for-6.13-rc1-tag

-----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEE8rQSAMVO+zA4DBdWxWXV+ddtWDsFAmdPMbMACgkQxWXV+ddt
 WDtFSA/+Kd61BPKaiZFF0yjOsjBOlu8WMHFLO5xa2ZRqV3HUzm6rdO4wUSq3Eyqg
 lrFszPfLA0REZIEdY2rqDKJduk1MZZg6NY7Pvn+/ByGOorM0Ym1BJoydtlUN5o2Y
 AWeaDxs6LBMwRqpai0+AkikufK/jK7QfhHci+Oo2XmOv1C39DQbkXO76SYh/yERt
 CcZNaSjm9DUlLxOOpkefxYvpPW4Uv4NSBfh9aymX/u1VxXqeuMuzPZqwZO+nwl/p
 M1yr9fcbrqh3yKC+JjhD7xmOJM3x4c2PmzSkQTdepOuAlQvuQ/iFD+Zc1YjY0XlT
 Fl938rdTKULgCarR5rdsXqdlFRnOprlgt0J1Pdf+GipTVU0EY3WU343HCz6h8pmG
 F/NPvlCahkEtk1UCguL92NOBeAf0adWhuYfKjkuxYuL5ZTvzsOl2ymF6Vlja0Y39
 VK4exjG4ilDESxweinJe53k3QLDwvUc2h0D291QVoo2X06dhXda8oOVK+6lR5mij
 zDtXqurjlCybT1W7op6RcWMY0TaA8IR3Bo9oU1YbVcctuc86/X8SERv1G8MQtXhh
 tevOVb/cKy7gEw9q73OtrH/J2EAklnsesnpH2sdt7WUEzOQPw9BIwEa/NRbjvoqn
 n2ts5KktBhNjM1s0elHBf7o4+OaTrTlTrp0HelXyONZucPrzDQ4=
 =4mFO
 -----END PGP SIGNATURE-----

Merge tag 'for-6.13-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:

 - add lockdep annotations for io_uring/encoded read integration, inode
   lock is held when returning to userspace

 - properly reflect experimental config option to sysfs

 - handle NULL root in case the rescue mode accepts invalid/damaged tree
   roots (rescue=ibadroot)

 - regression fix of a deadlock between transaction and extent locks

 - fix pending bio accounting bug in encoded read ioctl

 - fix NOWAIT mode when checking references for NOCOW files

 - fix use-after-free in a rb-tree cleanup in ref-verify debugging tool

* tag 'for-6.13-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: fix lockdep warnings on io_uring encoded reads
  btrfs: ref-verify: fix use-after-free after invalid ref action
  btrfs: add a sanity check for btrfs root in btrfs_search_slot()
  btrfs: don't loop for nowait writes when checking for cross references
  btrfs: sysfs: advertise experimental features only if CONFIG_BTRFS_EXPERIMENTAL=y
  btrfs: fix deadlock between transaction commits and extent locks
  btrfs: fix use-after-free in btrfs_encoded_read_endio()

fs/btrfs/ctree.c

@@ -2046,7 +2046,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		      const struct btrfs_key *key, struct btrfs_path *p,
 		      int ins_len, int cow)
 {
-	struct btrfs_fs_info *fs_info = root->fs_info;
+	struct btrfs_fs_info *fs_info;
 	struct extent_buffer *b;
 	int slot;
 	int ret;
@@ -2059,6 +2059,10 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	int min_write_lock_level;
 	int prev_cmp;
 
+	if (!root)
+		return -EINVAL;
+
+	fs_info = root->fs_info;
 	might_sleep();
 
 	lowest_level = p->lowest_level;

fs/btrfs/extent-tree.c

@@ -2422,7 +2422,7 @@ int btrfs_cross_ref_exist(struct btrfs_root *root, u64 objectid, u64 offset,
 			goto out;
 
 		ret = check_delayed_ref(root, path, objectid, offset, bytenr);
-	} while (ret == -EAGAIN);
+	} while (ret == -EAGAIN && !path->nowait);
 
 out:
 	btrfs_release_path(path);

fs/btrfs/inode.c

@@ -3063,6 +3063,19 @@ int btrfs_finish_one_ordered(struct btrfs_ordered_extent *ordered_extent)
 			goto out;
 	}
 
+	/*
+	 * If it's a COW write we need to lock the extent range as we will be
+	 * inserting/replacing file extent items and unpinning an extent map.
+	 * This must be taken before joining a transaction, as it's a higher
+	 * level lock (like the inode's VFS lock), otherwise we can run into an
+	 * ABBA deadlock with other tasks (transactions work like a lock,
+	 * depending on their current state).
+	 */
+	if (!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
+		clear_bits |= EXTENT_LOCKED;
+		lock_extent(io_tree, start, end, &cached_state);
+	}
+
 	if (freespace_inode)
 		trans = btrfs_join_transaction_spacecache(root);
 	else
@@ -3099,9 +3112,6 @@ int btrfs_finish_one_ordered(struct btrfs_ordered_extent *ordered_extent)
 		goto out;
 	}
 
