consider state-dependent leakage during sizing

[openroad-ci]

No description provided.

[github-actions]

clang-tidy review says "All clean, LGTM! 👍"

[github-actions]

clang-tidy review says "All clean, LGTM! 👍"

[rovinski]

What's the point of taking the average state-based leakage? For all PDKs I can remember, the non-state-based leakage is already calculated as the average of the state-based leakages. This would only help if the average is weighted by state probability.

[precisionmoon]

What's the point of taking the average state-based leakage? For all PDKs I can remember, the non-state-based leakage is already calculated as the average of the state-based leakages. This would only help if the average is weighted by state probability.

Correct, if a liberty has "cell_leakage_power" line, it is the average of the state-based leakages. Most public PDKs have the average leakage. However, many private PDK liberty files don't have this line. They only have multiple leakage values for different when conditions.