Formulating Security in Programming Languages

Hi,


        I came up with this notation in stating the goal of security in
programming languages:


        Programming Language Construct = Syntax + Semantics + Side Effect.


        N# semantics N# syntax = Syntax + Semantics + Side Effects


      Security = N# semantics N# syntax - Side Effects


        I will explain via a C example:


        N# (malloc-semantics) N# (malloc-syntax, free-opt-syntax) = Memory
Allocation + Memory Deallocation + Memory Leak


        In the above, the LHS says, apply the combined syntaxes of malloc
and optional-free to the semantics of malloc. The result is that
Memory Leak is a side effect. So, the goal of security is to remove
Memory Leaks.


      Does the approach specified in the above paragraphs occur to you as
brute force esp when generalizing over the entire grammar(small &
manageable) if not the semantics(the whole large LRM)?


      Is anyone interested in this topic ?


      Are there any other formal approaches to investigating security in
Programming Languages?


Sincerely,
Seima Rao.

Return to the comp.compilers page.
Search the comp.compilers archives again.