Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

Pull btrfs updates from Chris Mason:
 "This pull is mostly cleanups and fixes:

   - The raid5/6 cleanups from Zhao Lei fixup some long standing warts
     in the code and add improvements on top of the scrubbing support
     from 3.19.

   - Josef has round one of our ENOSPC fixes coming from large btrfs
     clusters here at FB.

   - Dave Sterba continues a long series of cleanups (thanks Dave), and
     Filipe continues hammering on corner cases in fsync and others

  This all was held up a little trying to track down a use-after-free in
  btrfs raid5/6.  It's not clear yet if this is just made easier to
  trigger with this pull or if its a new bug from the raid5/6 cleanups.
  Dave Sterba is the only one to trigger it so far, but he has a
  consistent way to reproduce, so we'll get it nailed shortly"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (68 commits)
  Btrfs: don't remove extents and xattrs when logging new names
  Btrfs: fix fsync data loss after adding hard link to inode
  Btrfs: fix BUG_ON in btrfs_orphan_add() when delete unused block group
  Btrfs: account for large extents with enospc
  Btrfs: don't set and clear delalloc for O_DIRECT writes
  Btrfs: only adjust outstanding_extents when we do a short write
  btrfs: Fix out-of-space bug
  Btrfs: scrub, fix sleep in atomic context
  Btrfs: fix scheduler warning when syncing log
  Btrfs: Remove unnecessary placeholder in btrfs_err_code
  btrfs: cleanup init for list in free-space-cache
  btrfs: delete chunk allocation attemp when setting block group ro
  btrfs: clear bio reference after submit_one_bio()
  Btrfs: fix scrub race leading to use-after-free
  Btrfs: add missing cleanup on sysfs init failure
  Btrfs: fix race between transaction commit and empty block group removal
  btrfs: add more checks to btrfs_read_sys_array
  btrfs: cleanup, rename a few variables in btrfs_read_sys_array
  btrfs: add checks for sys_chunk_array sizes
  btrfs: more superblock checks, lower bounds on devices and sectorsize/nodesize
  ...

1 files changed, 184 insertions, 50 deletions

diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 1a9585d4380a..9a37f8b39bae 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -453,11 +453,13 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans,
 insert:
 	btrfs_release_path(path);
 	/* try to insert the key into the destination tree */
+	path->skip_release_on_error = 1;
 	ret = btrfs_insert_empty_item(trans, root, path,
 				      key, item_size);
+	path->skip_release_on_error = 0;
 
 	/* make sure any existing item is the correct size */
-	if (ret == -EEXIST) {
+	if (ret == -EEXIST || ret == -EOVERFLOW) {
 		u32 found_size;
 		found_size = btrfs_item_size_nr(path->nodes[0],
 						path->slots[0]);
@@ -488,8 +490,20 @@ insert:
 		src_item = (struct btrfs_inode_item *)src_ptr;
 		dst_item = (struct btrfs_inode_item *)dst_ptr;
 
-		if (btrfs_inode_generation(eb, src_item) == 0)
+		if (btrfs_inode_generation(eb, src_item) == 0) {
+			struct extent_buffer *dst_eb = path->nodes[0];
+
+			if (S_ISREG(btrfs_inode_mode(eb, src_item)) &&
+			    S_ISREG(btrfs_inode_mode(dst_eb, dst_item))) {
+				struct btrfs_map_token token;
+				u64 ino_size = btrfs_inode_size(eb, src_item);
+
+				btrfs_init_map_token(&token);
+				btrfs_set_token_inode_size(dst_eb, dst_item,
+							   ino_size, &token);
+			}
 			goto no_copy;
+		}
 
 		if (overwrite_root &&
 		    S_ISDIR(btrfs_inode_mode(eb, src_item)) &&
@@ -844,7 +858,7 @@ out:
 static noinline int backref_in_log(struct btrfs_root *log,
 				   struct btrfs_key *key,
 				   u64 ref_objectid,
-				   char *name, int namelen)
+				   const char *name, int namelen)
 {
 	struct btrfs_path *path;
 	struct btrfs_inode_ref *ref;
@@ -1254,13 +1268,14 @@ out:
 }
 
