Update new_op_cn.md

fix  #132

doc/fluid/dev/new_op_cn.md

@@ -150,8 +150,9 @@ class MulOp : public framework::OperatorWithKernel {
 
  protected:
   void InferShape(const framework::InferShapeContext &ctx) const override {
-    auto dim0 = ctx.Input<Tensor>("X")->dims();
-    auto dim1 = ctx.Input<Tensor>("Y")->dims();
+    //never use Input or Output if you want a to get a LoDTensor.
+    auto dim0 = ctx.Input<LoDTensor>("X")->dims();
+    auto dim1 = ctx.Input<LoDTensor>("Y")->dims();
     PADDLE_ENFORCE_EQ(dim0.size(), 2,
                       "input X(%s) should be a tensor with 2 dims, a matrix",
                       ctx.op_.Input("X"));
@@ -161,7 +162,7 @@ class MulOp : public framework::OperatorWithKernel {
     PADDLE_ENFORCE_EQ(
         dim0[1], dim1[0],
         "First matrix's width must be equal with second matrix's height.");
-    ctx.Output<Tensor>("Out")->Resize({dim0[0], dim1[1]});
+    ctx.Output<LoDTensor>("Out")->Resize({dim0[0], dim1[1]});
   }
 };
 ```
@@ -201,6 +202,8 @@ MulOp(const std::string &type, const framework::VariableNameMap &inputs,
 - 与`InferShapeContext`相比，`ExecutionContext`增加了设备类型，同样可获取到输入输出和属性参数。
 - `Compute`函数里实现`OpKernel`的具体计算逻辑。
 
+Op的输入和输出可分别通过ExecutionContext::Input()和ExecutionContext::Output()获得。注意：若op的输入/输出的变量类型是LoDTensor（fluid默认所有的Tensor默认都是LoDTensor类型），请写成ExecutionContext::Input()和ExecutionContext::Output()，不要写ExecutionContext::Input()和ExecutionContext::Output()。因为若实际的变量类型为SelectedRows，Input()和Output()方法会将SelectedRows类型特化为Tensor，导致潜在的错误。
+
 下面是 `MulKernel` `Compute`的实现：
 
   ```cpp
@@ -208,9 +211,9 @@ MulOp(const std::string &type, const framework::VariableNameMap &inputs,
   class MulKernel : public framework::OpKernel {
   public:
   void Compute(const framework::ExecutionContext& context) const override {
-    auto* X = context.Input<Tensor>("X");
-    auto* Y = context.Input<Tensor>("Y");
-    auto* Z = context.Output<Tensor>("Out");
+    auto* X = context.Input<LoDTensor>("X");
+    auto* Y = context.Input<LoDTensor>("Y");
+    auto* Z = context.Output<LoDTensor>("Out");
     Z->mutable_data<T>(context.GetPlace());
     auto& device_context = context.template device_context<DeviceContext>();
     math::matmul<DeviceContext, T>(*X, false, *Y, false, 1, Z, 0, device_context);