Update bcachefs sources to 4a32728376a8 bcachefs: bcachefs_metadata_version_inode_depth

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

32 files changed, 713 insertions, 260 deletions

diff --git a/.bcachefs_revision b/.bcachefs_revision
index bbe1be9a..534b898f 100644
--- a/.bcachefs_revision
+++ b/.bcachefs_revision
@@ -1 +1 @@
-bc01863fb6eff06f7b028e4c5cd8850d21d7f10d
+4a32728376a81dd9e75851a49159ff33602840f4
diff --git a/bch_bindgen/src/bkey.rs b/bch_bindgen/src/bkey.rs
index 91d515b7..2a012508 100644
--- a/bch_bindgen/src/bkey.rs
+++ b/bch_bindgen/src/bkey.rs
@@ -50,6 +50,7 @@ pub enum BkeyValC<'a> {
     logged_op_truncate(&'a c::bch_logged_op_truncate),
     logged_op_finsert(&'a c::bch_logged_op_finsert),
     accounting(&'a c::bch_accounting),
+    inode_alloc_cursor(&'a c::bch_inode_alloc_cursor),
 }
 
 impl<'a, 'b> BkeySC<'a> {
@@ -106,6 +107,7 @@ impl<'a, 'b> BkeySC<'a> {
                 KEY_TYPE_logged_op_truncate => logged_op_truncate(transmute(self.v)),
                 KEY_TYPE_logged_op_finsert => logged_op_finsert(transmute(self.v)),
                 KEY_TYPE_accounting => accounting(transmute(self.v)),
+                KEY_TYPE_inode_alloc_cursor => inode_alloc_cursor(transmute(self.v)),
                 KEY_TYPE_MAX => unreachable!(),
             }
         }
diff --git a/include/linux/slab.h b/include/linux/slab.h
index 24df2cc5..4a6d015f 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -103,6 +103,8 @@ static inline void *kmalloc_array(size_t n, size_t size, gfp_t flags)
 	((size) != 0 && (n) > SIZE_MAX / (size)				\
 	 ? NULL : kmalloc((n) * (size), flags))
 
+#define kvcalloc(n, size, flags)	kvmalloc_array(n, size, flags|__GFP_ZERO)
+
 #define kvmalloc_array_noprof(...)	kvmalloc_array(__VA_ARGS__)
 
 #define kcalloc(n, size, flags)		kmalloc_array(n, size, flags|__GFP_ZERO)
diff --git a/include/linux/srcu.h b/include/linux/srcu.h
index b93cb8e3..7aa2b964 100644
--- a/include/linux/srcu.h
+++ b/include/linux/srcu.h
@@ -7,6 +7,11 @@
 
 typedef void (*rcu_callback_t)(struct rcu_head *head);
 
+static inline struct urcu_gp_poll_state get_state_synchronize_rcu()
+{
+	return start_poll_synchronize_rcu();
+}
+
 struct srcu_struct {
 };
 
@@ -17,36 +22,19 @@ static inline int srcu_read_lock(struct srcu_struct *ssp)
 	return 0;
 }
 
-static inline bool poll_state_synchronize_srcu(struct srcu_struct *ssp, unsigned long cookie)
-{
-	return true;
-}
-
-static inline unsigned long start_poll_synchronize_srcu(struct srcu_struct *ssp)
-{
-	return 0;
-}
-
-static inline unsigned long get_state_synchronize_srcu(struct srcu_struct *ssp)
-{
-	return 0;
-}
-
-#undef poll_state_synchronize_rcu
-static inline bool poll_state_synchronize_rcu(unsigned long cookie)
+static inline bool poll_state_synchronize_srcu(struct srcu_struct *ssp, struct urcu_gp_poll_state cookie)
 {
-	return false;
+	return poll_state_synchronize_rcu(cookie);
 }
 
-#undef start_poll_synchronize_rcu
-static inline unsigned long start_poll_synchronize_rcu()
+static inline struct urcu_gp_poll_state start_poll_synchronize_srcu(struct srcu_struct *ssp)
 {
-	return 0;
+	return start_poll_synchronize_rcu();
 }
 
-static inline unsigned long get_state_synchronize_rcu()
+static inline struct urcu_gp_poll_state get_state_synchronize_srcu(struct srcu_struct *ssp)
 {
-	return 0;
+	return get_state_synchronize_rcu();
 }
 
 static inline void synchronize_srcu_expedited(struct srcu_struct *ssp) {}
diff --git a/libbcachefs/backpointers.c b/libbcachefs/backpointers.c
index b53d98c7..72f37171 100644
--- a/libbcachefs/backpointers.c
+++ b/libbcachefs/backpointers.c
@@ -72,26 +72,14 @@ static bool extent_matches_bp(struct bch_fs *c,
 	const union bch_extent_entry *entry;
 	struct extent_ptr_decoded p;
 
-	rcu_read_lock();
 	bkey_for_each_ptr_decode(k.k, ptrs, p, entry) {
-		struct bpos bucket2;
 		struct bkey_i_backpointer bp2;
+		bch2_extent_ptr_to_bp(c, btree_id, level, k, p, entry, &bp2);
 
-		if (p.ptr.cached)
-			continue;
-
-		struct bch_dev *ca = bch2_dev_rcu(c, p.ptr.dev);
-		if (!ca)
-			continue;
-
-		bch2_extent_ptr_to_bp(c, ca, btree_id, level, k, p, entry, &bucket2, &bp2);
 		if (bpos_eq(bp.k->p, bp2.k.p) &&
-		    !memcmp(bp.v, &bp2.v, sizeof(bp2.v))) {
-			rcu_read_unlock();
+		    !memcmp(bp.v, &bp2.v, sizeof(bp2.v)))
 			return true;
-		}
 	}
-	rcu_read_unlock();
 
 	return false;
 }
@@ -183,9 +171,10 @@ err:
 
 static int bch2_backpointer_del(struct btree_trans *trans, struct bpos pos)
 {
-	return likely(!bch2_backpointers_no_use_write_buffer)
-	       ? bch2_btree_delete_at_buffered(trans, BTREE_ID_backpointers, pos)
-	       : bch2_btree_delete(trans, BTREE_ID_backpointers, pos, 0);
+	return (likely(!bch2_backpointers_no_use_write_buffer)
+		? bch2_btree_delete_at_buffered(trans, BTREE_ID_backpointers, pos)
+		: bch2_btree_delete(trans, BTREE_ID_backpointers, pos, 0)) ?:
+		 bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc);
 }
 
 static int bch2_backpointers_maybe_flush(struct btree_trans *trans,
@@ -197,20 +186,23 @@ static int bch2_backpointers_maybe_flush(struct btree_trans *trans,
 		: 0;
 }
 
-static void backpointer_target_not_found(struct btree_trans *trans,
-					 struct bkey_s_c_backpointer bp,
-					 struct bkey_s_c target_k)
+static int backpointer_target_not_found(struct btree_trans *trans,
+				  struct bkey_s_c_backpointer bp,
+				  struct bkey_s_c target_k,
+				  struct bkey_buf *last_flushed)
 {
 	struct bch_fs *c = trans->c;
 	struct printbuf buf = PRINTBUF;
+	int ret = 0;
 
 	/*
 	 * If we're using the btree write buffer, the backpointer we were
 	 * looking at may have already been deleted - failure to find what it
 	 * pointed to is not an error:
 	 */
-	if (likely(!bch2_backpointers_no_use_write_buffer))
-		return;
+	ret = bch2_backpointers_maybe_flush(trans, bp.s_c, last_flushed);
+	if (ret)
+		return ret;
 
 	prt_printf(&buf, "backpointer doesn't match %s it points to:\n  ",
 		   bp.v->level ? "btree node" : "extent");
@@ -218,18 +210,20 @@ static void backpointer_target_not_found(struct btree_trans *trans,
 	prt_printf(&buf, "\n  ");
 
 	bch2_bkey_val_to_text(&buf, c, target_k);
-	if (c->curr_recovery_pass >= BCH_RECOVERY_PASS_check_extents_to_backpointers)
-		bch_err_ratelimited(c, "%s", buf.buf);
-	else
-		bch2_trans_inconsistent(trans, "%s", buf.buf);
 
+	if (fsck_err(trans, backpointer_to_missing_ptr,
+		     "%s", buf.buf))
+		ret = bch2_backpointer_del(trans, bp.k->p);
+fsck_err:
 	printbuf_exit(&buf);
+	return ret;
 }
 
 struct bkey_s_c bch2_backpointer_get_key(struct btree_trans *trans,
 					 struct bkey_s_c_backpointer bp,
 					 struct btree_iter *iter,
-					 unsigned iter_flags)
+					 unsigned iter_flags,
+					 struct bkey_buf *last_flushed)
 {
 	struct bch_fs *c = trans->c;
 
@@ -253,22 +247,21 @@ struct bkey_s_c bch2_backpointer_get_key(struct btree_trans *trans,
 			return k;
 
 		bch2_trans_iter_exit(trans, iter);
-		backpointer_target_not_found(trans, bp, k);
-		return bkey_s_c_null;
+		int ret = backpointer_target_not_found(trans, bp, k, last_flushed);
+		return ret ? bkey_s_c_err(ret) : bkey_s_c_null;
 	} else {
-		struct btree *b = bch2_backpointer_get_node(trans, bp, iter);
+		struct btree *b = bch2_backpointer_get_node(trans, bp, iter, last_flushed);
+		if (IS_ERR_OR_NULL(b))
+			return ((struct bkey_s_c) { .k = ERR_CAST(b) });
 
-		if (IS_ERR_OR_NULL(b)) {
-			bch2_trans_iter_exit(trans, iter);
-			return IS_ERR(b) ? bkey_s_c_err(PTR_ERR(b)) : bkey_s_c_null;
-		}
 		return bkey_i_to_s_c(&b->key);
 	}
 }
 
 struct btree *bch2_backpointer_get_node(struct btree_trans *trans,
 					struct bkey_s_c_backpointer bp,
-					struct btree_iter *iter)
+					struct btree_iter *iter,
+					struct bkey_buf *last_flushed)
 {
 	struct bch_fs *c = trans->c;
 
@@ -293,8 +286,8 @@ struct btree *bch2_backpointer_get_node(struct btree_trans *trans,
 	if (btree_node_will_make_reachable(b)) {
 		b = ERR_PTR(-BCH_ERR_backpointer_to_overwritten_btree_node);
 	} else {
-		backpointer_target_not_found(trans, bp, bkey_i_to_s_c(&b->key));
-		b = NULL;
+		int ret = backpointer_target_not_found(trans, bp, bkey_i_to_s_c(&b->key), last_flushed);
+		b = ret ? ERR_PTR(ret) : NULL;
 	}
 err:
 	bch2_trans_iter_exit(trans, iter);
@@ -497,7 +490,7 @@ check_existing_bp:
 	struct bkey_s_c_backpointer other_bp = bkey_s_c_to_backpointer(bp_k);
 
 	struct bkey_s_c other_extent =
-		bch2_backpointer_get_key(trans, other_bp, &other_extent_iter, 0);
+		bch2_backpointer_get_key(trans, other_bp, &other_extent_iter, 0, &s->last_flushed);
 	ret = bkey_err(other_extent);
 	if (ret == -BCH_ERR_backpointer_to_overwritten_btree_node)
 		ret = 0;
@@ -578,31 +571,29 @@ static int check_extent_to_backpointers(struct btree_trans *trans,
 					struct bkey_s_c k)
 {
 	struct bch_fs *c = trans->c;
-	struct bkey_ptrs_c ptrs;
+	struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
 	const union bch_extent_entry *entry;
 	struct extent_ptr_decoded p;
-	int ret;
 
-	ptrs = bch2_bkey_ptrs_c(k);
 	bkey_for_each_ptr_decode(k.k, ptrs, p, entry) {
-		struct bpos bucket_pos;
-		struct bkey_i_backpointer bp;
-
 		if (p.ptr.cached)
 			continue;
 
+		if (p.ptr.dev == BCH_SB_MEMBER_INVALID)
+			continue;
+
 		rcu_read_lock();
 		struct bch_dev *ca = bch2_dev_rcu_noerror(c, p.ptr.dev);
-		if (ca)
-			bch2_extent_ptr_to_bp(c, ca, btree, level, k, p, entry, &bucket_pos, &bp);
+		bool check = ca && test_bit(PTR_BUCKET_NR(ca, &p.ptr), ca->bucket_backpointer_mismatches);
 		rcu_read_unlock();
 
-		if (!ca)
-			continue;
-
-		ret = check_bp_exists(trans, s, &bp, k);
-		if (ret)
-			return ret;
+		if (check) {
+			struct bkey_i_backpointer bp;
+			bch2_extent_ptr_to_bp(c, btree, level, k, p, entry, &bp);
+			int ret = check_bp_exists(trans, s, &bp, k);
+			if (ret)
+				return ret;
+		}
 	}
 
 	return 0;
@@ -811,26 +802,259 @@ static int bch2_check_extents_to_backpointers_pass(struct btree_trans *trans,
 	return 0;
 }
 
+enum alloc_sector_counter {
+	ALLOC_dirty,
+	ALLOC_cached,
+	ALLOC_stripe,
+	ALLOC_SECTORS_NR
+};
+
+static enum alloc_sector_counter data_type_to_alloc_counter(enum bch_data_type t)
+{
+	switch (t) {
+	case BCH_DATA_btree:
+	case BCH_DATA_user:
+		return ALLOC_dirty;
+	case BCH_DATA_cached:
+		return ALLOC_cached;
+	case BCH_DATA_stripe:
+		return ALLOC_stripe;
+	default:
+		BUG();
+	}
+}
+
+static int check_bucket_backpointer_mismatch_one(struct btree_trans *trans, struct bkey_s_c alloc_k,
+						 struct bkey_buf *last_flushed)
+{
+	int ret = 0;
+	struct bch_alloc_v4 a_convert;
+	const struct bch_alloc_v4 *a = bch2_alloc_to_v4(alloc_k, &a_convert);
+
+	if (a->data_type == BCH_DATA_sb ||
+	    a->data_type == BCH_DATA_journal ||
+	    a->data_type == BCH_DATA_parity)
+		return 0;
+
+	u32 sectors[ALLOC_SECTORS_NR];
+	memset(sectors, 0, sizeof(sectors));
+
+	struct bch_dev *ca = bch2_dev_bucket_tryget_noerror(trans->c, alloc_k.k->p);
+	if (!ca)
+		return 0;
+
+	struct btree_iter iter;
+	struct bkey_s_c bp_k;
+	for_each_btree_key_max_norestart(trans, iter, BTREE_ID_backpointers,
+				bucket_pos_to_bp_start(ca, alloc_k.k->p),
+				bucket_pos_to_bp_end(ca, alloc_k.k->p), 0, bp_k, ret) {
+		if (bp_k.k->type != KEY_TYPE_backpointer)
+			continue;
+
+		struct bkey_s_c_backpointer bp = bkey_s_c_to_backpointer(bp_k);
+
+		if (bp.v->bucket_gen != a->gen)
+			continue;
+
+		sectors[data_type_to_alloc_counter(bp.v->data_type)] += bp.v->bucket_len;
+	};
+	bch2_trans_iter_exit(trans, &iter);
+	if (ret)
+		goto err;
+
+	/* Cached pointers don't have backpointers: */
+
+	if (sectors[ALLOC_dirty] != a->dirty_sectors ||
+	    sectors[ALLOC_stripe] != a->stripe_sectors) {
+		ret = bch2_backpointers_maybe_flush(trans, alloc_k, last_flushed);
+		if (ret)
+			goto err;
+
+		__set_bit(alloc_k.k->p.offset, ca->bucket_backpointer_mismatches);
+	}
+err:
+	bch2_dev_put(ca);
+	return ret;
+}
+
+static int check_bucket_backpointer_mismatches(struct btree_trans *trans,
+					       struct bkey_buf *last_flushed)
+{
+	return for_each_btree_key(trans, iter, BTREE_ID_alloc,
+				     POS_MIN, BTREE_ITER_prefetch, k, ({
+		check_bucket_backpointer_mismatch_one(trans, k, last_flushed);
+	}));
+}
+
+static bool backpointer_node_has_missing(struct bch_fs *c, struct bkey_s_c k)
+{
+	switch (k.k->type) {
+	case KEY_TYPE_btree_ptr_v2: {
+		bool ret = false;
+
+		rcu_read_lock();
+		struct bpos pos = bkey_s_c_to_btree_ptr_v2(k).v->min_key;
+		while (pos.inode <= k.k->p.inode) {
+			struct bch_dev *ca = bch2_dev_rcu_noerror(c, pos.inode);
+			if (!ca)
+				goto next;
+
+			struct bpos bucket = bp_pos_to_bucket(ca, pos);
+			bucket.offset = find_next_bit(ca->bucket_backpointer_mismatches,
+						      ca->mi.nbuckets, bucket.offset);
+			if (bucket.offset == ca->mi.nbuckets)
+				goto next;
+
+			ret = bpos_le(bucket_pos_to_bp_end(ca, bucket), k.k->p);
+			if (ret)
+				break;
+next:
+			pos = SPOS(pos.inode + 1, 0, 0);
+		}
+		rcu_read_unlock();
+
+		return ret;
+	}
+	case KEY_TYPE_btree_ptr:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static int btree_node_get_and_pin(struct btree_trans *trans, struct bkey_i *k,
+				  enum btree_id btree, unsigned level)
+{
+	struct btree_iter iter;
+	bch2_trans_node_iter_init(trans, &iter, btree, k->k.p, 0, level, 0);
+	struct btree *b = bch2_btree_iter_peek_node(&iter);
+	int ret = PTR_ERR_OR_ZERO(b);
+	if (ret)
+		goto err;
+
+	if (b)
+		bch2_node_pin(trans->c, b);
+err:
+	bch2_trans_iter_exit(trans, &iter);
+	return ret;
+}
+
+static int bch2_pin_backpointer_nodes_with_missing(struct btree_trans *trans,
+					struct bpos start, struct bpos *end)
+{
+	struct bch_fs *c = trans->c;
+	int ret = 0;
+
+	struct bkey_buf tmp;
+	bch2_bkey_buf_init(&tmp);
+
+	bch2_btree_cache_unpin(c);
+
+	*end = SPOS_MAX;
+
+	s64 mem_may_pin = mem_may_pin_bytes(c);
+	struct btree_iter iter;
+	bch2_trans_node_iter_init(trans, &iter, BTREE_ID_backpointers, start,
+				  0, 1, BTREE_ITER_prefetch);
+	ret = for_each_btree_key_continue(trans, iter, 0, k, ({
+		if (!backpointer_node_has_missing(c, k))
+			continue;
+
+		mem_may_pin -= c->opts.btree_node_size;
+		if (mem_may_pin <= 0)
+			break;
+
+		bch2_bkey_buf_reassemble(&tmp, c, k);
+		struct btree_path *path = btree_iter_path(trans, &iter);
+
+		BUG_ON(path->level != 1);
+
+		bch2_btree_node_prefetch(trans, path, tmp.k, path->btree_id, path->level - 1);
+	}));
+	if (ret)
+		return ret;
+
+	struct bpos pinned = SPOS_MAX;
+	mem_may_pin = mem_may_pin_bytes(c);
+	bch2_trans_node_iter_init(trans, &iter, BTREE_ID_backpointers, start,
+				  0, 1, BTREE_ITER_prefetch);
+	ret = for_each_btree_key_continue(trans, iter, 0, k, ({
+		if (!backpointer_node_has_missing(c, k))
+			continue;
+
+		mem_may_pin -= c->opts.btree_node_size;
+		if (mem_may_pin <= 0) {
+			*end = pinned;
+			break;
+		}
+
+		bch2_bkey_buf_reassemble(&tmp, c, k);
+		struct btree_path *path = btree_iter_path(trans, &iter);
+
+		BUG_ON(path->level != 1);
+
+		int ret2 = btree_node_get_and_pin(trans, tmp.k, path->btree_id, path->level - 1);
+
+		if (!ret2)
+			pinned = tmp.k->k.p;
+
+		ret;
+	}));
+	if (ret)
+		return ret;
+
+	return ret;
+}
+
 int bch2_check_extents_to_backpointers(struct bch_fs *c)
 {
+	int ret = 0;
+
+	/*
+	 * Can't allow devices to come/go/resize while we have bucket bitmaps
+	 * allocated
+	 */
+	lockdep_assert_held(&c->state_lock);
+
+	for_each_member_device(c, ca) {
+		BUG_ON(ca->bucket_backpointer_mismatches);
+		ca->bucket_backpointer_mismatches = kvcalloc(BITS_TO_LONGS(ca->mi.nbuckets),
+							     sizeof(unsigned long),
+							     GFP_KERNEL);
+		if (!ca->bucket_backpointer_mismatches) {
+			bch2_dev_put(ca);
+			ret = -BCH_ERR_ENOMEM_backpointer_mismatches_bitmap;
+			goto err_free_bitmaps;
+		}
+	}
+
 	struct btree_trans *trans = bch2_trans_get(c);
 	struct extents_to_bp_state s = { .bp_start = POS_MIN };
-	int ret;
 
 	bch2_bkey_buf_init(&s.last_flushed);
 	bkey_init(&s.last_flushed.k->k);
 
+	ret = check_bucket_backpointer_mismatches(trans, &s.last_flushed);
+	if (ret)
+		goto err;
+
+	u64 nr_buckets = 0, nr_mismatches = 0;
+	for_each_member_device(c, ca) {
+		nr_buckets += ca->mi.nbuckets;
+		nr_mismatches += bitmap_weight(ca->bucket_backpointer_mismatches, ca->mi.nbuckets);
+	}
+
+	if (!nr_mismatches)
+		goto err;
+
+	bch_info(c, "scanning for missing backpointers in %llu/%llu buckets",
+		 nr_mismatches, nr_buckets);
+
 	while (1) {
-		struct bbpos end;
-		ret = bch2_get_btree_in_memory_pos(trans,
-				BIT_ULL(BTREE_ID_backpointers),
-				BIT_ULL(BTREE_ID_backpointers),
-				BBPOS(BTREE_ID_backpointers, s.bp_start), &end);
+		ret = bch2_pin_backpointer_nodes_with_missing(trans, s.bp_start, &s.bp_end);
 		if (ret)
 			break;
 