 static int insert_orphan_item(struct btrfs_trans_handle *trans,
-			      struct btrfs_root *root, u64 offset)
+			      struct btrfs_root *root, u64 ino)
 {
 	int ret;
-	ret = btrfs_find_item(root, NULL, BTRFS_ORPHAN_OBJECTID,
-			offset, BTRFS_ORPHAN_ITEM_KEY, NULL);
-	if (ret > 0)
-		ret = btrfs_insert_orphan_item(trans, root, offset);
+
+	ret = btrfs_insert_orphan_item(trans, root, ino);
+	if (ret == -EEXIST)
+		ret = 0;
+
 	return ret;
 }
 
@@ -1287,6 +1302,7 @@ static int count_inode_extrefs(struct btrfs_root *root,
 		leaf = path->nodes[0];
 		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
 		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
+		cur_offset = 0;
 
 		while (cur_offset < item_size) {
 			extref = (struct btrfs_inode_extref *) (ptr + cur_offset);
@@ -1302,7 +1318,7 @@ static int count_inode_extrefs(struct btrfs_root *root,
 	}
 	btrfs_release_path(path);
 
-	if (ret < 0)
+	if (ret < 0 && ret != -ENOENT)
 		return ret;
 	return nlink;
 }
@@ -1394,9 +1410,6 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans,
 	nlink = ret;
 
 	ret = count_inode_extrefs(root, inode, path);
-	if (ret == -ENOENT)
-		ret = 0;
-
 	if (ret < 0)
 		goto out;
 
@@ -1557,6 +1570,30 @@ static noinline int insert_one_name(struct btrfs_trans_handle *trans,
 }
 
 /*
+ * Return true if an inode reference exists in the log for the given name,
+ * inode and parent inode.
+ */
+static bool name_in_log_ref(struct btrfs_root *log_root,
+			    const char *name, const int name_len,
+			    const u64 dirid, const u64 ino)
+{
+	struct btrfs_key search_key;
+
+	search_key.objectid = ino;
+	search_key.type = BTRFS_INODE_REF_KEY;
+	search_key.offset = dirid;
+	if (backref_in_log(log_root, &search_key, dirid, name, name_len))
+		return true;
+
+	search_key.type = BTRFS_INODE_EXTREF_KEY;
+	search_key.offset = btrfs_extref_hash(dirid, name, name_len);
+	if (backref_in_log(log_root, &search_key, dirid, name, name_len))
+		return true;
+
+	return false;
+}
+
+/*
  * take a single entry in a log directory item and replay it into
  * the subvolume.
  *
@@ -1666,10 +1703,17 @@ out:
 	return ret;
 
 insert:
+	if (name_in_log_ref(root->log_root, name, name_len,
+			    key->objectid, log_key.objectid)) {
+		/* The dentry will be added later. */
+		ret = 0;
+		update_size = false;
+		goto out;
+	}
 	btrfs_release_path(path);
 	ret = insert_one_name(trans, root, path, key->objectid, key->offset,
 			      name, name_len, log_type, &log_key);
-	if (ret && ret != -ENOENT)
+	if (ret && ret != -ENOENT && ret != -EEXIST)
 		goto out;
 	update_size = false;
 	ret = 0;
@@ -2164,7 +2208,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 		parent = path->nodes[*level];
 		root_owner = btrfs_header_owner(parent);
 
-		next = btrfs_find_create_tree_block(root, bytenr, blocksize);
+		next = btrfs_find_create_tree_block(root, bytenr);
 		if (!next)
 			return -ENOMEM;
 
@@ -2416,8 +2460,8 @@ static void wait_for_writer(struct btrfs_trans_handle *trans,
 		mutex_unlock(&root->log_mutex);
 		if (atomic_read(&root->log_writers))
 			schedule();
-		mutex_lock(&root->log_mutex);
 		finish_wait(&root->log_writer_wait, &wait);
+		mutex_lock(&root->log_mutex);
 	}
 }
 
