Lucent Technologies DSP1611, Information Manual

Page: 409 / 443

Information Manual DSP1611/17/18/27/28/29 DIGITAL SIGNAL PROCESSOR
April 1998 Instruction Set Summary
Lucent Technologies Inc.
DRAFT COPY
B-28
F1 y = Y x = *pt++[i]
(multiply/ALU operation with parallel load of x and y registers)
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:
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 Section 3.3, Arithmetic and Precision , for the options of
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 01 - r1 10 - r2 11 - r3
F1 Operation F1 Operation F1 Operation
0000 aD = pp = x * y 0110 nop 1011 aS – y
0001 aD = aS + pp = x * y 0111 aD = aS – p 1100 aD = y
0010 p = x * y 1000 aD = aS | y 1101 aD = aS + y
0011 aD = aS – pp = x * y 1001 aD = aS ^ y 1110 aD = aS & y
0100 aD = p 1010 aS & y 1111 aD = aS – y
0101 aD = aS + p