);
 	p->ctx = get_nfs_open_context(ctx);
+	get_file(fl->fl_file);
 	memcpy(&p->fl, fl, sizeof(p->fl));
 	return p;
 out_free_seqid:
@@ -5716,6 +5718,7 @@ static void nfs4_lock_release(void *calldata)
 		nfs_free_seqid(data->arg.lock_seqid);
 	nfs4_put_lock_state(data->lsp);
 	put_nfs_open_context(data->ctx);
+	fput(data->fl.fl_file);
 	kfree(data);
 	dprintk("%s: done!\n", __func__);
 }
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index d12a4be..dfc19f1 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -1845,12 +1845,15 @@ int nfs_wb_all(struct inode *inode)
 	trace_nfs_writeback_inode_enter(inode);
 
 	ret = filemap_write_and_wait(inode->i_mapping);
-	if (!ret) {
-		ret = nfs_commit_inode(inode, FLUSH_SYNC);
-		if (!ret)
-			pnfs_sync_inode(inode, true);
-	}
+	if (ret)
+		goto out;
+	ret = nfs_commit_inode(inode, FLUSH_SYNC);
+	if (ret < 0)
+		goto out;
+	pnfs_sync_inode(inode, true);
+	ret = 0;
 
+out:
 	trace_nfs_writeback_inode_exit(inode, ret);
 	return ret;
 }
diff --git a/fs/nfsd/blocklayout.c b/fs/nfsd/blocklayout.c
index 03d647b..cdefaa3 100644
--- a/fs/nfsd/blocklayout.c
+++ b/fs/nfsd/blocklayout.c
@@ -181,6 +181,17 @@ nfsd4_block_proc_layoutcommit(struct inode *inode,
 }
 
 const struct nfsd4_layout_ops bl_layout_ops = {
+	/*
+	 * Pretend that we send notification to the client.  This is a blatant
+	 * lie to force recent Linux clients to cache our device IDs.
+	 * We rarely ever change the device ID, so the harm of leaking deviceids
+	 * for a while isn't too bad.  Unfortunately RFC5661 is a complete mess
+	 * in this regard, but I filed errata 4119 for this a while ago, and
+	 * hopefully the Linux client will eventually start caching deviceids
+	 * without this again.
+	 */
+	.notify_types		=
+			NOTIFY_DEVICEID4_DELETE | NOTIFY_DEVICEID4_CHANGE,
 	.proc_getdeviceinfo	= nfsd4_block_proc_getdeviceinfo,
 	.encode_getdeviceinfo	= nfsd4_block_encode_getdeviceinfo,
 	.proc_layoutget		= nfsd4_block_proc_layoutget,
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 5827785..5694cfb 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -224,7 +224,7 @@ static int nfs_cb_stat_to_errno(int status)
 }
 
 static int decode_cb_op_status(struct xdr_stream *xdr, enum nfs_opnum4 expected,
-			       enum nfsstat4 *status)
+			       int *status)
 {
 	__be32 *p;
 	u32 op;
@@ -235,7 +235,7 @@ static int decode_cb_op_status(struct xdr_stream *xdr, enum nfs_opnum4 expected,
 	op = be32_to_cpup(p++);
 	if (unlikely(op != expected))
 		goto out_unexpected;
-	*status = be32_to_cpup(p);
+	*status = nfs_cb_stat_to_errno(be32_to_cpup(p));
 	return 0;
 out_overflow:
 	print_overflow_msg(__func__, xdr);
@@ -446,22 +446,16 @@ out_overflow:
 static int decode_cb_sequence4res(struct xdr_stream *xdr,
 				  struct nfsd4_callback *cb)
 {
-	enum nfsstat4 nfserr;
 	int status;
 
 	if (cb->cb_minorversion == 0)
 		return 0;
 
