Run two bpf_conformance tests in parallel and compare the results.

[Alan-Jowett]

This pull request introduces a new script run_corpus.py to compare the outputs of two bpf_conformance plugins against a corpus of BPF programs. The script ensures that the plugins produce equivalent results, which is crucial for verifying plugin consistency.

Key changes include:

New Script Addition:

• Script Purpose and Description:

  • Added a script to run two bpf_conformance plugins against a corpus of BPF programs and compare their results. (libfuzzer/run_corpus.py)

• Command Line Argument Parsing:

  • Implemented argument parsing to read plugin paths, options, and corpus directory from the command line. (libfuzzer/run_corpus.py)

• Corpus File Handling:

  • Added logic to enumerate and validate files in the corpus directory. (libfuzzer/run_corpus.py)

• Plugin Execution and Comparison:

  • Implemented the execution of both plugins with the corpus files and compared their outputs. Differences in outputs are reported. (libfuzzer/run_corpus.py)

• Debugging and Error Handling:

  • Added debug flag for detailed output and comprehensive error handling to manage file and process-related errors. (libfuzzer/run_corpus.py)

[coveralls]

coverage: 79.171% (-0.08%) from 79.25%
when pulling 9525129 on Alan-Jowett:cross_check
into 78729ed on iovisor:main.