-		s.bp_end = end.pos;
-
 		if ( bpos_eq(s.bp_start, POS_MIN) &&
 		    !bpos_eq(s.bp_end, SPOS_MAX))
 			bch_verbose(c, "%s(): alloc info does not fit in ram, running in multiple passes with %zu nodes per pass",
@@ -855,10 +1079,15 @@ int bch2_check_extents_to_backpointers(struct bch_fs *c)
 
 		s.bp_start = bpos_successor(s.bp_end);
 	}
+err:
 	bch2_trans_put(trans);
 	bch2_bkey_buf_exit(&s.last_flushed, c);
-
 	bch2_btree_cache_unpin(c);
+err_free_bitmaps:
+	for_each_member_device(c, ca) {
+		kvfree(ca->bucket_backpointer_mismatches);
+		ca->bucket_backpointer_mismatches = NULL;
+	}
 
 	bch_err_fn(c, ret);
 	return ret;
@@ -874,39 +1103,21 @@ static int check_one_backpointer(struct btree_trans *trans,
 		return 0;
 
 	struct bkey_s_c_backpointer bp = bkey_s_c_to_backpointer(bp_k);
-	struct bch_fs *c = trans->c;
 	struct bbpos pos = bp_to_bbpos(*bp.v);
-	struct printbuf buf = PRINTBUF;
 
 	if (bbpos_cmp(pos, start) < 0 ||
 	    bbpos_cmp(pos, end) > 0)
 		return 0;
 
 	struct btree_iter iter;
-	struct bkey_s_c k = bch2_backpointer_get_key(trans, bp, &iter, 0);
+	struct bkey_s_c k = bch2_backpointer_get_key(trans, bp, &iter, 0, last_flushed);
 	int ret = bkey_err(k);
 	if (ret == -BCH_ERR_backpointer_to_overwritten_btree_node)
 		return 0;
 	if (ret)
 		return ret;
 
-	if (!k.k) {
-		ret = bch2_backpointers_maybe_flush(trans, bp.s_c, last_flushed);
-		if (ret)
-			goto out;
-
-		if (fsck_err(trans, backpointer_to_missing_ptr,
-			     "backpointer for missing %s\n  %s",
-			     bp.v->level ? "btree node" : "extent",
-			     (bch2_bkey_val_to_text(&buf, c, bp.s_c), buf.buf))) {
-			ret = bch2_backpointer_del(trans, bp.k->p);
-			goto out;
-		}
-	}
-out:
-fsck_err:
 	bch2_trans_iter_exit(trans, &iter);
-	printbuf_exit(&buf);
 	return ret;
 }
 
@@ -922,9 +1133,8 @@ static int bch2_check_backpointers_to_extents_pass(struct btree_trans *trans,
 	bkey_init(&last_flushed.k->k);
 	progress_init(&progress, trans->c, BIT_ULL(BTREE_ID_backpointers));
 
-	int ret = for_each_btree_key_commit(trans, iter, BTREE_ID_backpointers,
-				  POS_MIN, BTREE_ITER_prefetch, k,
-				  NULL, NULL, BCH_TRANS_COMMIT_no_enospc, ({
+	int ret = for_each_btree_key(trans, iter, BTREE_ID_backpointers,
+				     POS_MIN, BTREE_ITER_prefetch, k, ({
 			progress_update_iter(trans, &progress, &iter, "backpointers_to_extents");
 			check_one_backpointer(trans, start, end, k, &last_flushed);
 	}));
diff --git a/libbcachefs/backpointers.h b/libbcachefs/backpointers.h
index caffc684..060dad15 100644
--- a/libbcachefs/backpointers.h
+++ b/libbcachefs/backpointers.h
@@ -140,45 +140,29 @@ static inline enum bch_data_type bch2_bkey_ptr_data_type(struct bkey_s_c k,
 	}
 }
 
-static inline void __bch2_extent_ptr_to_bp(struct bch_fs *c, struct bch_dev *ca,
+static inline void bch2_extent_ptr_to_bp(struct bch_fs *c,
 			   enum btree_id btree_id, unsigned level,
 			   struct bkey_s_c k, struct extent_ptr_decoded p,
 			   const union bch_extent_entry *entry,
-			   struct bpos *bucket, struct bkey_i_backpointer *bp,
-			   u64 sectors)
+			   struct bkey_i_backpointer *bp)
 {
-	u32 bucket_offset;
-	*bucket = PTR_BUCKET_POS_OFFSET(ca, &p.ptr, &bucket_offset);
-
-	u64 bp_bucket_offset = ((u64) bucket_offset << MAX_EXTENT_COMPRESS_RATIO_SHIFT) + p.crc.offset;
-
 	bkey_backpointer_init(&bp->k_i);
-	bp->k.p = bucket_pos_to_bp(ca, *bucket, bp_bucket_offset);
+	bp->k.p = POS(p.ptr.dev, ((u64) p.ptr.offset << MAX_EXTENT_COMPRESS_RATIO_SHIFT) + p.crc.offset);
 	bp->v	= (struct bch_backpointer) {
 		.btree_id	= btree_id,
 		.level		= level,
 		.data_type	= bch2_bkey_ptr_data_type(k, p, entry),
 		.bucket_gen	= p.ptr.gen,
-		.bucket_len	= sectors,
+		.bucket_len	= ptr_disk_sectors(level ? btree_sectors(c) : k.k->size, p),
 		.pos		= k.k->p,
 	};
 }
 
-static inline void bch2_extent_ptr_to_bp(struct bch_fs *c, struct bch_dev *ca,
-			   enum btree_id btree_id, unsigned level,
-			   struct bkey_s_c k, struct extent_ptr_decoded p,
-			   const union bch_extent_entry *entry,
-			   struct bpos *bucket_pos, struct bkey_i_backpointer *bp)
-{
-	u64 sectors = ptr_disk_sectors(level ? btree_sectors(c) : k.k->size, p);
-
-	__bch2_extent_ptr_to_bp(c, ca, btree_id, level, k, p, entry, bucket_pos, bp, sectors);
-}
-
+struct bkey_buf;
 struct bkey_s_c bch2_backpointer_get_key(struct btree_trans *, struct bkey_s_c_backpointer,
-					 struct btree_iter *, unsigned);
+					 struct btree_iter *, unsigned, struct bkey_buf *);
 struct btree *bch2_backpointer_get_node(struct btree_trans *, struct bkey_s_c_backpointer,
-					struct btree_iter *);
+					struct btree_iter *, struct bkey_buf *);
 
 int bch2_check_btree_backpointers(struct bch_fs *);
 int bch2_check_extents_to_backpointers(struct bch_fs *);
diff --git a/libbcachefs/bcachefs.h b/libbcachefs/bcachefs.h
index b12c9c78..58d42b3f 100644
--- a/libbcachefs/bcachefs.h
+++ b/libbcachefs/bcachefs.h
@@ -557,6 +557,8 @@ struct bch_dev {
 	unsigned long		*buckets_nouse;
 	struct rw_semaphore	bucket_lock;
 
+	unsigned long		*bucket_backpointer_mismatches;
+
 	struct bch_dev_usage __percpu	*usage;
 
 	/* Allocator: */
@@ -759,6 +761,8 @@ struct bch_fs {
 		__uuid_t	user_uuid;
 
 		u16		version;
+		u16		version_incompat;
+		u16		version_incompat_allowed;
 		u16		version_min;
 		u16		version_upgrade_complete;
 
diff --git a/libbcachefs/bcachefs_format.h b/libbcachefs/bcachefs_format.h
index cef22c15..f140c336 100644
--- a/libbcachefs/bcachefs_format.h
+++ b/libbcachefs/bcachefs_format.h
@@ -680,7 +680,9 @@ struct bch_sb_field_ext {
 	x(rebalance_work_acct_fix,	BCH_VERSION(1, 12))		\
 	x(inode_has_child_snapshots,	BCH_VERSION(1, 13))		\
 	x(backpointer_bucket_gen,	BCH_VERSION(1, 14))		\
-	x(disk_accounting_big_endian,	BCH_VERSION(1, 15))
+	x(disk_accounting_big_endian,	BCH_VERSION(1, 15))		\
+	x(reflink_p_may_update_opts,	BCH_VERSION(1, 16))		\
+	x(inode_depth,			BCH_VERSION(1, 17))
 
 enum bcachefs_metadata_version {
 	bcachefs_metadata_version_min = 9,
@@ -845,6 +847,9 @@ LE64_BITMASK(BCH_SB_VERSION_UPGRADE_COMPLETE,
 					struct bch_sb, flags[5],  0, 16);
 LE64_BITMASK(BCH_SB_ALLOCATOR_STUCK_TIMEOUT,
 					struct bch_sb, flags[5], 16, 32);
+LE64_BITMASK(BCH_SB_VERSION_INCOMPAT,	struct bch_sb, flags[5], 32, 48);
+LE64_BITMASK(BCH_SB_VERSION_INCOMPAT_ALLOWED,
+					struct bch_sb, flags[5], 48, 64);
 
 static inline __u64 BCH_SB_COMPRESSION_TYPE(const struct bch_sb *sb)
 {
@@ -897,21 +902,22 @@ static inline void SET_BCH_SB_BACKGROUND_COMPRESSION_TYPE(struct bch_sb *sb, __u
 	x(new_varint,			15)	\
 	x(journal_no_flush,		16)	\
 	x(alloc_v2,			17)	\
-	x(extents_across_btree_nodes,	18)
+	x(extents_across_btree_nodes,	18)	\
+	x(incompat_version_field,	19)
 
 #define BCH_SB_FEATURES_ALWAYS				\
-	((1ULL << BCH_FEATURE_new_extent_overwrite)|	\
-	 (1ULL << BCH_FEATURE_extents_above_btree_updates)|\
-	 (1ULL << BCH_FEATURE_btree_updates_journalled)|\
-	 (1ULL << BCH_FEATURE_alloc_v2)|\
-	 (1ULL << BCH_FEATURE_extents_across_btree_nodes))
+	(BIT_ULL(BCH_FEATURE_new_extent_overwrite)|	\
+	 BIT_ULL(BCH_FEATURE_extents_above_btree_updates)|\
+	 BIT_ULL(BCH_FEATURE_btree_updates_journalled)|\
+	 BIT_ULL(BCH_FEATURE_alloc_v2)|\
+	 BIT_ULL(BCH_FEATURE_extents_across_btree_nodes))
 
 #define BCH_SB_FEATURES_ALL				\
 	(BCH_SB_FEATURES_ALWAYS|			\
-	 (1ULL << BCH_FEATURE_new_siphash)|		\
-	 (1ULL << BCH_FEATURE_btree_ptr_v2)|		\
-	 (1ULL << BCH_FEATURE_new_varint)|		\
-	 (1ULL << BCH_FEATURE_journal_no_flush))
+	 BIT_ULL(BCH_FEATURE_new_siphash)|		\
+	 BIT_ULL(BCH_FEATURE_btree_ptr_v2)|		\
+	 BIT_ULL(BCH_FEATURE_new_varint)|		\
+	 BIT_ULL(BCH_FEATURE_journal_no_flush))
 
 enum bch_sb_feature {
 #define x(f, n) BCH_FEATURE_##f,
diff --git a/libbcachefs/btree_cache.c b/libbcachefs/btree_cache.c
index 1117be90..b00c6a20 100644
--- a/libbcachefs/btree_cache.c
+++ b/libbcachefs/btree_cache.c
@@ -222,7 +222,6 @@ void bch2_node_pin(struct bch_fs *c, struct btree *b)
 	struct btree_cache *bc = &c->btree_cache;
 
 	mutex_lock(&bc->lock);
-	BUG_ON(!__btree_node_pinned(bc, b));
 	if (b != btree_node_root(c, b) && !btree_node_pinned(b)) {
 		set_btree_node_pinned(b);
 		list_move(&b->list, &bc->live[1].list);
diff --git a/libbcachefs/btree_journal_iter.c b/libbcachefs/btree_journal_iter.c
index de3db161..6d25e3f8 100644
--- a/libbcachefs/btree_journal_iter.c
+++ b/libbcachefs/btree_journal_iter.c
@@ -259,7 +259,7 @@ int bch2_journal_key_insert_take(struct bch_fs *c, enum btree_id id,
 		 * Ensure these keys are done last by journal replay, to unblock
 		 * journal reclaim:
 		 */
-		.journal_seq	= U32_MAX,
+		.journal_seq	= U64_MAX,
 	};
 	struct journal_keys *keys = &c->journal_keys;
 	size_t idx = bch2_journal_key_search(keys, id, level, k->k.p);
diff --git a/libbcachefs/buckets.c b/libbcachefs/buckets.c
index afd35c93..1e43ece2 100644
--- a/libbcachefs/buckets.c
+++ b/libbcachefs/buckets.c
@@ -574,8 +574,10 @@ static int bch2_trigger_pointer(struct btree_trans *trans,
 	struct printbuf buf = PRINTBUF;
 	int ret = 0;
 
