legacy_request_state_context_var.reset(contextvar_token)


@app.middleware("http")
async def custom_middleware(
    request: Request, call_next: Callable[[Request], Awaitable[Response]]
):
    response = await call_next(request)
    response.headers["custom"] = "foo"
    return response


@app.get("/user", dependencies=[Depends(set_up_request_state_dependency)])
def get_user():

			return
		}
		received := string(b)
		if received != "message one\n" {
			t.Errorf(`server: expected: "message one\n", got: %v`, received)
			return
		}

		// Wait for more than read timeout to simulate processing.
		time.Sleep(3 * time.Second)

		_, terr = deadlineconn.Read(b)
		if terr != nil {
			t.Errorf("failed to read from client. %v", terr)
			return
		}

```

Enabling blocking will cause the ring buffer to behave like a buffered [io.Pipe](https://pkg.go.dev/io#Pipe).

Regular Reads will block until data is available, but not wait for a full buffer. 
Writes will block until there is space available and writes bigger than the buffer will wait for reads to make space.

`TryRead` and `TryWrite` are still available for non-blocking reads and writes.

    return queryEarliestAvailable(nowMicros) - timeoutMicros <= nowMicros;
  }

  /**
   * Reserves next ticket and returns the wait time that the caller must wait for.
   *
   * @return the required wait time, never negative
   */
  final long reserveAndGetWaitLength(int permits, long nowMicros) {
    long momentAvailable = reserveEarliestAvailable(permits, nowMicros);

        // this is not too great as it depends on timing/clockskew
        // first we need to obtain a ticket, therefor need valid credentials
        // then we need to wait until the ticket is expired
        int wait = 10 * 1000;
        long princExp = start + ( wait / 2 );
        Subject s = getInitiatorSubject(getTestUser(), getTestUserPassword(), getTestUserDomainRequired(), princExp);

    override fun coordinatorNotify(taskRunner: TaskRunner) {
      taskRunner.condition.signal()
    }

    /**
     * Wait a duration in nanoseconds. Unlike [java.lang.Object.wait] this interprets 0 as
     * "don't wait" instead of "wait forever".
     */
    @Throws(InterruptedException::class)
    @Suppress("PLATFORM_CLASS_MAPPED_TO_KOTLIN")
    override fun coordinatorWait(

   * numElements} elements are not available, it will wait for them up to the specified timeout.
   *
   * @param q the blocking queue to be drained
   * @param buffer where to add the transferred elements
   * @param numElements the number of elements to be waited for
   * @param timeout how long to wait before giving up
   * @return the number of elements transferred

   * Last, but not least: consider a RateLimiter with rate of 1 permit per second, currently
   * completely unused, and an expensive acquire(100) request comes. It would be nonsensical to just
   * wait for 100 seconds, and /then/ start the actual task. Why wait without doing anything? A much
   * better approach is to /allow/ the request right away (as if it was an acquire(1) request

  {
    tensorflow::mutex_lock l(execution_mutex_);
    while (execution_state_ != ExecutionState::kIdle) {
      // If there's already a pending execution, wait until Join finishes before
      // starting on the next operation.
      finished_join_.wait(l);
    }
    context_ = context;
    operation_name_ = operation_name;
    step_id_ = step_id;
    op_inputs_ = inputs;
    attributes_ = attributes;

        "requestBody": {
            "content": {"application/x-yaml": {"schema": Item.model_json_schema()}},
            "required": True,
        },
    },
)
async def create_item(request: Request):
    raw_body = await request.body()
    try:
        data = yaml.safe_load(raw_body)
    except yaml.YAMLError:
        raise HTTPException(status_code=422, detail="Invalid YAML")
    try:
        item = Item.model_validate(data)