+ * new feature: the iterative spilling/coloring register allocator developed
+ by Alexandra Barchunova during Google Summer of Code 2013 is now merged
+ in. By default, it only activates for functions that optimize with
+ (speed 3) and (> speed compilation-speed), but setting
+ sb-regalloc:*register-allocation-method* to :iterative forces its
+ execution. The previous behaviour can be obtained by instead setting that
+ variable to :greedy. Thanks again to Google for their support, and, more
+ crucially, to Alexandra Barchunova for her hard work.