* amortized) constant time operations. Expected in the hashtable sense (depends on the hash
 * function doing a good job of distributing the elements to the buckets to a distribution not far
 * from uniform), and amortized since some operations can trigger a hash table resize.
 *
 * <p>As compared with {@link java.util.LinkedHashMap}, this structure places significantly reduced
 * load on the garbage collector by only using a constant number of internal objects.

			}
			b[k] = a[k]
		}
	}
}

func TestEndianBaseCompare(t *testing.T) {
	// This test compares byte slices that are almost identical, except one
	// difference that for some j, a[j]>b[j] and a[j+1]<b[j+1]. If the implementation
	// compares large chunks with wrong endianness, it gets wrong result.
	// no vector register is larger than 512 bytes for now
	const maxLength = 512

   * CASes when performing an update operation (see method
   * retryUpdate). Upon a collision, if the table size is less than
   * the capacity, it is doubled in size unless some other thread
   * holds the lock. If a hashed slot is empty, and lock is
   * available, a new Cell is created. Otherwise, if the slot
   * exists, a CAS is tried.  Retries proceed by "double hashing",

                }
              },
              directExecutor());
      for (int i = 0; i < 50_000; i++) {
        if (i % DIRECT_EXECUTIONS_PER_THREAD == 0) {
          // after some number of iterations, switch threads
          unused =
              serializer.submit(
                  new Callable<@Nullable Void>() {
                    @Override
                    public @Nullable Void call() {

    when(backingMap.get(KEY)).thenReturn(new AtomicInteger(INITIAL_COUNT));

    assertThrows(IllegalArgumentException.class, () -> multiset.add(KEY, COUNT_TO_ADD));
  }

  /**
   * Simulate some of the races that can happen on add. We can't easily simulate the race that
   * happens when an {@link AtomicInteger#compareAndSet} fails, but we can simulate the case where

    disabled.setAccessible(true);
    disabled.set(null, true);

    // Now make a parallel FRQ and an associated FinalizableWeakReference to an object, in order to
    // exercise some classes from the parallel ClassLoader.
    AtomicReference<Object> sepFrqA =
        new AtomicReference<Object>(sepFrqC.getDeclaredConstructor().newInstance());

```Python hl_lines="14"
{!../../docs_src/background_tasks/tutorial001.py!}
```

`.add_task()` は以下の引数を受け取ります:

* バックグラウンドで実行されるタスク関数 (`write_notification`)。
* タスク関数に順番に渡す必要のある引数の列 (`email`)。
* タスク関数に渡す必要のあるキーワード引数 (`message="some notification"`)。

## 依存性注入

`BackgroundTasks` の使用は依存性注入システムでも機能し、様々な階層 (*path operations 関数*、依存性 (依存可能性)、サブ依存性など) で `BackgroundTasks` 型のパラメーターを宣言できます。

                  ? new CancellationException("Future.cancel() was called.")
                  : t;
          releaseShared(finalState);
        } else if (getState() == COMPLETING) {
          // If some other thread is currently completing the future, block until
          // they are done so we can guarantee completion.
          acquireShared(-1);
        }
        return doCompletion;
      }
    }

依赖项就是一个函数，且可以使用与*路径操作函数*相同的参数：

```Python hl_lines="8-11"
{!../../docs_src/dependencies/tutorial001.py!}
```

大功告成。

只用了**2 行**代码。

依赖项函数的形式和结构与*路径操作函数*一样。

因此，可以把依赖项当作没有「装饰器」（即，没有 `@app.get("/some-path")` ）的路径操作函数。

依赖项可以返回各种内容。

本例中的依赖项预期接收如下参数：

* 类型为 `str` 的可选查询参数 `q`
* 类型为 `int` 的可选查询参数 `skip`，默认值是 `0`
* 类型为 `int` 的可选查询参数 `limit`，默认值是 `100`

然后，依赖项函数返回包含这些值的 `dict`。

### 导入 `Depends`

	// These goroutines yield during local work, so that switching from
	// a blocked goroutine to other goroutines is profitable.
	// As a matter of fact, this benchmark still triggers some spinning in the mutex.
	m := NewDRWMutex(ds, "")
	var acc0, acc1 uint64
	b.SetParallelism(4)
	b.RunParallel(func(pb *testing.PB) {
		c := make(chan bool)
		var data [4 << 10]uint64
		for i := 0; pb.Next(); i++ {