feMul(v, x, y)
	return v
}

// Square sets v = x * x, and returns v.
func (v *Element) Square(x *Element) *Element {
	feSquare(v, x)
	return v
}

// Mult32 sets v = x * y, and returns v.
func (v *Element) Mult32(x *Element, y uint32) *Element {
	x0lo, x0hi := mul51(x.l0, y)
	x1lo, x1hi := mul51(x.l1, y)
	x2lo, x2hi := mul51(x.l2, y)
	x3lo, x3hi := mul51(x.l3, y)
	x4lo, x4hi := mul51(x.l4, y)

// 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))
(Mul16 (Mul16 i:(Const16 <t>) z) x) && (z.Op != OpConst16 && x.Op != OpConst16) => (Mul16 i (Mul16 <t> x z))

	// 	x1:x0 := r.Uint64()
	// 	0:hi, lo1:lo0 := bits.Mul64(x1:x0, 0:n)
	// Writing out the multiplication in terms of bits.Mul32 allows
	// using direct hardware instructions and avoiding
	// the computations involving these zeros.
	x := r.Uint64()
	lo1a, lo0 := bits.Mul32(uint32(x), n)
	hi, lo1b := bits.Mul32(uint32(x>>32), n)
	lo1, c := bits.Add32(lo1a, lo1b, 0)
	hi += c
	if lo1 == 0 && lo0 < uint32(n) {
		n64 := uint64(n)

	if _W >= 32 {
		runModWTests(t, modWTests32)
	}
	if _W >= 64 {
		runModWTests(t, modWTests64)
	}
}

var montgomeryTests = []struct {
	x, y, m      string
	k0           uint64
	out32, out64 string
}{
	{
		"0xffffffffffffffffffffffffffffffffffffffffffffffffe",
		"0xffffffffffffffffffffffffffffffffffffffffffffffffe",
		"0xfffffffffffffffffffffffffffffffffffffffffffffffff",
		1,

				continue
			}
			d := auxIntToInt32(v_1.AuxInt)
			v.reset(OpConst32)
			v.AuxInt = int32ToAuxInt(c + d)
			return true
		}
		break
	}
	// match: (Add32 <t> (Mul32 x y) (Mul32 x z))
	// result: (Mul32 x (Add32 <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 != OpMul32 {
				continue
			}
			_ = v_0.Args[1]

	{name: "Sub32F", argLength: 2},
	{name: "Sub64F", argLength: 2},

	{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},

(Add(64|32)F ...) => (FADD(D|S) ...)

(Sub(Ptr|64|32|16|8) ...) => (SUB ...)
(Sub(64|32)F ...) => (FSUB(D|S) ...)

(Mul64 ...) => (MUL  ...)
(Mul64uhilo ...) => (LoweredMuluhilo ...)
(Mul64uover ...) => (LoweredMuluover ...)
(Mul32 ...) => (MULW ...)
(Mul16 x y) => (MULW (SignExt16to32 x) (SignExt16to32 y))
(Mul8 x y)  => (MULW (SignExt8to32 x)  (SignExt8to32 y))
(Mul(64|32)F ...) => (FMUL(D|S) ...)

(Div(64|32)F ...) => (FDIV(D|S) ...)

	{"hex float _", "{{print +0x_1.e_0p+0_2}}", "7.5", tVal, true},
	{"HEX float", "{{print +0X1.EP+2}}", "7.5", tVal, true},
	{"print multi", "{{print 1_2_3_4 7.5_00_00_00}}", "1234 7.5", tVal, true},
	{"print multi2", "{{print 1234 0x0_1.e_0p+02}}", "1234 7.5", tVal, true},

	// Fixed bugs.
	// Must separate dot and receiver; otherwise args are evaluated with dot set to variable.

	{"hex float _", "{{print +0x_1.e_0p+0_2}}", "7.5", tVal, true},
	{"HEX float", "{{print +0X1.EP+2}}", "7.5", tVal, true},
	{"print multi", "{{print 1_2_3_4 7.5_00_00_00}}", "1234 7.5", tVal, true},
	{"print multi2", "{{print 1234 0x0_1.e_0p+02}}", "1234 7.5", tVal, true},

	// Fixed bugs.
	// Must separate dot and receiver; otherwise args are evaluated with dot set to variable.

		{"LeadingZeros64", Func, 9},
		{"LeadingZeros8", Func, 9},
		{"Len", Func, 9},
		{"Len16", Func, 9},
		{"Len32", Func, 9},
		{"Len64", Func, 9},
		{"Len8", Func, 9},
		{"Mul", Func, 12},
		{"Mul32", Func, 12},
		{"Mul64", Func, 12},
		{"OnesCount", Func, 9},
		{"OnesCount16", Func, 9},
		{"OnesCount32", Func, 9},
		{"OnesCount64", Func, 9},
		{"OnesCount8", Func, 9},
		{"Rem", Func, 14},