Q. When a developer writes functional coverage, with what intentions they write. Do they write such a way to get 99-100% coverage Or they code so rigid that coverage can never be maximum.
Q. If coverage is 100%, then is it shoes that developer has coded weak coverage? Or 100% coverage is good for developer and verification guy both?
In reply to tech_savvy:
I don’t think I will have the best answer for the question, but here is my recommendation
Answer for Question 1:
- Write functional coverage as much as you think that no more areas in the design needs any additional “functional coverage”.
- This means, you may get functional coverage items that will never hit
- In this case, you will analyze all items that are never hit. Then you will exclude it with a comment why you decided to exclude it.
- You will keep iterating the following 2 points:
- Excluding items that will never hit, or are not important or hard to be hit (and don’t forget to document your analysis)
- Running more tests
- Till you get 100% coverage
Answer for Question 2:
- I think 100% coverage with exclusion and documentation for each excluded item is good for both.