Closed marlonjames closed 3 months ago
I think this was broken by 7dd41079090dcf. There's still a few failures if I run the cocotb testsuite with the current master but I'm using #857 to track that. Once that's fixed I'll add cocotb's full testsuite to the CI to avoid regressions in the future (at the moment it only runs the examples).
I also implemented vhpiParamDecl
for vhpi_handle()
and vhpiGenerateIndexP
for vhpi_get
but cocotb doesn't seem to be using those at the moment.
Thanks for the fix! cocotb runs its regression suite on the master branch of several open source simulators weekly: https://github.com/cocotb/cocotb/actions/workflows/experimental.yml
In
v1.11.3
, iterating forvhpiDecls
,vhpiSigDecls
, orvhpiConstDecls
from a reference handle of kindvhpiForGenerateK
would give an iterator handle that then gives declaration handles.On master, these iterators no longer return any handles to declarations when iterating with
vhpi_scan
.I don't have a minimal example, but the VHDL used by the cocotb regression test suite is here:
https://github.com/cocotb/cocotb/blob/785e0896930039023db3c2e2be47bed5375326c5/tests/designs/array_module/array_module.vhd#L149-L156
If you want, I can try to put together a test for NVC.