Another solution is to split the dependent member into a separate sub-module and have both existing modules depend on the new sub-module.

Suppose that we have two sub-modules set up in a way that they use functions from each other. Consider the scenario depicted in the diagram as follows:

The func1 function from the first sub-module uses func6 from the other sub-module. And, the func4 function from the other sub-module needs to call the func3 function from the first module. Clearly, there is a high coupling between these two modules.

Considering the dependencies between these modules, it looks like a cycle as the first sub-module depends on the second sub-module and vice versa. To fix that, we can introduce a new sub-module to break the cycle as follows:

Breaking the cycle has the benefit of a cleaner dependency graph. It also makes the code easier to understand.