We present the design and early implementation of p4rl, a system that uses reinforcement learning-guided fuzz testing to execute the verification of P4 switches automatically at runtime. p4rl system uses our novel user-friendly query language, p4q to conveniently specify the intended properties in simple conditional statements (if-else) and check the actual runtime behavior of the P4 switch against such properties. In p4rl, user-specified p4q queries with the control plane configuration, Agent, and the Reward System guide the fuzzing process to trigger runtime bugs automatically during Agent training. To illustrate the strength of p4rl, we developed and evaluated an early prototype of p4rl system that executes runtime verification of a P4 network device, e.g., L3 (Layer-3) switch. Our initial results are promising and show that p4rl automatically detects diverse bugs while outperforming the baseline approach.

We design, develop, and evaluate P6, an automated approach to (a) detect, (b) localize, and (c) patch software bugs in P4 programs. Bugs are reported via a violation of pre-specified expected behavior that is captured by P6. P6 is based on machine learning-guided fuzzing that tests P4 switch non-intrusively, i.e., without modifying the P4 program for detecting runtime bugs. This enables an automated and real-time localization and patching of bugs. We used a P6 prototype to detect and patch existing bugs in various publicly available P4 application programs deployed on two different switch platforms: behavioral model (bmv2) and Tofino. Our evaluation shows that P6 significantly outperforms bug detection baselines while generating fewer packets and patches bugs in large P4 programs such as switch.p4 without triggering any regressions.

No abstract