在DSP上实现DCT的三角波/方波/正弦波的显示代码(2）

1万 · 2014-11-12 15:51

*********************************************************************
;          *
; PROCESS FOURTH 8 INPUTS. X(3,0)...X(3,7)  ->  Y(3,0)...Y(3,7)  *
;          *
*********************************************************************

      LD    *AR1+, 4, A          ; + (16)*(X0)
      ADD *AR1+, 4, A          ; + (16)*(X1)
      ADD *AR1+, 4, A          ; + (16)*(X2)
      ADD *AR1+, 4, A          ; + (16)*(X3)
ADD *AR1+, 4, A          ; + (16)*(X4)
ADD *AR1+, 4, A          ; + (16)*(X5)
ADD *AR1+, 4, A          ; + (16)*(X6)
ADD *AR1, 4, A          ; + (16)*(X7)
STL A, Y30             ; = Y30

      RPTZ A, #7
      MACP *AR1-, COEF_F1, A
      SFTA A, 4
      MAR *AR1+                ; (64)[A*X0+B*X1+C*X2-D*X3-D*X4
      ADD ROUND1, A          ; -C*X5-B*X6-A*X7] + 4*ROUND1
      STH A, 2, Y31          ; = Y31

      LD    ROUND1, -4, A       ; (ROUND1)/16
      LD    E_P6, T
      MAC *AR1+0, A          ; + X0*E_P6
      MAS *AR1+, A             ; - X3*E_P6
      MAS *AR1+0, A          ; - X4*E_P6
      MAC *AR1-, A             ; + X7*E_P6
      LD    F_P6, T
MAC *AR1-, A             ; + X6*F_P6
      MAS *AR1-0, A          ; - X5*F_P6
      MAS *AR1-, A             ; - X2*F_P6
      MAC *AR1-, A             ; + X1*F_P6
      STH A, 6, Y32          ; Multiply prev summation by 16
                                 ; = Y32

      RPTZ A, #7
      MACP *AR1+, COEFF2, A    ; 64(B*X0-D*X1-A*X2-C*X3+C*X4+A*X5
      SFTA A, 4
      MAR *AR1-
      ADD ROUND1, A          ; +D*X6-B*X7) + 4*ROUND1
      STH A, 2, Y33          ; = Y33

LD *AR1-, 4, A          ; X7*16
      SUB *AR1-, 4, A          ; - X6*16
      SUB *AR1-, 4, A          ; - X5*16
      ADD *AR1-, 4, A          ; + X4*16
      ADD *AR1-, 4, A          ; + X3*16
      SUB *AR1-, 4, A          ; - X2*16
      SUB *AR1-, 4, A          ; - X1*16
      ADD *AR1, 4, A          ; + X0*16
      STL A, Y34                ; = Y34

      RPTZ  A, #7
      MACP  *AR1+, COEFF3, A    ; 8(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1-
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y35             ; = Y35

LD ROUND1, -4, A       ; (ROUND1)/16
      LD F_P6, T
      MAC *AR1-0, A             ; + F_P6*X7
      MAS *AR1-, A             ; - F_P6*X4
      MAS *AR1-0, A             ; - F_P6*X3
      MAC *AR1+, A             ; + F_P6*X0
      LD E_P6, T
      MAS *AR1+, A             ; - E_P6*X1
      MAC *AR1+0, A             ; + E_P6*X2
      MAC *AR1+, A             ; + E_P6*X5
      MAS *AR1+, A             ; - E_P6*X6
      STH A, 6, Y36             ; Multiply prev summation by 16
                                 ; = Y36
      RPTZ A, #7
      MACP  *AR1-, COEF_F4, A    ; 64(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1+
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y37             ; = Y37

