Debugging of Optimized Behavioral Speci

Symbolic debuggers are system development tools that can accelerate the validation speed of behavioral speciications by allowing a user to interact with an executing code at the source level. In response to a user query, the debugger must be able to retrieve and display the value of a source variable in a manner consistent with what the user expects with respect to the source statement where execution has halted. However, when a behavioral speciication has been optimized using transformations, values of variables may either be inaccessible in the run-time state or inconsistent with what the user expects. In this paper, we address the problem that pertains to the retrieval of source values for globally optimized behav-ioral speciications. We have developed a set of techniques that, given a behavioral speciication CDFG, enforce computation of a selected subset Vcut of user variables such that (i) all other variables v 2 CDFG can be computed from Vcut and (ii) this enforcement has minimal impact on the optimization potential of the computation. The implementation of the new debugging approach poses several optimization tasks. We have formulated the optimization tasks and developed heuristics to solve them. The eeectiveness of the proposed approach has been demonstrated on a set of benchmark designs. 1 Introduction Functional debugging of hardware and software systems has emerged as a dominant step with respect to time and cost of the development process. For example, debugging (architecture and functional veriication) of the UltraSPARC-I took twice as long as its design Yan95]. The diiculty of verifying designs is likely to worsen in the future, since the key technological trends indicate that the percentage of controllable and observable variables in designs will steadily decrease. The Intel development strategy team foresees that a major design concern for their year-2006 microprocessor will be the need to exhaustively test all possible computational and compatibility combinations Yu96]. Symbolic debuggers are system development tools that can accelerate the validation speed of behavioral speciica-tions by allowing a user to interact with an executing code at the source level Hen82]. Symbolic debugging must ensure that in response to a user inquiry, the debugger is able to retrieve and display the value of a source variable in a manner consistent with what the user expects with respect to a breakpoint in the source code. The application of code optimization techniques usually makes symbolic debugging harder. While code optimization techniques such as transformations …