不安全的。允许抛出检查异常而不声明它们。

这在处理反射或AOP的某些情况下非常有用。

假设您为用户定义的接口构建了通用代理。在特殊情况下，用户可以通过在接口中声明异常来指定由实现抛出的异常。这是我所知道的唯一方法，在接口的动态实现中引发一个受控异常。

import org.junit.Test;
/** need to allow forbidden references! */ import sun.misc.Unsafe;

/**
 * Demonstrate how to throw an undeclared checked exception.
 * This is a hack, because it uses the forbidden Class {@link sun.misc.Unsafe}.
 */
public class ExceptionTest {

    /**
     * A checked exception.
     */
    public static class MyException extends Exception {
        private static final long serialVersionUID = 5960664994726581924L;
    }

    /**
     * Throw the Exception.
     */
    @SuppressWarnings("restriction")
    public static void throwUndeclared() {
        getUnsafe().throwException(new MyException());
    }

    /**
     * Return an instance of {@link sun.misc.Unsafe}.
     * @return THE instance
     */
    @SuppressWarnings("restriction")
    private static Unsafe getUnsafe() {
        try {

            Field singleoneInstanceField = Unsafe.class.getDeclaredField("theUnsafe");
            singleoneInstanceField.setAccessible(true);
            return (Unsafe) singleoneInstanceField.get(null);

        } catch (IllegalArgumentException e) {
            throw createExceptionForObtainingUnsafe(e);
        } catch (SecurityException e) {
            throw createExceptionForObtainingUnsafe(e);
        } catch (NoSuchFieldException e) {
            throw createExceptionForObtainingUnsafe(e);
        } catch (IllegalAccessException e) {
            throw createExceptionForObtainingUnsafe(e);
        }
    }

    private static RuntimeException createExceptionForObtainingUnsafe(final Throwable cause) {
        return new RuntimeException("error while obtaining sun.misc.Unsafe", cause);
    }


    /**
     * scenario: test that an CheckedException {@link MyException} can be thrown
     * from an method that not declare it.
     */
    @Test(expected = MyException.class)
    public void testUnsingUnsaveToThrowCheckedException() {
        throwUndeclared();
    }
}

为了有效地复制内存(至少对于短块，复制速度比System.arraycopy()快);由Java LZF和Snappy编解码器使用。它们使用'getLong'和'putLong'，这比逐字节复制要快;在复制16/32/64字节块时尤其有效。

有趣的是，我甚至从未听说过这个类(这可能是一件好事，真的)。

我想到的一件事是使用Unsafe#setMemory将包含敏感信息的缓冲区归零(密码、密钥等)。您甚至可以对“不可变”对象的字段执行此操作(然后，我再次假设普通的反射也可以在这里执行此操作)。但我不是安全专家，所以对此持保留态度。

我们使用Unsafe实现了数组、HashMaps、TreeMaps等大型集合。为了避免/最小化碎片，我们使用dlmalloc over unsafe的概念实现内存分配器。这帮助我们获得了并发性方面的性能。

它使用的一个例子是random方法，该方法调用不安全的方法来更改种子。

这个网站也有它的一些用途。

使用它来有效地访问和分配大量的内存，例如在您自己的体素引擎中!(例如《我的世界》风格的游戏。)

In my experience, the JVM is often unable to eliminate bounds-checking in place you truly need it. For example, if you're iterating over a large array, but the actual memory access is tucked underneath a non-virtual* method call in the loop, the JVM may still perform a bounds check with each array access, rather than once just before the loop. Thus, for potentially large performance gains, you can eliminate JVM bounds-checking inside the loop via a method which employs sun.misc.Unsafe to access the memory directly, making sure to do any bounds-checking yourself at the correct places. (You are gonna bounds check at some level, right?) *by non-virtual, I mean the JVM shouldn't have to dynamically resolve whatever your particular method is, because you've correctly guaranteed that class/method/instance are some combination of static/final/what-have-you.

对于我自己开发的体素引擎来说，这在块生成和序列化期间(在我同时读取/写入整个数组的低位置)带来了显著的性能提升。结果可能会有所不同，但如果缺乏界限消除是您的问题，那么这将解决它。

There are some potentially major problems with this: specifically, when you provide the ability to access memory without bounds-checking to clients of your interface, they will probably abuse it. (Don't forget that hackers can also be clients of your interface... especially in the case of a voxel engine written in Java.) Thus, you should either design your interface in a way such that memory access cannot be abused, or you should be extremely careful to validate user-data before it can ever, ever mingle with your dangerous interface. Considering the catastrophic things a hacker can do with unchecked memory access, it's probably best to take both approaches.