      MAR *+AR1(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs

*********************************************************************
;          *
; PROCESS FIFTH 8 INPUTS.    X(4,0)...X(4,7)  ->  Y(4,0)...Y(4,7)  *
;          *
*********************************************************************
      LD    *AR1+, 4, A          ; + (16)*(X0)
      ADD *AR1+, 4, A          ; + (16)*(X1)
      ADD *AR1+, 4, A          ; + (16)*(X2)
      ADD *AR1+, 4, A          ; + (16)*(X3)
ADD *AR1+, 4, A          ; + (16)*(X4)
ADD *AR1+, 4, A          ; + (16)*(X5)
ADD *AR1+, 4, A          ; + (16)*(X6)
ADD *AR1, 4, A          ; + (16)*(X7)
STL A, Y40             ; = Y40

      RPTZ A, #7
      MACP *AR1-, COEF_F1, A
      SFTA A, 4
      MAR *AR1+                ; (64)[A*X0+B*X1+C*X2-D*X3-D*X4
      ADD ROUND1, A          ; -C*X5-B*X6-A*X7] + 4*ROUND1
      STH A, 2, Y41          ; = Y41

      LD    ROUND1, -4, A       ; (ROUND1)/16
      LD    E_P6, T
      MAC *AR1+0, A          ; + X0*E_P6
      MAS *AR1+, A             ; - X3*E_P6
      MAS *AR1+0, A          ; - X4*E_P6
      MAC *AR1-, A             ; + X7*E_P6
      LD    F_P6, T
MAC *AR1-, A             ; + X6*F_P6
      MAS *AR1-0, A          ; - X5*F_P6
      MAS *AR1-, A             ; - X2*F_P6
      MAC *AR1-, A             ; + X1*F_P6
      STH A, 6, Y42          ; Multiply prev summation by 16
                                 ; = Y42

      RPTZ A, #7
      MACP *AR1+, COEFF2, A    ; 64(B*X0-D*X1-A*X2-C*X3+C*X4+A*X5
      SFTA A, 4
      MAR *AR1-
      ADD ROUND1, A          ; +D*X6-B*X7) + 4*ROUND1
      STH A, 2, Y43          ; = Y43

LD *AR1-, 4, A          ; X7*16
      SUB *AR1-, 4, A          ; - X6*16
      SUB *AR1-, 4, A          ; - X5*16
      ADD *AR1-, 4, A          ; + X4*16
      ADD *AR1-, 4, A          ; + X3*16
      SUB *AR1-, 4, A          ; - X2*16
      SUB *AR1-, 4, A          ; - X1*16
      ADD *AR1, 4, A          ; + X0*16
      STL A, Y44                ; = Y44

      RPTZ  A, #7
      MACP  *AR1+, COEFF3, A    ; 8(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1-
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y45             ; = Y45

LD ROUND1, -4, A       ; (ROUND1)/16
      LD F_P6, T
      MAC *AR1-0, A             ; + F_P6*X7
      MAS *AR1-, A             ; - F_P6*X4
      MAS *AR1-0, A             ; - F_P6*X3
      MAC *AR1+, A             ; + F_P6*X0
      LD E_P6, T
      MAS *AR1+, A             ; - E_P6*X1
      MAC *AR1+0, A             ; + E_P6*X2
      MAC *AR1+, A             ; + E_P6*X5
      MAS *AR1+, A             ; - E_P6*X6
      STH A, 6, Y46             ; Multiply prev summation by 16
                                 ; = Y46
      RPTZ A, #7
      MACP  *AR1-, COEF_F4, A    ; 64(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1+
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y47             ; = Y47
ERROR:
      MAR *+AR1(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs
      NOP
      NOP
      NOP
      NOP
*********************************************************************
;          *
; PROCESS SIXTH 8 INPUTS.    X(5,0)...X(5,7)  ->  Y(5,0)...Y(5,7)  *
;          *
*********************************************************************
      LD    *AR1+, 4, A          ; + (16)*(X0)
      ADD *AR1+, 4, A          ; + (16)*(X1)
      ADD *AR1+, 4, A          ; + (16)*(X2)
      ADD *AR1+, 4, A          ; + (16)*(X3)
ADD *AR1+, 4, A          ; + (16)*(X4)
ADD *AR1+, 4, A          ; + (16)*(X5)
ADD *AR1+, 4, A          ; + (16)*(X6)
ADD *AR1, 4, A          ; + (16)*(X7)
STL A, Y50             ; = Y50

