FULL PRODUCT VERSION :
JDK 10, VM 10+46

FULL OS VERSION :
Microsoft Windows [Version 10.0.16299.309]

A DESCRIPTION OF THE PROBLEM :
32 bit integer dot product is no longer vectorised. This represents a performance regression.

THE PROBLEM WAS REPRODUCIBLE WITH -Xint FLAG: Did not try

THE PROBLEM WAS REPRODUCIBLE WITH -server FLAG: Yes

REGRESSION. Last worked in version 9.0.1

STEPS TO FOLLOW TO REPRODUCE THE PROBLEM :
Run the benchmark with JMH

EXPECTED VERSUS ACTUAL BEHAVIOR :
There should be similar performance between the two versions:

# JMH version: 1.19
# VM version: JDK 9.0.1, VM 9.0.1+11
Benchmark (size) Mode Cnt Score Error Units
DotProduct.dotProduct 1024 thrpt 10 2.998 ± 0.028 ops/us

# JMH version: 1.19
# VM version: JDK 10, VM 10+46
Benchmark (size) Mode Cnt Score Error Units
DotProduct.dotProduct 1024 thrpt 10 1.907 ± 0.063 ops/us
REPRODUCIBILITY :
This bug can be reproduced always.

---------- BEGIN SOURCE ----------
import org.openjdk.jmh.annotations.*;

import java.util.concurrent.ThreadLocalRandom;
import java.util.concurrent.TimeUnit;

@State(Scope.Thread)
@OutputTimeUnit(TimeUnit.MICROSECONDS)
public class DotProduct {

  private int[] left;
  private int[] right;

  @Param({"1024"})
  int size;

  @Setup(Level.Trial)
  public void init() {
    left = new int[size];
    right = new int[size];
    for (int i = 0; i < size; ++i) {
      left[i] = ThreadLocalRandom.current().nextInt();
      right[i] = ThreadLocalRandom.current().nextInt();
    }
  }

  @Benchmark
  public int dotProduct() {
    int dp = 0;
    for (int i = 0; i < left.length && i < right.length; ++i) {
      dp += left[i] * right[i ];
    }
    return dp;
  }

}
---------- END SOURCE ----------