* CacheLoader#reload}.
   *
   * <p>As the default implementation of {@link CacheLoader#reload} is synchronous, it is
   * recommended that users of this method override {@link CacheLoader#reload} with an asynchronous
   * implementation; otherwise refreshes will be performed during unrelated cache read and write
   * operations.
   *

      CacheLoader<String, String> loader =
          new CacheLoader<String, String>() {
            @Override
            public String load(String key) {
              return key + "Load";
            }

            @Override
            public ListenableFuture<String> reload(String key, String oldValue) {
              return refreshExecutor.submit(
                  () -> {
                    reloadStarted.countDown();

      CacheLoader<String, String> loader =
          new CacheLoader<String, String>() {
            @Override
            public String load(String key) {
              return key + "Load";
            }

            @Override
            public ListenableFuture<String> reload(String key, String oldValue) {
              return refreshExecutor.submit(
                  () -> {
                    reloadStarted.countDown();

   * CacheLoader#reload}.
   *
   * <p>As the default implementation of {@link CacheLoader#reload} is synchronous, it is
   * recommended that users of this method override {@link CacheLoader#reload} with an asynchronous
   * implementation; otherwise refreshes will be performed during unrelated cache read and write
   * operations.
   *

        @Override
        protected ArrayList<LockGraphNode> initialValue() {
          return Lists.<LockGraphNode>newArrayListWithCapacity(3);
        }
      };

  /**
   * A Throwable used to record a stack trace that illustrates an example of a specific lock
   * acquisition ordering. The top of the stack trace is truncated such that it starts with the
   * acquisition of the lock in question, e.g.
   *
   * <pre>

     * at table entries if it is 0.
     *
     * - All (synchronized) write operations should write to the "count" field after structurally
     * changing any bin. The operations must not take any action that could even momentarily cause a
     * concurrent read operation to see inconsistent data. This is made easier by the nature of the

          }
          collide = false;
          continue; // Retry with expanded table
        }
        h ^= h << 13; // Rehash
        h ^= h >>> 17;
        h ^= h << 5;
        hc[0] = h; // Record index for next time
      } else if (busy == 0 && cells == as && casBusy()) {
        boolean init = false;
        try { // Initialize table
          if (cells == as) {
            Cell[] rs = new Cell[2];

    checkNotNull(reason);
    /*
     * All elements of `futures` are completed, or this future has already completed and read
     * `futures` into a local variable (in preparation for propagating cancellation to them). In
     * either case, no one needs to read `futures` for cancellation purposes later. (And
     * cancellation purposes are the main reason to access `futures`, as discussed in its docs.)
     */

       * fastest (more than twice as fast as the Java implementation, vs ~1.5x with non-final static
       * fields, on x86_32) under the Hotspot server compiler. The reason is obviously that the
       * non-final fields need to be reloaded inside the loop.
       *
       * And, no, defining (final or not) local variables out of the loop still isn't as good
       * because the null check on the theUnsafe object remains inside the loop and

    // done before submitting more of them. This interleaving
    // plus the exception mechanics account for messiness of main
    // loop.

    try {
      // Record exceptions so that if we fail to obtain any
      // result, we can throw the last exception we got.
      ExecutionException ee = null;
      long lastTime = timed ? System.nanoTime() : 0;