In Interpreter.IsSubType(...) function, both the sub-type and the super-type gets converted from StaticType to sema.Type (see below). This adds an unnecessary overhead.
Having a separate sub-type checking implementation for runtime (static types) could eliminate this overhead. We could define the sub-typing rules in some declarative format, and then generate the two implementations from that, to make them consistent/in-sync.
Issue to be solved
In
Interpreter.IsSubType(...)function, both the sub-type and the super-type gets converted fromStaticTypetosema.Type(see below). This adds an unnecessary overhead.https://github.com/onflow/cadence/blob/b09231fe0216ecb024fbfc909f213bed53a2d3bb/interpreter/interpreter.go#L4066 https://github.com/onflow/cadence/blob/b09231fe0216ecb024fbfc909f213bed53a2d3bb/interpreter/interpreter.go#L4039
Suggested Solution
Having a separate sub-type checking implementation for runtime (static types) could eliminate this overhead. We could define the sub-typing rules in some declarative format, and then generate the two implementations from that, to make them consistent/in-sync.