F1 Y = Y X = *Pt++[I - Lucent Technologies DSP1617 Information Manual

Information Manual
April 1998

F1 y = Y x = *pt++[i]

perform operation F1 and in parallel, perform the following data moves:
(y) ← (*rM); then
modify rM; then
(x) ← (*pt); then
(pt) = (pt)+ [1 or i]
This instruction performs the following operations effectively in parallel:
1. The operation F1 is performed. The possible operations for F1 are as follows:
F1 Operation
0000 aD = pp = x * y
0001 aD = aS + pp = x * y
0010 p = x * y
0011 aD = aS – pp = x * y
0100 aD = p
0101 aD = aS + p
The value of S can be zero to select a0 or one to select a1. The value of D can be zero to select a0 or one to
select a1. Flags are modified based on the value computed by the DAU.
Note: For all diadic operations involving the y register, y is sign-extended to 36 bits before performing the
operation (including logical operations). (See
shifting the output of the p register into aS in the above operations.)
2. Access the Y-space location pointed to by rM, and write this value into the y register. rM is specified by the
two most significant bits of the Y field:
00 - r0
Lucent Technologies Inc.
DSP1611/17/18/27/28/29 DIGITAL SIGNAL PROCESSOR
(multiply/ALU operation with parallel load of x and y registers)
F1
0110
0111
1000
1001
1010
01 - r1
DRAFT COPY
Operation
nop 1011
aD = aS – p 1100
aD = aS | y 1101
aD = aS ^ y 1110
aS & y 1111
Section 3.3, Arithmetic and
10 - r2
Instruction Set Summary
F1 Operation
aS – y
aD = y
aD = aS + y
aD = aS & y
aD = aS – y
Precision, for the options of
11 - r3
B-28