Leveraging Hardware Probes and Optimizations for Accelerating Fuzz Testing of Heterogeneous Applications
There is a growing interest in the computer architecture community to incorporate heterogeneity and specialization to improve performance. Developers can create heterogeneous applications that consist of both host code and kernel code, where compute-intensive kernels can be offloaded from CPU to hardware accelerators. Testing such applications on real heterogeneous architectures is extremely challenging as kernels are black boxes, providing no information about the kernels’ internal execution to diagnose issues such as silent hangs or unexpected results. Additionally, inputs for heterogeneous applications are often large matrices, leading to a vast search space for identifying bug-revealing inputs.
We propose a novel fuzz testing technique, HFuzz, to enable efficient testing on real heterogeneous architectures. HFuzz aims to increase both the observability of hardware kernels and testing efficiency through a three-pronged approach. First, HFuzz automatically generates test guidance by inserting device-side in-kernel hardware probes in addition to host-side software monitors. Second, it performs rapid input space exploration by offloading compute-intensive input mutations to hardware kernels. Third, HFuzz parallelizes fuzzing and enables fast on-chip memory access, by utilizing four FPGA-level optimizations including loop unrolling, shannonization, data preloading, and dynamic kernel sharing.
We evaluate HFuzz on seven open-source OneAPI subjects from Intel. HFuzz speeds up fuzz testing by 4.7Ă— with HW-accelerated input space exploration. By incorporating HW probes in tandem with SW monitors, HFuzz finds 33 defects within 4 hours and reveals 25 unique, unexpected behavior symptoms that could not be found by SW-based monitoring alone. HFuzz is the first to design hardware optimizations to accelerate fuzz testing.
Wed 6 DecDisplayed time zone: Pacific Time (US & Canada) change
16:00 - 18:00 | FuzzingResearch Papers at Golden Gate C1 Chair(s): Shaukat Ali Simula Research Laboratory and Oslo Metropolitan University | ||
16:00 15mTalk | Enhancing Coverage-guided Fuzzing via Phantom Program Research Papers Mingyuan Wu Southern University of Science and Technology and the University of Hong Kong, Kunqiu Chen Southern University of Science and Technology, Qi Luo Southern University of Science and Technology, Jiahong Xiang Southern University of Science and Technology, Ji Qi The University of Hong Kong, Junjie Chen Tianjin University, Heming Cui University of Hong Kong, Yuqun Zhang Southern University of Science and Technology Media Attached | ||
16:15 15mTalk | Co-Dependence Aware Fuzzing for Dataflow-based Big Data Analytics Research Papers Ahmad Humayun Virginia Tech, Miryung Kim University of California at Los Angeles, USA, Muhammad Ali Gulzar Virginia Tech Pre-print Media Attached | ||
16:30 15mTalk | SJFuzz: Seed & Mutator Scheduling for JVM Fuzzing Research Papers Mingyuan Wu Southern University of Science and Technology and the University of Hong Kong, Yicheng Ouyang University of Illinois at Urbana-Champaign, Minghai Lu Southern University of Science and Technology, Junjie Chen Tianjin University, Yingquan Zhao College of Intelligence and Computing, Tianjin University, Heming Cui University of Hong Kong, Guowei Yang University of Queensland, Yuqun Zhang Southern University of Science and Technology Media Attached | ||
16:45 15mTalk | Metamong: Detecting Render-update Bugs in Web Browsers through Fuzzing Research Papers Suhwan Song Seoul National University, South Korea, Byoungyoung Lee Seoul National University, South Korea Media Attached | ||
17:00 15mTalk | Property-based Fuzzing for Finding Data Manipulation Errors in Android Apps Research Papers Jingling Sun East China Normal University, Ting Su East China Normal University, Jiayi Jiang East China Normal University, Jue Wang Nanjing University, Geguang Pu East China Normal University, Zhendong Su ETH Zurich Media Attached | ||
17:15 15mTalk | Leveraging Hardware Probes and Optimizations for Accelerating Fuzz Testing of Heterogeneous Applications Research Papers Jiyuan Wang University of California at Los Angeles, Qian Zhang University of California, Riverside, Hongbo Rong Intel Labs, Guoqing Harry Xu University of California at Los Angeles, Miryung Kim University of California at Los Angeles, USA Pre-print Media Attached | ||
17:30 15mTalk | NaNofuzz: A Usable Tool for Automatic Test Generation Research Papers Matthew C. Davis Carnegie Mellon University, Sangheon Choi Rose-Hulman Institute of Technology, Sam Estep Carnegie Mellon University, Brad A. Myers Carnegie Mellon University, Joshua Sunshine Carnegie Mellon University Link to publication DOI Media Attached | ||
17:45 15mTalk | [Remote] A Generative and Mutational Approach for Synthesizing Bug-exposing Test Cases to Guide Compiler Fuzzing Research Papers Guixin Ye Northwest University, Tianmin Hu Northwest University, Zhanyong Tang Northwest University, Zhenye Fan Northwest University, Shin Hwei Tan Concordia University, Bo Zhang Tencent Security Platform Department, Wenxiang Qian Tencent Security Platform Department, Zheng Wang University of Leeds, UK Media Attached |