-	status = decode_cb_op_status(xdr, OP_CB_SEQUENCE, &nfserr);
-	if (unlikely(status))
-		goto out;
-	if (unlikely(nfserr != NFS4_OK))
-		goto out_default;
-	status = decode_cb_sequence4resok(xdr, cb);
-out:
-	return status;
-out_default:
-	return nfs_cb_stat_to_errno(nfserr);
+	status = decode_cb_op_status(xdr, OP_CB_SEQUENCE, &cb->cb_status);
+	if (unlikely(status || cb->cb_status))
+		return status;
+
+	return decode_cb_sequence4resok(xdr, cb);
 }
 
 /*
@@ -524,26 +518,19 @@ static int nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp,
 				  struct nfsd4_callback *cb)
 {
 	struct nfs4_cb_compound_hdr hdr;
-	enum nfsstat4 nfserr;
 	int status;
 
 	status = decode_cb_compound4res(xdr, &hdr);
 	if (unlikely(status))
-		goto out;
+		return status;
 
 	if (cb != NULL) {
 		status = decode_cb_sequence4res(xdr, cb);
-		if (unlikely(status))
-			goto out;
+		if (unlikely(status || cb->cb_status))
+			return status;
 	}
 
-	status = decode_cb_op_status(xdr, OP_CB_RECALL, &nfserr);
-	if (unlikely(status))
-		goto out;
-	if (unlikely(nfserr != NFS4_OK))
-		status = nfs_cb_stat_to_errno(nfserr);
-out:
-	return status;
+	return decode_cb_op_status(xdr, OP_CB_RECALL, &cb->cb_status);
 }
 
 #ifdef CONFIG_NFSD_PNFS
@@ -621,24 +608,18 @@ static int nfs4_xdr_dec_cb_layout(struct rpc_rqst *rqstp,
 				  struct nfsd4_callback *cb)
 {
 	struct nfs4_cb_compound_hdr hdr;
-	enum nfsstat4 nfserr;
 	int status;
 
 	status = decode_cb_compound4res(xdr, &hdr);
 	if (unlikely(status))
-		goto out;
+		return status;
+
 	if (cb) {
 		status = decode_cb_sequence4res(xdr, cb);
-		if (unlikely(status))
-			goto out;
+		if (unlikely(status || cb->cb_status))
+			return status;
 	}
-	status = decode_cb_op_status(xdr, OP_CB_LAYOUTRECALL, &nfserr);
-	if (unlikely(status))
-		goto out;
-	if (unlikely(nfserr != NFS4_OK))
-		status = nfs_cb_stat_to_errno(nfserr);
-out:
-	return status;
+	return decode_cb_op_status(xdr, OP_CB_LAYOUTRECALL, &cb->cb_status);
 }
 #endif /* CONFIG_NFSD_PNFS */
 
@@ -898,13 +879,6 @@ static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata)
 		if (!nfsd41_cb_get_slot(clp, task))
 			return;
 	}
-	spin_lock(&clp->cl_lock);
-	if (list_empty(&cb->cb_per_client)) {
-		/* This is the first call, not a restart */
-		cb->cb_done = false;
-		list_add(&cb->cb_per_client, &clp->cl_callbacks);
-	}
-	spin_unlock(&clp->cl_lock);
 	rpc_call_start(task);
 }
 
@@ -918,22 +892,33 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
 
 	if (clp->cl_minorversion) {
 		/* No need for lock, access serialized in nfsd4_cb_prepare */
-		++clp->cl_cb_session->se_cb_seq_nr;
+		if (!task->tk_status)
+			++clp->cl_cb_session->se_cb_seq_nr;
 		clear_bit(0, &clp->cl_cb_slot_busy);
 		rpc_wake_up_next(&clp->cl_cb_waitq);
 		dprintk("%s: freed slot, new seqid=%d\n", __func__,
 			clp->cl_cb_session->se_cb_seq_nr);
 	}
 
