for-6.7-rc1-tag

-----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEE8rQSAMVO+zA4DBdWxWXV+ddtWDsFAmVSO50ACgkQxWXV+ddt
 WDuiyg/7BZviFAyiQMAzpA319qRJZ+EemfTdF/k69q4axGYuvqVdXKnpOV44AR4I
 dKcHLOPpDZIxsh8lFytkm1UEAHptw1v7A64c+gcdjGK0tAA7aKbw/1nmNysowT23
 L0v2+34hkBUfG8A3uVgOwL1rjItEX5Fl54slpVsazSqlEbKrqC4MGNjmqdp3IOeC
 qfXTgkvkXmm8s8NyoJybKewM9Aw0tmK0jkAFHA+2sgcZPYKXjqWGv9KUOsXnCx5o
 3kPWIRT1sj4q2qzrgP14Q12O6qPLZ2/0oTBhi6nhj8+N1yiH+USS5zBITegF+w2n
 leQeVHtyBYHlPYQSQlCIZy7+10gkePvs+JmoAuL8YFISnGYnvOZqCeArlV7cnNI3
 CQt7ZBER5Dqw78Y756usUhpYrLWa9kOpcPVRmjJ/R62+TY1FkkyY7irETbn5EGjI
 NlhEa4PMYeYpAOccoxWEm9tIiiVD1abURhVBdn3Znfcb1Sv/lrGBlo9DYGFCxbBh
 xU1JP7sly8w0aPLqCbn1X3VY8dXp+CeYz4FQabHjQA/zr9lF08/pRYj3haAbYAyH
 0KphXurwz/YqY+LmRg7SbQ/KMgBAiBV8Qk9JyNvdvaQbnYnq7CWdpoHcpZu3mvpb
 HLGoXew58kZaSfxLHlcT5wwYlbq0rooXRstuFg2+BBcOFOMCQfw=
 =GM+1
 -----END PGP SIGNATURE-----

Merge tag 'for-6.7-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:

 - fix potential overflow in returned value from SEARCH_TREE_V2
   ioctl on 32bit architecture

 - zoned mode fixes:

     - drop unnecessary write pointer check for RAID0/RAID1/RAID10
       profiles, now it works because of raid-stripe-tree

     - wait for finishing the zone when direct IO needs a new
       allocation

 - simple quota fixes:

     - pass correct owning root pointer when cleaning up an
       aborted transaction

     - fix leaking some structures when processing delayed refs

     - change key type number of BTRFS_EXTENT_OWNER_REF_KEY,
       reorder it before inline refs that are supposed to be
       sorted, keeping the original number would complicate a lot
       of things; this change needs an updated version of
       btrfs-progs to work and filesystems need to be recreated

 - fix error pointer dereference after failure to allocate fs
   devices

 - fix race between accounting qgroup extents and removing a
   qgroup

* tag 'for-6.7-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: make OWNER_REF_KEY type value smallest among inline refs
  btrfs: fix qgroup record leaks when using simple quotas
  btrfs: fix race between accounting qgroup extents and removing a qgroup
  btrfs: fix error pointer dereference after failure to allocate fs devices
  btrfs: make found_logical_ret parameter mandatory for function queue_scrub_stripe()
  btrfs: get correct owning_root when dropping snapshot
  btrfs: zoned: wait for data BG to be finished on direct IO allocation
  btrfs: zoned: drop no longer valid write pointer check
  btrfs: directly return 0 on no error code in btrfs_insert_raid_extent()
  btrfs: use u64 for buffer sizes in the tree search ioctls

fs/btrfs/ctree.c

@@ -432,7 +432,7 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 	if (btrfs_block_can_be_shared(trans, root, buf)) {
 		ret = btrfs_lookup_extent_info(trans, fs_info, buf->start,
 					       btrfs_header_level(buf), 1,
-					       &refs, &flags);
+					       &refs, &flags, NULL);
 		if (ret)
 			return ret;
 		if (unlikely(refs == 0)) {

fs/btrfs/delayed-ref.c

@@ -1041,7 +1041,7 @@ int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	}
 
-	if (btrfs_qgroup_enabled(fs_info) && !generic_ref->skip_qgroup) {
+	if (btrfs_qgroup_full_accounting(fs_info) && !generic_ref->skip_qgroup) {
 		record = kzalloc(sizeof(*record), GFP_NOFS);
 		if (!record) {
 			kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref);
@@ -1144,7 +1144,7 @@ int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	}
 
-	if (btrfs_qgroup_enabled(fs_info) && !generic_ref->skip_qgroup) {
+	if (btrfs_qgroup_full_accounting(fs_info) && !generic_ref->skip_qgroup) {
 		record = kzalloc(sizeof(*record), GFP_NOFS);
 		if (!record) {
 			kmem_cache_free(btrfs_delayed_data_ref_cachep, ref);

fs/btrfs/extent-tree.c

@@ -102,7 +102,8 @@ int btrfs_lookup_data_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len)
  */
 int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info, u64 bytenr,
-			     u64 offset, int metadata, u64 *refs, u64 *flags)
+			     u64 offset, int metadata, u64 *refs, u64 *flags,
+			     u64 *owning_root)
 {
 	struct btrfs_root *extent_root;
 	struct btrfs_delayed_ref_head *head;
@@ -114,6 +115,7 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 	u32 item_size;
 	u64 num_refs;
 	u64 extent_flags;
+	u64 owner = 0;
 	int ret;
 
 	/*
@@ -167,6 +169,8 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 					    struct btrfs_extent_item);
 			num_refs = btrfs_extent_refs(leaf, ei);
 			extent_flags = btrfs_extent_flags(leaf, ei);
+			owner = btrfs_get_extent_owner_root(fs_info, leaf,
+							    path->slots[0]);
 		} else {
 			ret = -EUCLEAN;
 			btrfs_err(fs_info,
@@ -226,6 +230,8 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 		*refs = num_refs;
 	if (flags)
 		*flags = extent_flags;
+	if (owning_root)
+		*owning_root = owner;
 out_free:
 	btrfs_free_path(path);
 	return ret;
@@ -5234,7 +5240,7 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans,
 		/* We don't lock the tree block, it's OK to be racy here */
 		ret = btrfs_lookup_extent_info(trans, fs_info, bytenr,
 					       wc->level - 1, 1, &refs,
-					       &flags);
+					       &flags, NULL);
 		/* We don't care about errors in readahead. */
 		if (ret < 0)
 			continue;
@@ -5301,7 +5307,8 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
 		ret = btrfs_lookup_extent_info(trans, fs_info,
 					       eb->start, level, 1,
 					       &wc->refs[level],
-					       &wc->flags[level]);
+					       &wc->flags[level],
+					       NULL);
 		BUG_ON(ret == -ENOMEM);
 		if (ret)
 			return ret;
@@ -5391,6 +5398,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 	u64 bytenr;
 	u64 generation;
 	u64 parent;
+	u64 owner_root = 0;
 	struct btrfs_tree_parent_check check = { 0 };
 	struct btrfs_key key;
 	struct btrfs_ref ref = { 0 };
@@ -5434,7 +5442,8 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_lookup_extent_info(trans, fs_info, bytenr, level - 1, 1,
 				       &wc->refs[level - 1],
-				       &wc->flags[level - 1]);
+				       &wc->flags[level - 1],
+				       &owner_root);
 	if (ret < 0)
 		goto out_unlock;
 
@@ -5567,8 +5576,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 		find_next_key(path, level, &wc->drop_progress);
 
 		btrfs_init_generic_ref(&ref, BTRFS_DROP_DELAYED_REF, bytenr,
-				       fs_info->nodesize, parent,
-				       btrfs_header_owner(next));
+				       fs_info->nodesize, parent, owner_root);
 		btrfs_init_tree_ref(&ref, level - 1, root->root_key.objectid,
 				    0, false);
 		ret = btrfs_free_extent(trans, &ref);
@@ -5635,7 +5643,8 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
 			ret = btrfs_lookup_extent_info(trans, fs_info,
 						       eb->start, level, 1,
 						       &wc->refs[level],
-						       &wc->flags[level]);
+						       &wc->flags[level],
+						       NULL);
 			if (ret < 0) {
 				btrfs_tree_unlock_rw(eb, path->locks[level]);
 				path->locks[level] = 0;
@@ -5880,7 +5889,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref, int for_reloc)
 			ret = btrfs_lookup_extent_info(trans, fs_info,
 						path->nodes[level]->start,
 						level, 1, &wc->refs[level],
-						&wc->flags[level]);
+						&wc->flags[level], NULL);
 			if (ret < 0) {
 				err = ret;
 				goto out_end_trans;

fs/btrfs/extent-tree.h

@@ -99,7 +99,8 @@ u64 btrfs_cleanup_ref_head_accounting(struct btrfs_fs_info *fs_info,
 int btrfs_lookup_data_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len);
 int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info, u64 bytenr,
-			     u64 offset, int metadata, u64 *refs, u64 *flags);
+			     u64 offset, int metadata, u64 *refs, u64 *flags,
+			     u64 *owner_root);
 int btrfs_pin_extent(struct btrfs_trans_handle *trans, u64 bytenr, u64 num,
 		     int reserved);
 int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans,

