// func addMulVVW1024(z, x *uint, y uint) (c uint)
TEXT ·addMulVVW1024(SB),$0-32
	MOV	$16, X30
	JMP	addMulVVWx(SB)

// func addMulVVW1536(z, x *uint, y uint) (c uint)
TEXT ·addMulVVW1536(SB),$0-32
	MOV	$24, X30
	JMP	addMulVVWx(SB)

// func addMulVVW2048(z, x *uint, y uint) (c uint)
TEXT ·addMulVVW2048(SB),$0-32
	MOV	$32, X30
	JMP	addMulVVWx(SB)

TEXT addMulVVWx(SB),NOFRAME|NOSPLIT,$0
	MOV	z+0(FP), X5

	//	_111111 = _111 + _111000
	//	x12     = _111111 << 6 + _111111
	//	x24     = x12 << 12 + x12
	//	x30     = x24 << 6 + _111111
	//	x31     = 2*x30 + 1
	//	x32     = 2*x31 + 1
	//	x63     = x32 << 31 + x31
	//	x126    = x63 << 63 + x63
	//	x252    = x126 << 126 + x126
	//	x255    = x252 << 3 + _111
	//	i397    = ((x255 << 33 + x32) << 94 + x30) << 2
	//	return    1 + i397
	//

	var z = new(P384Element).Set(e)

 * Calling into the gc tool chain, where all registers are caller save.
 * Called from standard ARM EABI, where x19-x29 are callee-save, so they
 * must be saved explicitly, along with x30 (LR).
 */
.globl EXT(crosscall1)
EXT(crosscall1):
	.cfi_startproc
	stp x29, x30, [sp, #-96]!
	.cfi_def_cfa_offset 96
	.cfi_offset 29, -96
	.cfi_offset 30, -88
	mov x29, sp
	.cfi_def_cfa_register 29
	stp x19, x20, [sp, #80]
	.cfi_offset 19, -16

	MOVBU	2(X12), X18
	MOVBU	3(X12), X20
	MOVBU	4(X12), X22
	MOVBU	5(X12), X24
	MOVBU	6(X12), X28
	MOVBU	7(X12), X30
	BNE	X8, X9, cmp1a
	BNE	X15, X16, cmp1b
	BNE	X17, X18, cmp1c
	BNE	X19, X20, cmp1d
	BNE	X21, X22, cmp1e
	BNE	X23, X24, cmp1f
	BNE	X25, X28, cmp1g
	BNE	X29, X30, cmp1h
	ADD	$8, X10
	ADD	$8, X12
	SUB	$8, X5
	BGE	X5, X6, compare8_unaligned
	BEQZ	X5, cmp_len

	x23 := (uint8(x22) & 0xff)
	x24 := (x22 >> 8)
	x25 := (uint8(x24) & 0xff)
	x26 := (x24 >> 8)
	x27 := (uint8(x26) & 0xff)
	x28 := (x26 >> 8)
	x29 := (uint8(x28) & 0xff)
	x30 := (x28 >> 8)
	x31 := (uint8(x30) & 0xff)
	x32 := uint8((x30 >> 8))
	x33 := (uint8(x2) & 0xff)
	x34 := (x2 >> 8)
	x35 := (uint8(x34) & 0xff)
	x36 := (x34 >> 8)
	x37 := (uint8(x36) & 0xff)
	x38 := (x36 >> 8)

	var x26 uint64
	var x27 uint64
	x26, x27 = bits.Add64(x25, x22, uint64(0x0))
	x28 := (uint64(p256Uint1(x27)) + x23)
	var x30 uint64
	_, x30 = bits.Add64(x11, x24, uint64(0x0))
	var x31 uint64
	var x32 uint64
	x31, x32 = bits.Add64(x13, x26, uint64(p256Uint1(x30)))
	var x33 uint64
	var x34 uint64
	x33, x34 = bits.Add64(x15, x28, uint64(p256Uint1(x32)))
	var x35 uint64
	var x36 uint64

	x26, x25 = bits.Mul64(x19, 0x5812631a5cf5d3ed)
	var x27 uint64
	var x28 uint64
	x27, x28 = bits.Add64(x26, x23, uint64(0x0))
	var x30 uint64
	_, x30 = bits.Add64(x11, x25, uint64(0x0))
	var x31 uint64
	var x32 uint64
	x31, x32 = bits.Add64(x13, x27, uint64(fiatScalarUint1(x30)))
	var x33 uint64
	var x34 uint64
	x33, x34 = bits.Add64(x15, (uint64(fiatScalarUint1(x28)) + x24), uint64(fiatScalarUint1(x32)))

	CTZ	X24, X25				// 931c1c60
	CTZW	X25, X26				// 1b9d1c60
	MAX	X26, X28, X29				// b36eae0b
	MAX	X26, X28				// 336eae0b
	MAXU	X28, X29, X30				// 33ffce0b
	MAXU	X28, X29				// b3fece0b
	MIN	X29, X30, X5				// b342df0b
	MIN	X29, X30				// 334fdf0b
	MINU	X30, X5, X6				// 33d3e20b
	MINU	X30, X5					// b3d2e20b
	ORN	X6, X7, X8				// 33e46340
	ORN	X6, X7					// b3e36340
	SEXTB	X16, X17				// 93184860

	x29 := (uint64(p384Uint1(x28)) + x8)
	var x30 uint64
	_, x30 = bits.Mul64(x17, 0x100000001)
	var x32 uint64
	var x33 uint64
	x33, x32 = bits.Mul64(x30, 0xffffffffffffffff)
	var x34 uint64
	var x35 uint64
	x35, x34 = bits.Mul64(x30, 0xffffffffffffffff)
	var x36 uint64
	var x37 uint64
	x37, x36 = bits.Mul64(x30, 0xffffffffffffffff)
	var x38 uint64
	var x39 uint64

	MOV X18, 112(X2)
	MOV X19, 120(X2)
	MOV X20, 128(X2)
	MOV X21, 136(X2)
	MOV X22, 144(X2)
	MOV X23, 152(X2)
	MOV X24, 160(X2)
	MOV X25, 168(X2)
	MOV X26, 176(X2)
	MOV X28, 184(X2)
	MOV X29, 192(X2)
	MOV X30, 200(X2)
	MOVD F0, 208(X2)
	MOVD F1, 216(X2)
	MOVD F2, 224(X2)
	MOVD F3, 232(X2)
	MOVD F4, 240(X2)
	MOVD F5, 248(X2)
	MOVD F6, 256(X2)
	MOVD F7, 264(X2)
	MOVD F8, 272(X2)
	MOVD F9, 280(X2)