      RPTZ A, #7
      MACP *AR1-, COEF_F1, A
      SFTA A, 4
      MAR *AR1+                ; (64)[A*X0+B*X1+C*X2-D*X3-D*X4
      ADD ROUND1, A          ; -C*X5-B*X6-A*X7] + 4*ROUND1
      STH A, 2, Y51          ; = Y51

      LD    ROUND1, -4, A       ; (ROUND1)/16
      LD    E_P6, T
      MAC *AR1+0, A          ; + X0*E_P6
      MAS *AR1+, A             ; - X3*E_P6
      MAS *AR1+0, A          ; - X4*E_P6
      MAC *AR1-, A             ; + X7*E_P6
      LD    F_P6, T
MAC *AR1-, A             ; + X6*F_P6
      MAS *AR1-0, A          ; - X5*F_P6
      MAS *AR1-, A             ; - X2*F_P6
      MAC *AR1-, A             ; + X1*F_P6
      STH A, 6, Y52          ; Multiply prev summation by 16
                                 ; = Y52

      RPTZ A, #7
      MACP *AR1+, COEFF2, A    ; 64(B*X0-D*X1-A*X2-C*X3+C*X4+A*X5
      SFTA A, 4
      MAR *AR1-
      ADD ROUND1, A          ; +D*X6-B*X7) + 4*ROUND1
      STH A, 2, Y53          ; = Y53

LD *AR1-, 4, A          ; X7*16
      SUB *AR1-, 4, A          ; - X6*16
      SUB *AR1-, 4, A          ; - X5*16
      ADD *AR1-, 4, A          ; + X4*16
      ADD *AR1-, 4, A          ; + X3*16
      SUB *AR1-, 4, A          ; - X2*16
      SUB *AR1-, 4, A          ; - X1*16
      ADD *AR1, 4, A          ; + X0*16
      STL A, Y54                ; = Y54

      RPTZ  A, #7
      MACP  *AR1+, COEFF3, A    ; 8(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1-
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y55             ; = Y55
      LD ROUND1, -4, A       ; (ROUND1)/16
      LD F_P6, T
      MAC *AR1-0, A             ; + F_P6*X7
      MAS *AR1-, A             ; - F_P6*X4
      MAS *AR1-0, A             ; - F_P6*X3
      MAC *AR1+, A             ; + F_P6*X0
      LD E_P6, T
      MAS *AR1+, A             ; - E_P6*X1
      MAC *AR1+0, A             ; + E_P6*X2
      MAC *AR1+, A             ; + E_P6*X5
      MAS *AR1+, A             ; - E_P6*X6
      STH A, 6, Y56             ; Multiply prev summation by 16
                                 ; = Y56
      RPTZ A, #7
      MACP  *AR1-, COEF_F4, A    ; 64(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1+
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y57             ; = Y57

      MAR *+AR1(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs
*********************************************************************
;          *
; PROCESS SEVENTH 8 INPUTS. X(6,0)...X(6,7)  ->  Y(6,0)...Y(6,7)  *
;          *
*********************************************************************
      LD    *AR1+, 4, A          ; + (16)*(X0)
      ADD *AR1+, 4, A          ; + (16)*(X1)
      ADD *AR1+, 4, A          ; + (16)*(X2)
      ADD *AR1+, 4, A          ; + (16)*(X3)
ADD *AR1+, 4, A          ; + (16)*(X4)
ADD *AR1+, 4, A          ; + (16)*(X5)
ADD *AR1+, 4, A          ; + (16)*(X6)
ADD *AR1, 4, A          ; + (16)*(X7)
STL A, Y60             ; = Y60

