-- zero and can therefore represent both zero and negative
-- values. They have 8 bits of precision and in \"roundtrip\"
-- conversions the total error never exceeds 1 part in 225 except for
-- zero and can therefore represent both zero and negative
-- values. They have 8 bits of precision and in \"roundtrip\"
-- conversions the total error never exceeds 1 part in 225 except for
--
-- This function returns the log2 for the specified 32-bit unsigned
-- value. The maximum value allowed is about 0xff800000 and returns
--
-- This function returns the log2 for the specified 32-bit unsigned
-- value. The maximum value allowed is about 0xff800000 and returns