Estimating the Cost of Native Method Calls for Resource-bounded Functional Programming Languages

We address the problem of applying resource-bounded functional programming languages in practice on object-oriented virtual machines which include calls to native methods coded in low-level languages without garbage collection support. We consider the application of a functional language with a high-level type system which incorporates measures of heap space consumption in types on such an execution platform. We supplement the syntactic type inference procedure of the functional language with a separate analysis which estimates the costs of memory leaks incurred by calls to garbage collection-ignorant functions.