      RPTZ A, #7
      MACP *AR1-, COEF_F1, A
      SFTA A, 4
      MAR *AR1+                ; (64)[A*X0+B*X1+C*X2-D*X3-D*X4
      ADD ROUND1, A          ; -C*X5-B*X6-A*X7] + 4*ROUND1
      STH A, 2, Y61          ; = Y61

      LD    ROUND1, -4, A       ; (ROUND1)/16
      LD    E_P6, T
      MAC *AR1+0, A          ; + X0*E_P6
      MAS *AR1+, A             ; - X3*E_P6
      MAS *AR1+0, A          ; - X4*E_P6
      MAC *AR1-, A             ; + X7*E_P6
      LD    F_P6, T
MAC *AR1-, A             ; + X6*F_P6
      MAS *AR1-0, A          ; - X5*F_P6
      MAS *AR1-, A             ; - X2*F_P6
      MAC *AR1-, A             ; + X1*F_P6
      STH A, 6, Y62          ; Multiply prev summation by 16
                                 ; = Y62

      RPTZ A, #7
      MACP *AR1+, COEFF2, A    ; 64(B*X0-D*X1-A*X2-C*X3+C*X4+A*X5
      SFTA A, 4
      MAR *AR1-
      ADD ROUND1, A          ; +D*X6-B*X7) + 4*ROUND1
      STH A, 2, Y63          ; = Y63

LD *AR1-, 4, A          ; X7*16
      SUB *AR1-, 4, A          ; - X6*16
      SUB *AR1-, 4, A          ; - X5*16
      ADD *AR1-, 4, A          ; + X4*16
      ADD *AR1-, 4, A          ; + X3*16
      SUB *AR1-, 4, A          ; - X2*16
      SUB *AR1-, 4, A          ; - X1*16
      ADD *AR1, 4, A          ; + X0*16
      STL A, Y64                ; = Y64

      RPTZ  A, #7
      MACP  *AR1+, COEFF3, A    ; 8(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1-
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y65             ; = Y65

1万 · 2014-11-12 15:53

LD ROUND1, -4, A       ; (ROUND1)/16
      LD F_P6, T
      MAC *AR1-0, A             ; + F_P6*X7
      MAS *AR1-, A             ; - F_P6*X4
      MAS *AR1-0, A             ; - F_P6*X3
      MAC *AR1+, A             ; + F_P6*X0
      LD E_P6, T
      MAS *AR1+, A             ; - E_P6*X1
      MAC *AR1+0, A             ; + E_P6*X2
      MAC *AR1+, A             ; + E_P6*X5
      MAS *AR1+, A             ; - E_P6*X6
      STH A, 6, Y66             ; Multiply prev summation by 16
                                 ; = Y66
      RPTZ A, #7
      MACP  *AR1-, COEF_F4, A    ; 64(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1+
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y67             ; = Y67

      MAR *+AR1(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs
*********************************************************************
;          *
; PROCESS EIGHTH 8 INPUTS. X(7,0)...X(7,7)  ->  Y(7,0)...Y(7,7)  *
;          *
*********************************************************************
      LD    *AR1+, 4, A          ; + (16)*(X0)
      ADD *AR1+, 4, A          ; + (16)*(X1)
      ADD *AR1+, 4, A          ; + (16)*(X2)
      ADD *AR1+, 4, A          ; + (16)*(X3)
ADD *AR1+, 4, A          ; + (16)*(X4)
ADD *AR1+, 4, A          ; + (16)*(X5)
ADD *AR1+, 4, A          ; + (16)*(X6)
ADD *AR1, 4, A          ; + (16)*(X7)
STL A, Y70             ; = Y70

      RPTZ A, #7
      MACP *AR1-, COEF_F1, A
      SFTA A, 4
      MAR *AR1+                ; (64)[A*X0+B*X1+C*X2-D*X3-D*X4
      ADD ROUND1, A          ; -C*X5-B*X6-A*X7] + 4*ROUND1
      STH A, 2, Y71          ; = Y71

