Annotate memcpy with performance information

src/util.s

@@ -8,18 +8,18 @@ wait_for_vblank::
     jr nz, wait_for_vblank
     ret
 
-; Copy data between two regions
+; Copy data between two regions (takes 40*N+12 cycles)
 ; @param bc Pointer to the destination region
 ; @param hl Pointer to the source region
 ; @param d  Size (in bytes) to copy. Must be >0
 ; @destroy a, b, c, d, h, l
 memcpy::
-    ld a, [hli]
-    ld [bc], a
-    inc bc
-    dec d
-    jr nz, memcpy
-    ret
+    ld a, [hli] ; 8
+    ld [bc], a ; 8
+    inc bc ; 8
+    dec d ; 4
+    jr nz, memcpy ; 12/8
+    ret ; 16
 
 ; Fill a memory region with a value
 ; @param hl Pointer to the destination region