separate ( Generic_Elementary_Functions ) function Exp( X : Float_Type ) return Float_Type is -- On input, X is a floating-point value in Float_Type; -- On output, the value of exp(X) (the natural exponential of X) is returned. -- The bulk of the computations are performed by the procedure -- KP_Exp (Y, M, Z1, Z2) which returns exp(Y) in M, Z1, and Z2 -- where -- exp(Y) = 2**M * ( Z1 + Z2 ) -- M of integer value, and Z1 only has at most 12 significant bits. Result : Float_Type; Y, Z1, Z2 : Common_Float; M, J : Common_Int; Two_to : constant array ( Common_Int range -3..3 ) of Common_Float := ( 0.125, 0.25, 0.5, 1.0, 2.0, 4.0, 8.0 ); Large_Threshold : constant Common_Float := 4.0 * Common_Float(Float_Type'Safe_Emax) * 6.931471806E-1; Small_Threshold : constant Common_Float := Float_Type'Base'Epsilon; begin -- Filter out exceptional cases. Y := Common_Float( X ); if abs(Y) >= Large_Threshold then if ( Y > 0.0 ) then raise Constraint_Error; else return( Float_Type( 0.0 ) ); end if; elsif abs(Y) <= SMALL_THRESHOLD then return( Float_Type( 1.0 + Y ) ); end if; -- Get the values of M, Z1, and Z2 so that the natural exponential of Y -- can be calculated by Exp(Y) = 2**M * (Z1 + Z2) KP_Exp (Y, M, Z1, Z2); case Common_Float'Machine_Radix is when 2 => Y := Z1 + Z2; when others => J := M rem 4; M := (M - J) / 4; Z1 := Z1 * Two_to(J); Z2 := Z2 * Two_to(J); Y := Z1 + Z2; end case; Result := Float_Type( Scale( Y, M ) ); return( Result ); end Exp;