Java 随机数生成器 Random & SecureRandom

Java 里提供了一些用于生成随机数的工具类，这里分析一下其实现原理，以及他们之间的区别、使用场景。

java.util.Random

Random 是比较常用的随机数生成类，它的基本信息在类的注释里都写到了，下面是 JDK8 里该类的注释：

1. /**
    * An instance of this class is used to generate a stream of
    * pseudorandom numbers. The class uses a 48-bit seed, which is
    * modified using a linear congruential formula. (See Donald Knuth,
    * <i>The Art of Computer Programming, Volume 2</i>, Section 3.2.1.)
    * <p>
    * If two instances of {@code Random} are created with the same
    * seed, and the same sequence of method calls is made for each, they
    * will generate and return identical sequences of numbers. In order to
    * guarantee this property, particular algorithms are specified for the
    * class {@code Random}. Java implementations must use all the algorithms
    * shown here for the class {@code Random}, for the sake of absolute
    * portability of Java code. However, subclasses of class {@code Random}
    * are permitted to use other algorithms, so long as they adhere to the
    * general contracts for all the methods.
    * <p>
    * The algorithms implemented by class {@code Random} use a
    * {@code protected} utility method that on each invocation can supply
    * up to 32 pseudorandomly generated bits.
    * <p>
    * Many applications will find the method {@link Math#random} simpler to use.
    *
    * <p>Instances of {@code java.util.Random} are threadsafe.
    * However, the concurrent use of the same {@code java.util.Random}
    * instance across threads may encounter contention and consequent
    * poor performance. Consider instead using
    * {@link java.util.concurrent.ThreadLocalRandom} in multithreaded
    * designs.
    *
    * <p>Instances of {@code java.util.Random} are not cryptographically
    * secure.  Consider instead using {@link java.security.SecureRandom} to
    * get a cryptographically secure pseudo-random number generator for use
    * by security-sensitive applications.
    *
    * @author  Frank Yellin
    * @since   1.0
    */
   

   翻译一下，主要有以下几点：

   Random 类使用线性同余法 linear congruential formula 来生成伪随机数。
   两个 Random 实例，如果使用相同的种子 seed，那他们产生的随机数序列也是一样的。
   Random 是线程安全的，你的程序如果对性能要求比较高的话，推荐使用 ThreadLocalRandom。
   Random 不是密码学安全的，加密相关的推荐使用 SecureRandom。
   Random 的基本用法如下所示：

   Random random = new Random();
   int r = random.nextInt(); // 生成一个随机数
   

   从下面的源码中可以看到，Random 的默认使用当前系统时钟来生成种子 seed。

       private static final AtomicLong seedUniquifier = new AtomicLong(8682522807148012L);
       public Random() {
           this(seedUniquifier() ^ System.nanoTime());
       }
   
       public Random(long seed) {
           if (getClass() == Random.class)
               this.seed = new AtomicLong(initialScramble(seed));
           else {
               // subclass might have overriden setSeed
               this.seed = new AtomicLong();
               setSeed(seed);
           }
       }
   
       private static long seedUniquifier() {
           for (;;) {
               long current = seedUniquifier.get();
               long next = current * 181783497276652981L;
               if (seedUniquifier.compareAndSet(current, next))
                   return next;
           }
       }
   

