Kaleb Barrett

I was not able to reproduce. 1 million test generations takes a couple seconds on my machine regardless if it's 1.9.2, 2.0.0, stable/2.0, or master. Please provide a full reproducer:...

Ah I see, this is specifically with `TESTCASE` being used. This is due to how cocotb handles test generation and filtering in 2.0. In 1.X `TESTCASE` would cause the RegressionManager...

Sounds like this is two requests then: make it fast and provide a test listing functionality. I'll open a new issue for the second one. @tymonx Yeah pytest will provide...

~~@cmarqu I don't understand why only this PR has the problem. It seems like it goes back to #4734.~~ Fixed in #4765. Rebased and fixed.

On master the `test_mixed_language` test, when run with pytest, does the wrong thing and loads into the VPI when running a VHDL test. I'm not sure why fixing that issue...

So AFAICT the issue is with how sources are built on Questa. The Makefile builds all VHDL sources at once, then the Verilog, while the runner builds them in the...

Is there something in particular you think is missing? That document you linked references both pyest and the makefiles, maybe not completely consistently. The Runners dont support qrun flow yet....

With #5160 in we could revisit this.

There seems to be some info left out of the issue, the "With the following test code:" section looked left accidentally half-blank. Crap. I guess my memory failed me there....

Yes, you still need that call to `_make_sim_handle`. ``` def get_elem(logic_array: LogicArrayObject, idx: int) -> LogicObject: return cocotb.handle._make_sim_object(logic_array._handle.get_handle_by_index(idx))