!48873 Handle the problem that the subgraph is eliminated, when the subgraph has side effects.

Merge pull request !48873 from Margaret_wangrui/side_effect_none

mindspore/ccsrc/pipeline/jit/action.cc

@@ -1,5 +1,5 @@
 /**
- * Copyright 2019-2022 Huawei Technologies Co., Ltd
+ * Copyright 2019-2023 Huawei Technologies Co., Ltd
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -507,6 +507,16 @@ bool HasIsolatedSideEffectNode(const FuncGraphPtr &func_graph) {
       }
     }
   } else {
+    // Process call function
+    if (isolated_node->isa<CNode>()) {
+      auto first_input = isolated_node->cast<CNodePtr>()->input(0);
+      if (IsValueNode<FuncGraph>(first_input)) {
+        auto func = GetValueNode<FuncGraphPtr>(first_input);
+        if (IsSideEffectCNode(func->output()) || HasIsolatedSideEffectNode(func)) {
+          return true;
+        }
+      }
+    }
     if (IsSideEffectCNode(isolated_node)) {
       MS_LOG(DEBUG) << "Single isolated side-effect node: " << isolated_node->DebugString();
       return true;

tests/st/auto_monad/test_auto_monad.py

@@ -21,6 +21,7 @@ import numpy as np
 import mindspore as ms
 import mindspore.ops.operations as P
 import mindspore.nn as nn
+import mindspore.common.dtype as mstype
 from mindspore.nn import Cell
 from mindspore.nn import ReLU, BatchNorm2d, Conv2d, ParameterUpdate
 from mindspore.nn import Momentum, SoftmaxCrossEntropyWithLogits
@@ -1890,3 +1891,33 @@ def test_print_in_constant_returned_func():
 
     patterns = {'x:\n(1, 2, 3, 4, 5)'}
     check_output(cap.output, patterns)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.platform_arm_ascend_training
+@pytest.mark.platform_x86_ascend_training
+@pytest.mark.env_onecard
+def test_return_none_with_side_effect():
+    """
+    Feature: Support None.
+    Description: Support None is the output of_function with side effect.
+    Expectation: No exception.
+    """
+    class Net(nn.Cell):
+        def __init__(self):
+            super().__init__()
+            self.param = Parameter(Tensor([5], dtype=mstype.int32), name='name_a')
+
+        def update_param(self, weight):  # pylint: disable=R1711
+            self.param = 2 * weight
+            return None
+
+        def construct(self, weight):
+            self.update_param(weight)
+            return self.param
+
+    net = Net()
+    input_x = Tensor([2], dtype=mstype.int32)
+    res = net(input_x)
+    assert res == 4