Not completely. I can still imagine a few overheads: - code size is bigger, taking up space in the cache and memory; - instructions need to be decoded (whethe

Not completely. I can still imagine a few overheads: - code size is bigger, taking up space in the cache and memory; - instructions need to be decoded (whether this affects performance depends on surrounding code); - this branch will take a slot in the branch predictor state (same here).