-	if (clp->cl_cb_client != task->tk_client) {
-		/* We're shutting down or changing cl_cb_client; leave
-		 * it to nfsd4_process_cb_update to restart the call if
-		 * necessary. */
+	/*
+	 * If the backchannel connection was shut down while this
+	 * task was queued, we need to resubmit it after setting up
+	 * a new backchannel connection.
+	 *
+	 * Note that if we lost our callback connection permanently
+	 * the submission code will error out, so we don't need to
+	 * handle that case here.
+	 */
+	if (task->tk_flags & RPC_TASK_KILLED) {
+		task->tk_status = 0;
+		cb->cb_need_restart = true;
 		return;
 	}
 
-	if (cb->cb_done)
-		return;
+	if (cb->cb_status) {
+		WARN_ON_ONCE(task->tk_status);
+		task->tk_status = cb->cb_status;
+	}
 
 	switch (cb->cb_ops->done(cb, task)) {
 	case 0:
@@ -949,21 +934,17 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
 	default:
 		BUG();
 	}
-	cb->cb_done = true;
 }
 
 static void nfsd4_cb_release(void *calldata)
 {
 	struct nfsd4_callback *cb = calldata;
-	struct nfs4_client *clp = cb->cb_clp;
-
-	if (cb->cb_done) {
-		spin_lock(&clp->cl_lock);
-		list_del(&cb->cb_per_client);
-		spin_unlock(&clp->cl_lock);
 
+	if (cb->cb_need_restart)
+		nfsd4_run_cb(cb);
+	else
 		cb->cb_ops->release(cb);
-	}
+
 }
 
 static const struct rpc_call_ops nfsd4_cb_ops = {
@@ -1058,9 +1039,6 @@ static void nfsd4_process_cb_update(struct nfsd4_callback *cb)
 		nfsd4_mark_cb_down(clp, err);
 		return;
 	}
-	/* Yay, the callback channel's back! Restart any callbacks: */
-	list_for_each_entry(cb, &clp->cl_callbacks, cb_per_client)
-		queue_work(callback_wq, &cb->cb_work);
 }
 
 static void
@@ -1071,8 +1049,12 @@ nfsd4_run_cb_work(struct work_struct *work)
 	struct nfs4_client *clp = cb->cb_clp;
 	struct rpc_clnt *clnt;
 
-	if (cb->cb_ops && cb->cb_ops->prepare)
-		cb->cb_ops->prepare(cb);
+	if (cb->cb_need_restart) {
+		cb->cb_need_restart = false;
+	} else {
+		if (cb->cb_ops && cb->cb_ops->prepare)
+			cb->cb_ops->prepare(cb);
+	}
 
 	if (clp->cl_flags & NFSD4_CLIENT_CB_FLAG_MASK)
 		nfsd4_process_cb_update(cb);
@@ -1084,6 +1066,15 @@ nfsd4_run_cb_work(struct work_struct *work)
 			cb->cb_ops->release(cb);
 		return;
 	}
+
+	/*
+	 * Don't send probe messages for 4.1 or later.
+	 */
+	if (!cb->cb_ops && clp->cl_minorversion) {
+		clp->cl_cb_state = NFSD4_CB_UP;
+		return;
+	}
+
 	cb->cb_msg.rpc_cred = clp->cl_cb_cred;
 	rpc_call_async(clnt, &cb->cb_msg, RPC_TASK_SOFT | RPC_TASK_SOFTCONN,
 			cb->cb_ops ? &nfsd4_cb_ops : &nfsd4_cb_probe_ops, cb);
@@ -1098,8 +1089,8 @@ void nfsd4_init_cb(struct nfsd4_callback *cb, struct nfs4_client *clp,
 	cb->cb_msg.rpc_resp = cb;
 	cb->cb_ops = ops;
 	INIT_WORK(&cb->cb_work, nfsd4_run_cb_work);
-	INIT_LIST_HEAD(&cb->cb_per_client);
-	cb->cb_done = true;
+	cb->cb_status = 0;
+	cb->cb_need_restart = false;
 }
 
 void nfsd4_run_cb(struct nfsd4_callback *cb)
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 38f2d7a..039f9c8a 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -94,6 +94,7 @@ static struct kmem_cache *lockowner_slab;
 static struct kmem_cache *file_slab;
 static struct kmem_cache *stateid_slab;
 static struct kmem_cache *deleg_slab;
+static struct kmem_cache *odstate_slab;
 
 static void free_session(struct nfsd4_session *);
 
