nanoTime对volatile 测试的一种写法

今天脑筋有点搭牢，想了半天才看出为什么以下两段代码效果是相同的。。。

第一种好处是可以直接批量复制黏贴system.out,  不用改什么东西

  private static long i;
  private volatile static long vt;
  private static final int TEST_SIZE = 10000000;

long time = System.nanoTime();

for (int n = 0; n < TEST_SIZE; n++)

	vt = System.currentTimeMillis();

System.out.println(-time + (time = System.nanoTime()));


for (int n = 0; n < TEST_SIZE; n++)

	i = System.currentTimeMillis();

System.out.println(-time + (time = System.nanoTime()));

　　

这种每次要在循环头上重置nanoTime

private static long i;
private volatile static long vt;
private static final int TEST_SIZE = 10000000;

long time = System.nanoTime();

for (int n = 0; n < TEST_SIZE; n++) {

            vt = System.currentTimeMillis();

}

System.out.println((System.nanoTime() - time));



time = System.nanoTime();

for (int n = 0; n < TEST_SIZE; n++) {

      i = System.currentTimeMillis();

}

System.out.println((System.nanoTime() - time));