-	clear_bits |= EXTENT_LOCKED;
-	lock_extent(io_tree, start, end, &cached_state);
-
 	if (test_bit(BTRFS_ORDERED_COMPRESSED, &ordered_extent->flags))
 		compress_type = ordered_extent->compress_type;
 	if (test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) {
@@ -9089,7 +9099,7 @@ static void btrfs_encoded_read_endio(struct btrfs_bio *bbio)
 		 */
 		WRITE_ONCE(priv->status, bbio->bio.bi_status);
 	}
-	if (atomic_dec_return(&priv->pending) == 0) {
+	if (atomic_dec_and_test(&priv->pending)) {
 		int err = blk_status_to_errno(READ_ONCE(priv->status));
 
 		if (priv->uring_ctx) {

fs/btrfs/ioctl.c

@@ -4751,6 +4751,9 @@ static void btrfs_uring_read_finished(struct io_uring_cmd *cmd, unsigned int iss
 	size_t page_offset;
 	ssize_t ret;
 
+	/* The inode lock has already been acquired in btrfs_uring_read_extent.  */
+	btrfs_lockdep_inode_acquire(inode, i_rwsem);
+
 	if (priv->err) {
 		ret = priv->err;
 		goto out;
@@ -4859,6 +4862,13 @@ static int btrfs_uring_read_extent(struct kiocb *iocb, struct iov_iter *iter,
 	 * and inode and freeing the allocations.
 	 */
 
+	/*
+	 * We're returning to userspace with the inode lock held, and that's
+	 * okay - it'll get unlocked in a worker thread.  Call
+	 * btrfs_lockdep_inode_release() to avoid confusing lockdep.
+	 */
+	btrfs_lockdep_inode_release(inode, i_rwsem);
+
 	return -EIOCBQUEUED;
 
 out_fail:

fs/btrfs/locking.h

@@ -128,6 +128,16 @@ enum btrfs_lockdep_trans_states {
 #define btrfs_lockdep_release(owner, lock)					\
 	rwsem_release(&owner->lock##_map, _THIS_IP_)
 
+/*
+ * Used to account for the fact that when doing io_uring encoded I/O, we can
+ * return to userspace with the inode lock still held.
+ */
+#define btrfs_lockdep_inode_acquire(owner, lock)				\
+	rwsem_acquire_read(&owner->vfs_inode.lock.dep_map, 0, 0, _THIS_IP_)
+
+#define btrfs_lockdep_inode_release(owner, lock)				\
+	rwsem_release(&owner->vfs_inode.lock.dep_map, _THIS_IP_)
+
 /*
  * Macros for the transaction states wait events, similar to the generic wait
  * event macros.

fs/btrfs/ref-verify.c

@@ -857,6 +857,7 @@ int btrfs_ref_tree_mod(struct btrfs_fs_info *fs_info,
 "dropping a ref for a root that doesn't have a ref on the block");
 			dump_block_entry(fs_info, be);
 			dump_ref_action(fs_info, ra);
+			rb_erase(&ref->node, &be->refs);
 			kfree(ref);
 			kfree(ra);
 			goto out_unlock;

fs/btrfs/sysfs.c

@@ -295,7 +295,7 @@ BTRFS_FEAT_ATTR_INCOMPAT(simple_quota, SIMPLE_QUOTA);
 #ifdef CONFIG_BLK_DEV_ZONED
 BTRFS_FEAT_ATTR_INCOMPAT(zoned, ZONED);
 #endif
-#ifdef CONFIG_BTRFS_DEBUG
+#ifdef CONFIG_BTRFS_EXPERIMENTAL
 /* Remove once support for extent tree v2 is feature complete */
 BTRFS_FEAT_ATTR_INCOMPAT(extent_tree_v2, EXTENT_TREE_V2);
 /* Remove once support for raid stripe tree is feature complete. */
@@ -329,7 +329,7 @@ static struct attribute *btrfs_supported_feature_attrs[] = {
 #ifdef CONFIG_BLK_DEV_ZONED
 	BTRFS_FEAT_ATTR_PTR(zoned),
 #endif
-#ifdef CONFIG_BTRFS_DEBUG
+#ifdef CONFIG_BTRFS_EXPERIMENTAL
 	BTRFS_FEAT_ATTR_PTR(extent_tree_v2),
 	BTRFS_FEAT_ATTR_PTR(raid_stripe_tree),
 #endif