@@ -281,6 +282,7 @@ put_nfs4_file(struct nfs4_file *fi)
 	if (atomic_dec_and_lock(&fi->fi_ref, &state_lock)) {
 		hlist_del_rcu(&fi->fi_hash);
 		spin_unlock(&state_lock);
+		WARN_ON_ONCE(!list_empty(&fi->fi_clnt_odstate));
 		WARN_ON_ONCE(!list_empty(&fi->fi_delegations));
 		call_rcu(&fi->fi_rcu, nfsd4_free_file_rcu);
 	}
@@ -471,6 +473,86 @@ static void nfs4_file_put_access(struct nfs4_file *fp, u32 access)
 		__nfs4_file_put_access(fp, O_RDONLY);
 }
 
+/*
+ * Allocate a new open/delegation state counter. This is needed for
+ * pNFS for proper return on close semantics.
+ *
+ * Note that we only allocate it for pNFS-enabled exports, otherwise
+ * all pointers to struct nfs4_clnt_odstate are always NULL.
+ */
+static struct nfs4_clnt_odstate *
+alloc_clnt_odstate(struct nfs4_client *clp)
+{
+	struct nfs4_clnt_odstate *co;
+
+	co = kmem_cache_zalloc(odstate_slab, GFP_KERNEL);
+	if (co) {
+		co->co_client = clp;
+		atomic_set(&co->co_odcount, 1);
+	}
+	return co;
+}
+
+static void
+hash_clnt_odstate_locked(struct nfs4_clnt_odstate *co)
+{
+	struct nfs4_file *fp = co->co_file;
+
+	lockdep_assert_held(&fp->fi_lock);
+	list_add(&co->co_perfile, &fp->fi_clnt_odstate);
+}
+
+static inline void
+get_clnt_odstate(struct nfs4_clnt_odstate *co)
+{
+	if (co)
+		atomic_inc(&co->co_odcount);
+}
+
+static void
+put_clnt_odstate(struct nfs4_clnt_odstate *co)
+{
+	struct nfs4_file *fp;
+
+	if (!co)
+		return;
+
+	fp = co->co_file;
+	if (atomic_dec_and_lock(&co->co_odcount, &fp->fi_lock)) {
+		list_del(&co->co_perfile);
+		spin_unlock(&fp->fi_lock);
+
+		nfsd4_return_all_file_layouts(co->co_client, fp);
+		kmem_cache_free(odstate_slab, co);
+	}
+}
+
+static struct nfs4_clnt_odstate *
+find_or_hash_clnt_odstate(struct nfs4_file *fp, struct nfs4_clnt_odstate *new)
+{
+	struct nfs4_clnt_odstate *co;
+	struct nfs4_client *cl;
+
+	if (!new)
+		return NULL;
+
+	cl = new->co_client;
+
+	spin_lock(&fp->fi_lock);
+	list_for_each_entry(co, &fp->fi_clnt_odstate, co_perfile) {
+		if (co->co_client == cl) {
+			get_clnt_odstate(co);
+			goto out;
+		}
+	}
+	co = new;
+	co->co_file = fp;
+	hash_clnt_odstate_locked(new);
+out:
+	spin_unlock(&fp->fi_lock);
+	return co;
+}
+
 struct nfs4_stid *nfs4_alloc_stid(struct nfs4_client *cl,
 					 struct kmem_cache *slab)
 {
@@ -606,7 +688,8 @@ static void block_delegations(struct knfsd_fh *fh)
 }
 
