dataflow-analysis-framework

Static program analysis is implemented by defining custom lattices and transfer functions to solve dataflow equations. This framework handles both forward and backward analyses, enabling you to build analyses like variable liveness across a control flow graph.

Dataflow analysis in compiler optimization identifies runtime behaviors statically to enable transformations like dead code elimination. By analyzing variable liveness across the control flow graph, compilers can safely remove unused instructions and improve performance.

You can define custom lattices for specific program verification domains by specifying the required domain properties. The framework supports defining these custom lattices to accurately model various analysis states for static program verification.

Dataflow equations are solved for both forward and backward analyses using worklist algorithms. This approach efficiently propagates information through the control flow graph, supporting 'may' and 'must' analysis variations for static program analysis.

The framework supports both 'may' and 'must' analysis variations for static program analysis. It handles the implementation of both forward and backward analyses to accurately compute the properties of program states.