It is inherently double precision, and 1.0 is perfectly represented
anyway.
Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>