&& x.Uses == 1
  && is20Bit(int64(i)-12)
  && setPos(v, x.Pos)
  && clobber(x)
  => (STM4 [i-12] {s} p w0 w1 w2 w3 mem)
(STM2 [i] {s} p w2 w3 x:(STM2 [i-8] {s} p w0 w1 mem))
  && x.Uses == 1
  && is20Bit(int64(i)-8)
  && setPos(v, x.Pos)
  && clobber(x)
  => (STM4 [i-8] {s} p w0 w1 w2 w3 mem)
// 64-bit
(MOVDstore [i] {s} p w1 x:(MOVDstore [i-8] {s} p w0 mem))
  && p.Op != OpSB

			asm:            "STMY",
			faultOnNilArg0: true,
			symEffect:      "Write",
			clobberFlags:   true, // TODO(mundaym): currently uses AGFI to handle large offsets
		},
		{
			name:      "STM4",
			argLength: 6,
			reg: regInfo{inputs: []regMask{
				ptrsp,
				buildReg("R1"),
				buildReg("R2"),
				buildReg("R3"),
				buildReg("R4"),
				0,
			}},
			aux:            "SymOff",

		v.AddArg5(p, w0, w1, w2, mem)
		return true
	}
	// match: (MOVWstore [i] {s} p w3 x:(STM3 [i-12] {s} p w0 w1 w2 mem))
	// cond: x.Uses == 1 && is20Bit(int64(i)-12) && setPos(v, x.Pos) && clobber(x)
	// result: (STM4 [i-12] {s} p w0 w1 w2 w3 mem)
	for {
		i := auxIntToInt32(v.AuxInt)
		s := auxToSym(v.Aux)
		p := v_0
		w3 := v_1
		x := v_2

				{1, 2},     // R1
				{2, 4},     // R2
				{3, 8},     // R3
				{0, 56318}, // R1 R2 R3 R4 R5 R6 R7 R8 R9 R11 R12 R14 SP
			},
		},
	},
	{
		name:           "STM4",
		auxType:        auxSymOff,
		argLen:         6,
		clobberFlags:   true,
		faultOnNilArg0: true,
		symEffect:      SymWrite,
		asm:            s390x.ASTMY,
		reg: regInfo{