A optimisation of line_up_inv.s by moving some logic out of loop

wip-hugo/routines/memory/line_down.s

@@ -28,7 +28,6 @@
     STA Y_end
     ;;~~~~~~~~~~
 
-
     ;;We need to clear this memory
     LDA #$00
     STA <V

wip-hugo/routines/memory/line_up.s

@@ -62,7 +62,6 @@
      STA <D
      Add_16 >D, <D, #$00, #$01
      Sub_16 >D, <D, dx, #$00
-
 for_x:
     jsr pixel_draw
     ;;Increment X until X_pos = X_end and Y_pos = Y_end

wip-hugo/routines/memory/line_up_inv.s

@@ -13,13 +13,10 @@
     X_pos = $FC
     Y_pos = $FB
     dx_2 = $0607
-    dy = $08
+    dy = $0c
+    V = $0809
     D = $0a0b
 
-    ;;Set values
-    LDA #$00
-    STA $FD ; for pixel_draw
-
     ;example values ~~~~~ SHOULD BE PRECOMPILED
     LDA #$60
     STA X_pos
@@ -31,37 +28,39 @@
     STA Y_end
     ;;~~~~~~~~~~
 
-    ;;2*dx = 2*(X_pos - X_end)
+    ;We need to clear this memory
+    LDA #$00
+    STA <V
+    STA <dx_2
+    STA $FD ; for pixel_draw
+
+    ;; V = 2*(dy -dx)
+    ;; where: dy = Y_pos - Y_end, dx = X_end - X_start
     LDA X_pos
     SEC
     SBC X_end
-    STA >dx_2
-    LDA #$00
-    STA <dx_2
-    mult_16 >dx_2, <dx_2
-
-    ;;dy = (Y_end -Y_pos)
+    STA >V
+    STA >dx_2;  >dy_2 = dy. Needed for dy_2 (not for V)
     LDA Y_end
     SEC
     SBC Y_pos
     STA dy
+    SEC
+    SBC >V
+    STA >V; <V = dx - dy
+    mult_16 >V, <V; V = 2*(dx -dy)
 
-    ;;D = 2*dx - dy + 2*255. 0 <=D <= 4*255
-    ;;Our D is bigger then wikipedia to not use negative numbers
+    ;dy_2 = dy*2
+    mult_16 >dx_2, <dx_2
+
+    ;;D = 2*dx - dy + 2*255
+    ;;Our D is bigger then wikipedia because D is unsigned.
     LDA >dx_2
     STA >D
     LDA <dx_2
     STA <D
     Add_16 >D, <D, #$00, #$01
     Sub_16 >D, <D, dy, #$00
-
-
-    ;;We wont use dy, only 2*dy from now on
-    dy_2 = $0809 ;Lets reuse dy register ($08)
-    LDA #$00
-    STA <dy_2
-    mult_16 >dy_2, <dy_2
-
 for_y:
     jsr pixel_draw
     ;;Increment Y until Y_pos = Y_end and X_pos = X_end
@@ -76,10 +75,10 @@ for_y:
     BEQ case_2
 case_1:
     DEC X_pos
-    Sub_16 >D, <D, >dy_2, <dy_2; D = D - 2*dx
+    Sub_16 >D, <D, >V, <V; D = D - V
+    JMP for_y
 case_2:
     Add_16 >D, <D, >dx_2, <dx_2;D = D + 2*dy
-last:
     JMP for_y
 end:
 .endproc

wip-hugo/source.s

@@ -1,6 +1,6 @@
 .include "STARTUP.s"
 .include "macros/16aritmatic.s"
-.include "routines/memory/line_up.s"
+.include "routines/memory/line_up_inv.s"
 
 loop_:
 jmp loop_