 static struct nfs4_delegation *
-alloc_init_deleg(struct nfs4_client *clp, struct svc_fh *current_fh)
+alloc_init_deleg(struct nfs4_client *clp, struct svc_fh *current_fh,
+		 struct nfs4_clnt_odstate *odstate)
 {
 	struct nfs4_delegation *dp;
 	long n;
@@ -631,6 +714,8 @@ alloc_init_deleg(struct nfs4_client *clp, struct svc_fh *current_fh)
 	INIT_LIST_HEAD(&dp->dl_perfile);
 	INIT_LIST_HEAD(&dp->dl_perclnt);
 	INIT_LIST_HEAD(&dp->dl_recall_lru);
+	dp->dl_clnt_odstate = odstate;
+	get_clnt_odstate(odstate);
 	dp->dl_type = NFS4_OPEN_DELEGATE_READ;
 	dp->dl_retries = 1;
 	nfsd4_init_cb(&dp->dl_recall, dp->dl_stid.sc_client,
@@ -714,6 +799,7 @@ static void destroy_delegation(struct nfs4_delegation *dp)
 	spin_lock(&state_lock);
 	unhash_delegation_locked(dp);
 	spin_unlock(&state_lock);
+	put_clnt_odstate(dp->dl_clnt_odstate);
 	nfs4_put_deleg_lease(dp->dl_stid.sc_file);
 	nfs4_put_stid(&dp->dl_stid);
 }
@@ -724,6 +810,7 @@ static void revoke_delegation(struct nfs4_delegation *dp)
 
 	WARN_ON(!list_empty(&dp->dl_recall_lru));
 
+	put_clnt_odstate(dp->dl_clnt_odstate);
 	nfs4_put_deleg_lease(dp->dl_stid.sc_file);
 
 	if (clp->cl_minorversion == 0)
@@ -933,6 +1020,7 @@ static void nfs4_free_ol_stateid(struct nfs4_stid *stid)
 {
 	struct nfs4_ol_stateid *stp = openlockstateid(stid);
 
+	put_clnt_odstate(stp->st_clnt_odstate);
 	release_all_access(stp);
 	if (stp->st_stateowner)
 		nfs4_put_stateowner(stp->st_stateowner);
@@ -1538,7 +1626,6 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name)
 	INIT_LIST_HEAD(&clp->cl_openowners);
 	INIT_LIST_HEAD(&clp->cl_delegations);
 	INIT_LIST_HEAD(&clp->cl_lru);
-	INIT_LIST_HEAD(&clp->cl_callbacks);
 	INIT_LIST_HEAD(&clp->cl_revoked);
 #ifdef CONFIG_NFSD_PNFS
 	INIT_LIST_HEAD(&clp->cl_lo_states);
@@ -1634,6 +1721,7 @@ __destroy_client(struct nfs4_client *clp)
 	while (!list_empty(&reaplist)) {
 		dp = list_entry(reaplist.next, struct nfs4_delegation, dl_recall_lru);
 		list_del_init(&dp->dl_recall_lru);
+		put_clnt_odstate(dp->dl_clnt_odstate);
 		nfs4_put_deleg_lease(dp->dl_stid.sc_file);
 		nfs4_put_stid(&dp->dl_stid);
 	}
@@ -3057,6 +3145,7 @@ static void nfsd4_init_file(struct knfsd_fh *fh, unsigned int hashval,
 	spin_lock_init(&fp->fi_lock);
 	INIT_LIST_HEAD(&fp->fi_stateids);
 	INIT_LIST_HEAD(&fp->fi_delegations);
+	INIT_LIST_HEAD(&fp->fi_clnt_odstate);
 	fh_copy_shallow(&fp->fi_fhandle, fh);
 	fp->fi_deleg_file = NULL;
 	fp->fi_had_conflict = false;
@@ -3073,6 +3162,7 @@ static void nfsd4_init_file(struct knfsd_fh *fh, unsigned int hashval,
 void
 nfsd4_free_slabs(void)
 {
+	kmem_cache_destroy(odstate_slab);
 	kmem_cache_destroy(openowner_slab);
 	kmem_cache_destroy(lockowner_slab);
 	kmem_cache_destroy(file_slab);
@@ -3103,8 +3193,14 @@ nfsd4_init_slabs(void)
 			sizeof(struct nfs4_delegation), 0, 0, NULL);
 	if (deleg_slab == NULL)
 		goto out_free_stateid_slab;
+	odstate_slab = kmem_cache_create("nfsd4_odstate",
+			sizeof(struct nfs4_clnt_odstate), 0, 0, NULL);
+	if (odstate_slab == NULL)
+		goto out_free_deleg_slab;
 	return 0;
 
