Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into…

… support_prim_in_new_ir
PaddlePaddle · Aug 17, 2023 · 95642e3 · 95642e3
2 parents c161003 + dcfe2f1
commit 95642e3
Show file tree

Hide file tree

Showing 22 changed files with 295 additions and 351 deletions.
diff --git a/paddle/fluid/framework/new_executor/interpreter/static_build.cc b/paddle/fluid/framework/new_executor/interpreter/static_build.cc
@@ -474,6 +474,13 @@ void FakeInitializeOutputsForFunctionKernel(
                       ? DataType::INT64
                       : in_dtype;
             }
+          } else if (op_type == "searchsorted") {
+            bool out_int32 = op.Attr<bool>("out_int32");
+            if (out_int32) {
+              dtype = DataType::INT32;
+            } else {
+              dtype = DataType::INT64;
+            }
           } else {
             VLOG(4) << "Get dtype result from InferMeta";
             RuntimeInferShapeContext infer_shape_ctx(op, runtime_ctx);

diff --git a/paddle/fluid/framework/new_executor/program_interpreter.cc b/paddle/fluid/framework/new_executor/program_interpreter.cc
@@ -47,7 +47,6 @@ ProgramInterpreter::ProgramInterpreter(const platform::Place& place,
 
   static_build_ = FLAGS_new_executor_static_build &&
                   !FLAGS_new_executor_use_cuda_graph &&
-                  !execution_config.used_for_control_flow_op &&
                   interpreter::BlockCanBeStaticBuilt(block);
 
   exception_notifier_ = main_thread_blocker_.RegisterEvent(kExceptionCaught);

diff --git a/paddle/fluid/operators/collective/c_allreduce_op.h b/paddle/fluid/operators/collective/c_allreduce_op.h
@@ -265,11 +265,6 @@ class CAllReduceOpCUDAKernel : public framework::OpKernel<T> {
     if (map->has(rid)) {
       // Use ProcessGroup
       distributed::ProcessGroup* pg = map->get(rid);
-      std::vector<phi::DenseTensor> in_tensor;
-      std::vector<phi::DenseTensor> out_tensor;
-      in_tensor.push_back(*in);
-      out_tensor.push_back(*out);
-
       distributed::AllreduceOptions opts;
       switch (red_type) {
         case kRedSum:
@@ -293,7 +288,7 @@ class CAllReduceOpCUDAKernel : public framework::OpKernel<T> {
               "Invalid reduce type: %d", red_type));
       }
 
-      auto task = pg->AllReduce(in_tensor, out_tensor, opts);
+      auto task = pg->AllReduce(out, *in, opts, false, true);
       task->Wait();
       return;
     }

diff --git a/paddle/fluid/prim/utils/utils.cc b/paddle/fluid/prim/utils/utils.cc
@@ -55,6 +55,13 @@ void PrimCommonUtils::AddSkipCompOps(const std::string& op_type) {
   StaticCompositeContext::Instance().AddSkipCompOps(op_type);
 }
 
+void PrimCommonUtils::SetPrimBackwardBlacklist(
+    const std::unordered_set<std::string>& op_types) {
+  for (const auto& item : op_types) {
+    StaticCompositeContext::Instance().AddSkipCompOps(item);
+  }
+}
+
 void PrimCommonUtils::RemoveSkipCompOps(const std::string& op_type) {
   StaticCompositeContext::Instance().RemoveSkipCompOps(op_type);
 }

diff --git a/paddle/fluid/prim/utils/utils.h b/paddle/fluid/prim/utils/utils.h
@@ -30,6 +30,8 @@ class PrimCommonUtils {
   static void SetAllPrimEnabled(bool enabled);
   static size_t CheckSkipCompOps(const std::string& op_type);
   static void AddSkipCompOps(const std::string& op_type);
+  static void SetPrimBackwardBlacklist(
+      const std::unordered_set<std::string>& op_types);
   static void RemoveSkipCompOps(const std::string& op_type);
   static void SetTargetGradName(const std::map<std::string, std::string>& m);
 };