[MSLITE][Develop] return RET_INFET_INVALID when infer_flag is false for op infershape

mindspore/lite/src/ops/addn.cc

@@ -86,7 +86,7 @@ int AddN::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> outputs
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output->set_shape(input->shape());
 

mindspore/lite/src/ops/argmax.cc

@@ -74,7 +74,7 @@ int ArgMax::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int> output_shape(input->shape());
   auto input_shape_size = input->shape().size();

mindspore/lite/src/ops/argmin.cc

@@ -72,7 +72,7 @@ int ArgMin::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite::Te
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape_size = input->shape().size();
   auto axis = GetAxis() < 0 ? GetAxis() + input_shape_size : GetAxis();

mindspore/lite/src/ops/arithmetic.cc

@@ -45,7 +45,7 @@ int Arithmetic::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite
   output->set_format(format);
   output->set_data_type(input0->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (input_shape0.size() > 10 || input_shape1.size() > 10) {
     int wrong_dim = input_shape0.size() > input_shape1.size() ? input_shape0.size() : input_shape1.size();

mindspore/lite/src/ops/arithmetic_self.cc

@@ -33,7 +33,7 @@ int ArithmeticSelf::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output->set_shape(input->shape());
   return RET_OK;

mindspore/lite/src/ops/audio_spectrogram.cc

@@ -77,7 +77,7 @@ int AudioSpectrogram::InferShape(std::vector<Tensor *> inputs_, std::vector<Tens
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != 2) {

mindspore/lite/src/ops/batch_to_space.cc

@@ -98,7 +98,7 @@ int BatchToSpace::InferShape(std::vector<lite::Tensor *> inputs, std::vector<lit
   outputs[0]->set_format(input->format());
   outputs[0]->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != kDimension_4d) {

mindspore/lite/src/ops/broadcast_to.cc

@@ -80,7 +80,7 @@ int BroadcastTo::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *>
   outputs[0]->set_format(input->format());
   outputs[0]->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int32_t> dst_shape(GetDstShape());
   for (size_t i = 0; i < dst_shape.size(); ++i) {

mindspore/lite/src/ops/cast.cc

@@ -93,7 +93,7 @@ int Cast::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
 
   output->set_data_type(static_cast<TypeId>(GetDstT()));
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   if (GetSrcT() != 0 && input->data_type() != GetSrcT()) {

mindspore/lite/src/ops/concat.cc

@@ -98,7 +98,7 @@ int Concat::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output->set_data_type(input0->data_type());
   output->set_format(input0->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto input0_shape = inputs_.at(0)->shape();

mindspore/lite/src/ops/constant_of_shape.cc

@@ -83,7 +83,7 @@ int ConstantOfShape::InferShape(std::vector<Tensor *> inputs_, std::vector<Tenso
   out_tensor->set_data_type(static_cast<TypeId>(GetDataType()));
   out_tensor->set_format(in_tensor->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto in_data = reinterpret_cast<int *>(in_tensor->data_c());
   if (in_data == nullptr) {

mindspore/lite/src/ops/conv2d.cc

@@ -391,7 +391,7 @@ int Conv2D::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   pad_r_ = GetPadRight();
 
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto in_shape = input_tensor->shape();
   int input_h = in_shape.at(1);

mindspore/lite/src/ops/crop.cc

@@ -71,7 +71,7 @@ int Crop::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> outputs
   outputs[0]->set_format(inputs[0]->format());
   outputs[0]->set_data_type(inputs[0]->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   outputs[0]->set_shape(inputs[1]->shape());
   return RET_OK;

mindspore/lite/src/ops/deconv2d.cc

@@ -317,7 +317,7 @@ int DeConv2D::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite::
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   int32_t input_h = input->Height();
   int32_t input_w = input->Width();

mindspore/lite/src/ops/dedepthwise_conv2d.cc

@@ -138,7 +138,7 @@ int DeDepthwiseConv2D::InferShape(std::vector<lite::Tensor *> inputs_, std::vect
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto in_shape = input->shape();
   int input_h = in_shape.at(1);

mindspore/lite/src/ops/depth_to_space.cc

@@ -73,7 +73,7 @@ int DepthToSpace::InferShape(std::vector<lite::Tensor *> inputs, std::vector<lit
   outputs[0]->set_data_type(input->data_type());
   outputs[0]->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != kDimension_4d) {

mindspore/lite/src/ops/depthwise_conv2d.cc

@@ -219,7 +219,7 @@ int DepthwiseConv2D::InferShape(std::vector<lite::Tensor *> inputs_, std::vector
   pad_r_ = GetPadRight();
 
