1. disable reuse SELECTED_ROWS type variable (#16150)

2. remove lod check in reshape op
test=develop

paddle/fluid/framework/details/memory_optimize_helper.cc

@@ -337,7 +337,6 @@ bool NodeCanReused(const VarDesc& node) {
   auto type = node.GetType();
   // only these types holds bulk of gpu memory
   if (!(type == proto::VarType::LOD_TENSOR ||
-        type == proto::VarType::SELECTED_ROWS ||
         type == proto::VarType::LOD_TENSOR_ARRAY)) {
     return false;
   }

paddle/fluid/operators/reshape_op.cc

@@ -219,14 +219,6 @@ class ReshapeKernel {
           std::vector<int>(shape_data, shape_data + shape_tensor->numel());
       out_dims = ReshapeOp::ValidateShape(shape, in->dims());
     }
-    if (!in->lod().empty()) {
-      PADDLE_ENFORCE_EQ(
-          out_dims[0], in->dims()[0],
-          "Reshape operator cannot reshape an input sequence batch "
-          "into an output sequence batch that has a different "
-          "number of time steps. Please consider using "
-          "sequence_reshape op.");
-    }
 
     out->mutable_data(ctx.GetPlace(), in->type());
     framework::TensorCopy(