IEEE-754 Floating point software simulation

Features

(and thier absence)

• all functions are standard compliant from IEEE-754-2019
• signaling exceptions is not supported
• only rounding TOWARD_ZERO is supported
• no magic values in code
• build for c++20

Implementation

under the hood it uses internal represetation struct IEEE754Fields with extended precision, so there is no need for special handling of subnormal values, pack() and unpack() functions take care of that.

To implemented

• diffrent roundings
• remainder(x,y)
• from_characters(str) / from_hex_characters(str)
• into_characters(str) / from_hex_characters(str)