Texas Instruments TMS320C6000 Programmer's Manual page 185

Software Pipelining
Example 6–29. Assembly Code for Floating-Point Dot Product (Software Pipelined
With No Extraneous Loads)
MVK
|| ZERO
|| ZERO
|| LDDW
|| LDDW
LDDW
|| LDDW
LDDW
|| LDDW
LDDW
|| LDDW
||[A1] SUB
LDDW
|| LDDW
||[A1] B
||[A1] SUB
LDDW
|| LDDW
|| MPYSP
|| MPYSP
||[A1] B
||[A1] SUB
LDDW
|| LDDW
|| MPYSP
|| MPYSP
||[A1] B
||[A1] SUB
LDDW
|| LDDW
|| MPYSP
|| MPYSP
||[A1] B
||[A1] SUB
LDDW
|| LDDW
|| MPYSP
|| MPYSP
||[A1] B
||[A1] SUB
6-48
.S1 41,A1
.L1 A8
.L2 B8
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.S1 A1,1,A1
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.S2 LOOP
.S1 A1,1,A1
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.M1X A6,B6,A5
.M2X A7,B7,B5
.S2 LOOP
.S1 A1,1,A1
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.M1X A6,B6,A5
.M2X A7,B7,B5
.S2 LOOP
.S1 A1,1,A1
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.M1X A6,B6,A5
.M2X A7,B7,B5
.S2 LOOP
.S1 A1,1,A1
.D1 A4++,A7:A6
.D2 B4++,B7:B6
.M1X A6,B6,A5
.M2X A7,B7,B5
.S2 LOOP
.S1 A1,1,A1
; set up loop counter
; sum0 = 0
; sum1 = 0
; load ai & ai + 1 from memory
; load bi & bi + 1 from memory
;* load ai & ai + 1 from memory
;* load bi & bi + 1 from memory
;** load ai & ai + 1 from memory
;** load bi & bi + 1 from memory
;*** load ai & ai + 1 from memory
;*** load bi & bi + 1 from memory
; decrement loop counter
;**** load ai & ai + 1 from memory
;**** load bi & bi + 1 from memory
; branch to loop
;* decrement loop counter
;***** load ai & ai + 1 from memory
;***** load bi & bi + 1 from memory
; pi = a0 b0
; pi1 = a1 b1
;* branch to loop
;** decrement loop counter
;****** load ai & ai + 1 from memory
;****** load bi & bi + 1 from memory
;* pi = a0 b0
;* pi1 = a1 b1
;** branch to loop
;*** decrement loop counter
;******* load ai & ai + 1 from memory
;******* load bi & bi + 1 from memory
;** pi = a0 b0
;** pi1 = a1 b1
;*** branch to loop
;**** decrement loop counter
;******** load ai & ai + 1 from memory
;******** load bi & bi + 1 from memory
;*** pi = a0 b0
;*** pi1 = a1 b1
;**** branch to loop
;***** decrement loop counter