Computer Hardware User's Guide

Floating-Point Conversion (IEEE Std. 754)
5-15
Data Formats and Floating-Point Operation
Figure 5–15. TMS320C3x Single-Precision 2s-Complement Floating-Point Format
e
f
31 23 22 024
s
Note: Same format as for the ’C4x
In comparison, Figure 5–15 shows the the ‘C3x 2s-complement floating-point
format. In this format, two cases can be used to define value
v
of a number:
1) If
e
= –128 then
v
= 0
2) If
e
–128 then
v
=
ss.f
2
2
e
where:
s
= sign bit
e
= the exponent field
f
= the fraction field
For this representation,
e
is treated as a 2s-complement integer. The fraction
and sign bit form a normalized 2s-complement mantissa.
Note: Differentiating Symbols for IEEE and TMS320C3x Formats
To differentiate between the symbols that define these two formats, all IEEE
fields are subscripted with an IEEE (for example,
e
IEEE
,
s
IEEE
, and so forth).
Similarly, all 2s-complement fields are subscripted with 2 (that is,
e
2
,
s
2
,
f
2
).
5.4.1 Converting IEEE Format to 2s-Complement TMS320C3x Floating-Point Format
The most common conversion is the IEEE-to-2s-complement format. This
conversion is done according to rules in Table 5–1.
Table 5–1. Converting IEEE Format to 2s-Complement Floating-Point Format
If these values are present Then these values equal
Description Case e
IEEE
s
IEEE
f
IEEE
e
2
s
2
f
2
max neg 1 255 1 any 7Fh 1 00 0000h
max pos 2 255 0 any 7Fh 0 7F FFFFh
30<e
IEEE
<255 0 f
IEEE
e
IEEE
7Fh 0 f
IEEE
40<e
IEEE
<255 1 0e
IEEE
7Fh 1 f
IEEE
+1
50<e
IEEE
<255 1 0 e
IEEE
80h 1 0
zero
6 0 any any 80h 0 00 0000h
f
IEEE
= 1s complement of f
IEEE