   java.Security.SecureRandom

   介绍 Random 类时提到过，要生成加密基本的随机数应该使用 SecureRandom 类，该类信息如下所示：

   /**
    * This class provides a cryptographically strong random number
    * generator (RNG).
    *
    * <p>A cryptographically strong random number
    * minimally complies with the statistical random number generator tests
    * specified in <a href="http://csrc.nist.gov/cryptval/140-2.htm">
    * <i>FIPS 140-2, Security Requirements for Cryptographic Modules</i></a>,
    * section 4.9.1.
    * Additionally, SecureRandom must produce non-deterministic output.
    * Therefore any seed material passed to a SecureRandom object must be
    * unpredictable, and all SecureRandom output sequences must be
    * cryptographically strong, as described in
    * <a href="http://www.ietf.org/rfc/rfc1750.txt">
    * <i>RFC 1750: Randomness Recommendations for Security</i></a>.
    *
    * <p>A caller obtains a SecureRandom instance via the
    * no-argument constructor or one of the {@code getInstance} methods:
    *
    * <pre>
    *      SecureRandom random = new SecureRandom();
    * </pre>
    *
    * <p> Many SecureRandom implementations are in the form of a pseudo-random
    * number generator (PRNG), which means they use a deterministic algorithm
    * to produce a pseudo-random sequence from a true random seed.
    * Other implementations may produce true random numbers,
    * and yet others may use a combination of both techniques.
    *
    * <p> Typical callers of SecureRandom invoke the following methods
    * to retrieve random bytes:
    *
    * <pre>
    *      SecureRandom random = new SecureRandom();
    *      byte bytes[] = new byte[20];
    *      random.nextBytes(bytes);
    * </pre>
    *
    * <p> Callers may also invoke the {@code generateSeed} method
    * to generate a given number of seed bytes (to seed other random number
    * generators, for example):
    * <pre>
    *      byte seed[] = random.generateSeed(20);
    * </pre>
    *
    * Note: Depending on the implementation, the {@code generateSeed} and
    * {@code nextBytes} methods may block as entropy is being gathered,
    * for example, if they need to read from /dev/random on various Unix-like
    * operating systems.
    */
   

   主要有以下几点：

   该类提供了能满足加密要求的强随机数生成器。
   传递给 SecureRandom 种子必须是不可预测的，seed 使用不当引发的安全漏洞可以看看 比特币电子钱包漏洞。
   一般使用默认的种子生成策略就行，对应 Linux 里面就是 /dev/random 和 /dev/urandom。其实现原理是：操作系统收集了一些随机事件，比如鼠标点击，键盘点击等等，SecureRandom 使用这些随机事件作为种子。
   使用 /dev/random 来生成种子时，可能会因为熵不够而阻塞，性能比较差。
   SecureRandom 用法如下所示：

   SecureRandom random = new SecureRandom();
   byte[] data = random.nextBytes(16);
   

   下面我们看看其内部实现：

       synchronized public void nextBytes(byte[] bytes) {
           secureRandomSpi.engineNextBytes(bytes);
       }
       public SecureRandom() {
           super(0);
           getDefaultPRNG(false, null);
       }
       private void getDefaultPRNG(boolean setSeed, byte[] seed) {
           String prng = getPrngAlgorithm();
           if (prng == null) {
               // bummer, get the SUN implementation
               prng = "SHA1PRNG";
               this.secureRandomSpi = new sun.security.provider.SecureRandom();
               this.provider = Providers.getSunProvider();
               if (setSeed) {
                   this.secureRandomSpi.engineSetSeed(seed);
               }
           } else {
               try {
                   SecureRandom random = SecureRandom.getInstance(prng);
                   this.secureRandomSpi = random.getSecureRandomSpi();
                   this.provider = random.getProvider();
                   if (setSeed) {
                       this.secureRandomSpi.engineSetSeed(seed);
                   }
               } catch (NoSuchAlgorithmException nsae) {
                   // never happens, because we made sure the algorithm exists
                   throw new RuntimeException(nsae);
               }
           }
           if (getClass() == SecureRandom.class) {
               this.algorithm = prng;
           }
       }
   

   在 mac 环境下使用 JDK8 测试时发现，默认使用了 NativePRNG 而非 SHA1PRNG，但是 NativePRNG 其实还是在 sun.security.provider.SecureRandom 的基础上做了一些封装。

   在 sun.security.provider.SeedGenerator 类里，可以看到 seed 是利用 /dev/random 或 /dev/urandom 来生成的，启动应用程序时可以通过参数 -Djava.security.egd=file:/dev/urandom 来指定 seed 源。

       static {
           String var0 = SunEntries.getSeedSource();
           if (!var0.equals("file:/dev/random") && !var0.equals("file:/dev/urandom")) {
               if (var0.length() != 0) {
                   try {
                       instance = new SeedGenerator.URLSeedGenerator(var0);
                       if (debug != null) {
                           debug.println("Using URL seed generator reading from " + var0);
                       }
                   } catch (IOException var2) {
                       if (debug != null) {
                           debug.println("Failed to create seed generator with " + var0 + ": " + var2.toString());
                       }
                   }
               }
           } else {
               try {
                   instance = new NativeSeedGenerator(var0);
                   if (debug != null) {
                       debug.println("Using operating system seed generator" + var0);
                   }
               } catch (IOException var3) {
                   if (debug != null) {
                       debug.println("Failed to use operating system seed generator: " + var3.toString());
                   }
               }
           }
   
           if (instance == null) {
               if (debug != null) {
                   debug.println("Using default threaded seed generator");
               }
   
               instance = new SeedGenerator.ThreadedSeedGenerator();
           }
   
       }
   

   在 Random 类里，多个实例设置相同的seed，产生的随机数序列也是一样的。而 SecureRandom 则不同，运行下面的代码：

   public class RandomTest {
       public static void main(String[] args) {
           byte[] seed = "hello".getBytes();
           for (int i = 0; i < 10; ++i) {
               SecureRandom secureRandom = new SecureRandom(seed);
               System.out.println(secureRandom.nextInt());
           }
       }
   }
   