@@ -3219,7 +3263,8 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans,
 static void fill_inode_item(struct btrfs_trans_handle *trans,
 			    struct extent_buffer *leaf,
 			    struct btrfs_inode_item *item,
-			    struct inode *inode, int log_inode_only)
+			    struct inode *inode, int log_inode_only,
+			    u64 logged_isize)
 {
 	struct btrfs_map_token token;
 
@@ -3232,7 +3277,7 @@ static void fill_inode_item(struct btrfs_trans_handle *trans,
 		 * to say 'update this inode with these values'
 		 */
 		btrfs_set_token_inode_generation(leaf, item, 0, &token);
-		btrfs_set_token_inode_size(leaf, item, 0, &token);
+		btrfs_set_token_inode_size(leaf, item, logged_isize, &token);
 	} else {
 		btrfs_set_token_inode_generation(leaf, item,
 						 BTRFS_I(inode)->generation,
@@ -3245,19 +3290,19 @@ static void fill_inode_item(struct btrfs_trans_handle *trans,
 	btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token);
 	btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token);
 
-	btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item),
+	btrfs_set_token_timespec_sec(leaf, &item->atime,
 				     inode->i_atime.tv_sec, &token);
-	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item),
+	btrfs_set_token_timespec_nsec(leaf, &item->atime,
 				      inode->i_atime.tv_nsec, &token);
 
-	btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item),
+	btrfs_set_token_timespec_sec(leaf, &item->mtime,
 				     inode->i_mtime.tv_sec, &token);
-	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item),
+	btrfs_set_token_timespec_nsec(leaf, &item->mtime,
 				      inode->i_mtime.tv_nsec, &token);
 
-	btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item),
+	btrfs_set_token_timespec_sec(leaf, &item->ctime,
 				     inode->i_ctime.tv_sec, &token);
-	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item),
+	btrfs_set_token_timespec_nsec(leaf, &item->ctime,
 				      inode->i_ctime.tv_nsec, &token);
 
 	btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode),
@@ -3284,7 +3329,7 @@ static int log_inode_item(struct btrfs_trans_handle *trans,
 		return ret;
 	inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
 				    struct btrfs_inode_item);
-	fill_inode_item(trans, path->nodes[0], inode_item, inode, 0);
+	fill_inode_item(trans, path->nodes[0], inode_item, inode, 0, 0);
 	btrfs_release_path(path);
 	return 0;
 }
@@ -3293,7 +3338,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 			       struct inode *inode,
 			       struct btrfs_path *dst_path,
 			       struct btrfs_path *src_path, u64 *last_extent,
-			       int start_slot, int nr, int inode_only)
+			       int start_slot, int nr, int inode_only,
+			       u64 logged_isize)
 {
 	unsigned long src_offset;
 	unsigned long dst_offset;
@@ -3350,7 +3396,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 						    dst_path->slots[0],
 						    struct btrfs_inode_item);
 			fill_inode_item(trans, dst_path->nodes[0], inode_item,
-					inode, inode_only == LOG_INODE_EXISTS);
+					inode, inode_only == LOG_INODE_EXISTS,
+					logged_isize);
 		} else {
 			copy_extent_buffer(dst_path->nodes[0], src, dst_offset,
 					   src_offset, ins_sizes[i]);
@@ -3902,6 +3949,33 @@ process:
 	return ret;
 }
 
