$M$ is an irreducible $R$ module $\iff$ $M$ is a cyclic module and every nonzero element is a generator.
($\rightarrow$) If $M$ is an irreducible $R$-module then it's obvious that $M$ is a cylclic module where every nonzero element is a generator, otherwise $mR <_R M$ would be a submodule of M.
($\leftarrow$) So, this would be obvious if we knew that every $M$ submodule is somehow related to the cyclic submodules of $M$, like a direct product or something. If $M$ was free then this would be easy, but as it is this suprises me, I feel like there should be some weird counterexample where there is some module $M$ with some submodule not related to its cyclic submodules. Why is this not possible?