Kaleb Barrett
Kaleb Barrett
So what about this for TestCases? ```python class TestStuff(cocotb.TestCase): fail_fast = True async def setUp() -> None: ... @cocotb.test() async def test1() ->None: ... @cocotb.test() async def test2() -> None:...
I think we decided to do this in 1.7 instead.
Supporting 3.7+ allows us to * drop the _py_compat library as everything in there is for compat with < 3.7 * allows the use of `from __future__ import annotations` to...
To best guide the improvement, what exactly do you think is missing or needs improvement?
I'm not sure if filter functions is the best way. When you "don't care" you don't want to filter all of those tests, you want to keep one test; and...
Please provide a minimum reproducible failing example. GHDL uses VPI, which does not contain enough information to map all VHDL features, so I'm not surprised if something isn't being discovered...
Change your timeunit to something reasonable. All cocotb tests end after a 1 step Timer to ensure the next test starts in a normal time phase. Your steps are enormous...
I don't see `timeunit 1s` in that documentation.
> and registers the callbacks as a repeated callback, which seems to be different than other simulators. The behavior of only servicing callbacks once per registry makes sense to me....
Since this is potentially a bug in Verilator, should this be removed from 1.5?