A Calculus of Chaos in Stochastic Compilation: Engineering in the Cause of Mathematics

EasyChair Preprint 1612

Abstract

An unexpected result from an open project to develop a `chaotic' compiler for ANSI C is described here: an information entropy calculus for stochastically compiled programs. A stochastic compiler produces randomly different object codes every time it is applied to the same source code. The calculus quantifies the entropy introduced into the run-time program traces by the compiler, allowing the strategy for maximum entropy to be characterised and that the compiler implements it to be verified. The analysis aims to support the argument that, on a suitable run-time platform and via any polynomial time method, the programmer's intention is unreadable from the trace, as word length n tends to infinity.

Keyphrases: ANSI-C, Compilers, Program calculus, information entropy, machine code

@booklet{EasyChair:1612,
  author    = {Peter T. Breuer and Simon Pickin},
  title     = {A Calculus of Chaos in Stochastic Compilation: Engineering in the Cause of Mathematics},
  howpublished = {EasyChair Preprint 1612},
  year      = {EasyChair, 2019}}