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto in_shape = input->shape();
   int input_h = in_shape.at(1);

mindspore/lite/src/ops/detection_post_process.cc

@@ -190,7 +190,7 @@ int DetectionPostProcess::InferShape(std::vector<lite::Tensor *> inputs_, std::v
   num_det->set_format(boxes->format());
   num_det->set_data_type(kNumberTypeFloat32);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   const auto max_detections = GetMaxDetections();
   const auto max_classes_per_detection = GetMaxClassesPerDetection();

mindspore/lite/src/ops/dropout.cc

@@ -84,7 +84,7 @@ int Dropout::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   auto output0 = outputs_.front();
   MS_ASSERT(output0 != nullptr);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output0->set_shape(input->shape());
   output0->set_data_type(input->data_type());

mindspore/lite/src/ops/dropout_grad.cc

@@ -87,7 +87,7 @@ int DropoutGrad::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *>
   auto output = outputs_.front();
   MS_ASSERT(output != nullptr);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output->set_shape(input->shape());
   output->set_data_type(input->data_type());

mindspore/lite/src/ops/embedding_lookup.cc

@@ -70,7 +70,7 @@ int EmbeddingLookup::InferShape(std::vector<Tensor *> inputs_, std::vector<Tenso
   output->set_format(params_->format());
   output->set_data_type(params_->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto embedding_shape = params_->shape();

mindspore/lite/src/ops/expand_dims.cc

@@ -103,7 +103,7 @@ int ExpandDims::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *>
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   int dim = GetDim();
   if (dim < 0) {

mindspore/lite/src/ops/fft_imag.cc

@@ -43,7 +43,7 @@ int FftImag::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   output->set_data_type(TypeId::kNumberTypeFloat32);
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   input_shape.pop_back();

mindspore/lite/src/ops/fft_real.cc

@@ -43,7 +43,7 @@ int FftReal::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   output->set_data_type(TypeId::kNumberTypeFloat32);
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   input_shape.pop_back();

mindspore/lite/src/ops/fill.cc

@@ -71,7 +71,7 @@ int Fill::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> output_shape;

mindspore/lite/src/ops/flatten.cc

@@ -40,7 +40,7 @@ int Flatten::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto input_shape = input->shape();

mindspore/lite/src/ops/flatten_grad.cc

@@ -39,7 +39,7 @@ int FlattenGrad::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *>
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto input_shape = input->shape();

mindspore/lite/src/ops/full_connection.cc

@@ -70,7 +70,7 @@ int FullConnection::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<
   auto output = outputs_.front();
   MS_ASSERT(output != nullptr);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if ((GetHasBias() && inputs_.size() != kMultiNum) || (!GetHasBias() && inputs_.size() != kDoubleNum)) {
     MS_LOG(ERROR) << "Input tensors num error";

mindspore/lite/src/ops/gather.cc

@@ -114,7 +114,7 @@ int Gather::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   int axis = GetAxis();

mindspore/lite/src/ops/gather_nd.cc

@@ -66,7 +66,7 @@ int GatherNd::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> ou
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto in_shape = input->shape();
   int in_rank = in_shape.size();

mindspore/lite/src/ops/layer_norm.cc

@@ -103,7 +103,7 @@ int LayerNorm::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite:
     }
   }
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   output->set_shape(input_shape);

mindspore/lite/src/ops/lstm.cc

@@ -68,7 +68,7 @@ int Lstm::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
     outputs_.at(i)->set_format(input->format());
   }
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> in_shape = input->shape();

mindspore/lite/src/ops/matmul.cc

@@ -101,7 +101,7 @@ int MatMul::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output->set_data_type(input0->data_type());
   output->set_format(input0->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> a_shape = input0->shape();

mindspore/lite/src/ops/maximum_grad.cc

@@ -93,7 +93,7 @@ int MaximumGrad::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *>
   MS_ASSERT(dx1 != nullptr);
   MS_ASSERT(dx2 != nullptr);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto inShape0 = x1->shape();

mindspore/lite/src/ops/mean.cc

@@ -77,7 +77,7 @@ int Mean::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (this->primitive_ == nullptr) {
     return RET_NULL_PTR;

mindspore/lite/src/ops/mfcc.cc

@@ -59,7 +59,7 @@ int Mfcc::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != 3) {

mindspore/lite/src/ops/nchw2nhwc.cc

@@ -48,7 +48,7 @@ int Nchw2Nhwc::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite:
   output->set_format(schema::Format::Format_NHWC);
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int> nchw_shape = input->shape();
   if (nchw_shape.size() != 4) {