-	u64 abs_sectors = ptr_disk_sectors(level ? btree_sectors(c) : k.k->size, p);
-	*sectors = insert ? abs_sectors : -abs_sectors;
+	struct bkey_i_backpointer bp;
+	bch2_extent_ptr_to_bp(c, btree_id, level, k, p, entry, &bp);
+
+	*sectors = insert ? bp.v.bucket_len : -(s64) bp.v.bucket_len;
 
 	struct bch_dev *ca = bch2_dev_tryget(c, p.ptr.dev);
 	if (unlikely(!ca)) {
@@ -584,9 +586,7 @@ static int bch2_trigger_pointer(struct btree_trans *trans,
 		goto err;
 	}
 
-	struct bpos bucket;
-	struct bkey_i_backpointer bp;
-	__bch2_extent_ptr_to_bp(trans->c, ca, btree_id, level, k, p, entry, &bucket, &bp, abs_sectors);
+	struct bpos bucket = PTR_BUCKET_POS(ca, &p.ptr);
 
 	if (flags & BTREE_TRIGGER_transactional) {
 		struct bkey_i_alloc_v4 *a = bch2_trans_start_alloc_update(trans, bucket, 0);
diff --git a/libbcachefs/ec.c b/libbcachefs/ec.c
index 250e7389..6b297f90 100644
--- a/libbcachefs/ec.c
+++ b/libbcachefs/ec.c
@@ -1276,7 +1276,8 @@ static int ec_stripe_update_extent(struct btree_trans *trans,
 				   struct bch_dev *ca,
 				   struct bpos bucket, u8 gen,
 				   struct ec_stripe_buf *s,
-				   struct bkey_s_c_backpointer bp)
+				   struct bkey_s_c_backpointer bp,
+				   struct bkey_buf *last_flushed)
 {
 	struct bch_stripe *v = &bkey_i_to_stripe(&s->key)->v;
 	struct bch_fs *c = trans->c;
@@ -1293,7 +1294,7 @@ static int ec_stripe_update_extent(struct btree_trans *trans,
 		struct btree_iter node_iter;
 		struct btree *b;
 
-		b = bch2_backpointer_get_node(trans, bp, &node_iter);
+		b = bch2_backpointer_get_node(trans, bp, &node_iter, last_flushed);
 		bch2_trans_iter_exit(trans, &node_iter);
 
 		if (!b)
@@ -1307,7 +1308,7 @@ static int ec_stripe_update_extent(struct btree_trans *trans,
 		return -EIO;
 	}
 
-	k = bch2_backpointer_get_key(trans, bp, &iter, BTREE_ITER_intent);
+	k = bch2_backpointer_get_key(trans, bp, &iter, BTREE_ITER_intent, last_flushed);
 	ret = bkey_err(k);
 	if (ret)
 		return ret;
@@ -1374,6 +1375,10 @@ static int ec_stripe_update_bucket(struct btree_trans *trans, struct ec_stripe_b
 
 	struct bpos bucket_pos = PTR_BUCKET_POS(ca, &ptr);
 
+	struct bkey_buf last_flushed;
+	bch2_bkey_buf_init(&last_flushed);
+	bkey_init(&last_flushed.k->k);
+
 	ret = for_each_btree_key_max_commit(trans, bp_iter, BTREE_ID_backpointers,
 			bucket_pos_to_bp_start(ca, bucket_pos),
 			bucket_pos_to_bp_end(ca, bucket_pos), 0, bp_k,
@@ -1387,9 +1392,10 @@ static int ec_stripe_update_bucket(struct btree_trans *trans, struct ec_stripe_b
 			continue;
 
 		ec_stripe_update_extent(trans, ca, bucket_pos, ptr.gen, s,
-					bkey_s_c_to_backpointer(bp_k));
+					bkey_s_c_to_backpointer(bp_k), &last_flushed);
 	}));
 
+	bch2_bkey_buf_exit(&last_flushed, c);
 	bch2_dev_put(ca);
 	return ret;
 }
diff --git a/libbcachefs/errcode.h b/libbcachefs/errcode.h
index 47387f7d..924d2400 100644
--- a/libbcachefs/errcode.h
+++ b/libbcachefs/errcode.h
@@ -54,6 +54,7 @@
 	x(ENOMEM,			ENOMEM_compression_bounce_read_init)	\
 	x(ENOMEM,			ENOMEM_compression_bounce_write_init)	\
 	x(ENOMEM,			ENOMEM_compression_workspace_init)	\
+	x(ENOMEM,			ENOMEM_backpointer_mismatches_bitmap)	\
 	x(EIO,				compression_workspace_not_initialized)	\
 	x(ENOMEM,			ENOMEM_bucket_gens)			\
 	x(ENOMEM,			ENOMEM_buckets_nouse)			\
diff --git a/libbcachefs/fs-common.c b/libbcachefs/fs-common.c
index dcaa47f6..d2966b67 100644
--- a/libbcachefs/fs-common.c
+++ b/libbcachefs/fs-common.c
@@ -172,6 +172,10 @@ int bch2_create_trans(struct btree_trans *trans,
 		new_inode->bi_dir_offset	= dir_offset;
 	}
 
+	if (S_ISDIR(mode) &&
+	    !new_inode->bi_subvol)
+		new_inode->bi_depth = dir_u->bi_depth + 1;
+
 	inode_iter.flags &= ~BTREE_ITER_all_snapshots;
 	bch2_btree_iter_set_snapshot(&inode_iter, snapshot);
 
@@ -512,6 +516,15 @@ int bch2_rename_trans(struct btree_trans *trans,
 		dst_dir_u->bi_nlink++;
 	}
 
+	if (S_ISDIR(src_inode_u->bi_mode) &&
+	    !src_inode_u->bi_subvol)
+		src_inode_u->bi_depth = dst_dir_u->bi_depth + 1;
+
+	if (mode == BCH_RENAME_EXCHANGE &&
+	    S_ISDIR(dst_inode_u->bi_mode) &&
+	    !dst_inode_u->bi_subvol)
+		dst_inode_u->bi_depth = src_dir_u->bi_depth + 1;
+
 	if (dst_inum.inum && is_subdir_for_nlink(dst_inode_u)) {
 		dst_dir_u->bi_nlink--;
 		src_dir_u->bi_nlink += mode == BCH_RENAME_EXCHANGE;
diff --git a/libbcachefs/fs-io.c b/libbcachefs/fs-io.c
index 33d0e708..94bf34b9 100644
--- a/libbcachefs/fs-io.c
+++ b/libbcachefs/fs-io.c
@@ -906,11 +906,18 @@ loff_t bch2_remap_file_range(struct file *file_src, loff_t pos_src,
 	bch2_mark_pagecache_unallocated(src, pos_src >> 9,
 				   (pos_src + aligned_len) >> 9);
 
+	/*
+	 * XXX: we'd like to be telling bch2_remap_range() if we have
+	 * permission to write to the source file, and thus if io path option
+	 * changes should be propagated through the copy, but we need mnt_idmap
+	 * from the pathwalk, awkward
+	 */
 	ret = bch2_remap_range(c,
 			       inode_inum(dst), pos_dst >> 9,
 			       inode_inum(src), pos_src >> 9,
 			       aligned_len >> 9,
-			       pos_dst + len, &i_sectors_delta);
+			       pos_dst + len, &i_sectors_delta,
+			       false);
 	if (ret < 0)
 		goto err;
 
diff --git a/libbcachefs/fsck.c b/libbcachefs/fsck.c
index 1a5a0711..d67a3f52 100644
--- a/libbcachefs/fsck.c
+++ b/libbcachefs/fsck.c
@@ -2791,6 +2791,48 @@ struct pathbuf_entry {
 
 typedef DARRAY(struct pathbuf_entry) pathbuf;
 
+static int bch2_bi_depth_renumber_one(struct btree_trans *trans, struct pathbuf_entry *p,
+				      u32 new_depth)
+{
+	struct btree_iter iter;
+	struct bkey_s_c k = bch2_bkey_get_iter(trans, &iter, BTREE_ID_inodes,
+					       SPOS(0, p->inum, p->snapshot), 0);
+
+	struct bch_inode_unpacked inode;
+	int ret = bkey_err(k) ?:
+		!bkey_is_inode(k.k) ? -BCH_ERR_ENOENT_inode
+		: bch2_inode_unpack(k, &inode);
+	if (ret)
+		goto err;
+
+	if (inode.bi_depth != new_depth) {
+		inode.bi_depth = new_depth;
+		ret = __bch2_fsck_write_inode(trans, &inode) ?:
+			bch2_trans_commit(trans, NULL, NULL, 0);
+	}
+err:
+	bch2_trans_iter_exit(trans, &iter);
+	return ret;
+}
+
+static int bch2_bi_depth_renumber(struct btree_trans *trans, pathbuf *path, u32 new_bi_depth)
+{
+	u32 restart_count = trans->restart_count;
+	int ret = 0;
+
+	darray_for_each_reverse(*path, i) {
+		ret = nested_lockrestart_do(trans,
+				bch2_bi_depth_renumber_one(trans, i, new_bi_depth));
+		bch_err_fn(trans->c, ret);
+		if (ret)
+			break;
+
+		new_bi_depth++;
+	}
+
+	return ret ?: trans_was_restarted(trans, restart_count);
+}
+
 static bool path_is_dup(pathbuf *p, u64 inum, u32 snapshot)
 {
 	darray_for_each(*p, i)
@@ -2800,22 +2842,20 @@ static bool path_is_dup(pathbuf *p, u64 inum, u32 snapshot)
 	return false;
 }
 
-static int check_path(struct btree_trans *trans, pathbuf *p, struct bkey_s_c inode_k)
+static int check_path_loop(struct btree_trans *trans, struct bkey_s_c inode_k)
 {
 	struct bch_fs *c = trans->c;
 	struct btree_iter inode_iter = {};
-	struct bch_inode_unpacked inode;
+	pathbuf path = {};
 	struct printbuf buf = PRINTBUF;
 	u32 snapshot = inode_k.k->p.snapshot;
+	bool redo_bi_depth = false;
+	u32 min_bi_depth = U32_MAX;
 	int ret = 0;
 
-	p->nr = 0;
-
+	struct bch_inode_unpacked inode;
 	BUG_ON(bch2_inode_unpack(inode_k, &inode));
 
-	if (!S_ISDIR(inode.bi_mode))
-		return 0;
-
 	while (!inode.bi_subvol) {
 		struct btree_iter dirent_iter;
 		struct bkey_s_c_dirent d;
@@ -2839,7 +2879,7 @@ static int check_path(struct btree_trans *trans, pathbuf *p, struct bkey_s_c ino
 
 		bch2_trans_iter_exit(trans, &dirent_iter);
 
-		ret = darray_push(p, ((struct pathbuf_entry) {
+		ret = darray_push(&path, ((struct pathbuf_entry) {
 			.inum		= inode.bi_inum,
 			.snapshot	= snapshot,
 		}));
@@ -2851,22 +2891,32 @@ static int check_path(struct btree_trans *trans, pathbuf *p, struct bkey_s_c ino
 		bch2_trans_iter_exit(trans, &inode_iter);
 		inode_k = bch2_bkey_get_iter(trans, &inode_iter, BTREE_ID_inodes,
 					     SPOS(0, inode.bi_dir, snapshot), 0);
+
+		struct bch_inode_unpacked parent_inode;
 		ret = bkey_err(inode_k) ?:
 			!bkey_is_inode(inode_k.k) ? -BCH_ERR_ENOENT_inode
-			: bch2_inode_unpack(inode_k, &inode);
+			: bch2_inode_unpack(inode_k, &parent_inode);
 		if (ret) {
 			/* Should have been caught in dirents pass */
 			bch_err_msg(c, ret, "error looking up parent directory");
 			break;
 		}
 
