Waveform-Based Bug Localization for SoC Functional Debugging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Debugging functional bugs in system-on-chip (SoC) level is time-consuming and resource-intensive, requiring efficient methods to identify deviations in behavior patterns and locate the relevant source code for errors.
Innovation Solution
An automated method and system using an electronic design automation (EDA) tool to analyze waveforms of signals from a functioning and faulty design version, extract behavior patterns, identify deviations, and present corresponding code and time windows for debugging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If manual debugging methods are used to identify functional bugs in SoC design, then debugging accuracy can be achieved, but the debugging time and resource consumption increase significantly
Solution Approach 1:
The system performs self-service by automatically comparing waveforms and identifying bug locations without requiring manual intervention. The EDA tool autonomously extracts behavior patterns, compares them between good and bad designs, and presents localized bug information to users, eliminating the need for manual debugging efforts.
Solution Approach 2:
The patent replaces manual mechanical debugging processes with automated electronic analysis. Instead of manually examining waveforms and identifying deviations, the system uses processors to automatically compare behavior patterns in waveforms, identify deviations, and locate bugs, substituting human effort with computational automation.
2Productivity
If automated waveform analysis is implemented to identify behavior pattern deviations, then debugging speed increases, but system complexity increases
Solution Approach 1:
The patent introduces behavior patterns as an intermediary layer between raw waveforms and bug identification. Instead of directly comparing complex waveforms, the system extracts behavior patterns that represent characteristic signal behaviors, then compares these simplified patterns to identify deviations. This intermediary abstraction reduces the complexity of the comparison process while maintaining debugging effectiveness.
Solution Approach 2:
The system changes the parameters of analysis by transforming raw waveform data into behavior pattern representations. This parameter transformation simplifies the data structure being compared, making the automated analysis process more efficient and manageable while improving debugging productivity.
3Measurement precision
If detailed behavior pattern comparison is performed across all signals, then bug localization accuracy improves, but computational resources consumed increase
Solution Approach 1:
The patent extracts only the relevant behavior patterns from the complete waveform data that are necessary for bug identification. Instead of analyzing all signal parameters in detail, the system identifies and extracts characteristic behavior patterns that are indicative of functional correctness, then compares only these extracted patterns between good and bad designs, reducing computational overhead while maintaining localization accuracy.
Data Source
AI summary
An automated debugging method includes obtaining waveforms of signals recorded during execution of one or a plurality of tests on a first properly functioning version of a design and on a second faulty version of the design, extracting behavior patterns of the signals in the waveforms of the first design and behavior patterns of the signals in the waveforms of the second design, identifying one or more deviations in signals of the extracted behavior patterns, identifying a corresponding simulation time or time window and presenting to a user via an output device code of the second design pertaining to a deviation of the one or more deviations in corresponding behavior patterns of the extracted behavior patterns and the time or time window during which that deviation occurred.