Kaleb Barrett
Kaleb Barrett
Whose decision is this waiting on? I think the first solution is the least disruptive and I would also like to see this rule split between `Union` and `Optional` as...
Perhaps `Optional` *shouldn't* mean the same thing as `X | None`, but instead pass X, or don't give an argument, giving a warning otherwise. Not that I expect this to...
Questa FLI passes `test_get_root_handle`, not fails. Questa VHPI fails `test_package_access_vhdl` and `test_get_root_handle` causes a crash so it should be skipped. I have no idea why the discovery tests are different...
Our CI only tests released versions of simulators, so we will have to wait until the version of NVC with the fix is released before integrating this PR. But I...
~~It would be trivial to add something to try to construct any value into `BinaryValue` at the beginning of the assignment operation. However, `BinaryValue` isn't the greatest type, there are...
Closed by #3777.
Oh wait... the last two comments only apply after #3777. Whoops... I will probably try to finish out #3777 before merging this.
For `test_iteration.recursive_discovery` Questa and Xcelium VPI now find 283 instead of 275. For Riviera failures: `test_array_simple.test_struct_unpacked` and `test_struct.test_struct_format` now passes.
Cocotb does not put limits on the size of objects, we simply return whatever the VPI gives us. This is probably an upstream issue, try opening an issue against Verilator.
Upstream issue reported.