Created by: AoZhang

• 环境    1）PaddlePaddle版本：1.6.2    2）CPU：Intel(R) Xeon(R) CPU E5-2620 v2    3）GPU：no    4）系统环境：CentOS 6.3 ，Python版本 3.6.5

• 复现信息：

x = fluid.layers.data(name='x', shape=[3, 3], dtype='int64', append_batch_size=False)
index = fluid.layers.data(name='index', shape=[2, 2], dtype='int64', append_batch_size=False)
update = fluid.layers.data(name='update', shape=[2], dtype='int64', append_batch_size=False)

output = fluid.layers.scatter_nd_add(x, index, update)

in_data = np.array([[1, 1, 1], [2, 2, 1], [3, 3, 1]]).astype(np.int64)
index_data = np.array([[1, 1], [0, 1]]).astype(np.int64)
update_data = np.array([5, 3]).astype(np.int64)

exe.run(feed={'x':in_data, "index":index_data, "update":update_data}, fetch_list=[output])

• 问题描述： x 和 update 数据类型为 float32 时可以正常运行，改成int32、int64 都会报错。错误信息如下

--------------------------------------------
C++ Call Stacks (More useful to developers):
--------------------------------------------
0   std::string paddle::platform::GetTraceBackString<std::string const&>(std::string const&, char const*, int)
1   paddle::platform::EnforceNotMet::EnforceNotMet(std::string const&, char const*, int)
2   paddle::framework::Tensor::Slice(long, long) const
3   std::enable_if<!std::is_floating_point<long>::value, void>::type paddle::operators::elementwise_inner_add<long, long>(paddle::framework::ExecutionContext const&, long const*, long const*, long*, paddle::framework::Tensor const&, paddle::framework::Tensor*, int const&, long const&, int const&, unsigned long const&)
4   void paddle::operators::ScatterNdAdd<long, long>(paddle::framework::ExecutionContext const&, paddle::framework::Tensor const&, paddle::framework::Tensor const&, paddle::framework::Tensor*)
5   paddle::operators::ScatterNdAddOpKernel<long>::Compute(paddle::framework::ExecutionContext const&) const
6   std::_Function_handler<void (paddle::framework::ExecutionContext const&), paddle::framework::OpKernelRegistrarFunctor<paddle::platform::CPUPlace, false, 2ul, paddle::operators::ScatterNdAddOpKernel<float>, paddle::operators::ScatterNdAddOpKernel<double>, paddle::operators::ScatterNdAddOpKernel<long>, paddle::operators::ScatterNdAddOpKernel<int>, paddle::operators::ScatterNdAddOpKernel<unsigned char> >::operator()(char const*, char const*, int) const::{lambda(paddle::framework::ExecutionContext const&)#1}>::_M_invoke(std::_Any_data const&, paddle::framework::ExecutionContext const&)
7   paddle::framework::OperatorWithKernel::RunImpl(paddle::framework::Scope const&, paddle::platform::Place const&, paddle::framework::RuntimeContext*) const
8   paddle::framework::OperatorWithKernel::RunImpl(paddle::framework::Scope const&, paddle::platform::Place const&) const
9   paddle::framework::OperatorBase::Run(paddle::framework::Scope const&, paddle::platform::Place const&)
10  paddle::framework::Executor::RunPreparedContext(paddle::framework::ExecutorPrepareContext*, paddle::framework::Scope*, bool, bool, bool)
11  paddle::framework::Executor::Run(paddle::framework::ProgramDesc const&, paddle::framework::Scope*, int, bool, bool, std::vector<std::string, std::allocator<std::string> > const&, bool)

------------------------------------------
Python Call Stacks (More useful to users):
------------------------------------------
  File "/xxx/python3.6/site-packages/paddle/fluid/framework.py", line 2488, in append_op
    attrs=kwargs.get("attrs", None))
  File "/xxx/python3.6/site-packages/paddle/fluid/layer_helper.py", line 43, in append_op
    return self.main_program.current_block().append_op(*args, **kwargs)
  File "/xxx/python3.6/site-packages/paddle/fluid/layers/nn.py", line 11104, in scatter_nd_add
    outputs={"Out": output})
  File "./script/cookbook.py", line 455, in scatter_nd_add
    output = fluid.layers.scatter_nd_add(x, index, update)
  File "./script/cookbook.py", line 581, in <module>
    fn()

----------------------
Error Message Summary:
----------------------
Error: The end row index is out of bound.
  [Hint: Expected end_idx <= dims_[0], but received end_idx:5 > dims_[0]:3.] at (/paddle/paddle/fluid/framework/tensor.cc:79)
  [operator < scatter_nd_add > error]