[HELP] How to check the correctness of benchmark execution?

[sunggg]

Hi, all.
I'm working on analyzing V8 JS engine and making some modification to understand the code structure.
I want to make sure my modification does not harm the correctness of program execution, but it seems like there is no output from the benchmark that can be compared with.

Is there anyone who knows how to measure the correctness?
Any advice/idea/suggestion will be greatly helpful.
Thank you.