mindspore/lite/src/ops/nhwc2nchw.cc

@@ -49,7 +49,7 @@ int Nhwc2Nchw::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite:
   output->set_format(schema::Format::Format_NCHW);
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int> nhwc_shape = input->shape();
   if (nhwc_shape.size() != 4) {

mindspore/lite/src/ops/one_hot.cc

@@ -117,7 +117,7 @@ int OneHot::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> outpu
   output->set_data_type(on_value->data_type());
   output->set_format(on_value->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   const auto input_shape = input->shape();
   int input_rank = static_cast<int>(input_shape.size());

mindspore/lite/src/ops/pad.cc

@@ -83,7 +83,7 @@ int Pad::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> outputs)
   output->set_format(input->format());
   output->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> paddings;

mindspore/lite/src/ops/pooling.cc

@@ -181,7 +181,7 @@ int Pooling::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   output->set_data_type(input->data_type());
   output->set_format(schema::Format::Format_NHWC);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   int input_h = input->shape().at(1);
   int input_w = input->shape().at(2);

mindspore/lite/src/ops/power.cc

@@ -115,7 +115,7 @@ int Power::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> output
   output_tensor->set_data_type(x_tensor->data_type());
   output_tensor->set_format(x_tensor->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (exp_tensor != nullptr) {
     if ((exp_tensor->shape().size() > 1 && exp_tensor->shape() != x_tensor->shape()) ||

mindspore/lite/src/ops/prior_box.cc

@@ -140,7 +140,7 @@ int PriorBox::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> ou
   output->set_data_type(kNumberTypeFloat32);
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<float> different_aspect_ratios{1.0f};
   auto aspect_ratios = GetAspectRatios();

mindspore/lite/src/ops/quant_dtype_cast.cc

@@ -63,7 +63,7 @@ int QuantDTypeCast::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor
   output->set_data_type(static_cast<TypeId>(GetDstT()));
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output->set_shape(input->shape());
   return RET_OK;

mindspore/lite/src/ops/range.cc

@@ -71,7 +71,7 @@ int Range::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outpu
   }
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   int shape_size = 0;

mindspore/lite/src/ops/rank.cc

@@ -45,7 +45,7 @@ int Rank::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int> in_shape(1, 1);
   output->set_shape(in_shape);

mindspore/lite/src/ops/reduce.cc

@@ -157,7 +157,7 @@ int Reduce::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (this->primitive_ == nullptr) {
     return RET_NULL_PTR;

mindspore/lite/src/ops/reshape.cc

@@ -173,7 +173,7 @@ int Reshape::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> out_shape;

mindspore/lite/src/ops/resize.cc

@@ -131,7 +131,7 @@ int Resize::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite::Te
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> output_shape;

mindspore/lite/src/ops/return.cc

@@ -72,7 +72,7 @@ int Return::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (this->primitive_ == nullptr) {
     return RET_NULL_PTR;

mindspore/lite/src/ops/reverse_sequence.cc

@@ -66,7 +66,7 @@ int ReverseSequence::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output->set_shape(input->shape());
   return RET_OK;

mindspore/lite/src/ops/rfft.cc

@@ -54,7 +54,7 @@ int Rfft::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output->set_data_type(TypeId::kNumberTypeComplex64);
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   input_shape[input_shape.size() - 1] = GetFftLength() / 2 + 1;

mindspore/lite/src/ops/roi_pooling.cc

@@ -79,7 +79,7 @@ int ROIPooling::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *>
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto new_h = GetPooledH();

mindspore/lite/src/ops/scatter_nd.cc

@@ -62,7 +62,7 @@ int ScatterND::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> o
   output->set_data_type(update->data_type());
   output->set_format(update->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto shape_data = reinterpret_cast<int *>(shape->MutableData());
   std::vector<int> out_shape(shape_data, shape_data + shape->ElementsNum());

mindspore/lite/src/ops/shape.cc

@@ -44,7 +44,7 @@ int Shape::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outpu
   out_tensor->set_data_type(kNumberTypeInt32);
   out_tensor->set_format(schema::Format::Format_NHWC);
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int> out_shape;
   out_shape.push_back(static_cast<int>(in_tensor->shape().size()));

mindspore/lite/src/ops/slice.cc

@@ -175,7 +175,7 @@ int Slice::InferShape(std::vector<lite::Tensor *> inputs, std::vector<lite::Tens
   outputs[0]->set_data_type(input->data_type());
   outputs[0]->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   std::vector<int32_t> slice_begin(GetBegin());

