(Xor8  (Xor8  i:(Const8  <t>) z) x) && (z.Op != OpConst8  && x.Op != OpConst8)  => (Xor8  i (Xor8  <t> z x))

// x * (D * z) = D * (x * z)
(Mul64 (Mul64 i:(Const64 <t>) z) x) && (z.Op != OpConst64 && x.Op != OpConst64) => (Mul64 i (Mul64 <t> x z))
(Mul32 (Mul32 i:(Const32 <t>) z) x) && (z.Op != OpConst32 && x.Op != OpConst32) => (Mul32 i (Mul32 <t> x z))

	{name: "Mul8", argLength: 2, commutative: true}, // arg0 * arg1
	{name: "Mul16", argLength: 2, commutative: true},
	{name: "Mul32", argLength: 2, commutative: true},
	{name: "Mul64", argLength: 2, commutative: true},
	{name: "Mul32F", argLength: 2, commutative: true},
	{name: "Mul64F", argLength: 2, commutative: true},

	{name: "Div32F", argLength: 2}, // arg0 / arg1
	{name: "Div64F", argLength: 2},

	{name: "Hmul32", argLength: 2, commutative: true},

(Sub(64|Ptr) ...) => (SUB ...)
(Sub(32|16|8) ...) => (SUBW ...)
(Sub32F x y) => (Select0 (FSUBS x y))
(Sub64F x y) => (Select0 (FSUB x y))

(Mul64 ...) => (MULLD ...)
(Mul(32|16|8) ...) => (MULLW ...)
(Mul32F ...) => (FMULS ...)
(Mul64F ...) => (FMUL ...)
(Mul64uhilo ...) => (MLGR ...)

(Div32F ...) => (FDIVS ...)
(Div64F ...) => (FDIV ...)

(Div64 x y) => (DIVD x y)
(Div64u ...) => (DIVDU ...)

	x6 := arg1[0]
	var x7 uint64
	var x8 uint64
	x8, x7 = bits.Mul64(x6, arg2[5])
	var x9 uint64
	var x10 uint64
	x10, x9 = bits.Mul64(x6, arg2[4])
	var x11 uint64
	var x12 uint64
	x12, x11 = bits.Mul64(x6, arg2[3])
	var x13 uint64
	var x14 uint64
	x14, x13 = bits.Mul64(x6, arg2[2])
	var x15 uint64
	var x16 uint64
	x16, x15 = bits.Mul64(x6, arg2[1])
	var x17 uint64
	var x18 uint64

	x2 := arg1[2]
	x3 := arg1[3]
	x4 := arg1[0]
	var x5 uint64
	var x6 uint64
	x6, x5 = bits.Mul64(x4, arg2[3])
	var x7 uint64
	var x8 uint64
	x8, x7 = bits.Mul64(x4, arg2[2])
	var x9 uint64
	var x10 uint64
	x10, x9 = bits.Mul64(x4, arg2[1])
	var x11 uint64
	var x12 uint64
	x12, x11 = bits.Mul64(x4, arg2[0])
	var x13 uint64
	var x14 uint64
	x13, x14 = bits.Add64(x12, x9, uint64(0x0))
	var x15 uint64

				continue
			}
			d := auxIntToInt64(v_1.AuxInt)
			v.reset(OpConst64)
			v.AuxInt = int64ToAuxInt(c + d)
			return true
		}
		break
	}
	// match: (Add64 <t> (Mul64 x y) (Mul64 x z))
	// result: (Mul64 x (Add64 <t> y z))
	for {
		t := v.Type
		for _i0 := 0; _i0 <= 1; _i0, v_0, v_1 = _i0+1, v_1, v_0 {
			if v_0.Op != OpMul64 {
				continue
			}
			_ = v_0.Args[1]

	r0 := uint128{"r0", GP64(), GP64()}
	mul64(r0, 1, l0, l0)
	addMul64(r0, 38, l1, l4)
	addMul64(r0, 38, l2, l3)

	// r1 = 2×l0×l1 + 19×2×l2×l4 + 19×l3×l3
	r1 := uint128{"r1", GP64(), GP64()}
	mul64(r1, 2, l0, l1)
	addMul64(r1, 38, l2, l4)
	addMul64(r1, 19, l3, l3)

	// r2 = = 2×l0×l2 + l1×l1 + 19×2×l3×l4
	r2 := uint128{"r2", GP64(), GP64()}
	mul64(r2, 2, l0, l2)
	addMul64(r2, 1, l1, l1)

// bits.Mul64 and bits.Add64 intrinsics.
type uint128 struct {
	lo, hi uint64
}

// mul64 returns a * b.
func mul64(a, b uint64) uint128 {
	hi, lo := bits.Mul64(a, b)
	return uint128{lo, hi}
}

// addMul64 returns v + a * b.
func addMul64(v uint128, a, b uint64) uint128 {
	hi, lo := bits.Mul64(a, b)
	lo, c := bits.Add64(lo, v.lo, 0)

	x9 := arg1[0]
	var x10 uint64
	var x11 uint64
	x11, x10 = bits.Mul64(x9, arg2[8])
	var x12 uint64
	var x13 uint64
	x13, x12 = bits.Mul64(x9, arg2[7])
	var x14 uint64
	var x15 uint64
	x15, x14 = bits.Mul64(x9, arg2[6])
	var x16 uint64
	var x17 uint64
	x17, x16 = bits.Mul64(x9, arg2[5])
	var x18 uint64
	var x19 uint64
	x19, x18 = bits.Mul64(x9, arg2[4])
	var x20 uint64
	var x21 uint64

		// carry propagation in a separate step, as if we wrote two digit sums
		// at first (the 128-bit limbs), and then carried the tens all at once.

		h0r0 := mul64(h0, r0)
		h1r0 := mul64(h1, r0)
		h2r0 := mul64(h2, r0)
		h0r1 := mul64(h0, r1)
		h1r1 := mul64(h1, r1)
		h2r1 := mul64(h2, r1)

		// Since h2 is known to be at most 7 (5 + 1 + 1), and r0 and r1 have their