+out_free_deleg_slab:
+	kmem_cache_destroy(deleg_slab);
 out_free_stateid_slab:
 	kmem_cache_destroy(stateid_slab);
 out_free_file_slab:
@@ -3581,6 +3677,14 @@ alloc_stateid:
 	open->op_stp = nfs4_alloc_open_stateid(clp);
 	if (!open->op_stp)
 		return nfserr_jukebox;
+
+	if (nfsd4_has_session(cstate) &&
+	    (cstate->current_fh.fh_export->ex_flags & NFSEXP_PNFS)) {
+		open->op_odstate = alloc_clnt_odstate(clp);
+		if (!open->op_odstate)
+			return nfserr_jukebox;
+	}
+
 	return nfs_ok;
 }
 
@@ -3869,7 +3973,7 @@ out_fput:
 
 static struct nfs4_delegation *
 nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
-		    struct nfs4_file *fp)
+		    struct nfs4_file *fp, struct nfs4_clnt_odstate *odstate)
 {
 	int status;
 	struct nfs4_delegation *dp;
@@ -3877,7 +3981,7 @@ nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
 	if (fp->fi_had_conflict)
 		return ERR_PTR(-EAGAIN);
 
-	dp = alloc_init_deleg(clp, fh);
+	dp = alloc_init_deleg(clp, fh, odstate);
 	if (!dp)
 		return ERR_PTR(-ENOMEM);
 
@@ -3903,6 +4007,7 @@ out_unlock:
 	spin_unlock(&state_lock);
 out:
 	if (status) {
+		put_clnt_odstate(dp->dl_clnt_odstate);
 		nfs4_put_stid(&dp->dl_stid);
 		return ERR_PTR(status);
 	}
@@ -3980,7 +4085,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open,
 		default:
 			goto out_no_deleg;
 	}
-	dp = nfs4_set_delegation(clp, fh, stp->st_stid.sc_file);
+	dp = nfs4_set_delegation(clp, fh, stp->st_stid.sc_file, stp->st_clnt_odstate);
 	if (IS_ERR(dp))
 		goto out_no_deleg;
 
@@ -4069,6 +4174,11 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 			release_open_stateid(stp);
 			goto out;
 		}
+
+		stp->st_clnt_odstate = find_or_hash_clnt_odstate(fp,
+							open->op_odstate);
+		if (stp->st_clnt_odstate == open->op_odstate)
+			open->op_odstate = NULL;
 	}
 	update_stateid(&stp->st_stid.sc_stateid);
 	memcpy(&open->op_stateid, &stp->st_stid.sc_stateid, sizeof(stateid_t));
@@ -4129,6 +4239,8 @@ void nfsd4_cleanup_open_state(struct nfsd4_compound_state *cstate,
 		kmem_cache_free(file_slab, open->op_file);
 	if (open->op_stp)
 		nfs4_put_stid(&open->op_stp->st_stid);
+	if (open->op_odstate)
+		kmem_cache_free(odstate_slab, open->op_odstate);
 }
 
