.program argb_tx

.side_set 1 opt

.wrap_target
    pull           ; Get data from FIFO
    mov y, osr     ; Store bit count in Y (lower 8 bits)
    pull           ; Get first data word
    mov x, osr     ; Store in X
    set y, 24      ; Remaining bits in first word (32-8)

bit_loop:
    out pins, 1    ; Shift 1 bit to pins (side-set 0)
    jmp y--, delay ; Decrement Y, handle delay

    pull           ; Load new data when Y underflows
    mov x, osr
    set y, 31      ; Reset bit counter for new word

delay:
    jmp x--, nobit ; Check if current bit is set [1]
    nop            ; [1] Align timing

    ; Bit=1 pattern (0.8µs high, 0.45µs low)
    set pins, 1 [99]  ; 1 + 99 = 100 cycles (0.8µs)
    set pins, 0 [55]  ; 1 + 55 = 56 cycles (0.448µs)
    jmp end_bit

nobit:
    ; Bit=0 pattern (0.4µs high, 0.85µs low)
    set pins, 1 [49]  ; 50 cycles (0.4µs)
    set pins, 0 [105] ; 106 cycles (0.848µs)

end_bit:
    jmp bit_loop      ; Total per bit: 156 cycles (1.248µs)

post_delay:
    set pins, 0       ; 70µs low period
    set y, 8749       ; (70µs / 8ns) - 1 = 8749 cycles
delay_loop:
    jmp y--, delay_loop
    irq 0             ; Signal completion
.wrap