We observe a performance bottleneck in byte reversing code in Apache ORC on Tez framework as shown below. We could optimize this typical code by using pattern matching in ppc.ad.

public float readFloat(InputStream in) throws IOException {
  readFully(in, readBuffer, 0, 4);
  int val = (((readBuffer[0] & 0xff) << 0)
          + ((readBuffer[1] & 0xff) << 8)
          + ((readBuffer[2] & 0xff) << 16)
          + ((readBuffer[3] & 0xff) << 24));
  return Float.intBitsToFloat(val);
}

https://github.com/apache/orc/blob/master/java/core/src/java/org/apache/orc/impl/SerializationUtils.java