fs/btrfs/inode.c

@@ -6983,8 +6983,15 @@ static struct extent_map *btrfs_new_extent_direct(struct btrfs_inode *inode,
 	int ret;
 
 	alloc_hint = get_extent_allocation_hint(inode, start, len);
+again:
 	ret = btrfs_reserve_extent(root, len, len, fs_info->sectorsize,
 				   0, alloc_hint, &ins, 1, 1);
+	if (ret == -EAGAIN) {
+		ASSERT(btrfs_is_zoned(fs_info));
+		wait_on_bit_io(&inode->root->fs_info->flags, BTRFS_FS_NEED_ZONE_FINISH,
+			       TASK_UNINTERRUPTIBLE);
+		goto again;
+	}
 	if (ret)
 		return ERR_PTR(ret);
 

fs/btrfs/ioctl.c

@@ -1528,7 +1528,7 @@ static noinline int key_in_sk(struct btrfs_key *key,
 static noinline int copy_to_sk(struct btrfs_path *path,
 			       struct btrfs_key *key,
 			       struct btrfs_ioctl_search_key *sk,
-			       size_t *buf_size,
+			       u64 *buf_size,
 			       char __user *ubuf,
 			       unsigned long *sk_offset,
 			       int *num_found)
@@ -1660,7 +1660,7 @@ static noinline int copy_to_sk(struct btrfs_path *path,
 
 static noinline int search_ioctl(struct inode *inode,
 				 struct btrfs_ioctl_search_key *sk,
-				 size_t *buf_size,
+				 u64 *buf_size,
 				 char __user *ubuf)
 {
 	struct btrfs_fs_info *info = btrfs_sb(inode->i_sb);
@@ -1733,7 +1733,7 @@ static noinline int btrfs_ioctl_tree_search(struct inode *inode,
 	struct btrfs_ioctl_search_args __user *uargs = argp;
 	struct btrfs_ioctl_search_key sk;
 	int ret;
-	size_t buf_size;
+	u64 buf_size;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -1763,8 +1763,8 @@ static noinline int btrfs_ioctl_tree_search_v2(struct inode *inode,
 	struct btrfs_ioctl_search_args_v2 __user *uarg = argp;
 	struct btrfs_ioctl_search_args_v2 args;
 	int ret;
-	size_t buf_size;
-	const size_t buf_limit = SZ_16M;
+	u64 buf_size;
+	const u64 buf_limit = SZ_16M;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;

fs/btrfs/qgroup.c

@@ -1888,7 +1888,7 @@ int btrfs_qgroup_trace_extent_nolock(struct btrfs_fs_info *fs_info,
 	u64 bytenr = record->bytenr;
 
 	if (!btrfs_qgroup_full_accounting(fs_info))
-		return 0;
+		return 1;
 
 	lockdep_assert_held(&delayed_refs->lock);
 	trace_btrfs_qgroup_trace_extent(fs_info, record);
@@ -2874,13 +2874,19 @@ int btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans, u64 bytenr,
 	qgroup_update_counters(fs_info, &qgroups, nr_old_roots, nr_new_roots,
 			       num_bytes, seq);
 
+	/*
+	 * We're done using the iterator, release all its qgroups while holding
+	 * fs_info->qgroup_lock so that we don't race with btrfs_remove_qgroup()
+	 * and trigger use-after-free accesses to qgroups.
+	 */
+	qgroup_iterator_nested_clean(&qgroups);
+
 	/*
 	 * Bump qgroup_seq to avoid seq overlap
 	 */
 	fs_info->qgroup_seq += max(nr_old_roots, nr_new_roots) + 1;
 	spin_unlock(&fs_info->qgroup_lock);
 out_free:
-	qgroup_iterator_nested_clean(&qgroups);
 	ulist_free(old_roots);
 	ulist_free(new_roots);
 	return ret;

fs/btrfs/raid-stripe-tree.c

@@ -145,7 +145,7 @@ int btrfs_insert_raid_extent(struct btrfs_trans_handle *trans,
 		btrfs_put_bioc(bioc);
 	}
 
-	return ret;
+	return 0;
 }
 
 int btrfs_get_raid_extent_offset(struct btrfs_fs_info *fs_info,

fs/btrfs/scrub.c

@@ -1868,6 +1868,9 @@ static int queue_scrub_stripe(struct scrub_ctx *sctx, struct btrfs_block_group *
 	 */
 	ASSERT(sctx->cur_stripe < SCRUB_TOTAL_STRIPES);
 
