Merge tag 'for-6.17-rc6-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:

 - in zoned mode, turn assertion to proper code when reserving space in
   relocation block group

 - fix search key of extended ref (hardlink) when replaying log

 - fix initialization of file extent tree on filesystems without
   no-holes feature

 - add harmless data race annotation to block group comparator

* tag 'for-6.17-rc6-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: annotate block group access with data_race() when sorting for reclaim
  btrfs: initialize inode::file_extent_tree after i_mode has been set
  btrfs: zoned: fix incorrect ASSERT in btrfs_zoned_reserve_data_reloc_bg()
  btrfs: fix invalid extref key setup when replaying dentry

Author: torvalds

fs/btrfs/block-group.c

@@ -1795,7 +1795,14 @@ static int reclaim_bgs_cmp(void *unused, const struct list_head *a,
 	bg1 = list_entry(a, struct btrfs_block_group, bg_list);
 	bg2 = list_entry(b, struct btrfs_block_group, bg_list);
 
-	return bg1->used > bg2->used;
+	/*
+	 * Some other task may be updating the ->used field concurrently, but it
+	 * is not serious if we get a stale value or load/store tearing issues,
+	 * as sorting the list of block groups to reclaim is not critical and an
+	 * occasional imperfect order is ok. So silence KCSAN and avoid the
+	 * overhead of locking or any other synchronization.
+	 */
+	return data_race(bg1->used > bg2->used);
 }
 
 static inline bool btrfs_should_reclaim(const struct btrfs_fs_info *fs_info)

fs/btrfs/delayed-inode.c

@@ -1843,7 +1843,6 @@ static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
 
 int btrfs_fill_inode(struct btrfs_inode *inode, u32 *rdev)
 {
-	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 	struct btrfs_delayed_node *delayed_node;
 	struct btrfs_inode_item *inode_item;
 	struct inode *vfs_inode = &inode->vfs_inode;
@@ -1864,8 +1863,6 @@ int btrfs_fill_inode(struct btrfs_inode *inode, u32 *rdev)
 	i_uid_write(vfs_inode, btrfs_stack_inode_uid(inode_item));
 	i_gid_write(vfs_inode, btrfs_stack_inode_gid(inode_item));
 	btrfs_i_size_write(inode, btrfs_stack_inode_size(inode_item));
-	btrfs_inode_set_file_extent_range(inode, 0,
-			round_up(i_size_read(vfs_inode), fs_info->sectorsize));
 	vfs_inode->i_mode = btrfs_stack_inode_mode(inode_item);
 	set_nlink(vfs_inode, btrfs_stack_inode_nlink(inode_item));
 	inode_set_bytes(vfs_inode, btrfs_stack_inode_nbytes(inode_item));

fs/btrfs/inode.c

@@ -3885,10 +3885,6 @@ static int btrfs_read_locked_inode(struct btrfs_inode *inode, struct btrfs_path
 	bool filled = false;
 	int first_xattr_slot;
 
-	ret = btrfs_init_file_extent_tree(inode);
-	if (ret)
-		goto out;
-
 	ret = btrfs_fill_inode(inode, &rdev);
 	if (!ret)
 		filled = true;
@@ -3920,8 +3916,6 @@ static int btrfs_read_locked_inode(struct btrfs_inode *inode, struct btrfs_path
 	i_uid_write(vfs_inode, btrfs_inode_uid(leaf, inode_item));
 	i_gid_write(vfs_inode, btrfs_inode_gid(leaf, inode_item));
 	btrfs_i_size_write(inode, btrfs_inode_size(leaf, inode_item));
-	btrfs_inode_set_file_extent_range(inode, 0,
-			round_up(i_size_read(vfs_inode), fs_info->sectorsize));
 
 	inode_set_atime(vfs_inode, btrfs_timespec_sec(leaf, &inode_item->atime),
 			btrfs_timespec_nsec(leaf, &inode_item->atime));
@@ -3953,6 +3947,11 @@ static int btrfs_read_locked_inode(struct btrfs_inode *inode, struct btrfs_path
 	btrfs_set_inode_mapping_order(inode);
 
 cache_index:
+	ret = btrfs_init_file_extent_tree(inode);
+	if (ret)
+		goto out;
+	btrfs_inode_set_file_extent_range(inode, 0,
+			round_up(i_size_read(vfs_inode), fs_info->sectorsize));
 	/*
 	 * If we were modified in the current generation and evicted from memory
 	 * and then re-read we need to do a full sync since we don't have any

fs/btrfs/tree-log.c

@@ -1964,7 +1964,7 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 
 	search_key.objectid = log_key.objectid;
 	search_key.type = BTRFS_INODE_EXTREF_KEY;
-	search_key.offset = key->objectid;
+	search_key.offset = btrfs_extref_hash(key->objectid, name.name, name.len);
 	ret = backref_in_log(root->log_root, &search_key, key->objectid, &name);
 	if (ret < 0) {
 		goto out;

fs/btrfs/zoned.c

@@ -2582,9 +2582,9 @@ void btrfs_zoned_reserve_data_reloc_bg(struct btrfs_fs_info *fs_info)
 			spin_lock(&space_info->lock);
 			space_info->total_bytes -= bg->length;
 			space_info->disk_total -= bg->length * factor;
+			space_info->disk_total -= bg->zone_unusable;
 			/* There is no allocation ever happened. */
 			ASSERT(bg->used == 0);
-			ASSERT(bg->zone_unusable == 0);
 			/* No super block in a block group on the zoned setup. */
 			ASSERT(bg->bytes_super == 0);
 			spin_unlock(&space_info->lock);