+		min_bi_depth = parent_inode.bi_depth;
+
+		if (parent_inode.bi_depth < inode.bi_depth &&
+		    min_bi_depth < U16_MAX)
+			break;
+
+		inode = parent_inode;
 		snapshot = inode_k.k->p.snapshot;
+		redo_bi_depth = true;
 
-		if (path_is_dup(p, inode.bi_inum, snapshot)) {
+		if (path_is_dup(&path, inode.bi_inum, snapshot)) {
 			/* XXX print path */
 			bch_err(c, "directory structure loop");
 
-			darray_for_each(*p, i)
+			darray_for_each(path, i)
 				pr_err("%llu:%u", i->inum, i->snapshot);
 			pr_err("%llu:%u", inode.bi_inum, snapshot);
 
@@ -2879,12 +2929,21 @@ static int check_path(struct btree_trans *trans, pathbuf *p, struct bkey_s_c ino
 				ret = reattach_inode(trans, &inode);
 				bch_err_msg(c, ret, "reattaching inode %llu", inode.bi_inum);
 			}
+
+			redo_bi_depth = false;
 			break;
 		}
 	}
+
+	if (inode.bi_subvol)
+		min_bi_depth = 0;
+
+	if (redo_bi_depth)
+		ret = bch2_bi_depth_renumber(trans, &path, min_bi_depth);
 out:
 fsck_err:
 	bch2_trans_iter_exit(trans, &inode_iter);
+	darray_exit(&path);
 	printbuf_exit(&buf);
 	bch_err_fn(c, ret);
 	return ret;
@@ -2896,24 +2955,20 @@ fsck_err:
  */
 int bch2_check_directory_structure(struct bch_fs *c)
 {
-	pathbuf path = { 0, };
-	int ret;
-
-	ret = bch2_trans_run(c,
+	int ret = bch2_trans_run(c,
 		for_each_btree_key_commit(trans, iter, BTREE_ID_inodes, POS_MIN,
 					  BTREE_ITER_intent|
 					  BTREE_ITER_prefetch|
 					  BTREE_ITER_all_snapshots, k,
 					  NULL, NULL, BCH_TRANS_COMMIT_no_enospc, ({
-			if (!bkey_is_inode(k.k))
+			if (!S_ISDIR(bkey_inode_mode(k)))
 				continue;
 
 			if (bch2_inode_flags(k) & BCH_INODE_unlinked)
 				continue;
 
-			check_path(trans, &path, k);
+			check_path_loop(trans, k);
 		})));
-	darray_exit(&path);
 
 	bch_err_fn(c, ret);
 	return ret;
diff --git a/libbcachefs/inode.h b/libbcachefs/inode.h
index 927c8759..5bca6950 100644
--- a/libbcachefs/inode.h
+++ b/libbcachefs/inode.h
@@ -219,6 +219,20 @@ static inline u32 bch2_inode_flags(struct bkey_s_c k)
 	}
 }
 
+static inline unsigned bkey_inode_mode(struct bkey_s_c k)
+{
+	switch (k.k->type) {
+	case KEY_TYPE_inode:
+		return le16_to_cpu(bkey_s_c_to_inode(k).v->bi_mode);
+	case KEY_TYPE_inode_v2:
+		return le16_to_cpu(bkey_s_c_to_inode_v2(k).v->bi_mode);
+	case KEY_TYPE_inode_v3:
+		return INODEv3_MODE(bkey_s_c_to_inode_v3(k).v);
+	default:
+		return 0;
+	}
+}
+
 /* i_nlink: */
 
 static inline unsigned nlink_bias(umode_t mode)
diff --git a/libbcachefs/inode_format.h b/libbcachefs/inode_format.h
index 7928d0c6..be1e7476 100644
--- a/libbcachefs/inode_format.h
+++ b/libbcachefs/inode_format.h
@@ -101,7 +101,8 @@ struct bch_inode_generation {
 	x(bi_dir_offset,		64)	\
 	x(bi_subvol,			32)	\
 	x(bi_parent_subvol,		32)	\
-	x(bi_nocow,			8)
+	x(bi_nocow,			8)	\
+	x(bi_depth,			32)
 
 /* subset of BCH_INODE_FIELDS */
 #define BCH_INODE_OPTS()			\
diff --git a/libbcachefs/journal_io.c b/libbcachefs/journal_io.c
index 1627f3e1..bb69d808 100644
--- a/libbcachefs/journal_io.c
+++ b/libbcachefs/journal_io.c
@@ -2036,8 +2036,9 @@ CLOSURE_CALLBACK(bch2_journal_write)
 		struct printbuf buf = PRINTBUF;
 		buf.atomic++;
 
-		prt_printf(&buf, bch2_fmt(c, "Unable to allocate journal write at seq %llu: %s"),
+		prt_printf(&buf, bch2_fmt(c, "Unable to allocate journal write at seq %llu for %zu sectors: %s"),
 					  le64_to_cpu(w->data->seq),
+					  vstruct_sectors(w->data, c->block_bits),
 					  bch2_err_str(ret));
 		__bch2_journal_debug_to_text(&buf, j);
 		spin_unlock(&j->lock);
diff --git a/libbcachefs/journal_reclaim.c b/libbcachefs/journal_reclaim.c
index 1aabbbe3..b7936ad3 100644
--- a/libbcachefs/journal_reclaim.c
+++ b/libbcachefs/journal_reclaim.c
@@ -140,6 +140,7 @@ static struct journal_space __journal_space_available(struct journal *j, unsigne
 	struct bch_fs *c = container_of(j, struct bch_fs, journal);
 	unsigned pos, nr_devs = 0;
 	struct journal_space space, dev_space[BCH_SB_MEMBERS_MAX];
+	unsigned min_bucket_size = U32_MAX;
 
 	BUG_ON(nr_devs_want > ARRAY_SIZE(dev_space));
 
@@ -148,6 +149,8 @@ static struct journal_space __journal_space_available(struct journal *j, unsigne
 		if (!ca->journal.nr)
 			continue;
 
+		min_bucket_size = min(min_bucket_size, ca->mi.bucket_size);
+
 		space = journal_dev_space_available(j, ca, from);
 		if (!space.next_entry)
 			continue;
@@ -167,7 +170,9 @@ static struct journal_space __journal_space_available(struct journal *j, unsigne
 	 * We sorted largest to smallest, and we want the smallest out of the
 	 * @nr_devs_want largest devices:
 	 */
-	return dev_space[nr_devs_want - 1];
+	space = dev_space[nr_devs_want - 1];
+	space.next_entry = min(space.next_entry, min_bucket_size);
+	return space;
 }
 
 void bch2_journal_space_available(struct journal *j)
diff --git a/libbcachefs/logged_ops.c b/libbcachefs/logged_ops.c
index 60e00702..1ac51af1 100644
--- a/libbcachefs/logged_ops.c
+++ b/libbcachefs/logged_ops.c
@@ -63,8 +63,9 @@ fsck_err:
 int bch2_resume_logged_ops(struct bch_fs *c)
 {
 	int ret = bch2_trans_run(c,
-		for_each_btree_key(trans, iter,
-				   BTREE_ID_logged_ops, POS_MIN,
+		for_each_btree_key_max(trans, iter,
+				   BTREE_ID_logged_ops,
+				   POS(LOGGED_OPS_INUM, 0), POS(LOGGED_OPS_INUM, U64_MAX),
 				   BTREE_ITER_prefetch, k,
 			resume_logged_op(trans, &iter, k)));
 	bch_err_fn(c, ret);
@@ -74,9 +75,8 @@ int bch2_resume_logged_ops(struct bch_fs *c)
 static int __bch2_logged_op_start(struct btree_trans *trans, struct bkey_i *k)
 {
 	struct btree_iter iter;
-	int ret;
-
-	ret = bch2_bkey_get_empty_slot(trans, &iter, BTREE_ID_logged_ops, POS_MAX);
+	int ret = bch2_bkey_get_empty_slot(trans, &iter,
+				 BTREE_ID_logged_ops, POS(LOGGED_OPS_INUM, U64_MAX));
 	if (ret)
 		return ret;
 
diff --git a/libbcachefs/logged_ops_format.h b/libbcachefs/logged_ops_format.h
index 6a4bf712..0b370a96 100644
--- a/libbcachefs/logged_ops_format.h
+++ b/libbcachefs/logged_ops_format.h
@@ -2,6 +2,8 @@
 #ifndef _BCACHEFS_LOGGED_OPS_FORMAT_H
 #define _BCACHEFS_LOGGED_OPS_FORMAT_H
 
+#define LOGGED_OPS_INUM		0
+
 struct bch_logged_op_truncate {
 	struct bch_val		v;
 	__le32			subvol;
diff --git a/libbcachefs/move.c b/libbcachefs/move.c
index 46017546..a50f27f4 100644
--- a/libbcachefs/move.c
+++ b/libbcachefs/move.c
@@ -22,6 +22,7 @@
 #include "keylist.h"
 #include "move.h"
 #include "rebalance.h"
+#include "reflink.h"
 #include "replicas.h"
 #include "snapshot.h"
 #include "super-io.h"
@@ -389,6 +390,7 @@ err:
 
 static struct bch_io_opts *bch2_move_get_io_opts(struct btree_trans *trans,
 			  struct per_snapshot_io_opts *io_opts,
+			  struct bpos extent_pos, /* extent_iter, extent_k may be in reflink btree */
 			  struct btree_iter *extent_iter,
 			  struct bkey_s_c extent_k)
 {
@@ -400,12 +402,12 @@ static struct bch_io_opts *bch2_move_get_io_opts(struct btree_trans *trans,
 	if (extent_k.k->type == KEY_TYPE_reflink_v)
 		goto out;
 
-	if (io_opts->cur_inum != extent_k.k->p.inode) {
+	if (io_opts->cur_inum != extent_pos.inode) {
 		io_opts->d.nr = 0;
 
-		ret = for_each_btree_key(trans, iter, BTREE_ID_inodes, POS(0, extent_k.k->p.inode),
+		ret = for_each_btree_key(trans, iter, BTREE_ID_inodes, POS(0, extent_pos.inode),
 					 BTREE_ITER_all_snapshots, k, ({
-			if (k.k->p.offset != extent_k.k->p.inode)
+			if (k.k->p.offset != extent_pos.inode)
 				break;
 
 			if (!bkey_is_inode(k.k))
@@ -419,7 +421,7 @@ static struct bch_io_opts *bch2_move_get_io_opts(struct btree_trans *trans,
 
 			darray_push(&io_opts->d, e);
 		}));
-		io_opts->cur_inum = extent_k.k->p.inode;
+		io_opts->cur_inum = extent_pos.inode;
 	}
 
 	ret = ret ?: trans_was_restarted(trans, restart_count);
@@ -525,9 +527,15 @@ static int bch2_move_data_btree(struct moving_context *ctxt,
 	struct per_snapshot_io_opts snapshot_io_opts;
 	struct bch_io_opts *io_opts;
 	struct bkey_buf sk;
-	struct btree_iter iter;
+	struct btree_iter iter, reflink_iter = {};
 	struct bkey_s_c k;
 	struct data_update_opts data_opts;
+	/*
+	 * If we're moving a single file, also process reflinked data it points
+	 * to (this includes propagating changed io_opts from the inode to the
+	 * extent):
+	 */
+	bool walk_indirect = start.inode == end.inode;
 	int ret = 0, ret2;
 
 	per_snapshot_io_opts_init(&snapshot_io_opts, c);
@@ -547,6 +555,8 @@ static int bch2_move_data_btree(struct moving_context *ctxt,
 		bch2_ratelimit_reset(ctxt->rate);
 
 	while (!bch2_move_ratelimit(ctxt)) {
+		struct btree_iter *extent_iter = &iter;
+
 		bch2_trans_begin(trans);
 
 		k = bch2_btree_iter_peek(&iter);
@@ -565,10 +575,36 @@ static int bch2_move_data_btree(struct moving_context *ctxt,
 		if (ctxt->stats)
 			ctxt->stats->pos = BBPOS(iter.btree_id, iter.pos);
 
