How Well Industry-level Cause Bisection Works in Real-world - A Study on Linux Kernel.

The artifact repository for

Kangzheng Gu, Yuan Zhang, Jiajun Cao, Xin Tan, and Min Yang. 2024. How Well Industry-Level Cause Bisection Works in Real-World: A Study on Linux Kernel. In Companion Proceedings of the 32nd ACM International Conference on the Foundations of Software Engineering (FSE Companion ’24)

Prerequisites

• python environment

  conda env create -f syzbot_bisect_env.yaml -n syzbot_bisect
  

• srcml

  1. download from https://www.srcml.org/#download
  2. config .bashrc

      export PATH=/PATH/TO/srcml/bin:$PATH
      export LD_LIBRARY_PATH=/PATH/TO/srcml/lib:$LD_LIBRARY_PATH
  

• create a MYSQL database using syzbot_data/syzbot_bug_info.sql

• assign the global variables in the analyses/config.py according to your database configuration.

   DATA_DIR = '../syzbot_data'
   DB_IP = '1.1.1.1'
   DB_PORT = 6603
   DB_USER = 'a'
   DB_PWD = 'a'
   DB_NAME = 'a'
  

Dataset (./syzbot_data)

The data we use to perform the study.

Statistics (./analyses)

Before continuing, you should check the paths and database address in config.py

• gt_filter.py

  Filter out the unreasonable fixes tags.

  See details in the code comments.

• ground_truth.py

  Build the ground truth.

  Note: You can directly load the syzbot_data/syzbot_bug_info.sql into your database, instead of building from scratch.

• data.py

  Calculate the following statistics:

  1. overall performance;

  2. impact on bug-fixing time;

  3. distribution of result commits

• efficiency_analysis.py

  Calculate the following statistics:

  1. Avg building time and testing time for each commit;

  2. analysis of the tested commits who cost more time than avg time, i.e. expected commits to test VS. actual.

• failure_cause_analysis.py

  Analyze the failure causes (C1/C2/T1/T2/T3)

• dreamutil.py

  Extract the files and functions modified by a commit.

  Only support C/C++.

  Parsing grammar is based on srcml.

• file_maintainer.py

  Obtain the maintainer information of guilty file, and output to maintainers_crash_report.json

• relation_between_two_commits.py

  Examine the relationship between the result commit and patch commit, including the developer assignment and code location (line, func, file), respectively for correct/incorrect bisection result.

• time_limit.py

  Calculate the avg number of versions tested if a timeout occurs.

• dataset_dist.py

  Count the version distribution of ground-truth commit and crash commit.