Authors: David F. Bacon, Perry Cheng, Daniel Frampton, and David Grove.
Citation: IBM Research Report RC24162.
Debugging the timing behavior of real-time systems is notoriously difficult, and with a new generation of complex real-time systems whose size is measured in tens of millions of lines of code, the difficulty is increasing enormously. We have developed TuningFork, a tool especially designed for visualization, analysis, and debugging of large-scale real-time systems. The system is capable of recording high-frequency events at sub-microsecond resolution with almost no perturbation to the application. The visualization tool is capable of viewing system activity online in real-time, and users can simultaneously explore the data interactively. TuningFork has allowed us to find numerous timing bugs and anomalies, including unexpected scheduling behavior, clock resets, delayed lock release, and introduction of non-constant-time functions.
PDF