!5517 gpu kernel_info_setter code review

Merge pull request !5517 from limingqi107/master

mindspore/ccsrc/backend/session/gpu_session.cc

@@ -49,10 +49,10 @@ using AnfAlgo = mindspore::session::AnfRuntimeAlgorithm;
 
 void GPUSession::SelectKernel(const std::shared_ptr<KernelGraph> &kernel_graph) const {
   MS_EXCEPTION_IF_NULL(kernel_graph);
-  bool in_black_list = CheckInModeBlackList(kernel_graph);
+  bool graph_format_transform = IsSupportFormatTransform(kernel_graph);
   for (const auto &kernel_node : kernel_graph->execution_order()) {
     MS_EXCEPTION_IF_NULL(kernel_node);
-    device::gpu::SetKernelInfo(kernel_node, in_black_list);
+    device::gpu::SetKernelInfo(kernel_node, graph_format_transform);
   }
 }
 
@@ -76,7 +76,7 @@ void GPUSession::Optimize(const std::shared_ptr<KernelGraph> &kernel_graph) {
   pm->AddPass(std::make_shared<opt::ReplaceBNGradCastFusion>());
   pm->AddPass(std::make_shared<opt::ReplaceMomentumCastFusion>());
   pm->AddPass(std::make_shared<opt::ReplaceAddNFusion>());
-  if (!CheckInModeBlackList(kernel_graph) && context_ptr->get_param<int>(MS_CTX_EXECUTION_MODE) != kPynativeMode) {
+  if (IsSupportFormatTransform(kernel_graph) && context_ptr->get_param<int>(MS_CTX_EXECUTION_MODE) != kPynativeMode) {
     pm->AddPass(std::make_shared<opt::BatchNormReluFusion>());
     pm->AddPass(std::make_shared<opt::BatchNormReluGradFusion>());
     pm->AddPass(std::make_shared<opt::BatchNormAddReluFusion>());
@@ -193,14 +193,14 @@ void GPUSession::Execute(const std::shared_ptr<KernelGraph> &kernel_graph) const
   }
 }
 
-bool GPUSession::CheckInModeBlackList(const std::shared_ptr<KernelGraph> &kernel_graph) const {
+bool GPUSession::IsSupportFormatTransform(const std::shared_ptr<KernelGraph> &kernel_graph) const {
   auto kernels = kernel_graph->execution_order();
   size_t conv_cnt = 0;
   size_t bn_cnt = 0;
   for (const auto &kernel : kernels) {
     auto kernel_name = AnfAlgo::GetCNodeName(kernel);
     if (kernel_name == prim::kPrimLayerNorm->name()) {
-      return true;
+      return false;
     }
     if (kernel_name == prim::kPrimConv2D->name()) {
       conv_cnt++;
@@ -210,9 +210,9 @@ bool GPUSession::CheckInModeBlackList(const std::shared_ptr<KernelGraph> &kernel
     }
   }
   if (conv_cnt == kConv2dCount && bn_cnt == kFusedBatchNormCount) {
-    return true;
+    return false;
   }
-  return false;
+  return true;
 }
 
 GraphId GPUSession::CompileGraph(const AnfNodePtrList &lst, const AnfNodePtrList &outputs) {

mindspore/ccsrc/backend/session/gpu_session.h

@@ -67,7 +67,7 @@ class GPUSession : public SessionBasic {
 
   void Execute(const std::shared_ptr<KernelGraph> &kernel_graph) const;
 
-  bool CheckInModeBlackList(const std::shared_ptr<KernelGraph> &kernel_graph) const;
+  bool IsSupportFormatTransform(const std::shared_ptr<KernelGraph> &kernel_graph) const;
 
 #ifdef ENABLE_DEBUGGER
   void Dump(const std::shared_ptr<KernelGraph> &kernel_graph) const;

mindspore/ccsrc/runtime/device/gpu/gpu_kernel_runtime.cc

@@ -404,7 +404,9 @@ void GPUKernelRuntime::ClearGraphRuntimeResource(uint32_t graph_id, const std::v
   // Release the kernel resource.
   for (const auto &kernel : execution_order) {
     auto kernel_mod = AnfAlgo::GetKernelMod(kernel);
-    MS_EXCEPTION_IF_NULL(kernel_mod);
+    if (kernel_mod == nullptr) {
+      continue;
+    }
     kernel_mod->ReleaseResource();
   }
 }

mindspore/ccsrc/runtime/device/gpu/kernel_info_setter.cc

@@ -176,9 +176,18 @@ bool IsNeedProcessFormatInfo(const CNodePtr &kernel_node, const std::vector<Type
   if (inputs_type.size() == 0) {
     return false;
   }
-  auto input_shape = AnfAlgo::GetPrevNodeOutputInferShape(kernel_node, 0);
-  if (input_shape.size() != 4) {
-    return false;
+  auto inputs_format_position = iter->second.first;
+  // If input position is empty, then insert all the input positions, because the input numbers of this op are variable.
+  if (inputs_format_position.size() == 0) {
+    for (size_t input_index = 0; input_index < AnfAlgo::GetInputTensorNum(kernel_node); input_index++) {
+      inputs_format_position.push_back(input_index);
+    }
+  }
+  for (const auto &input_format_position : inputs_format_position) {
+    auto input_shape = AnfAlgo::GetPrevNodeOutputInferShape(kernel_node, input_format_position);
+    if (input_shape.size() != 4) {
+      return false;
+    }
   }
   return true;
 }
@@ -223,7 +232,7 @@ void UpdateKernelFormatInfo(const CNodePtr &kernel_node, const std::vector<TypeI
 }
 }  // namespace
 
-void SetKernelInfo(const CNodePtr &kernel_node, bool in_black_list) {
+void SetKernelInfo(const CNodePtr &kernel_node, bool graph_format_transform) {
   std::vector<std::string> inputs_format;
   std::vector<TypeId> inputs_type;
   for (size_t input_index = 0; input_index < AnfAlgo::GetInputTensorNum(kernel_node); ++input_index) {
@@ -237,7 +246,7 @@ void SetKernelInfo(const CNodePtr &kernel_node, bool in_black_list) {
     outputs_type.push_back(AnfAlgo::GetOutputInferDataType(kernel_node, output_index));
   }
   std::string origin_data_format = kOpFormat_DEFAULT;
-  if (!in_black_list && IsNeedProcessFormatInfo(kernel_node, inputs_type)) {
+  if (graph_format_transform && IsNeedProcessFormatInfo(kernel_node, inputs_type)) {
     UpdateKernelFormatInfo(kernel_node, inputs_type, &inputs_format, &outputs_format, &origin_data_format);
   }
   std::shared_ptr<KernelBuildInfo::KernelBuildInfoBuilder> builder =

mindspore/ccsrc/runtime/device/gpu/kernel_info_setter.h

@@ -53,7 +53,7 @@ static std::map<std::string, std::pair<std::vector<size_t>, std::vector<size_t>>
   {prim::kPrimAddN->name(), {{}, {0}}},
 };
 
-void SetKernelInfo(const CNodePtr &kernel_node, bool in_black_list = false);
+void SetKernelInfo(const CNodePtr &kernel_node, bool graph_format_transform = false);
 
 class KernelAttr {
  public: