1 Tests a regression in float addition for optimizing. The second argument 2 could be now be a constant for floating point numbers. 3