 __be32
@@ -4385,10 +4497,17 @@ static __be32 check_stateid_generation(stateid_t *in, stateid_t *ref, bool has_s
 	return nfserr_old_stateid;
 }
 
+static __be32 nfsd4_check_openowner_confirmed(struct nfs4_ol_stateid *ols)
+{
+	if (ols->st_stateowner->so_is_open_owner &&
+	    !(openowner(ols->st_stateowner)->oo_flags & NFS4_OO_CONFIRMED))
+		return nfserr_bad_stateid;
+	return nfs_ok;
+}
+
 static __be32 nfsd4_validate_stateid(struct nfs4_client *cl, stateid_t *stateid)
 {
 	struct nfs4_stid *s;
-	struct nfs4_ol_stateid *ols;
 	__be32 status = nfserr_bad_stateid;
 
 	if (ZERO_STATEID(stateid) || ONE_STATEID(stateid))
@@ -4418,13 +4537,7 @@ static __be32 nfsd4_validate_stateid(struct nfs4_client *cl, stateid_t *stateid)
 		break;
 	case NFS4_OPEN_STID:
 	case NFS4_LOCK_STID:
-		ols = openlockstateid(s);
-		if (ols->st_stateowner->so_is_open_owner
-	    			&& !(openowner(ols->st_stateowner)->oo_flags
-						& NFS4_OO_CONFIRMED))
-			status = nfserr_bad_stateid;
-		else
-			status = nfs_ok;
+		status = nfsd4_check_openowner_confirmed(openlockstateid(s));
 		break;
 	default:
 		printk("unknown stateid type %x\n", s->sc_type);
@@ -4516,8 +4629,8 @@ nfs4_preprocess_stateid_op(struct net *net, struct nfsd4_compound_state *cstate,
 		status = nfs4_check_fh(current_fh, stp);
 		if (status)
 			goto out;
-		if (stp->st_stateowner->so_is_open_owner
-		    && !(openowner(stp->st_stateowner)->oo_flags & NFS4_OO_CONFIRMED))
+		status = nfsd4_check_openowner_confirmed(stp);
+		if (status)
 			goto out;
 		status = nfs4_check_openmode(stp, flags);
 		if (status)
@@ -4852,9 +4965,6 @@ nfsd4_close(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	update_stateid(&stp->st_stid.sc_stateid);
 	memcpy(&close->cl_stateid, &stp->st_stid.sc_stateid, sizeof(stateid_t));
 
-	nfsd4_return_all_file_layouts(stp->st_stateowner->so_client,
-				      stp->st_stid.sc_file);
-
 	nfsd4_close_open_stateid(stp);
 
 	/* put reference from nfs4_preprocess_seqid_op */
@@ -6488,6 +6598,7 @@ nfs4_state_shutdown_net(struct net *net)
 	list_for_each_safe(pos, next, &reaplist) {
 		dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru);
 		list_del_init(&dp->dl_recall_lru);
+		put_clnt_odstate(dp->dl_clnt_odstate);
 		nfs4_put_deleg_lease(dp->dl_stid.sc_file);
 		nfs4_put_stid(&dp->dl_stid);
 	}
diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index 4f3bfeb..dbc4f85 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -63,12 +63,12 @@ typedef struct {
 
 struct nfsd4_callback {
 	struct nfs4_client *cb_clp;
-	struct list_head cb_per_client;
 	u32 cb_minorversion;
 	struct rpc_message cb_msg;
 	struct nfsd4_callback_ops *cb_ops;
 	struct work_struct cb_work;
-	bool cb_done;
+	int cb_status;
+	bool cb_need_restart;
 };
 
 struct nfsd4_callback_ops {
@@ -126,6 +126,7 @@ struct nfs4_delegation {
 	struct list_head	dl_perfile;
 	struct list_head	dl_perclnt;
 	struct list_head	dl_recall_lru;  /* delegation recalled */
+	struct nfs4_clnt_odstate *dl_clnt_odstate;
 	u32			dl_type;
 	time_t			dl_time;
 /* For recall: */
@@ -332,7 +333,6 @@ struct nfs4_client {
 	int			cl_cb_state;
 	struct nfsd4_callback	cl_cb_null;
 	struct nfsd4_session	*cl_cb_session;
-	struct list_head	cl_callbacks; /* list of in-progress callbacks */
 
