VANDPD X11, X7, K7, X8                             // 6251c50f54c3
	VANDPD -17(BP)(SI*2), X7, K7, X8                   // 6271c50f548475efffffff
	VANDPD 7(AX)(CX*2), X7, K7, X8                     // 6271c50f54844807000000
	VANDPD X22, X0, K7, X8                             // 6231fd0f54c6
	VANDPD X1, X0, K7, X8                              // 6271fd0f54c1

	var x878 uint64
	x878, x877 = bits.Mul64(x8, arg2[8])
	var x879 uint64
	var x880 uint64
	x880, x879 = bits.Mul64(x8, arg2[7])
	var x881 uint64
	var x882 uint64
	x882, x881 = bits.Mul64(x8, arg2[6])
	var x883 uint64
	var x884 uint64
	x884, x883 = bits.Mul64(x8, arg2[5])
	var x885 uint64
	var x886 uint64
	x886, x885 = bits.Mul64(x8, arg2[4])
	var x887 uint64
	var x888 uint64

	VPCMPGTB X26, X8, K1, K1                           // 62913d0964ca or 6291bd0964ca
	VPCMPGTB (BX), X8, K1, K1                          // 62f13d09640b or 62f1bd09640b
	VPCMPGTB -17(BP)(SI*1), X8, K1, K1                 // 62f13d09648c35efffffff or 62f1bd09648c35efffffff
	VPCMPGTB X26, X8, K1, K3                           // 62913d0964da or 6291bd0964da

	"SP",
	"BP",
	"SI",
	"DI",
	"R8",
	"R9",
	"R10",
	"R11",
	"R12",
	"R13",
	"g", // a.k.a. R14
	"R15",
	"X0",
	"X1",
	"X2",
	"X3",
	"X4",
	"X5",
	"X6",
	"X7",
	"X8",
	"X9",
	"X10",
	"X11",
	"X12",
	"X13",
	"X14",
	"X15", // constant 0 in ABIInternal

	// If you add registers, update asyncPreempt in runtime

	// pseudo-registers
	"SB",
}

#define shiftC1Left BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xff; BYTE $0x08 // PALIGNR $8, X7, X7
#define shiftC2Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xc0; BYTE $0x08 // PALIGNR $8, X8, X8
#define shiftC3Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xf6; BYTE $0x08 // PALIGNR $8, X14, X14

  // Build the shape of the refined type, if lhs is unranked it
  // will be directly the shape of the refined type, otherwise we merged by
  // taking the most specialized. This combines `10x?x?` and `?x?x8` into
  // `10x?x8`.
  if (!lhs_shape_type.hasRank()) {
    if (rhs_shape_type.hasRank()) {
      shape.append(rhs_shape_type.getShape().begin(),
                   rhs_shape_type.getShape().end());
      refined_shape = true;

// left shift
(Lsh(64|32|16|8)x64 <t> x y) && shiftIsBounded(v) => (SLL <t> x y)
(Lsh(64|32|16|8)x32 <t> x y) && shiftIsBounded(v) => (SLL <t> x y)
(Lsh(64|32|16|8)x16 <t> x y) && shiftIsBounded(v) => (SLL <t> x y)
(Lsh(64|32|16|8)x8  <t> x y) && shiftIsBounded(v) => (SLL <t> x y)

// signed right shift
(Rsh64x(64|32|16|8) <t> x y) && shiftIsBounded(v) => (SRA <t> x y)
(Rsh32x(64|32|16|8) <t> x y) && shiftIsBounded(v) => (SRA <t> (SignExt32to64 x) y)