Implement earlier perfromance improvments to line_up.s

wip-hugo/routines/line/line_up.s

@@ -12,6 +12,15 @@
 .proc line_up
         .include "line.inc"; Defines memory positions, ex X_pos
 
+        ;LDA #$20
+        ;STA X_pos
+        ;STA Y_pos
+        ;LDA #$10
+        ;STA Y_end
+        ;LDA #$40
+        ;STA X_end
+
+
         ;;We need to clear this memory
         LDA #$00
         STA <V
@@ -19,9 +28,7 @@
         STA $FD ; for pixel_draw
 
         ;; V = 2*(dx -dy)
-        ;; where: dy = Y_pos - Y_end, dx = X_end - X_start
-        ;; This logic is comented out because line.s does it woithout any
-        ;;extra cost. May be needed in the future so it will stay as coments!
+        ;; where: dy = Y_end - Y_start, dx = OO - X_start
         LDA Y_pos
         SEC
         SBC Y_end
@@ -32,12 +39,18 @@
         SBC X_pos
         STA dx
         SEC
-        SBC >dy_2
+        SBC >V
         STA >V; <V = dx - dy
         mult_16 >V, <V; V = 2*(dx -dy)
 
         ;dy_2 = dy*2
-        mult_16 >dy_2, <dy_2
+        mult_16 >dy_2, <dy_2, !
+        ;; D = 2*dy - dx
+        ;; In loop we have that D = D -V
+        ;; So D needs to be at least >=V.
+        ;; V_max = 00000001 11111111
+        ;; For us to work with unsigned numbers we add 00000001 11111111
+        ;; to V and the branch logic to V!
 
         ;;D = 2*dy - dx + 2*255
         ;;Our D is bigger then wikipedia because D is unsigned.
@@ -45,25 +58,43 @@
         STA >D
         LDA <dy_2
         STA <D
-        Add_16 >D, <D, #$ff, #$01
+        Add_16 >D, <D, #$ff, #$01, !
         Sub_16 >D, <D, dx, #$00
-for_x:
+
         jsr pixel_draw
-        ;;Increment X until X_pos = X_end and Y_pos = Y_end
-        INC X_pos
+        ;INY ; Dont want underflow
+        ;LDY #$09
+        ;Sub_16 >btp_mem_pos, <btp_mem_pos, #$40, #$01 ;-320
         LDX X_pos
+for_x:
+        LDA byte_to_paint
+        ORA (>btp_mem_pos), Y
+        STA (>btp_mem_pos), Y
+increment_pixel_x:
+        LSR byte_to_paint
+        BCC increment_pixel_x_end
+move_8px_left:
+        Add_16 >btp_mem_pos, <btp_mem_pos, #$08, #$00
+        LDA #%10000000
+        STA byte_to_paint
+increment_pixel_x_end:
+        INX
         CPX X_end
         BEQ end
-
-        ;;If D <  %00000010 00000000:  case_2
-        ;;else case 1.
         Lag_16 >D, <D, #$00, #$02, case_2
 case_1:
-        DEC Y_pos
-        Sub_16 >D, <D, >V, <V; D = D - V
-        JMP for_x
+        Sub_16 >D, <D, >V, <V
+decrement_y_pos:
+        DEY
+        CPY #$ff
+        BNE for_x
+move_8px_up:
+        LDY #$07
+        Sub_16 >btp_mem_pos, <btp_mem_pos, #$40, #$01 ;-320
+        jmp for_x
+decrement_y_pos_end:
 case_2:
-        Add_16 >D, <D, >dy_2, <dy_2;D = D + 2*dy
+        Add_16 >D, <D, >dy_2, <dy_2 ;D = D + 2*dy
         JMP for_x
 end:
         RTS