+static int logged_inode_size(struct btrfs_root *log, struct inode *inode,
+			     struct btrfs_path *path, u64 *size_ret)
+{
+	struct btrfs_key key;
+	int ret;
+
+	key.objectid = btrfs_ino(inode);
+	key.type = BTRFS_INODE_ITEM_KEY;
+	key.offset = 0;
+
+	ret = btrfs_search_slot(NULL, log, &key, path, 0, 0);
+	if (ret < 0) {
+		return ret;
+	} else if (ret > 0) {
+		*size_ret = i_size_read(inode);
+	} else {
+		struct btrfs_inode_item *item;
+
+		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				      struct btrfs_inode_item);
+		*size_ret = btrfs_inode_size(path->nodes[0], item);
+	}
+
+	btrfs_release_path(path);
+	return 0;
+}
+
 /* log a single inode in the tree log.
  * At least one parent directory for this inode must exist in the tree
  * or be logged already.
@@ -3939,6 +4013,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	bool fast_search = false;
 	u64 ino = btrfs_ino(inode);
 	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
+	u64 logged_isize = 0;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -3966,15 +4041,22 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 		max_key.type = (u8)-1;
 	max_key.offset = (u64)-1;
 
-	/* Only run delayed items if we are a dir or a new file */
+	/*
+	 * Only run delayed items if we are a dir or a new file.
+	 * Otherwise commit the delayed inode only, which is needed in
+	 * order for the log replay code to mark inodes for link count
+	 * fixup (create temporary BTRFS_TREE_LOG_FIXUP_OBJECTID items).
+	 */
 	if (S_ISDIR(inode->i_mode) ||
-	    BTRFS_I(inode)->generation > root->fs_info->last_trans_committed) {
+	    BTRFS_I(inode)->generation > root->fs_info->last_trans_committed)
 		ret = btrfs_commit_inode_delayed_items(trans, inode);
-		if (ret) {
-			btrfs_free_path(path);
-			btrfs_free_path(dst_path);
-			return ret;
-		}
+	else
+		ret = btrfs_commit_inode_delayed_inode(inode);
+
+	if (ret) {
+		btrfs_free_path(path);
+		btrfs_free_path(dst_path);
+		return ret;
 	}
 
 	mutex_lock(&BTRFS_I(inode)->log_mutex);
@@ -3988,22 +4070,56 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	if (S_ISDIR(inode->i_mode)) {
 		int max_key_type = BTRFS_DIR_LOG_INDEX_KEY;
 
-		if (inode_only == LOG_INODE_EXISTS)
-			max_key_type = BTRFS_XATTR_ITEM_KEY;
+		if (inode_only == LOG_INODE_EXISTS) {
+			max_key_type = BTRFS_INODE_EXTREF_KEY;
+			max_key.type = max_key_type;
+		}
 		ret = drop_objectid_items(trans, log, path, ino, max_key_type);
 	} else {
-		if (test_and_clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
-				       &BTRFS_I(inode)->runtime_flags)) {
-			clear_bit(BTRFS_INODE_COPY_EVERYTHING,
-				  &BTRFS_I(inode)->runtime_flags);
-			ret = btrfs_truncate_inode_items(trans, log,
-							 inode, 0, 0);
-		} else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING,
-					      &BTRFS_I(inode)->runtime_flags) ||
+		if (inode_only == LOG_INODE_EXISTS) {
+			/*
+			 * Make sure the new inode item we write to the log has
+			 * the same isize as the current one (if it exists).
+			 * This is necessary to prevent data loss after log
+			 * replay, and also to prevent doing a wrong expanding
+			 * truncate - for e.g. create file, write 4K into offset
+			 * 0, fsync, write 4K into offset 4096, add hard link,
+			 * fsync some other file (to sync log), power fail - if
+			 * we use the inode's current i_size, after log replay
+			 * we get a 8Kb file, with the last 4Kb extent as a hole
+			 * (zeroes), as if an expanding truncate happened,
+			 * instead of getting a file of 4Kb only.
+			 */
+			err = logged_inode_size(log, inode, path,
+						&logged_isize);
+			if (err)
+				goto out_unlock;
+		}
+		if (test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
+			     &BTRFS_I(inode)->runtime_flags)) {
+			if (inode_only == LOG_INODE_EXISTS) {
+				max_key.type = BTRFS_INODE_EXTREF_KEY;
+				ret = drop_objectid_items(trans, log, path, ino,
+							  max_key.type);
+			} else {
+				clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
+					  &BTRFS_I(inode)->runtime_flags);
+				clear_bit(BTRFS_INODE_COPY_EVERYTHING,
+					  &BTRFS_I(inode)->runtime_flags);
+				ret = btrfs_truncate_inode_items(trans, log,
+								 inode, 0, 0);
+			}
+		} else if (test_bit(BTRFS_INODE_COPY_EVERYTHING,
+				    &BTRFS_I(inode)->runtime_flags) ||
 			   inode_only == LOG_INODE_EXISTS) {
-			if (inode_only == LOG_INODE_ALL)
+			if (inode_only == LOG_INODE_ALL) {
+				clear_bit(BTRFS_INODE_COPY_EVERYTHING,
+					  &BTRFS_I(inode)->runtime_flags);
 				fast_search = true;
-			max_key.type = BTRFS_XATTR_ITEM_KEY;
+				max_key.type = BTRFS_XATTR_ITEM_KEY;
+			} else {
+				max_key.type = BTRFS_INODE_EXTREF_KEY;
+			}
 			ret = drop_objectid_items(trans, log, path, ino,
 						  max_key.type);
 		} else {
@@ -4047,7 +4163,8 @@ again:
 		}
 
