// result: (Leq16U (RotateLeft16 <typ.UInt16> (Add16 <typ.UInt16> (Mul16 <typ.UInt16> (Const16 <typ.UInt16> [int16(sdivisible16(c).m)]) x) (Const16 <typ.UInt16> [int16(sdivisible16(c).a)]) ) (Const16 <typ.UInt16> [int16(16-sdivisible16(c).k)]) ) (Const16 <typ.UInt16> [int16(sdivisible16(c).max)]) )
	for {
		for _i0 := 0; _i0 <= 1; _i0, v_0, v_1 = _i0+1, v_1, v_0 {
			x := v_0
			if v_1.Op != OpMul16 {
				continue
			}

		v.AuxInt = int64ToAuxInt(int64(uint64(c) << uint64(d)))
		v.AddArg(x)
		return true
	}
	// match: (ADDshiftLL <typ.UInt16> [8] (UBFX <typ.UInt16> [armBFAuxInt(8, 8)] x) x)
	// result: (REV16W x)
	for {
		if v.Type != typ.UInt16 || auxIntToInt64(v.AuxInt) != 8 || v_0.Op != OpARM64UBFX || v_0.Type != typ.UInt16 || auxIntToArm64BitField(v_0.AuxInt) != armBFAuxInt(8, 8) {
			break
		}
		x := v_0.Args[0]
		if x != v_1 {

For example:

```python
import tensorflow as tf
from tensorflow.python.ops import bitwise_ops
dtype_list = [tf.int8, tf.int16, tf.int32, tf.int64,
              tf.uint8, tf.uint16, tf.uint32, tf.uint64]

for dtype in dtype_list:
  lhs = tf.constant([0, 5, 3, 14], dtype=dtype)
  rhs = tf.constant([5, 0, 7, 11], dtype=dtype)

    *   Added 16-bit and 32-bit int support for the built-in op `bitcast`.
    *   Added 8-bit/16-bit/32-bit int/uint support for the built-in op `bitwise_xor`
    *   Added int16 indices support for built-in op `gather` and `gather_nd`.
    *   Added 8-bit/16-bit/32-bit int/uint support for the built-in op `right_shift`
    *   Added reference implementation for 16-bit int unquantized `add`.