+		if (walk_indirect &&
+		    k.k->type == KEY_TYPE_reflink_p &&
+		    REFLINK_P_MAY_UPDATE_OPTIONS(bkey_s_c_to_reflink_p(k).v)) {
+			struct bkey_s_c_reflink_p p = bkey_s_c_to_reflink_p(k);
+			s64 offset_into_extent	= iter.pos.offset - bkey_start_offset(k.k);
+
+			bch2_trans_iter_exit(trans, &reflink_iter);
+			k = bch2_lookup_indirect_extent(trans, &reflink_iter, &offset_into_extent, p, true, 0);
+			ret = bkey_err(k);
+			if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
+				continue;
+			if (ret)
+				break;
+
+			if (bkey_deleted(k.k))
+				goto next_nondata;
+
+			/*
+			 * XXX: reflink pointers may point to multiple indirect
+			 * extents, so don't advance past the entire reflink
+			 * pointer - need to fixup iter->k
+			 */
+			extent_iter = &reflink_iter;
+		}
+
 		if (!bkey_extent_is_direct_data(k.k))
 			goto next_nondata;
 
-		io_opts = bch2_move_get_io_opts(trans, &snapshot_io_opts, &iter, k);
+		io_opts = bch2_move_get_io_opts(trans, &snapshot_io_opts,
+						iter.pos, extent_iter, k);
 		ret = PTR_ERR_OR_ZERO(io_opts);
 		if (ret)
 			continue;
@@ -584,7 +620,7 @@ static int bch2_move_data_btree(struct moving_context *ctxt,
 		bch2_bkey_buf_reassemble(&sk, c, k);
 		k = bkey_i_to_s_c(sk.k);
 
-		ret2 = bch2_move_extent(ctxt, NULL, &iter, k, *io_opts, data_opts);
+		ret2 = bch2_move_extent(ctxt, NULL, extent_iter, k, *io_opts, data_opts);
 		if (ret2) {
 			if (bch2_err_matches(ret2, BCH_ERR_transaction_restart))
 				continue;
@@ -605,6 +641,7 @@ next_nondata:
 		bch2_btree_iter_advance(&iter);
 	}
 
+	bch2_trans_iter_exit(trans, &reflink_iter);
 	bch2_trans_iter_exit(trans, &iter);
 	bch2_bkey_buf_exit(&sk, c);
 	per_snapshot_io_opts_exit(&snapshot_io_opts);
@@ -675,6 +712,7 @@ int bch2_evacuate_bucket(struct moving_context *ctxt,
 	struct bkey_s_c k;
 	struct data_update_opts data_opts;
 	unsigned sectors_moved = 0;
+	struct bkey_buf last_flushed;
 	int ret = 0;
 
 	struct bch_dev *ca = bch2_dev_tryget(c, bucket.inode);
@@ -683,6 +721,8 @@ int bch2_evacuate_bucket(struct moving_context *ctxt,
 
 	trace_bucket_evacuate(c, &bucket);
 
+	bch2_bkey_buf_init(&last_flushed);
+	bkey_init(&last_flushed.k->k);
 	bch2_bkey_buf_init(&sk);
 
 	/*
@@ -724,7 +764,7 @@ int bch2_evacuate_bucket(struct moving_context *ctxt,
 		struct bkey_s_c_backpointer bp = bkey_s_c_to_backpointer(k);
 
 		if (!bp.v->level) {
-			k = bch2_backpointer_get_key(trans, bp, &iter, 0);
+			k = bch2_backpointer_get_key(trans, bp, &iter, 0, &last_flushed);
 			ret = bkey_err(k);
 			if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
 				continue;
@@ -782,7 +822,7 @@ int bch2_evacuate_bucket(struct moving_context *ctxt,
 		} else {
 			struct btree *b;
 
-			b = bch2_backpointer_get_node(trans, bp, &iter);
+			b = bch2_backpointer_get_node(trans, bp, &iter, &last_flushed);
 			ret = PTR_ERR_OR_ZERO(b);
 			if (ret == -BCH_ERR_backpointer_to_overwritten_btree_node)
 				continue;
@@ -820,6 +860,7 @@ err:
 	bch2_trans_iter_exit(trans, &bp_iter);
 	bch2_dev_put(ca);
 	bch2_bkey_buf_exit(&sk, c);
+	bch2_bkey_buf_exit(&last_flushed, c);
 	return ret;
 }
 
diff --git a/libbcachefs/rcu_pending.c b/libbcachefs/rcu_pending.c
index 67522aa3..bef2aa1b 100644
--- a/libbcachefs/rcu_pending.c
+++ b/libbcachefs/rcu_pending.c
@@ -25,21 +25,37 @@ enum rcu_pending_special {
 #define RCU_PENDING_KVFREE_FN		((rcu_pending_process_fn) (ulong) RCU_PENDING_KVFREE)
 #define RCU_PENDING_CALL_RCU_FN		((rcu_pending_process_fn) (ulong) RCU_PENDING_CALL_RCU)
 
-static inline unsigned long __get_state_synchronize_rcu(struct srcu_struct *ssp)
+#ifdef __KERNEL__
+typedef unsigned long			rcu_gp_poll_state_t;
+
+static inline bool rcu_gp_poll_cookie_eq(rcu_gp_poll_state_t l, rcu_gp_poll_state_t r)
+{
+	return l == r;
+}
+#else
+typedef struct urcu_gp_poll_state	rcu_gp_poll_state_t;
+
+static inline bool rcu_gp_poll_cookie_eq(rcu_gp_poll_state_t l, rcu_gp_poll_state_t r)
+{
+	return l.grace_period_id == r.grace_period_id;
+}
+#endif
+
+static inline rcu_gp_poll_state_t __get_state_synchronize_rcu(struct srcu_struct *ssp)
 {
 	return ssp
 		? get_state_synchronize_srcu(ssp)
 		: get_state_synchronize_rcu();
 }
 
-static inline unsigned long __start_poll_synchronize_rcu(struct srcu_struct *ssp)
+static inline rcu_gp_poll_state_t __start_poll_synchronize_rcu(struct srcu_struct *ssp)
 {
 	return ssp
 		? start_poll_synchronize_srcu(ssp)
 		: start_poll_synchronize_rcu();
 }
 
-static inline bool __poll_state_synchronize_rcu(struct srcu_struct *ssp, unsigned long cookie)
+static inline bool __poll_state_synchronize_rcu(struct srcu_struct *ssp, rcu_gp_poll_state_t cookie)
 {
 	return ssp
 		? poll_state_synchronize_srcu(ssp, cookie)
@@ -71,13 +87,13 @@ struct rcu_pending_seq {
 	GENRADIX(struct rcu_head *)	objs;
 	size_t				nr;
 	struct rcu_head			**cursor;
-	unsigned long			seq;
+	rcu_gp_poll_state_t		seq;
 };
 
 struct rcu_pending_list {
 	struct rcu_head			*head;
 	struct rcu_head			*tail;
-	unsigned long			seq;
+	rcu_gp_poll_state_t		seq;
 };
 
 struct rcu_pending_pcpu {
@@ -316,10 +332,10 @@ static void rcu_pending_rcu_cb(struct rcu_head *rcu)
 }
 
 static __always_inline struct rcu_pending_seq *
-get_object_radix(struct rcu_pending_pcpu *p, unsigned long seq)
+get_object_radix(struct rcu_pending_pcpu *p, rcu_gp_poll_state_t seq)
 {
 	darray_for_each_reverse(p->objs, objs)
-		if (objs->seq == seq)
+		if (rcu_gp_poll_cookie_eq(objs->seq, seq))
 			return objs;
 
 	if (darray_push_gfp(&p->objs, ((struct rcu_pending_seq) { .seq = seq }), GFP_ATOMIC))
@@ -329,7 +345,7 @@ get_object_radix(struct rcu_pending_pcpu *p, unsigned long seq)
 }
 
 static noinline bool
-rcu_pending_enqueue_list(struct rcu_pending_pcpu *p, unsigned long seq,
+rcu_pending_enqueue_list(struct rcu_pending_pcpu *p, rcu_gp_poll_state_t seq,
 			 struct rcu_head *head, void *ptr,
 			 unsigned long *flags)
 {
@@ -364,7 +380,7 @@ rcu_pending_enqueue_list(struct rcu_pending_pcpu *p, unsigned long seq,
 again:
 	for (struct rcu_pending_list *i = p->lists;
 	     i < p->lists + NUM_ACTIVE_RCU_POLL_OLDSTATE; i++) {
-		if (i->seq == seq) {
+		if (rcu_gp_poll_cookie_eq(i->seq, seq)) {
 			rcu_pending_list_add(i, head);
 			return false;
 		}
@@ -408,7 +424,7 @@ __rcu_pending_enqueue(struct rcu_pending *pending, struct rcu_head *head,
 	struct rcu_pending_pcpu *p;
 	struct rcu_pending_seq *objs;
 	struct genradix_node *new_node = NULL;
-	unsigned long seq, flags;
+	unsigned long flags;
 	bool start_gp = false;
 
 	BUG_ON((ptr != NULL) != (pending->process == RCU_PENDING_KVFREE_FN));
@@ -416,7 +432,7 @@ __rcu_pending_enqueue(struct rcu_pending *pending, struct rcu_head *head,
 	local_irq_save(flags);
 	p = this_cpu_ptr(pending->p);
 	spin_lock(&p->lock);
-	seq = __get_state_synchronize_rcu(pending->srcu);
+	rcu_gp_poll_state_t seq = __get_state_synchronize_rcu(pending->srcu);
 restart:
 	if (may_sleep &&
 	    unlikely(process_finished_items(pending, p, flags)))
@@ -478,9 +494,7 @@ start_gp:
 		 */
 		if (!p->cb_armed) {
 			p->cb_armed = true;
-			spin_unlock_irqrestore(&p->lock, flags);
 			__call_rcu(pending->srcu, &p->cb, rcu_pending_rcu_cb);
-			goto free_node;
 		} else {
 			__start_poll_synchronize_rcu(pending->srcu);
 		}
diff --git a/libbcachefs/recovery.c b/libbcachefs/recovery.c
index a342744f..7849916e 100644
--- a/libbcachefs/recovery.c
+++ b/libbcachefs/recovery.c
@@ -616,6 +616,7 @@ static bool check_version_upgrade(struct bch_fs *c)
 					 bch2_latest_compatible_version(c->sb.version));
 	unsigned old_version = c->sb.version_upgrade_complete ?: c->sb.version;
 	unsigned new_version = 0;
+	bool ret = false;
 
 	if (old_version < bcachefs_metadata_required_upgrade_below) {
 		if (c->opts.version_upgrade == BCH_VERSION_UPGRADE_incompatible ||
@@ -671,14 +672,32 @@ static bool check_version_upgrade(struct bch_fs *c)
 		}
 
 		bch_info(c, "%s", buf.buf);
+		printbuf_exit(&buf);
+
+		ret = true;
+	}
 
-		bch2_sb_upgrade(c, new_version);
+	if (new_version > c->sb.version_incompat &&
+	    c->opts.version_upgrade == BCH_VERSION_UPGRADE_incompatible) {
+		struct printbuf buf = PRINTBUF;
+
+		prt_str(&buf, "Now allowing incompatible features up to ");
+		bch2_version_to_text(&buf, new_version);
+		prt_str(&buf, ", previously allowed up to ");
+		bch2_version_to_text(&buf, c->sb.version_incompat_allowed);
+		prt_newline(&buf);
 
