A note on the cone conjecture for K3 surfaces in positive characteristic