Bug traces produced in simulation serve as the basis for patching the RTL code in order to fix a bug. It is important to prove that the patch covers all instances of the bug scenario; otherwise, the bug may return with a different valuation of the variables involved in the bug scenario. For large circuits, formal methods do not scale well enough to comprehensively eliminate the bug, and achieving adequate coverage in simulation and regression testing becomes expensive. This paper proposes formal methods for analyzing the control trace leading to the observed manifestation of the bug and verifying the robustness of the bug fix with respect to that control trace. We propose a classification of the bug fix based on the guarantee that our analysis can provide about the quality of the bug fix. Our method also prescribes the types of tests that are recommended to validate the bug fix on other types of scenarios. Since our methods are more scalable by orders of magnitude than model checking the entire design, we believe that the proposed formal methods hold immense promise in analyzing bug fixes in practice.

Bug-fixing in deeply embedded portions of the logic is typically accompanied by the postfacto addition of new assertions, which cover the bug scenario. Formally verifying the assertions defined over such deeply embedded portions of the logic is challenging because formal methods do not scale to the size of the entire logic. Verifying the assertion on the embedded logic in isolation typically throws up a large number of counterexamples, many of which are spurious because the scenarios they depict are not possible in the entire logic. In this paper, we introduce the notion of ranking the counterexamples so that only the most likely counterexamples are presented to the designer. Our ranking is based on assume properties mined from simulation traces of the entire logic. We define a metric to compute a belief for each assume property that is mined, and rank counterexamples based on their relationships with the mined assume properties. Experimental results demonstrate a remarkable correlation between the real counterexamples (if they exist) and the proposed ranking metric, thereby establishing the proposed method as a very promising verification approach.

Bug-fixing in deeply embedded portions of the logic is typically accompanied by the post-facto addition to new assertions which cover the bug scenario. Formally verifying properties defined over such deeply embedded portions of the logic is challenging because formal methods do not scale to the size of the entire logic, and verifying the property on the embedded logic in isolation typically throws up a large number of counterexamples, many of which are spurious because the scenarios they depict are not possible in the entire logic. In this paper we introduce the notion of ranking the counterexamples so that only the most likely counterexamples are presented to the designer. Our ranking is based on assume properties mined from simulation traces of the entire logic. We define a metric to compute a belief for each assume property that is mined, and rank counterexamples based on their conflicts with the mined assume properties. Experimental results demonstrate an amazing correlation between the real counterexamples (if they exist) and the proposed ranking metric, thereby establishing the proposed method as a very promising verification approach.

Automated methods for bug localization for hardware designs typically work on the design implementation to root-cause a given bug. This paper presents a novel debugging approach where instead of using the design implementation in the debugging process, we use causal deduction using formal properties scattered across the design to locate the bug. This has two advantages, namely, (a) the reasoning takes place in the property space instead of the state space of the implementation, which enhances scalability, and (b) new properties can be added in hindsight to perform what-if analysis, which is less expensive than modifying the implementation for each alternative. Experimental results demonstrate the scalability of the approach in debugging designs with large property suites.

Recent research has indicated ways of using UPF specifications for extracting valid low-level control sequences to express the transitions between the power states of individual domains. Today there is a disconnect between the high-level architectural power management strategy which relates multiple power domains and these low-level assertions for controlling individual power domains. In this paper we attempt to bridge this disconnect by leveraging the low-level per-domain assertions for translating architectural power intent properties into global assertions over low-level signals. We show that the inter-domain properties created in this manner can be formally verified over the global power management logic.

The advent of multi-core embedded processors has brought along new challenges for embedded system design. This paper presents an efficient, battery aware, code partitioning technique for a text to speech system, which is executed on a multi-core embedded processor. The system achieves significant performance improvements both in terms of execution time as well as battery lifetimes. The mentioned technique provides a new paradigm for battery aware embedded system design which can be easily extended to other applications