+		bch_info(c, "%s", buf.buf);
 		printbuf_exit(&buf);
-		return true;
+
+		ret = true;
 	}
 
-	return false;
+	if (ret)
+		bch2_sb_upgrade(c, new_version,
+				c->opts.version_upgrade == BCH_VERSION_UPGRADE_incompatible);
+
+	return ret;
 }
 
 int bch2_fs_recovery(struct bch_fs *c)
@@ -1078,7 +1097,7 @@ int bch2_fs_initialize(struct bch_fs *c)
 	bch2_check_version_downgrade(c);
 
 	if (c->opts.version_upgrade != BCH_VERSION_UPGRADE_none) {
-		bch2_sb_upgrade(c, bcachefs_metadata_version_current);
+		bch2_sb_upgrade(c, bcachefs_metadata_version_current, false);
 		SET_BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb, bcachefs_metadata_version_current);
 		bch2_write_super(c);
 	}
diff --git a/libbcachefs/recovery_passes_types.h b/libbcachefs/recovery_passes_types.h
index 94dc20ca..f967b23d 100644
--- a/libbcachefs/recovery_passes_types.h
+++ b/libbcachefs/recovery_passes_types.h
@@ -8,53 +8,59 @@
 #define PASS_ALWAYS		BIT(3)
 #define PASS_ONLINE		BIT(4)
 
+#ifdef CONFIG_BCACHEFS_DEBUG
+#define PASS_FSCK_DEBUG		BIT(1)
+#else
+#define PASS_FSCK_DEBUG		0
+#endif
+
 /*
  * Passes may be reordered, but the second field is a persistent identifier and
  * must never change:
  */
-#define BCH_RECOVERY_PASSES()							\
-	x(recovery_pass_empty,			41, PASS_SILENT)		\
-	x(scan_for_btree_nodes,			37, 0)				\
-	x(check_topology,			 4, 0)				\
-	x(accounting_read,			39, PASS_ALWAYS)		\
-	x(alloc_read,				 0, PASS_ALWAYS)		\
-	x(stripes_read,				 1, PASS_ALWAYS)		\
-	x(initialize_subvolumes,		 2, 0)				\
-	x(snapshots_read,			 3, PASS_ALWAYS)		\
-	x(check_allocations,			 5, PASS_FSCK)			\
-	x(trans_mark_dev_sbs,			 6, PASS_ALWAYS|PASS_SILENT)	\
-	x(fs_journal_alloc,			 7, PASS_ALWAYS|PASS_SILENT)	\
-	x(set_may_go_rw,			 8, PASS_ALWAYS|PASS_SILENT)	\
-	x(journal_replay,			 9, PASS_ALWAYS)		\
-	x(check_alloc_info,			10, PASS_ONLINE|PASS_FSCK)	\
-	x(check_lrus,				11, PASS_ONLINE|PASS_FSCK)	\
-	x(check_btree_backpointers,		12, PASS_ONLINE|PASS_FSCK)	\
-	x(check_backpointers_to_extents,	13, PASS_ONLINE|PASS_FSCK)	\
-	x(check_extents_to_backpointers,	14, PASS_ONLINE|PASS_FSCK)	\
-	x(check_alloc_to_lru_refs,		15, PASS_ONLINE|PASS_FSCK)	\
-	x(fs_freespace_init,			16, PASS_ALWAYS|PASS_SILENT)	\
-	x(bucket_gens_init,			17, 0)				\
-	x(reconstruct_snapshots,		38, 0)				\
-	x(check_snapshot_trees,			18, PASS_ONLINE|PASS_FSCK)	\
-	x(check_snapshots,			19, PASS_ONLINE|PASS_FSCK)	\
-	x(check_subvols,			20, PASS_ONLINE|PASS_FSCK)	\
-	x(check_subvol_children,		35, PASS_ONLINE|PASS_FSCK)	\
-	x(delete_dead_snapshots,		21, PASS_ONLINE|PASS_FSCK)	\
-	x(fs_upgrade_for_subvolumes,		22, 0)				\
-	x(check_inodes,				24, PASS_FSCK)			\
-	x(check_extents,			25, PASS_FSCK)			\
-	x(check_indirect_extents,		26, PASS_FSCK)			\
-	x(check_dirents,			27, PASS_FSCK)			\
-	x(check_xattrs,				28, PASS_FSCK)			\
-	x(check_root,				29, PASS_ONLINE|PASS_FSCK)	\
-	x(check_unreachable_inodes,		40, PASS_ONLINE|PASS_FSCK)	\
-	x(check_subvolume_structure,		36, PASS_ONLINE|PASS_FSCK)	\
-	x(check_directory_structure,		30, PASS_ONLINE|PASS_FSCK)	\
-	x(check_nlinks,				31, PASS_FSCK)			\
-	x(resume_logged_ops,			23, PASS_ALWAYS)		\
-	x(delete_dead_inodes,			32, PASS_ALWAYS)		\
-	x(fix_reflink_p,			33, 0)				\
-	x(set_fs_needs_rebalance,		34, 0)				\
+#define BCH_RECOVERY_PASSES()								\
+	x(recovery_pass_empty,			41, PASS_SILENT)			\
+	x(scan_for_btree_nodes,			37, 0)					\
+	x(check_topology,			 4, 0)					\
+	x(accounting_read,			39, PASS_ALWAYS)			\
+	x(alloc_read,				 0, PASS_ALWAYS)			\
+	x(stripes_read,				 1, PASS_ALWAYS)			\
+	x(initialize_subvolumes,		 2, 0)					\
+	x(snapshots_read,			 3, PASS_ALWAYS)			\
+	x(check_allocations,			 5, PASS_FSCK)				\
+	x(trans_mark_dev_sbs,			 6, PASS_ALWAYS|PASS_SILENT)		\
+	x(fs_journal_alloc,			 7, PASS_ALWAYS|PASS_SILENT)		\
+	x(set_may_go_rw,			 8, PASS_ALWAYS|PASS_SILENT)		\
+	x(journal_replay,			 9, PASS_ALWAYS)			\
+	x(check_alloc_info,			10, PASS_ONLINE|PASS_FSCK)		\
+	x(check_lrus,				11, PASS_ONLINE|PASS_FSCK)		\
+	x(check_btree_backpointers,		12, PASS_ONLINE|PASS_FSCK)		\
+	x(check_backpointers_to_extents,	13, PASS_ONLINE|PASS_FSCK_DEBUG)	\
+	x(check_extents_to_backpointers,	14, PASS_ONLINE|PASS_FSCK)		\
+	x(check_alloc_to_lru_refs,		15, PASS_ONLINE|PASS_FSCK)		\
+	x(fs_freespace_init,			16, PASS_ALWAYS|PASS_SILENT)		\
+	x(bucket_gens_init,			17, 0)					\
+	x(reconstruct_snapshots,		38, 0)					\
+	x(check_snapshot_trees,			18, PASS_ONLINE|PASS_FSCK)		\
+	x(check_snapshots,			19, PASS_ONLINE|PASS_FSCK)		\
+	x(check_subvols,			20, PASS_ONLINE|PASS_FSCK)		\
+	x(check_subvol_children,		35, PASS_ONLINE|PASS_FSCK)		\
+	x(delete_dead_snapshots,		21, PASS_ONLINE|PASS_FSCK)		\
+	x(fs_upgrade_for_subvolumes,		22, 0)					\
+	x(check_inodes,				24, PASS_FSCK)				\
+	x(check_extents,			25, PASS_FSCK)				\
+	x(check_indirect_extents,		26, PASS_FSCK)				\
+	x(check_dirents,			27, PASS_FSCK)				\
+	x(check_xattrs,				28, PASS_FSCK)				\
+	x(check_root,				29, PASS_ONLINE|PASS_FSCK)		\
+	x(check_unreachable_inodes,		40, PASS_ONLINE|PASS_FSCK)		\
+	x(check_subvolume_structure,		36, PASS_ONLINE|PASS_FSCK)		\
+	x(check_directory_structure,		30, PASS_ONLINE|PASS_FSCK)		\
+	x(check_nlinks,				31, PASS_FSCK)				\
+	x(resume_logged_ops,			23, PASS_ALWAYS)			\
+	x(delete_dead_inodes,			32, PASS_ALWAYS)			\
+	x(fix_reflink_p,			33, 0)					\
+	x(set_fs_needs_rebalance,		34, 0)
 
 /* We normally enumerate recovery passes in the order we run them: */
 enum bch_recovery_pass {
diff --git a/libbcachefs/reflink.c b/libbcachefs/reflink.c
index e1911b9b..93ba4f4e 100644
--- a/libbcachefs/reflink.c
+++ b/libbcachefs/reflink.c
@@ -482,7 +482,8 @@ int bch2_trigger_indirect_inline_data(struct btree_trans *trans,
 
 static int bch2_make_extent_indirect(struct btree_trans *trans,
 				     struct btree_iter *extent_iter,
-				     struct bkey_i *orig)
+				     struct bkey_i *orig,
+				     bool reflink_p_may_update_opts_field)
 {
 	struct bch_fs *c = trans->c;
 	struct btree_iter reflink_iter = { NULL };
@@ -548,6 +549,9 @@ static int bch2_make_extent_indirect(struct btree_trans *trans,
 
 	SET_REFLINK_P_IDX(&r_p->v, bkey_start_offset(&r_v->k));
 
+	if (reflink_p_may_update_opts_field)
+		SET_REFLINK_P_MAY_UPDATE_OPTIONS(&r_p->v, true);
+
 	ret = bch2_trans_update(trans, extent_iter, &r_p->k_i,
 				BTREE_UPDATE_internal_snapshot_node);
 err:
@@ -578,7 +582,8 @@ s64 bch2_remap_range(struct bch_fs *c,
 		     subvol_inum dst_inum, u64 dst_offset,
 		     subvol_inum src_inum, u64 src_offset,
 		     u64 remap_sectors,
-		     u64 new_i_size, s64 *i_sectors_delta)
+		     u64 new_i_size, s64 *i_sectors_delta,
+		     bool may_change_src_io_path_opts)
 {
 	struct btree_trans *trans;
 	struct btree_iter dst_iter, src_iter;
@@ -591,6 +596,8 @@ s64 bch2_remap_range(struct bch_fs *c,
 	struct bpos src_want;
 	u64 dst_done = 0;
 	u32 dst_snapshot, src_snapshot;
+	bool reflink_p_may_update_opts_field =
+		bch2_request_incompat_feature(c, bcachefs_metadata_version_reflink_p_may_update_opts);
 	int ret = 0, ret2 = 0;
 
 	if (!bch2_write_ref_tryget(c, BCH_WRITE_REF_reflink))
@@ -672,7 +679,8 @@ s64 bch2_remap_range(struct bch_fs *c,
 			src_k = bkey_i_to_s_c(new_src.k);
 
 			ret = bch2_make_extent_indirect(trans, &src_iter,
-						new_src.k);
+						new_src.k,
+						reflink_p_may_update_opts_field);
 			if (ret)
 				continue;
 
