Fix isBuffer

[akashlevy]

I discovered a bug when one of our tests failed on is_buffer property check on SKY130. It's clearly a buffer but is marked as a non-buffer because of the wells in SKY130. I've included a minimal example in the tests/ directory.

This PR adds a PortDirection enum value for wells (nwell, pwell, deepnwell, deeppwell) and ignores them in a couple places where they should be ignored (I just searched for isPowerGround() and evaluated whether wells should be ignored at those places).

Let me know if any changes are needed. Thanks!

[jjcherry56]

Well I guess we can argue about the need for a well port direction over using LibertyPort::isPwrGnd, which means pg_pin in liberty speak and is probably what you really mean for excluding wells.

[akashlevy]

The difference between isPowerGround and isPwrGnd was confusing me. I didn't notice isPwrGnd before, that seems to obviate the need for distinguishing wells.

Probably the best solution is to simply replace most instances of isPowerGround (which matches only power and ground) with isPwrGnd (which matches any valid pg pin). Shall I amend this to do that?

[akashlevy]

(Also isPwrGnd could be renamed to is isPGPin or something like that for more clarity)

[maliberty]

(Also isPwrGnd could be renamed to is isPGPin or something like that for more clarity)

I can't say I find that clearer and we are in the process of integrating with the current name.

[parallaxsw]

It could. But in OpenSTA speak it is not a pin; it is a port. And there is no point in having a Port in the name of a port member function because it is redundant. The header comment tries to make the pg_pin related functions clear.

…

On Fri, Nov 14, 2025 at 3:39 PM Akash Levy ***@***.***> wrote: *akashlevy* left a comment (parallaxsw/OpenSTA#329) <#329 (comment)> (Also isPwrGnd could be renamed to is isPGPin or something like that for more clarity) — Reply to this email directly, view it on GitHub <#329 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BAKEVHAQOFF6LAZU7GOTIAL34ZK3LAVCNFSM6AAAAACLU4WBHGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTKMZUHEZTQNRXG4> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[akashlevy]

Yeah, let's minimize changes; I don't care to argue for well as a port direction or isPwrGnd naming.

Let's just fix the issue at hand, which is isBuffer. One-line fix with simple test case demonstrating the issue