+	/* @found_logical_ret must be specified. */
+	ASSERT(found_logical_ret);
+
 	stripe = &sctx->stripes[sctx->cur_stripe];
 	scrub_reset_stripe(stripe);
 	ret = scrub_find_fill_first_stripe(bg, &sctx->extent_path,
@@ -1876,8 +1879,7 @@ static int queue_scrub_stripe(struct scrub_ctx *sctx, struct btrfs_block_group *
 	/* Either >0 as no more extents or <0 for error. */
 	if (ret)
 		return ret;
-	if (found_logical_ret)
-		*found_logical_ret = stripe->logical;
+	*found_logical_ret = stripe->logical;
 	sctx->cur_stripe++;
 
 	/* We filled one group, submit it. */
@@ -2080,7 +2082,7 @@ static int scrub_simple_mirror(struct scrub_ctx *sctx,
 
 	/* Go through each extent items inside the logical range */
 	while (cur_logical < logical_end) {
-		u64 found_logical;
+		u64 found_logical = U64_MAX;
 		u64 cur_physical = physical + cur_logical - logical_start;
 
 		/* Canceled? */
@@ -2115,6 +2117,8 @@ static int scrub_simple_mirror(struct scrub_ctx *sctx,
 		if (ret < 0)
 			break;
 
+		/* queue_scrub_stripe() returned 0, @found_logical must be updated. */
+		ASSERT(found_logical != U64_MAX);
 		cur_logical = found_logical + BTRFS_STRIPE_LEN;
 
 		/* Don't hold CPU for too long time */

fs/btrfs/volumes.c

@@ -748,13 +748,13 @@ static noinline struct btrfs_device *device_list_add(const char *path,
 
 	if (!fs_devices) {
 		fs_devices = alloc_fs_devices(disk_super->fsid);
+		if (IS_ERR(fs_devices))
+			return ERR_CAST(fs_devices);
+
 		if (has_metadata_uuid)
 			memcpy(fs_devices->metadata_uuid,
 			       disk_super->metadata_uuid, BTRFS_FSID_SIZE);
 
-		if (IS_ERR(fs_devices))
-			return ERR_CAST(fs_devices);
-
 		if (same_fsid_diff_dev) {
 			generate_random_uuid(fs_devices->fsid);
 			fs_devices->temp_fsid = true;

fs/btrfs/zoned.c

@@ -1661,13 +1661,6 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
 	}
 
 out:
-	if (cache->alloc_offset > fs_info->zone_size) {
-		btrfs_err(fs_info,
-			"zoned: invalid write pointer %llu in block group %llu",
-			cache->alloc_offset, cache->start);
-		ret = -EIO;
-	}
-
 	if (cache->alloc_offset > cache->zone_capacity) {
 		btrfs_err(fs_info,
 "zoned: invalid write pointer %llu (larger than zone capacity %llu) in block group %llu",

include/uapi/linux/btrfs_tree.h

@@ -219,6 +219,22 @@
  */
 #define BTRFS_METADATA_ITEM_KEY	169
 
+/*
+ * Special inline ref key which stores the id of the subvolume which originally
+ * created the extent. This subvolume owns the extent permanently from the
+ * perspective of simple quotas. Needed to know which subvolume to free quota
+ * usage from when the extent is deleted.
+ *
+ * Stored as an inline ref rather to avoid wasting space on a separate item on
+ * top of the existing extent item. However, unlike the other inline refs,
+ * there is one one owner ref per extent rather than one per extent.
+ *
+ * Because of this, it goes at the front of the list of inline refs, and thus
+ * must have a lower type value than any other inline ref type (to satisfy the
+ * disk format rule that inline refs have non-decreasing type).
+ */
+#define BTRFS_EXTENT_OWNER_REF_KEY	172
+
 #define BTRFS_TREE_BLOCK_REF_KEY	176
 
 #define BTRFS_EXTENT_DATA_REF_KEY	178
@@ -233,14 +249,6 @@
 
 #define BTRFS_SHARED_DATA_REF_KEY	184
 
-/*
- * Special inline ref key which stores the id of the subvolume which originally
- * created the extent. This subvolume owns the extent permanently from the
- * perspective of simple quotas. Needed to know which subvolume to free quota
- * usage from when the extent is deleted.
- */
-#define BTRFS_EXTENT_OWNER_REF_KEY	188
-
 /*
  * block groups give us hints into the extent allocation trees.  Which
  * blocks are free etc etc