 		ret = copy_items(trans, inode, dst_path, path, &last_extent,
-				 ins_start_slot, ins_nr, inode_only);
+				 ins_start_slot, ins_nr, inode_only,
+				 logged_isize);
 		if (ret < 0) {
 			err = ret;
 			goto out_unlock;
@@ -4071,7 +4188,7 @@ next_slot:
 		if (ins_nr) {
 			ret = copy_items(trans, inode, dst_path, path,
 					 &last_extent, ins_start_slot,
-					 ins_nr, inode_only);
+					 ins_nr, inode_only, logged_isize);
 			if (ret < 0) {
 				err = ret;
 				goto out_unlock;
@@ -4092,7 +4209,8 @@ next_slot:
 	}
 	if (ins_nr) {
 		ret = copy_items(trans, inode, dst_path, path, &last_extent,
-				 ins_start_slot, ins_nr, inode_only);
+				 ins_start_slot, ins_nr, inode_only,
+				 logged_isize);
 		if (ret < 0) {
 			err = ret;
 			goto out_unlock;
@@ -4273,6 +4391,9 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 	struct dentry *old_parent = NULL;
 	int ret = 0;
 	u64 last_committed = root->fs_info->last_trans_committed;
+	const struct dentry * const first_parent = parent;
+	const bool did_unlink = (BTRFS_I(inode)->last_unlink_trans >
+				 last_committed);
 
 	sb = inode->i_sb;
 
@@ -4328,7 +4449,6 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 		goto end_trans;
 	}
 
-	inode_only = LOG_INODE_EXISTS;
 	while (1) {
 		if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
 			break;
@@ -4337,8 +4457,22 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 		if (root != BTRFS_I(inode)->root)
 			break;
 
+		/*
+		 * On unlink we must make sure our immediate parent directory
+		 * inode is fully logged. This is to prevent leaving dangling
+		 * directory index entries and a wrong directory inode's i_size.
+		 * Not doing so can result in a directory being impossible to
+		 * delete after log replay (rmdir will always fail with error
+		 * -ENOTEMPTY).
+		 */
+		if (did_unlink && parent == first_parent)
+			inode_only = LOG_INODE_ALL;
+		else
+			inode_only = LOG_INODE_EXISTS;
+
 		if (BTRFS_I(inode)->generation >
-		    root->fs_info->last_trans_committed) {
+		    root->fs_info->last_trans_committed ||
+		    inode_only == LOG_INODE_ALL) {
 			ret = btrfs_log_inode(trans, root, inode, inode_only,
 					      0, LLONG_MAX, ctx);
 			if (ret)