mindspore/lite/src/ops/softmax.cc

@@ -86,7 +86,7 @@ int SoftMax::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (input->shape().size() > 5) {
     MS_LOG(ERROR) << "Softmax input dim must be less than 5, get " << input->shape().size();

mindspore/lite/src/ops/space_to_batch.cc

@@ -97,7 +97,7 @@ int SpaceToBatch::InferShape(std::vector<lite::Tensor *> inputs, std::vector<lit
   outputs[0]->set_data_type(input->data_type());
   outputs[0]->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != kDimension_4d) {

mindspore/lite/src/ops/space_to_batch_nd.cc

@@ -99,7 +99,7 @@ int SpaceToBatchND::InferShape(std::vector<lite::Tensor *> inputs, std::vector<l
   outputs[0]->set_data_type(input->data_type());
   outputs[0]->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != kDimension_4d) {

mindspore/lite/src/ops/space_to_depth.cc

@@ -74,7 +74,7 @@ int SpaceToDepth::InferShape(std::vector<lite::Tensor *> inputs, std::vector<lit
   outputs[0]->set_format(input->format());
   outputs[0]->set_data_type(input->data_type());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
   if (input_shape.size() != kDimension_4d) {

mindspore/lite/src/ops/sparse_to_dense.cc

@@ -64,7 +64,7 @@ int SparseToDense::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor
   outputs_[0]->set_format(input2->format());
 
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   if (this->primitive_ == nullptr) {
     return RET_NULL_PTR;

mindspore/lite/src/ops/split.cc

@@ -123,7 +123,7 @@ int Split::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outpu
     outputs_[i]->set_format(input->format());
   }
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   size_t split_dim = GetSplitDim() == -1 ? input->shape().size() - 1 : GetSplitDim();
   std::vector<int> input_shape = input->shape();

mindspore/lite/src/ops/squeeze.cc

@@ -105,7 +105,7 @@ int Squeeze::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> out
   outputs_.front()->set_data_type(in_tensor->data_type());
   outputs_.front()->set_format(in_tensor->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto in_shape = in_tensor->shape();
   std::vector<int> out_shape;

mindspore/lite/src/ops/stack.cc

@@ -83,7 +83,7 @@ int Stack::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> output
   outputs[0]->set_data_type(input0_data_type);
   outputs[0]->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto input_shape = input->shape();
 

mindspore/lite/src/ops/tile.cc

@@ -135,7 +135,7 @@ int Tile::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   std::vector<int> out_shape;

mindspore/lite/src/ops/topk.cc

@@ -69,7 +69,7 @@ int TopK::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> output
   output1->set_data_type(kNumberTypeInt32);
   output1->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   auto out_shape = input->shape();
   out_shape[out_shape.size() - 1] = GetK();

mindspore/lite/src/ops/transpose.cc

@@ -122,7 +122,7 @@ int Transpose::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> o
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   MS_ASSERT(inputs_.size() == kSingleNum || inputs_.size() == kDoubleNum);
   MS_ASSERT(outputs_.size() == kSingleNum);

mindspore/lite/src/ops/unique.cc

@@ -65,7 +65,7 @@ int Unique::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outp
   output1->set_format(input->format());
   output0->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output0->set_shape(input->shape());
   output1->set_shape(input->shape());

mindspore/lite/src/ops/unsqueeze.cc

@@ -78,7 +78,7 @@ int Unsqueeze::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> o
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
 
   auto dims = GetAxis();

mindspore/lite/src/ops/unstack.cc

@@ -66,7 +66,7 @@ int Unstack::InferShape(std::vector<Tensor *> inputs, std::vector<Tensor *> outp
     out->set_format(input->format());
   }
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   std::vector<int> output_shape;
   for (size_t i = 0; i < input_shape.size(); ++i) {

mindspore/lite/src/ops/where.cc

@@ -81,7 +81,7 @@ int Where::InferShape(std::vector<Tensor *> inputs_, std::vector<Tensor *> outpu
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   int num = input0->ElementsNum();
   int num1 = input1->ElementsNum();

mindspore/lite/src/ops/zeros_like.cc

@@ -56,7 +56,7 @@ int ZerosLike::InferShape(std::vector<lite::Tensor *> inputs_, std::vector<lite:
   output->set_data_type(input->data_type());
   output->set_format(input->format());
   if (!infer_flag()) {
-    return RET_OK;
+    return RET_INFER_INVALID;
   }
   output->set_shape(input->shape());
   return RET_OK;