This monograph focuses on optimal flow analysis of sequential and parallel programs. It studies algorithmic properties of various versions of the well-known constant-propagation problem, combining techniques from such diverse areas as linear algebra, computable ring theory, abstract interpretation, program verification and complexity theory. This combination of techniques is the key to further progress in automatic analysis and constant-propagation illustrates this point in a theoretical study.

Inhalt

1. Introduction.- 2. A Hierarchy of Constants.- 3. Deciding Constants by Effective Weakest Preconditions.- 4. Limits of Parallel Flow Analysis.- 5. Parallel Flow Graphs.- 6. Non-atomic Execution.- 7. Dependence Traces.- 8. Detecting Copy Constants and Eliminating Faint Code.- 9. Complexity in the Non-atomic Scenario.- 10. Conclusion.- A. A Primer on Constraint-Based Program Analysis.