grt: add prevent division by zero and overflow when computing resistance cost

[eder-matheus]

Summary

Check if default_resistance and final_resistance values are valid for the division performed later. This shouldn't impact any of our designs, but acts as a guard for cases where resistance is not defined for a specific routing/via layer.

Type of Change

• Bug fix

Impact

No impact on ORFS designs.

Verification

• I have verified that the local build succeeds (./etc/Build.sh).
• I have run the relevant tests and they pass.
• My code follows the repository's formatting guidelines.
• I have signed my commits (DCO).

Related Issues

[Link issues here]

[eder-matheus]

@jfgava FYI.

[github-actions]

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

[gemini-code-assist]

Code Review

This pull request introduces safety checks in getWireCost and getViaCost to prevent division by zero and handle large resistance values. Feedback highlights a remaining risk of integer overflow when the divisor is a very small positive number and suggests calculating the cost as a double with explicit clamping to BIG_INT before casting to an integer.

[Divinesoumyadip]

Nice fix @eder-matheus Just one thing I noticed that if final_resistance is a very small positive float (not zero but close to it), the ceil result could still overflow when implicitly cast to int.
Something like this might be safer:
return static_cast<int>(std::min(std::ceil(final_resistance / default_resistance), static_cast<double>(BIG_INT)));
Same edge case applies to the via cost path too. Rest looks clean.

[github-actions]

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

[maliberty]

Is this for some sort of benchmark? It would be odd to have no layer resistance in a production pdk.

[eder-matheus]

Is this for some sort of benchmark? It would be odd to have no layer resistance in a production pdk.

I faced this issue on a private PDK I'm working on. I added this guard to prevent crash/errors on it.