Describe the problem

There is a bug in type resolution which causes workload failures like #56081 (comment) and #56081 (comment)
Upon further investigation, it seems that primitive types (eg. those that exist in pg_types) are not being resolved correctly. We check for equivalence between types to predict errors. Since types aren't being resolved correctly, our predictions are wrong. Thus, the workload fails.

Expected behavior
Primitive types should be resolved correctly, so that we can predict type-related errors when performing ops like setColumnType and setColumnDefault.