Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.Sign up
Fuzzer: Internal error after other error on existing testcase #1559
Author Name: Eric Rippey
with test existing testcase at "test_regress/t/t_order_loop_bad.v", which is a regression test for this ancient bug:
%Error: Circular logic when ordering code (non-cutable edge loop)
So it diagnoses that there is a loop, which the test harness look for. But then it goes on to produce the internal error message.
I would assume that it's intended that when there's a construct that Verilator's design can't support it should just be an error given to the user and all paths that lead to "Internal Error" are meant to be unreachable. But I don't know this for sure.
Original Redmine Comment
In general as long as there is an %Error printed first then I'm going to consider a later internal error as being acceptable. Otherwise there's too much stuff that will need to be cleaned up, and while it would be nicer not to print the internal error, cleaning it up is of low value (relative to other fixes).
If the internal error comes first, or there's a core dump, then that's something we should fix.