@@ -690,6 +698,10 @@ s64 bch2_remap_range(struct bch_fs *c,
 				 bkey_start_offset(src_k.k));
 
 			SET_REFLINK_P_IDX(&dst_p->v, offset);
+
+			if (reflink_p_may_update_opts_field &&
+			    may_change_src_io_path_opts)
+				SET_REFLINK_P_MAY_UPDATE_OPTIONS(&dst_p->v, true);
 		} else {
 			BUG();
 		}
diff --git a/libbcachefs/reflink.h b/libbcachefs/reflink.h
index f119316a..1632780b 100644
--- a/libbcachefs/reflink.h
+++ b/libbcachefs/reflink.h
@@ -78,7 +78,8 @@ struct bkey_s_c bch2_lookup_indirect_extent(struct btree_trans *, struct btree_i
 					    bool, unsigned);
 
 s64 bch2_remap_range(struct bch_fs *, subvol_inum, u64,
-		     subvol_inum, u64, u64, u64, s64 *);
+		     subvol_inum, u64, u64, u64, s64 *,
+		     bool);
 
 int bch2_gc_reflink_done(struct bch_fs *);
 int bch2_gc_reflink_start(struct bch_fs *);
diff --git a/libbcachefs/reflink_format.h b/libbcachefs/reflink_format.h
index 53502627..92995e4f 100644
--- a/libbcachefs/reflink_format.h
+++ b/libbcachefs/reflink_format.h
@@ -19,6 +19,8 @@ struct bch_reflink_p {
 
 LE64_BITMASK(REFLINK_P_IDX,	struct bch_reflink_p, idx_flags,  0, 56);
 LE64_BITMASK(REFLINK_P_ERROR,	struct bch_reflink_p, idx_flags, 56, 57);
+LE64_BITMASK(REFLINK_P_MAY_UPDATE_OPTIONS,
+				struct bch_reflink_p, idx_flags, 57, 58);
 
 struct bch_reflink_v {
 	struct bch_val		v;
diff --git a/libbcachefs/sb-errors_format.h b/libbcachefs/sb-errors_format.h
index 3bbda181..4248ff3c 100644
--- a/libbcachefs/sb-errors_format.h
+++ b/libbcachefs/sb-errors_format.h
@@ -140,8 +140,8 @@ enum bch_fsck_flags {
 	x(backpointer_bucket_offset_wrong,			125,	0)		\
 	x(backpointer_level_bad,				294,	0)		\
 	x(backpointer_dev_bad,					297,	0)		\
-	x(backpointer_to_missing_device,			126,	0)		\
-	x(backpointer_to_missing_alloc,				127,	0)		\
+	x(backpointer_to_missing_device,			126,	FSCK_AUTOFIX)	\
+	x(backpointer_to_missing_alloc,				127,	FSCK_AUTOFIX)	\
 	x(backpointer_to_missing_ptr,				128,	FSCK_AUTOFIX)	\
 	x(lru_entry_at_time_0,					129,	FSCK_AUTOFIX)	\
 	x(lru_entry_to_invalid_bucket,				130,	FSCK_AUTOFIX)	\
diff --git a/libbcachefs/super-io.c b/libbcachefs/super-io.c
index 6a086c1c..912c3696 100644
--- a/libbcachefs/super-io.c
+++ b/libbcachefs/super-io.c
@@ -42,7 +42,7 @@ static const struct bch2_metadata_version bch2_metadata_versions[] = {
 #undef x
 };
 
-void bch2_version_to_text(struct printbuf *out, unsigned v)
+void bch2_version_to_text(struct printbuf *out, enum bcachefs_metadata_version v)
 {
 	const char *str = "(unknown version)";
 
@@ -55,7 +55,7 @@ void bch2_version_to_text(struct printbuf *out, unsigned v)
 	prt_printf(out, "%u.%u: %s", BCH_VERSION_MAJOR(v), BCH_VERSION_MINOR(v), str);
 }
 
-unsigned bch2_latest_compatible_version(unsigned v)
+enum bcachefs_metadata_version bch2_latest_compatible_version(enum bcachefs_metadata_version v)
 {
 	if (!BCH_VERSION_MAJOR(v))
 		return v;
@@ -69,6 +69,16 @@ unsigned bch2_latest_compatible_version(unsigned v)
 	return v;
 }
 
+void bch2_set_version_incompat(struct bch_fs *c, enum bcachefs_metadata_version version)
+{
+	mutex_lock(&c->sb_lock);
+	SET_BCH_SB_VERSION_INCOMPAT(c->disk_sb.sb,
+		max(BCH_SB_VERSION_INCOMPAT(c->disk_sb.sb), version));
+	c->disk_sb.sb->features[0] |= cpu_to_le64(BCH_FEATURE_incompat_version_field);
+	bch2_write_super(c);
+	mutex_unlock(&c->sb_lock);
+}
+
 const char * const bch2_sb_fields[] = {
 #define x(name, nr)	#name,
 	BCH_SB_FIELDS()
@@ -369,6 +379,12 @@ static int bch2_sb_validate(struct bch_sb_handle *disk_sb,
 		return -BCH_ERR_invalid_sb_features;
 	}
 
+	if (BCH_VERSION_MAJOR(le16_to_cpu(sb->version)) > BCH_VERSION_MAJOR(bcachefs_metadata_version_current) ||
+	    BCH_SB_VERSION_INCOMPAT(sb) > bcachefs_metadata_version_current) {
+		prt_printf(out, "Filesystem has incompatible version");
+		return -BCH_ERR_invalid_sb_features;
+	}
+
 	block_size = le16_to_cpu(sb->block_size);
 
 	if (block_size > PAGE_SECTORS) {
@@ -407,6 +423,18 @@ static int bch2_sb_validate(struct bch_sb_handle *disk_sb,
 		return -BCH_ERR_invalid_sb_time_precision;
 	}
 
+	/* old versions didn't know to downgrade this field */
+	if (BCH_SB_VERSION_INCOMPAT_ALLOWED(sb) > le16_to_cpu(sb->version))
+		SET_BCH_SB_VERSION_INCOMPAT_ALLOWED(sb, le16_to_cpu(sb->version));
+
+	if (BCH_SB_VERSION_INCOMPAT(sb) > BCH_SB_VERSION_INCOMPAT_ALLOWED(sb)) {
+		prt_printf(out, "Invalid version_incompat ");
+		bch2_version_to_text(out, BCH_SB_VERSION_INCOMPAT(sb));
+		prt_str(out, " > incompat_allowed ");
+		bch2_version_to_text(out, BCH_SB_VERSION_INCOMPAT_ALLOWED(sb));
+		return -BCH_ERR_invalid_sb_version;
+	}
+
 	if (!flags) {
 		/*
 		 * Been seeing a bug where these are getting inexplicably
@@ -520,6 +548,9 @@ static void bch2_sb_update(struct bch_fs *c)
 	c->sb.uuid		= src->uuid;
 	c->sb.user_uuid		= src->user_uuid;
 	c->sb.version		= le16_to_cpu(src->version);
+	c->sb.version_incompat	= BCH_SB_VERSION_INCOMPAT(src);
+	c->sb.version_incompat_allowed
+				= BCH_SB_VERSION_INCOMPAT_ALLOWED(src);
 	c->sb.version_min	= le16_to_cpu(src->version_min);
 	c->sb.version_upgrade_complete = BCH_SB_VERSION_UPGRADE_COMPLETE(src);
 	c->sb.nr_devices	= src->nr_devices;
@@ -1152,6 +1183,8 @@ bool bch2_check_version_downgrade(struct bch_fs *c)
 	 */
 	if (BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb) > bcachefs_metadata_version_current)
 		SET_BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb, bcachefs_metadata_version_current);
+	if (BCH_SB_VERSION_INCOMPAT_ALLOWED(c->disk_sb.sb) > bcachefs_metadata_version_current)
+		SET_BCH_SB_VERSION_INCOMPAT_ALLOWED(c->disk_sb.sb, bcachefs_metadata_version_current);
 	if (c->sb.version > bcachefs_metadata_version_current)
 		c->disk_sb.sb->version = cpu_to_le16(bcachefs_metadata_version_current);
 	if (c->sb.version_min > bcachefs_metadata_version_current)
@@ -1160,7 +1193,7 @@ bool bch2_check_version_downgrade(struct bch_fs *c)
 	return ret;
 }
 
-void bch2_sb_upgrade(struct bch_fs *c, unsigned new_version)
+void bch2_sb_upgrade(struct bch_fs *c, unsigned new_version, bool incompat)
 {
 	lockdep_assert_held(&c->sb_lock);
 
@@ -1170,6 +1203,10 @@ void bch2_sb_upgrade(struct bch_fs *c, unsigned new_version)
 
 	c->disk_sb.sb->version = cpu_to_le16(new_version);
 	c->disk_sb.sb->features[0] |= cpu_to_le64(BCH_SB_FEATURES_ALL);
+
+	if (incompat)
+		SET_BCH_SB_VERSION_INCOMPAT_ALLOWED(c->disk_sb.sb,
+			max(BCH_SB_VERSION_INCOMPAT_ALLOWED(c->disk_sb.sb), new_version));
 }
 
 static int bch2_sb_ext_validate(struct bch_sb *sb, struct bch_sb_field *f,
@@ -1334,6 +1371,14 @@ void bch2_sb_to_text(struct printbuf *out, struct bch_sb *sb,
 	bch2_version_to_text(out, le16_to_cpu(sb->version));
 	prt_newline(out);
 
+	prt_printf(out, "Incompatible features allowed:\t");
+	bch2_version_to_text(out, BCH_SB_VERSION_INCOMPAT_ALLOWED(sb));
+	prt_newline(out);
+
+	prt_printf(out, "Incompatible features in use:\t");
+	bch2_version_to_text(out, BCH_SB_VERSION_INCOMPAT(sb));
+	prt_newline(out);
+
 	prt_printf(out, "Version upgrade complete:\t");
 	bch2_version_to_text(out, BCH_SB_VERSION_UPGRADE_COMPLETE(sb));
 	prt_newline(out);
diff --git a/libbcachefs/super-io.h b/libbcachefs/super-io.h
index 90e7b176..f1ab4f94 100644
--- a/libbcachefs/super-io.h
+++ b/libbcachefs/super-io.h
@@ -18,8 +18,21 @@ static inline bool bch2_version_compatible(u16 version)
 		version >= bcachefs_metadata_version_min;
 }
 
-void bch2_version_to_text(struct printbuf *, unsigned);
-unsigned bch2_latest_compatible_version(unsigned);
+void bch2_version_to_text(struct printbuf *, enum bcachefs_metadata_version);
+enum bcachefs_metadata_version bch2_latest_compatible_version(enum bcachefs_metadata_version);
+
+void bch2_set_version_incompat(struct bch_fs *, enum bcachefs_metadata_version);
+
+static inline bool bch2_request_incompat_feature(struct bch_fs *c,
+						 enum bcachefs_metadata_version version)
+{
+	if (unlikely(version > c->sb.version_incompat)) {
+		if (version > c->sb.version_incompat_allowed)
+			return false;
+		bch2_set_version_incompat(c, version);
+	}
+	return true;
+}
 
 static inline size_t bch2_sb_field_bytes(struct bch_sb_field *f)
 {
@@ -94,7 +107,7 @@ static inline void bch2_check_set_feature(struct bch_fs *c, unsigned feat)
 }
 
 bool bch2_check_version_downgrade(struct bch_fs *);
-void bch2_sb_upgrade(struct bch_fs *, unsigned);
+void bch2_sb_upgrade(struct bch_fs *, unsigned, bool);
 
 void __bch2_sb_field_to_text(struct printbuf *, struct bch_sb *,
 			     struct bch_sb_field *);