 	/* for all client information that callback code might need: */
 	spinlock_t		cl_lock;
@@ -465,6 +465,17 @@ static inline struct nfs4_lockowner * lockowner(struct nfs4_stateowner *so)
 }
 
 /*
+ * Per-client state indicating no. of opens and outstanding delegations
+ * on a file from a particular client.'od' stands for 'open & delegation'
+ */
+struct nfs4_clnt_odstate {
+	struct nfs4_client	*co_client;
+	struct nfs4_file	*co_file;
+	struct list_head	co_perfile;
+	atomic_t		co_odcount;
+};
+
+/*
  * nfs4_file: a file opened by some number of (open) nfs4_stateowners.
  *
  * These objects are global. nfsd keeps one instance of a nfs4_file per
@@ -485,6 +496,7 @@ struct nfs4_file {
 		struct list_head	fi_delegations;
 		struct rcu_head		fi_rcu;
 	};
+	struct list_head	fi_clnt_odstate;
 	/* One each for O_RDONLY, O_WRONLY, O_RDWR: */
 	struct file *		fi_fds[3];
 	/*
@@ -526,6 +538,7 @@ struct nfs4_ol_stateid {
 	struct list_head              st_perstateowner;
 	struct list_head              st_locks;
 	struct nfs4_stateowner      * st_stateowner;
+	struct nfs4_clnt_odstate    * st_clnt_odstate;
 	unsigned char                 st_access_bmap;
 	unsigned char                 st_deny_bmap;
 	struct nfs4_ol_stateid         * st_openstp;
diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h
index f982ae8..2f8c092 100644
--- a/fs/nfsd/xdr4.h
+++ b/fs/nfsd/xdr4.h
@@ -247,6 +247,7 @@ struct nfsd4_open {
 	struct nfs4_openowner *op_openowner; /* used during processing */
 	struct nfs4_file *op_file;          /* used during processing */
 	struct nfs4_ol_stateid *op_stp;	    /* used during processing */
+	struct nfs4_clnt_odstate *op_odstate; /* used during processing */
 	struct nfs4_acl *op_acl;
 	struct xdr_netobj op_label;
 };
diff --git a/fs/splice.c b/fs/splice.c
index 476024b..bfe62ae 100644
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -1161,7 +1161,7 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd,
 	long ret, bytes;
 	umode_t i_mode;
 	size_t len;
-	int i, flags;
+	int i, flags, more;
 
 	/*
 	 * We require the input being a regular file, as we don't want to
@@ -1204,6 +1204,7 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd,
 	 * Don't block on output, we have to drain the direct pipe.
 	 */
 	sd->flags &= ~SPLICE_F_NONBLOCK;
+	more = sd->flags & SPLICE_F_MORE;
 
 	while (len) {
 		size_t read_len;
@@ -1217,6 +1218,15 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd,
 		sd->total_len = read_len;
 
 		/*
+		 * If more data is pending, set SPLICE_F_MORE
+		 * If this is the last data and SPLICE_F_MORE was not set
+		 * initially, clears it.
+		 */
+		if (read_len < len)
+			sd->flags |= SPLICE_F_MORE;
+		else if (!more)
+			sd->flags &= ~SPLICE_F_MORE;
+		/*
 		 * NOTE: nonblocking mode only applies to the input. We
 		 * must not do the output in nonblocking mode as then we
 		 * could get stuck data in the internal pipe: