wip-hugo/routines/line/line_down.s

@@ -43,6 +43,9 @@
         Add_16 >D, <D, #$ff, #$01, !
         Sub_16 >D, <D, dx, #$00
 
+        jsr pixel_draw ;;only used first pixel. after this relative position is abused
+        LDX X_pos
+
 selfmod:
         ;; Self modifying code. Makes LDA and SBC instructions each take 1 cycle less.
         ;; You can remove this if you run the loop without # at dy_2 and V.
@@ -63,9 +66,7 @@ selfmod:
         LDA <V
         STA case_1 +7
 end_selfmod:
-        LDY #$01
-        JSR pixel_draw ;;only used first pixel. after this relative position is abused
-        LDX X_pos
+
 for_x:
         ;; Paints A to address in |btp_mem_pos* + Y|
         ;; Y is pixel position in the chunk. Therefor it may be that Y = 0, 1, 2, 3, 4, ,5 ,6 ,7.

wip-hugo/routines/line/line_down_inv.s

@@ -39,7 +39,7 @@ selfmod:
         LDA <V
         STA case_1 +7
 end_selfmod:
-        LDY #$01
+
         jsr pixel_draw
         LDX Y_pos
 for_y:

wip-hugo/routines/line/line_up.s

@@ -36,13 +36,12 @@ selfmod:
         LDA <dy_2
         STA case_2 +7
         LDA >V
-        STA case_1 +3
+        STA case_1 +1
         LDA <V
-        STA case_1 +9
+        STA case_1 +7
 end_selfmod:
+
         jsr pixel_draw
-        LDY #$01
-        Sub_16 >btp_mem_pos, <btp_mem_pos, #$01, #$00 ;; Y has always a offset of at least 1
         LDX X_pos
 for_x:
         LDA byte_to_paint
@@ -61,12 +60,13 @@ increment_pixel_x_end:
         BEQ end
         Lag_16 >D, <D, #$00, #$02, case_2
 case_1:
-        Sub_16 >D, <D, #>V, #<V,!
+        Sub_16_A >D, <D, #>V, #<V, !
 decrement_y_pos:
         DEY
+        CPY #$ff
         BNE for_x
 move_8px_up:
-        LDY #$08
+        LDY #$07
         Sub_16 >btp_mem_pos, <btp_mem_pos, #$40, #$01, !
         jmp for_x
 decrement_y_pos_end:

wip-hugo/routines/line/line_up_inv.s

@@ -39,7 +39,7 @@
                 LDA <V
                 STA case_1 +7
         end_selfmod:
-        LDY #$01
+
         jsr pixel_draw
         LDX Y_pos
 for_y: