Motorola MVME166IG 用户手册

Floating Point Support

4-19

This format would appear in memory as:

A packed decimal number takes 12 bytes in memory.  

This format provides a convenient way to enter and display a floating point 
decimal number.  Internally, the number is assembled into a packed decimal 
number and then converted into a number of the specified data type.  

Entering data in this format requires the following fields: 

An optional sign bit (+ or -).  
One decimal digit followed by a decimal point.  
Up to 17 decimal digits (at least one must be entered).  
An optional Exponent field that consists of: 

An optional underscore.  
The Exponent field identifier, letter "E".  
An optional Exponent sign (+, -).  
From 1 to 3 decimal digits.  

For more information about the MC68040 floating point unit, refer to the 
M68040 Microprocessor User’s Manual.  

4-bit sign field

(4 binary digits)

16-bit exponent field

(4 hex digits)

68-bit mantissa field

(17 hex digits)