      LD    ROUND1, -4, A       ; (ROUND1)/16
      LD    E_P6, T
      MAC *AR1+0, A          ; + X0*E_P6
      MAS *AR1+, A             ; - X3*E_P6
      MAS *AR1+0, A          ; - X4*E_P6
      MAC *AR1-, A             ; + X7*E_P6
      LD    F_P6, T
MAC *AR1-, A             ; + X6*F_P6
      MAS *AR1-0, A          ; - X5*F_P6
      MAS *AR1-, A             ; - X2*F_P6
      MAC *AR1-, A             ; + X1*F_P6
      STH A, 6, Y72          ; Multiply prev summation by 16
                                 ; = Y72

      RPTZ A, #7
      MACP *AR1+, COEFF2, A    ; 64(B*X0-D*X1-A*X2-C*X3+C*X4+A*X5
      SFTA A, 4
      MAR *AR1-
      ADD ROUND1, A          ; +D*X6-B*X7) + 4*ROUND1
      STH A, 2, Y73          ; = Y73

LD *AR1-, 4, A          ; X7*16
      SUB *AR1-, 4, A          ; - X6*16
      SUB *AR1-, 4, A          ; - X5*16
      ADD *AR1-, 4, A          ; + X4*16
      ADD *AR1-, 4, A          ; + X3*16
      SUB *AR1-, 4, A          ; - X2*16
      SUB *AR1-, 4, A          ; - X1*16
      ADD *AR1, 4, A          ; + X0*16
      STL A, Y74                ; = Y74

      RPTZ  A, #7
      MACP  *AR1+, COEFF3, A    ; 8(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1-
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y75             ; = Y75

LD ROUND1, -4, A       ; (ROUND1)/16
      LD F_P6, T
      MAC *AR1-0, A             ; + F_P6*X7
      MAS *AR1-, A             ; - F_P6*X4
      MAS *AR1-0, A             ; - F_P6*X3
      MAC *AR1+, A             ; + F_P6*X0
      LD E_P6, T
      MAS *AR1+, A             ; - E_P6*X1
      MAC *AR1+0, A             ; + E_P6*X2
      MAC *AR1+, A             ; + E_P6*X5
      MAS *AR1+, A             ; - E_P6*X6
      STH A, 6, Y76             ; Multiply prev summation by 16
                                 ; = Y76
      RPTZ A, #7
      MACP  *AR1-, COEF_F4, A    ; 64(D*X0-C*X1+B*X2-A*X3+A*X4-B*X5
      SFTA  A, 4                ; +C*X6-D*X7)
      MAR *AR1+
      ADD ROUND1, A             ; + 4*ROUND1
      STH A, 2, Y77             ; = Y77

      MAR *+AR1(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs


;    ******************************************
;          *                                     *
;    * SECOND STEP :       *
;    *          *
;    *    CALCULATE OUTER TRANSFORM       *
;          *       (PROCESS INPUT COLUMNS)       *
;          *                                     *
;    ******************************************
DCT2 LD #PAGE25, DP

*********************************************************************
;          *
; PROCESS FIRST 8 INPUTS.    Y(0,0)...Y(7,0)  ->  Z(0,0)...Z(7,0)  *
;          *
*********************************************************************
      LD    ROUND3, A          ; + ROUND3
      ADD *AR2+, 9, A          ; + (Y0)
      ADD *AR2+, 9, A          ; + (Y1)
      ADD *AR2+, 9, A          ; + (Y2)
      ADD *AR2+, 9, A          ; + (Y3)
ADD *AR2+, 9, A          ; + (Y4)
ADD *AR2+, 9, A          ; + (Y5)
ADD *AR2+, 9, A          ; + (Y6)
ADD *AR2, 9, A          ; + (Y7)
STH A, Z00             ; = Z00

      RPTZ A, #7
      MACP *AR2-, COEF_F1, A
      SFTA A, -6
      MAR *AR2+                ; (128)[A*Y0+B*Y1+C*Y2-D*Y3-D*Y4
      ADD ROUND2, A          ; -C*Y5-B*Y6-A*Y7] + 2*ROUND2
      STH A, 1, Z10          ; = Z10
ZY20:
      LD    ROUND2, -6, A       ; (ROUND1)/64
      LD    E_P7, T
      MAC *AR2+0, A          ; + Y0*E_P7
      MAS *AR2+, A             ; - Y3*E_P7
      MAS *AR2+0, A          ; - Y4*E_P7
      MAC *AR2-, A             ; + Y7*E_P7
      LD    F_P7, T
MAC *AR2-, A             ; + Y6*F_P7
      MAS *AR2-0, A          ; - Y5*F_P7
      MAS *AR2-, A             ; - Y2*F_P7
      MAC *AR2-, A             ; + Y1*F_P7
      STH A, -5, Z20          ; Multiply prev summation by 128
                                 ; = Z20

      RPTZ A, #7
      MACP *AR2+, COEFF2, A    ; 128(B*Y0-D*Y1-A*Y2-C*Y3+C*Y4+A*Y5
      SFTA A, -6
      MAR *AR2-
      ADD ROUND2, A          ; +D*Y6-B*Y7) + 2*ROUND2
      STH A, 1, Z30          ; = Z30

      LD ROUND3, A
ADD *AR2-, 9, A          ; Y7
      SUB *AR2-, 9, A          ; - Y6
      SUB *AR2-, 9, A          ; - Y5
      ADD *AR2-, 9, A          ; + Y4
      ADD *AR2-, 9, A          ; + Y3
      SUB *AR2-, 9, A          ; - Y2
      SUB *AR2-, 9, A          ; - Y1
      ADD *AR2, 9, A          ; + Y0
      STH A, Z40                ; = Z40

      RPTZ  A, #7
      MACP  *AR2+, COEFF3, A    ; 218(D*Y0-C*Y1+B*Y2-A*Y3+A*Y4-B*Y5
      SFTA  A, -6                ; +C*Y6-D*Y7)
      MAR *AR2-
      ADD ROUND2, A             ; + 2*ROUND2
      STH A, 1, Z50             ; = Z50

      LD ROUND2, -6, A       ; (ROUND1)/64
      LD F_P7, T
      MAC *AR2-0, A             ; + F_P7*Y7
      MAS *AR2-, A             ; - F_P7*Y4
      MAS *AR2-0, A             ; - F_P7*Y3
      MAC *AR2+, A             ; + F_P7*Y0
      LD E_P7, T
      MAS *AR2+, A             ; - E_P7*Y1
      MAC *AR2+0, A             ; + E_P7*Y2
      MAC *AR2+, A             ; + E_P7*Y5
      MAS *AR2+, A             ; - E_P7*Y6
      STH A, -5, Z60          ; Multiply prev summation by 128
                                 ; = Z60
      RPTZ A, #7
      MACP  *AR2-, COEF_F4, A    ; 128(D*Y0-C*Y1+B*Y2-A*Y3+A*Y4-B*Y5
      SFTA  A, -6                ; +C*Y6-D*Y7)
      MAR *AR2+
      ADD ROUND2, A             ; + 2*ROUND2
      STH A, 1, Z70             ; = Z70

      MAR *+AR2(8)             ; Set AR2 to point to the next set
                                 ; of 8 inputs
*********************************************************************
;          *
; PROCESS SECOND 8 INPUTS. Y(0,1)...Y(7,1)  ->  Z(0,1)...Z(7,1)  *
;          *
*********************************************************************
      LD    ROUND3, A          ; + ROUND3
      ADD *AR2+, 9, A          ; + (Y0)
      ADD *AR2+, 9, A          ; + (Y1)
      ADD *AR2+, 9, A          ; + (Y2)
      ADD *AR2+, 9, A          ; + (Y3)
ADD *AR2+, 9, A          ; + (Y4)
ADD *AR2+, 9, A          ; + (Y5)
ADD *AR2+, 9, A          ; + (Y6)
ADD *AR2, 9, A          ; + (Y7)
STH A, Z01             ; = Z01

      RPTZ A, #7
      MACP *AR2-, COEF_F1, A
      SFTA A, -6
      MAR *AR2+                ; (128)[A*Y0+B*Y1+C*Y2-D*Y3-D*Y4
      ADD ROUND2, A          ; -C*Y5-B*Y6-A*Y7] + 2*ROUND2
      STH A, 1, Z11          ; = Z11

      LD    ROUND2, -6, A       ; (ROUND1)/64
      LD    E_P7, T
      MAC *AR2+0, A          ; + Y0*E_P7
      MAS *AR2+, A             ; - Y3*E_P7
      MAS *AR2+0, A          ; - Y4*E_P7
      MAC *AR2-, A             ; + Y7*E_P7
      LD    F_P7, T
MAC *AR2-, A             ; + Y6*F_P7
      MAS *AR2-0, A          ; - Y5*F_P7
      MAS *AR2-, A             ; - Y2*F_P7
      MAC *AR2-, A             ; + Y1*F_P7
      STH A, -5, Z21          ; Multiply prev summation by 128
                                 ; = Z21

      RPTZ A, #7
      MACP *AR2+, COEFF2, A    ; 128(B*Y0-D*Y1-A*Y2-C*Y3+C*Y4+A*Y5
      SFTA A, -6
      MAR *AR2-
      ADD ROUND2, A          ; +D*Y6-B*Y7) + 2*ROUND2
      STH A, 1, Z31          ; = Z31

      LD ROUND3, A
ADD *AR2-, 9, A          ; Y7
      SUB *AR2-, 9, A          ; - Y6
      SUB *AR2-, 9, A          ; - Y5
      ADD *AR2-, 9, A          ; + Y4
      ADD *AR2-, 9, A          ; + Y3
      SUB *AR2-, 9, A          ; - Y2
      SUB *AR2-, 9, A          ; - Y1
      ADD *AR2, 9, A          ; + Y0
      STH A, Z41                ; = Z41

      RPTZ  A, #7
      MACP  *AR2+, COEFF3, A    ; 218(D*Y0-C*Y1+B*Y2-A*Y3+A*Y4-B*Y5
      SFTA  A, -6                ; +C*Y6-D*Y7)
      MAR *AR2-
      ADD ROUND2, A             ; + 2*ROUND2
      STH A, 1, Z51             ; = Z51

      LD ROUND2, -6, A       ; (ROUND1)/64
      LD F_P7, T
      MAC *AR2-0, A             ; + F_P7*X7
      MAS *AR2-, A             ; - F_P7*X4
      MAS *AR2-0, A             ; - F_P7*X3
      MAC *AR2+, A             ; + F_P7*X0
      LD E_P7, T
      MAS *AR2+, A             ; - E_P7*X1
      MAC *AR2+0, A             ; + E_P7*X2
      MAC *AR2+, A             ; + E_P7*X5
      MAS *AR2+, A             ; - E_P7*X6
      STH A, -5, Z61          ; Multiply prev summation by 128
                                 ; = Z61
      RPTZ A, #7
      MACP  *AR2-, COEF_F4, A    ; 128(D*Y0-C*Y1+B*Y2-A*Y3+A*Y4-B*Y5
      SFTA  A, -6                ; +C*Y6-D*Y7)
      MAR *AR2+
      ADD ROUND2, A             ; + 2*ROUND2
      STH A, 1, Z71             ; = Z71

      MAR *+AR2(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs
*********************************************************************
;          *
; PROCESS THIRD 8 INPUTS.    Y(0,2)...Y(7,2)  ->  Z(0,2)...Z(7,2)  *
;          *
*********************************************************************
      LD    ROUND3, A          ; + ROUND3
      ADD *AR2+, 9, A          ; + (Y0)
      ADD *AR2+, 9, A          ; + (Y1)
      ADD *AR2+, 9, A          ; + (Y2)
      ADD *AR2+, 9, A          ; + (Y3)
ADD *AR2+, 9, A          ; + (Y4)
ADD *AR2+, 9, A          ; + (Y5)
ADD *AR2+, 9, A          ; + (Y6)
ADD *AR2, 9, A          ; + (Y7)
STH A, Z02             ; = Z02

      RPTZ A, #7
      MACP *AR2-, COEF_F1, A
      SFTA A, -6
      MAR *AR2+                ; (128)[A*Y0+B*Y1+C*Y2-D*Y3-D*Y4
      ADD ROUND2, A          ; -C*Y5-B*Y6-A*Y7] + 2*ROUND2
      STH A, 1, Z12          ; = Z12

      LD    ROUND2, -6, A       ; (ROUND1)/64
      LD    E_P7, T
      MAC *AR2+0, A          ; + Y0*E_P7
      MAS *AR2+, A             ; - Y3*E_P7
      MAS *AR2+0, A          ; - Y4*E_P7
      MAC *AR2-, A             ; + Y7*E_P7
      LD    F_P7, T
MAC *AR2-, A             ; + Y6*F_P7
      MAS *AR2-0, A          ; - Y5*F_P7
      MAS *AR2-, A             ; - Y2*F_P7
      MAC *AR2-, A             ; + Y1*F_P7
      STH A, -5, Z22          ; Multiply prev summation by 128
                                 ; = Z22

      RPTZ A, #7
      MACP *AR2+, COEFF2, A    ; 128(B*Y0-D*Y1-A*Y2-C*Y3+C*Y4+A*Y5
      SFTA A, -6
      MAR *AR2-
      ADD ROUND2, A          ; +D*Y6-B*Y7) + 2*ROUND2
      STH A, 1, Z32          ; = Z32

      LD ROUND3, A
ADD *AR2-, 9, A          ; Y7
      SUB *AR2-, 9, A          ; - Y6
      SUB *AR2-, 9, A          ; - Y5
      ADD *AR2-, 9, A          ; + Y4
      ADD *AR2-, 9, A          ; + Y3
      SUB *AR2-, 9, A          ; - Y2
      SUB *AR2-, 9, A          ; - Y1
      ADD *AR2, 9, A          ; + Y0
      STH A, Z42                ; = Z42

      RPTZ  A, #7
      MACP  *AR2+, COEFF3, A    ; 218(D*Y0-C*Y1+B*Y2-A*Y3+A*Y4-B*Y5
      SFTA  A, -6                ; +C*Y6-D*Y7)
      MAR *AR2-
      ADD ROUND2, A             ; + 2*ROUND2
      STH A, 1, Z52             ; = Z52

      LD ROUND2, -6, A       ; (ROUND1)/64
      LD F_P7, T
      MAC *AR2-0, A             ; + F_P7*Y7
      MAS *AR2-, A             ; - F_P7*Y4
      MAS *AR2-0, A             ; - F_P7*Y3
      MAC *AR2+, A             ; + F_P7*Y0
      LD E_P7, T
      MAS *AR2+, A             ; - E_P7*Y1
      MAC *AR2+0, A             ; + E_P7*Y2
      MAC *AR2+, A             ; + E_P7*Y5
      MAS *AR2+, A             ; - E_P7*Y6
      STH A, -5, Z62          ; Multiply prev summation by 128
                                 ; = Z62
      RPTZ A, #7
      MACP  *AR2-, COEF_F4, A    ; 128(D*Y0-C*Y1+B*Y2-A*Y3+A*Y4-B*Y5
      SFTA  A, -6                ; +C*Y6-D*Y7)
      MAR *AR2+
      ADD ROUND2, A             ; + 2*ROUND2
      STH A, 1, Z72             ; = Z72

      MAR *+AR2(8)             ; Set AR1 to point to the next set
                                 ; of 8 inputs