.global _start

.data

input:
.byte 5, 1,  2,  3,  4,  5,  6,  7
.byte 8, 9, 10, 11, 12, 13, 14, 15

sbox:
.byte 0xC, 0x5, 0x6, 0xB, 0x9, 0x0, 0xA, 0xD
.byte 0x3, 0xE, 0xF, 0x8, 0x4, 0x7, 0x1, 0x2

.text

// INPUT:  64 bits block (in r0), sbox (in r1)
// OUTPUT: 64 bits block (in r0)
// Processing: subtitute every nibble using the sbox
sub_block:
/*
    for (i = 0; i < 8; i++) {
        uint8_t c = input[i];
        uint8_t c_lo = c & 15;  // extract (keep) low nibble
        uint8_t c_hi = c >> 4;  // extract high nibble
        output[i]  = sbox[c_lo];        // store in low nibble
        output[i] |= sbox[c_hi] << 4;   // store in high nibble
    }
*/
    // loop through 8 bytes (64 bits) (use r3 as counter)
    // i = 0
    eor r3, r3
sub_block_loop:
    cmp r3, #8
    bge sub_block_loop_end

    // c = input[i] (r4 is low, r5 is high)
    ldrb r4, [r0, r3]
    mov r5, r4
    and r4, #15
    lsr r5, #4

    // sbox[c_lo]
    ldrb r4, [r1, r4]
    // sbox[c_hi]
    ldrb r5, [r1, r5]

    // output[i] = (hi << 4) | lo
    orr r4, r5
    strb r4, [r0, r3]

    add r3, #1
    b   sub_block_loop
sub_block_loop_end:

    // return
    mov pc, lr

_start:
    ldr r0, =input
    ldr r1, =sbox
    bl  sub_block
    b   forever

forever:
    b   forever