The result of a floating point operation cannot be represented perfectly. For example the value 1/3 cannot be represented in binary form exactly.
Most programs that handle math will ignore this exception since the accuracy error is quite minor and generally acceptable.
Floating-point inexact result
Linear Mode
