fix runtime shape

lite/kernels/mlu/bridges/graph.h

@@ -123,8 +123,8 @@ class Graph {
 
   void Compute(cnrtInvokeFuncParam_t forward_param,
                cnrtQueue_t que,
-               std::vector<std::shared_ptr<MLUTensor>> in,
-               std::vector<std::shared_ptr<MLUTensor>> out) {
+               const std::vector<std::shared_ptr<MLUTensor>>& in,
+               const std::vector<std::shared_ptr<MLUTensor>>& out) {
     std::vector<cnmlTensor_t> in_tensor;
     std::vector<cnmlTensor_t> out_tensor;
     input_addrs_.resize(in.size());

lite/kernels/mlu/subgraph_compute.h

@@ -238,7 +238,8 @@ class SubgraphEngine : public subgraph::Engine {
 
     for (size_t i = 0; i < origin_itensors_.size(); ++i) {
       paddle::lite::subgraph::mlu::MLUTensor tmp(
-          graph_input->at(i)->get_origin_shape());
+          origin_itensors_[i]->dims().Vectorize());
+          // graph_input->at(i)->get_origin_shape());
       tmp.set_mlu_dtype(graph_input->at(i)->dtype());
       tmp.set_mlu_ptr(const_cast<void*>(origin_itensors_[i]->raw_data()));
       graph_in.push_back(
@@ -251,7 +252,8 @@ class SubgraphEngine : public subgraph::Engine {
               ->mutable_data<typename paddle::lite::subgraph::mlu::FPTypeTraits<
                   Precision>::T>(TARGET(kMLU)));
       paddle::lite::subgraph::mlu::MLUTensor tmp(
-          graph_output->at(i)->get_origin_shape());
+          origin_otensors_[i]->dims().Vectorize());
+          // graph_output->at(i)->get_origin_shape());
       tmp.set_mlu_dtype(graph_output->at(i)->dtype());
       tmp.set_mlu_ptr(p_data);
       graph_out.push_back(