using "code coverage coloring" to help debuging

When I tried to investigate a defect, usually the first thing that I want to verify is the “runtime” execution flow. And if the “runtime” flow didn’t match the “expected” flow, I would be very close to the root cause of a defect. However, sometimes it’s not very easy to tell the “real/runtime” execution flow (especially when you can’t attach your debugger to the target assembly, or when the defect can’t be reproduced stably). I used to add more debug info output to the logs to help me out, and sometimes it could be a very tedious and time-consuming process. When I tried to investigate a defect, usually