[BUG] Ray executor fail on test_bincount_execution

[fyrestone]

Describe the bug A clear and concise description of what the bug is.

setup = <mars.deploy.oscar.session.SyncSession object at 0x1762c9520>

    def test_bincount_execution(setup):
        rs = np.random.RandomState(0)
        raw = rs.randint(0, 9, (100,))
        raw[raw == 3] = 0
        raw_weights = rs.rand(100)

        # test non-chunked
        a = tensor(raw)
        result = bincount(a).execute().fetch()
        expected = np.bincount(raw)
        np.testing.assert_array_equal(result, expected)

        weights = tensor(raw_weights)
        result = bincount(a, weights=weights).execute().fetch()
        expected = np.bincount(raw, weights=raw_weights)
        np.testing.assert_array_equal(result, expected)

        # test chunked
        a = tensor(raw, chunk_size=13)
>       result = bincount(a, chunk_size_limit=5).execute().fetch()

mars/tensor/statistics/tests/test_statistics_execution.py:528:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
mars/core/entity/tileables.py:462: in execute
    result = self.data.execute(session=session, **kw)
mars/core/entity/executable.py:144: in execute
    return execute(self, session=session, **kw)
mars/deploy/oscar/session.py:1888: in execute
    return session.execute(
mars/deploy/oscar/session.py:1682: in execute
    execution_info: ExecutionInfo = fut.result(
../../.pyenv/versions/3.8.13/lib/python3.8/concurrent/futures/_base.py:444: in result
    return self.__get_result()
../../.pyenv/versions/3.8.13/lib/python3.8/concurrent/futures/_base.py:389: in __get_result
    raise self._exception
mars/deploy/oscar/session.py:1868: in _execute
    await execution_info
../../.pyenv/versions/3.8.13/lib/python3.8/asyncio/tasks.py:695: in _wrap_awaitable
    return (yield from awaitable.__await__())
mars/deploy/oscar/session.py:105: in wait
    return await self._aio_task
mars/deploy/oscar/session.py:953: in _run_in_background
    raise task_result.error.with_traceback(task_result.traceback)
mars/services/task/supervisor/processor.py:372: in run
    await self._process_stage_chunk_graph(*stage_args)
mars/services/task/supervisor/processor.py:250: in _process_stage_chunk_graph
    chunk_to_result = await self._executor.execute_subtask_graph(
mars/services/task/execution/ray/executor.py:551: in execute_subtask_graph
    meta_list = await asyncio.gather(*output_meta_object_refs)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

awaitable = ObjectRef(1e360ffa862f8fe3ffffffffffffffffffffffff0100000001000000)

    @types.coroutine
    def _wrap_awaitable(awaitable):
        """Helper for asyncio.ensure_future().

        Wraps awaitable (an object with __await__) into a coroutine
        that will later be wrapped in a Task by ensure_future().
        """
>       return (yield from awaitable.__await__())
E       ray.exceptions.RayTaskError(TypeError): ray::execute_subtask() (pid=76452, ip=127.0.0.1)
E         File "/Users/po.lb/Work/mars/mars/services/task/execution/ray/executor.py", line 185, in execute_subtask
E           execute(context, chunk.op)
E         File "/Users/po.lb/Work/mars/mars/core/operand/core.py", line 491, in execute
E           result = executor(results, op)
E         File "/Users/po.lb/Work/mars/mars/tensor/statistics/bincount.py", line 214, in execute
E           op._execute_reduce(ctx, op)
E         File "/Users/po.lb/Work/mars/mars/tensor/statistics/bincount.py", line 193, in _execute_reduce
E           for input_key in op.inputs[0].op.source_keys:
E       TypeError: 'NoneType' object is not iterable

../../.pyenv/versions/3.8.13/lib/python3.8/asyncio/tasks.py:695: RayTaskError(TypeError)

It seems that there are some bugs in bin count operand when shuffle by FETCH_BY_INDEX mode.

To Reproduce To help us reproducing this bug, please provide information below:

1. Your Python version
2. The version of Mars you use
3. Versions of crucial packages, such as numpy, scipy and pandas
4. Full stack of the error.
5. Minimized code to reproduce the error.

Expected behavior A clear and concise description of what you expected to happen.

Additional context Add any other context about the problem here.