"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)