Saving Interlocks By Using The Tfr Instruction - Motorola DSP56600 Manual

Pipeline Interlocks
Data ALU Pipeline Interlocks
6-6 Optimizing DSP56300/DSP56600 Applications
move x:(r0)+,bb,y:(r4)+
_end
move a,y:(r4)+
move b,y:(r4)+
6.1.2.3

Saving Interlocks by Using the TFR Instruction.

The following C code adds a constant to two memory arrays, one in
X memory space and the other in Y memory space:
static int a[N],b[N];
int i;
for (i=0;i<N;i++)
{
b[i] = b[i]+c;}
for (i=0;i<N;i++)
{
a[i] = a[i]+c;}
The straightforward implementation of the code will execute in 8N
cycles:
move var_a,r4
move var_b,r0
move var_c,x0
do #N,_1Loop
move y:(r4),a
add x0,a
move a,y:(r4)+
_1Loop
do #N,_2Loop
move x:(r0),a
add x0,a
move a,x:(r0)+
_2Loop
By combining the two loops into one and using the TFR instruction,
an optimized implementation takes only 1.5 cycles for main loop
iteration summing up to 3N cycles for the whole task:
move var_a,r4
move var_b,r0
lua (r4)+,r5
lua (r0)+,r1
move var_c,x1
move x:(r0),b
add x1,b
do #N,_3Loop
add x1,a
add x1,b
tfr y1,a
_3Loop
;previous data
;write destination memory,
;read next data
;write last-1 word to
;destination memory
;write last word to destination
;memory
;a array in Y:memory space
;b array in X:memory space
;constant to add
;handle Y array
;read data word
;add constant
;store result and increment pointer
;handle X array
;read data word
;add constant
;store result and increment
;pointer
x:(r1)+,x0 y:(r4),a
b,x:(r0)+ x0,b
y:(r5)+,y1
x:(r1)+,x0 a,y:(r4)+
;a array in Y memory
;b array in X memory
;r5 = r4 + 1
;r1 = r0 + 1
MOTOROLA