   输出如下所示，每次运行产生的随机数都不一样。

   -2105877601
   1151182748
   1329080810
   -617594950
   2094315881
   -1649759687
   -1360561033
   -653424535
   -927058354
   -1577199965
   

   为什么呢？因为 engineSetSeed 方法设置 seed 时调用的是静态实例 INSTANCE 的 implSetSeed 方法，该方法通过 getMixedRandom 得到的 SecureRandom 来设置 seed，而这个 SecureRandom 初始化种子是系统的。

       private static final NativePRNG.RandomIO INSTANCE;
   	// in NativePRNG
       protected void engineSetSeed(byte[] var1) {
           INSTANCE.implSetSeed(var1);
       }
   		
           private void implSetSeed(byte[] var1) {
               Object var2 = this.LOCK_SET_SEED;
               synchronized(this.LOCK_SET_SEED) {
                   if (!this.seedOutInitialized) {
                       this.seedOutInitialized = true;
                       this.seedOut = (OutputStream)AccessController.doPrivileged(new PrivilegedAction<OutputStream>() {
                           public OutputStream run() {
                               try {
                                   return new FileOutputStream(RandomIO.this.seedFile, true);
                               } catch (Exception var2) {
                                   return null;
                               }
                           }
                       });
                   }
   
                   if (this.seedOut != null) {
                       try {
                           this.seedOut.write(var1);
                       } catch (IOException var5) {
                           throw new ProviderException("setSeed() failed", var5);
                       }
                   }
   
                   this.getMixRandom().engineSetSeed(var1);
               }
           }
   
           private SecureRandom getMixRandom() {
               SecureRandom var1 = this.mixRandom;
               if (var1 == null) {
                   Object var2 = this.LOCK_GET_BYTES;
                   synchronized(this.LOCK_GET_BYTES) {
                       var1 = this.mixRandom;
                       if (var1 == null) {
                           var1 = new SecureRandom();
   
                           try {
                               byte[] var3 = new byte[20];
                               readFully(this.nextIn, var3);
                               var1.engineSetSeed(var3);
                           } catch (IOException var5) {
                               throw new ProviderException("init failed", var5);
                           }
   
                           this.mixRandom = var1;
                       }
                   }
               }
   
               return var1;
           }
   
   

   在 sun.security.provider.SecureRandom.engineSetSeed 方法里，新种子的生成不仅和刚设置的 seed 有关，也和原来的种子（系统产生的 seed）有关。

   	// in sun.security.provider.SecureRandom 
       public synchronized void engineSetSeed(byte[] var1) {
           if (this.state != null) {
               this.digest.update(this.state);
   
               for(int var2 = 0; var2 < this.state.length; ++var2) {
                   this.state[var2] = 0;
               }
           }
   
           this.state = this.digest.digest(var1);
       }
   

   /dev/random 与 /dev/urandom
   在 Linux 操作系统中，有一个特殊的设备文件 /dev/random，可以用作随机数发生器或伪随机数发生器。

   在读取时，/dev/random 设备会返回小于熵池噪声总数的随机字节。/dev/random 可生成高随机性的公钥或一次性密码本。若熵池空了，对/dev/random的读操作将会被阻塞，直到从别的设备中收集到了足够的环境噪声为止。

   当然你也可以设置成不堵塞，当你在 open 的时候设置参数 O_NONBLOCK， 但是当你read 的时候，如果熵池空了，会返回 -1。

   /dev/random 的一个副本是 /dev/urandom （“unlocked”，非阻塞的随机数发生器），它会重复使用熵池中的数据以产生伪随机数据。这表示对/dev/urandom的读取操作不会产生阻塞，但其输出的熵可能小于 /dev/random 的。它可以作为生成较低强度密码的伪随机数生成器，不建议用于生成高强度长期密码。