Wires in an array cannot be a non-trivial function of another wire in the same array about cascade HOT 2 CLOSED

This is tough. The hanging you see is a spurious instance of a zero-time loop (combinational logic which feeds back into itself). I say spurious because cascade treats arrays as monolithic entities rather than as collections of individually nameable objects.

Test case 1 doesn't look like a zero time loop, but cascade thinks that whenever grid[1] changes, that means that every other element in grid (grid[0] included) has changed as well. The same is true of test case 2. It's just a quirk of the way that cascade evaluates changes to combinational logic that it doesn't fall into an infinite loop in this case as well.

So there's a couple of things that need to be fixed here:

1. One way or another, cascade's evaluation logic should be changed so that it doesn't fall into an infinite loop for non-zero-time loops.
2. We may also want to update the typechecker to detect zero time loops and either emit an error, or emit a warning and treat zero-time loops as undefined and replace them with terminating logic somewhere downstream.

from cascade.

Fixed in issue-138.

Actually not all that difficult. In sw_logic, all that's necessary is to notify a variable's monitor only when a call to a call to assign_value produces a change in value.

from cascade.