Faddp With 8-Bit Pixels; Faddp With 16-Bit Pixels; Faddp With 32-Bit Pixels - Intel i86W Manual

FLOATING-POINT INSTRUCTIONS
Figure 6-6 illustrates
faddp
when PS is set for 8-bit pixels. Since
faddp
adds 16-bit values
in this case, each value can be treated as a fixed-point real number with an 8-bit integer
portion and an 8-bit fractional portion. The real numbers are rounded to 8 bits by
truncation when they are loaded into the MERGE register. With each
faddp
instruction,
the MERGE register is shifted right by 8 bits. Two
faddp
instructions should be executed
consecutively, one to interpolate for even-numbered pixels, the next to interpolate for
odd-numbered pixels. The shifting of the MERGE register has the effect of merging the
results of the two
faddp
instructions.
63
47
31 15
o
fsrc1
I
I I
i
INT
I
FRAC
INT
I
FRAC
INT
I
FRAC
INT
FRAC
I
I I I
I
I I I
++++++
++++++
++++++ ++++++
63 47 31 15
o
INT
FRAC
INT
FRAC
INT
FRAC
INT
FRAC fsrc2
INT
FRAC
INT
FRAC
INT
FRAC
INT
FRAC fdest
o
MERGE
240329i
Figure 6-6. FADDP with a-Bit Pixels
6-33