bcachefs: Rhashtable based buckets_in_flight for copygc

Previously, copygc used a fifo for tracking buckets in flight - this had
the disadvantage of being fixed size, since we pass references to
elements into the move code.

This restructures it to be a hash table and linked list, since with
erasure coding we need to be able to pipeline across an arbitrary number
of buckets.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

fs/bcachefs/move_types.h

@@ -16,9 +16,20 @@ struct bch_move_stats {
 	atomic64_t		sectors_raced;
 };
 
-struct move_bucket_in_flight {
+struct move_bucket_key {
 	struct bpos		bucket;
 	u8			gen;
+};
+
+struct move_bucket {
+	struct move_bucket_key	k;
+	unsigned		sectors;
+};
+
+struct move_bucket_in_flight {
+	struct move_bucket_in_flight *next;
+	struct rhash_head	hash;
+	struct move_bucket	bucket;
 	atomic_t		count;
 };