More robust and dynamic way to obtain fuzz target info

[DonggeLiu]

1. Improve the existing logic of getting fuzz target binary name and source file path to update benchmark yaml.
2. Dynamically re-generate fuzz target info at runtime.
3. If dynamic generation fails, use the default one in benchmark yaml.
4. Record which binary and source file is used during the experiment to avoid confusion in debugging.

[DavidKorczynski]

Ref: https://github.com/google/oss-fuzz-gen/pull/524#issuecomment-2264668798

[DavidKorczynski]

I think we can also use FI data more here to increase reliability

[oliverchang]

@DavidKorczynski (or @arthurscchan ), can you help us remove the hardcoded fuzz target binary names from our benchmark YAML configs? This is a recurring source of regressions as targets change names.

e.g. https://llm-exp.oss-fuzz.com/Result-reports/scheduled/2024-08-31-weekly-all/sample/output-tinyxml2-_zn8tinyxml210xmlelement12setattributeepkcs2_/01.html