H ∞ -control by state feedback: an iterative algorithm and characterization of high-gain occurrence