The hash and equals methods of code objects are large, complex and unnecessary. #101346
Description
We should just compare code objects by identity.
Technically, they are immutable values but realistically, there is only on code object per source unit.
Attempting to maintain value semantics when the underlying code object is internally mutable (due to specialization and instrumentation) is cumbersome and probably buggy, and will only get worse with more optimizations.
We should just delete the hash and comparison methods and have identity comparison. It will be faster too.