Ultimate Java Performance / Optimization

Question

So, I ran some tests on different data structures and noticed that when I declared my final variables, I got 10-20% performance.

It really surprised me. I thought the last keyword was purely used to limit changes in variables, and optimization would determine if any variable has a constant value or not.

Here is an example:

import javafx.scene.input.KeyCode;
import java.util.*;

public class Main {
    static /*final*/ int LOOPS = Integer.MAX_VALUE / 100;

    static /*final*/ KeyCode[] keyCodes = KeyCode.values();

    public static void main(String[] args) {
        long startTime;
        long endTime;

        testEnumSet(); //warmup
        startTime = System.nanoTime();
        testEnumSet();
        endTime = System.nanoTime();
        System.out.println("  EnumSet: " + (endTime - startTime) + "ns");
    }

    static /*final*/ EnumSet<KeyCode> enumSet = EnumSet.noneOf(KeyCode.class);
    static void testEnumSet() {
        for (int i = 0; i < LOOPS; i++) {
            /*final*/ KeyCode add = getRandomKeyCode();
            if(!enumSet.contains(add)) enumSet.add(add);

            /*final*/ KeyCode remove = getRandomKeyCode();
            if(enumSet.contains(remove)) enumSet.remove(remove);
        }
    }

    /*final*/ static Random random = new Random();
    static KeyCode getRandomKeyCode() {
        return keyCodes[random.nextInt(keyCodes.length)];
    }
}

In the finals: .... EnumSet: 652 266 207ns
Without a finale:EnumSet: 802 121 596ns

It is constantly reproduced!

Why is there such a huge difference between code that uses final and code that doesn't? Why is it not optimized? And why, in any case, finally faster, what is the difference in the generated bytecode?

Answer 1

- , , , . , , .

, .

-, , , JIT .

final .

, final , . EnumSet , . , , .

, , , JVM JVM. , - , - .

final.