[P4Testgen]: Support action selectors and action profiles for P4Testgen v1model PTF tests

[Copilot]

• Add one-shot action profile and selector configuration support to P4Testgen.
• Make sure that the base_test.py PTF module supports valid action profile and selector configuration
• Do not require root for nanomsg PTF tests.

[jafingerhut]

If this thing converges to a solution on its own, that would be pleasantly surprising :-)

[fruffy]

If this thing converges to a solution on its own, that would be pleasantly surprising :-)

The problem with the free tier is that it has a time limit. Our builds take a while and so it always terminates before doing anything. I'll play around with it a bit more to see whether I can find a functioning approach.

[fruffy]

@jafingerhut I had to hand this over to a local GPT 5.3 model. Unfortunately, the online Git models are a bit brain-damaged, because they have a limited time-budget and less context allowed, from what I can tell.

Still, now we actually have action selector programming working for the PTF BMv2 framework. There is some question on the exact semantics for one-shot programming, I tried to follow the ones described here: https://p4lang.github.io/p4runtime/spec/main/P4Runtime-Spec.html#sec-oneshot

Weight must be non-zero. Maybe we should make this configurable.

[fruffy]

Marking as breaking change because we change the base-test API. I can also split this out into a separate PR.

[jafingerhut]

https://p4lang.github.io/p4runtime/spec/main/P4Runtime-Spec.html#sec-oneshot

Is there some benefit to allowing weights in one-shot action selector configuration to be 0? It is not clear to me what that might be.

[jafingerhut]

Marking as breaking change because we change the base-test API. I can also split this out into a separate PR.

I have only skimmed through the base_test.py changes so far, but in every case I checked I saw only new optional parameters added to existing method calls. Is there an example of a change in that file you consider breaking?

[fruffy]

I have only skimmed through the base_test.py changes so far, but in every case I checked I saw only new optional parameters added to existing method calls. Is there an example of a change in that file you consider breaking?

I went back and added _normalize_group_members to make sure we preserve API compatibility, but it might be overkill and make the internals more complicated than necessary. If we drop it the parameters become mandatory and breaking.

Also _validate_selector_weight will now throw errors.

[jafingerhut]

Looks reasonable to me, although I can't claim to have digested every line of change.