Generating constraint preserving testcases in the presence of dead-end constraints

摘要

Mechanisms for generating constraint preserving testcases in the presence of dead-end constraints are provided. A balance between precision and computational expense in generating the testcases is achieved by establishing a sliding window of constraint solving for a selected number of K time-steps in the future from a current time-step. The testcases solve for the constraints for the next K time-steps at every state of a netlist instead of just trying to solve the constraint for the present time-step. K is determined by determining, for each input, either a minimum length path depth or maximum length depth path from the input to the constraint. The largest depth value for the inputs to the netlist is then utilized as the depth for the netlist. This depth then is used to define the width of the sliding window of constraint solving.