Boundary-value problem