Fix test calc gradient (#37672)

* add scope_guard

* 1. fix control flow cases 2. fix calc_gradient

paddle/fluid/framework/new_executor/interpretercore_util.cc

@@ -16,6 +16,9 @@
 
 #include "paddle/fluid/framework/executor_gc_helper.h"
 #include "paddle/fluid/framework/new_executor/data_transfer.h"
+#include "paddle/fluid/operators/controlflow/conditional_block_op_helper.h"
+#include "paddle/fluid/operators/controlflow/recurrent_op_helper.h"
+#include "paddle/fluid/operators/controlflow/while_op_helper.h"
 
 namespace paddle {
 namespace framework {
@@ -127,6 +130,9 @@ void build_variable_scope(const framework::BlockDesc& block,
 
   for (auto& var_desc : block.AllVars()) {
     auto var_name = var_desc->Name();
+    // TODO(xiongkun): user may create a variable with name that exists before.
+    // under such circumstances, we should raise a error. Currently we can't
+    // get the var_desc of startup_program, so leave it later.
     if (var_name == framework::kEmptyVarName) {
       continue;
     }
@@ -149,7 +155,7 @@ void build_variable_scope(const framework::BlockDesc& block,
 }
 
 void create_all_ops(const framework::BlockDesc& block,
-                    std::vector<std::shared_ptr<OperatorBase>>* ops) {
+                    std::vector<std::unique_ptr<OperatorBase>>* ops) {
   for (auto& op : block.AllOps()) {
     VLOG(3) << "CreateOp from : " << op->Type();
 
@@ -164,7 +170,7 @@ void create_all_ops(const framework::BlockDesc& block,
     }
     auto op_base =
         info.Creator()(op->Type(), inputs_names, outputs_names, op_attr_map);
-    ops->emplace_back(std::shared_ptr<OperatorBase>(op_base));
+    ops->emplace_back(std::unique_ptr<OperatorBase>(op_base));
   }
 }
 
@@ -260,10 +266,24 @@ void build_op_func_list(const platform::Place& place,
   Scope* local_scope = use_local_scope ? var_scope->GetMutableLocalScope()
                                        : var_scope->GetMutableScope();
   auto& all_op_kernels = OperatorWithKernel::AllOpKernels();
+  std::vector<std::unique_ptr<OperatorBase>>
+      ops_unique;  // its elements will be moved to vec_func_list
+  // Step 1: create all ops for current block.
+  create_all_ops(block, &ops_unique);
+  // If gc is enabled and block size > 1
+  const ProgramDesc& main_program = *block.Program();
+  operators::PrepareSafeEagerDeletionOnConditionalOpAndConditionalGradOp(
+      main_program, block.ID(), ops_unique);
+  operators::PrepareSafeEagerDeletionOnWhileOpAndWhileGradOp(
+      main_program, block.ID(), ops_unique);
+  operators::PrepareSafeEagerDeletionOnRecurrentOpAndRecurrentGradOp(
+      main_program, block.ID(), ops_unique);
+
   std::vector<std::shared_ptr<OperatorBase>>
       ops;  // its elements will be moved to vec_func_list
-  // Step 1: create all ops for current block.
-  create_all_ops(block, &ops);
+  for (auto& op_unique : ops_unique) {
+    ops.emplace_back(std::move(op_unique));
+  }
   auto unused_var_map = get_unused_vars(block, ops);
 
   for (size_t i = 0; i < ops.size(); ++i) {

paddle/fluid/framework/new_executor/standalone_executor.cc

@@ -33,6 +33,7 @@ StandaloneExecutor::StandaloneExecutor(const platform::Place& place,
   if (scope) {
     auto name_list = scope->LocalVarNames();
     for (auto name : name_list) {
+      VLOG(4) << "Sync Variable from variable scope: " << name;
       auto v = scope->Var(name);
       if (!global_scope_.HasVar(name)) {
         global_scope_.AddVar(name, *v);
@@ -87,8 +88,9 @@ void StandaloneExecutor::BuildVariableScope(const framework::ProgramDesc& pdesc,
     if (var->Name() == framework::kEmptyVarName) {
       continue;
     }
-
     if (!var_scope->HasVar(var->Name())) {
+      VLOG(4) << "Create variable from startup_prog: "
+              << var->Proto()->SerializeAsString();
       var_scope->AddVar(var->Name(), var);
     }
   }

python/paddle/fluid/tests/unittests/test_calc_gradient.py

@@ -14,6 +14,7 @@
 
 from __future__ import print_function
 
+import paddle
 import unittest
 import numpy as np
 import paddle.fluid as fluid
@@ -83,6 +84,7 @@ class TestDoubleGrad(unittest.TestCase):
 
 class TestGradientWithPrune(unittest.TestCase):
     def test_prune(self):
+        with paddle.fluid.scope_guard(paddle.static.Scope()):
             x = fluid.data(name='x', shape=[3], dtype='float32')
             x.stop_gradient = False
             x1, x2, x3 = fluid.layers.split(x, dim=0, num_or_sections=3)