An operation resulted in a Floating-point underflow. If the
exponent of a numeric result is too small for the destination real
format, this exception is issued. This can occur with most arithmetic
instructions and some of the conversion instructions.
Floating-point underflow
Linear Mode
