LDAXRB (R19), R16       <=>      ldaxrb w16, [x19]
	NOOP                    <=>      nop

# Register mapping rules

1. All basic register names are written as Rn.

2. Go uses ZR as the zero register and RSP as the stack pointer.

3. Bn, Hn, Dn, Sn and Qn instructions are written as Fn in floating-point instructions and as Vn
in SIMD instructions.

# Argument mapping rules

000000f0  0f d0 ec dd d4 75 57 0d  3f 7e a3 a4 40 f7 67 d2  |.....uW.?~..@.g.|
00000100  22 ba 5f a1 38 0b ea 8e  7d 95 43 70 52 0f b0 5f  |"._.8...}.CpR.._|
00000110  ef 26 5a 52 a6 94 b4 69  89 e9 0e 4f f5 d8 60 1b  |.&ZR...i...O..`.|
00000120  d3 6a fd 74 8d 19 ce 6a  72 f1 c1 96 f9 86 66 3b  |.j.t...jr.....f;|
00000130  2b 38 b3 e3 76 4b fd 4a  82 3e f2 2c bc 4c 19 d7  |+8..vK.J.>.,.L..|

	AND	$~1, R3
	LSL	$3, R3
	SUB	R3, RSP

	// R4: size of stack arguments (n-8)*8
	// R5: &args[8]
	// R6: loop counter, from 0 to (n-8)*8
	// R7: scratch
	// R8: copy of RSP - (R2)(RSP) assembles as (R2)(ZR)
	SUB	$8, R0, R4
	LSL	$3, R4
	ADD	$(8*8), R12, R5
	MOVD	$0, R6
	MOVD	RSP, R8
stackargs:
	MOVD	(R6)(R5), R7
	MOVD	R7, (R6)(R8)
	ADD	$8, R6
	CMP	R6, R4
	BNE	stackargs

_8args:

00000230  ec 62 4b ba d1 94 98 e5  de 45 f4 92 22 79 9a 2e  |.bK......E.."y..|
00000240  b4 12 a6 f8 ab 8a 52 c3  df f8 77 a9 71 8b ff d6  |......R...w.q...|
00000250  2b bc 47 0b 63 5b 0f cf  f8 f1 86 7a 72 91 78 7b  |+.G.c[.....zr.x{|
00000260  b8 52 74 96 40 4c 08 f3  ca 46 ef d9 b9 6d 7a dc  |******@****.***.|
00000270  2c da f1 c2 c4 1f 19 38  84 cf 7d b7 5d f2 19 a7  |,......8..}.]...|

ks128Loop:
		VMOV	R7, V2.S[0]
		WORD	$0x4E030042       // TBL V3.B16, [V2.B16], V2.B16
		AESE	V0.B16, V2.B16    // Use AES to compute the SBOX
		EORW	R13, R4
		LSLW	$1, R13           // Compute next Rcon
		ANDSW	$0x100, R13, ZR
		CSELW	NE, R14, R13, R13 // Fake modulo
		SUBS	$1, R8
		VMOV	V2.S[0], R0
		EORW	R0, R4
		EORW	R4, R5
		EORW	R5, R6
		EORW	R6, R7
		STPW.P	(R4, R5), 8(R10)
		STPW.P	(R6, R7), 8(R10)
	BNE	ks128Loop