Differences

This shows you the differences between two versions of the page.

--- base:signed_8bit_16bit_addition [2017-10-26 07:30] – white_flame
+++ base:signed_8bit_16bit_addition [2018-09-28 03:58] – white_flame
@@ Line 1: / Line 1: @@
-To add a signed 8-bit delta to a 16-bit value, we need to sign-extend the delta to a full 16 bits.  We hold the high byte of the delta in .X and set it to $00 or $ff based on its sign:
+To add a signed 8-bit delta to a 16-bit value, we need to sign-extend the delta to a full 16 bits.  The low byte can be added as normal, but the upper byte needs to be $00 or $ff based on the sign of the low byte.
-<code asm>
+Using only the accumulator:
- ldx #$00  ; implied high byte of delta
+<code 6502tasm>
+ ; Standard low byte addition
+ clc
+ lda delta
+ adc value
+ sta value
+ ; Sign extend the high byte
+ lda delta
+ and #$80    ; Extract the sign bit
+ beq :+      ; If zero, add #$00 (+ carry)
+  lda #$ff   ; Else, add $ff (+ carry)
+:adc value+1
+ sta value+1
+</code>
+The following version is 2 bytes shorter & 2 cycles faster, using .X:
+<code 6502tasm>
+ ; Precalculate the sign-extended high byte in .X
+ ldx #$00
  lda delta
  bpl :+
-  dex      ; high byte becomes $ff to reflect negative delta
+  dex        ; decrement high byte to $ff for a negative delta
-:clc
+:
- adc value ; normal 16-bit addition
+ ; Normal 16-bit addition
+ clc
+ adc value   ; .A still holds delta
  sta value
- txa
+ txa         ; .X is the high byte
  adc value+1
  sta value+1
 </code>
 ----
 White Flame