reduce pass traversals if no pre-ad python pass exists

mindspore/ccsrc/frontend/optimizer/pass_group.h

@@ -49,6 +49,7 @@ class PassGroup {
   bool Run(const FuncGraphPtr &func_graph, const std::vector<PythonPassPtr> &passes, const MatchResultPtr &res) const;
   std::string name() const { return name_; }
   void SetRunOnlyOnce(bool run_only_once) { run_only_once_ = run_only_once; }
+  size_t size() { return passes_.size(); }
 
  private:
   const std::string name_;

mindspore/ccsrc/pipeline/jit/action.cc

@@ -301,7 +301,12 @@ bool OptimizeAction(const ResourcePtr &res, const std::vector<PassItem> &passes)
   return true;
 }
 
-bool OptInlineAction(const ResourcePtr &res) { return OptimizeAction(res, kInlinePasses); }
+bool OptInlineAction(const ResourcePtr &res) {
+  if (opt::python_pass::PyPassManager::GetInstance()->GetPassGroup(opt::python_pass::Phase::PREAD)->size() != 0) {
+    return OptimizeAction(res, kInlinePasses);
+  }
+  return true;
+}
 
 bool GeOptimizeAction(const ResourcePtr &res) { return OptimizeAction(res, kGePasses); }
 

mindspore/graph_utils/python_pass/__init__.py

@@ -13,14 +13,14 @@
 # limitations under the License.
 # ============================================================================
 """Reference for python pass registration."""
-from .python_pass_register import registe_pass, unregiste_pass, gen_new_parameter, cancel_new_parameter, _set_renorm,\
-    _set_reopt
+from .python_pass_register import registe_pass, unregiste_pass, gen_new_parameter, cancel_new_parameter, set_renorm,\
+    set_reopt
 
 __all__ = [
     "registe_pass",
     "unregiste_pass",
     "gen_new_parameter",
     "cancel_new_parameter",
-    "_set_renorm",
-    "_set_reopt"
+    "set_renorm",
+    "set_reopt"
     ]

mindspore/graph_utils/python_pass/python_pass_register.py

@@ -23,8 +23,8 @@ __all__ = [
     "unregiste_pass",
     "gen_new_parameter",
     "cancel_new_parameter",
-    "_set_renorm",
-    "_set_reopt"
+    "set_renorm",
+    "set_reopt"
 ]
 class PyPassManager(PyPassManager_):
     r"""
@@ -162,7 +162,7 @@ def cancel_new_parameter(pattern):
     ppm = PyPassManager()
     ppm.unregiste(pattern.para_name)
 
-def _set_renorm(should_renorm):
+def set_renorm(should_renorm):
     """
     Set whether or not to do renormalization after modified graph in python pass(es).
 
@@ -176,7 +176,7 @@ def _set_renorm(should_renorm):
     ppm = PyPassManager()
     ppm.set_renorm(should_renorm)
 
-def _set_reopt(do_reopt):
+def set_reopt(do_reopt):
     """
     Set whether or not to do optimization after modified graph in python pass(es).
 

tests/ut/python/optimizer/test_python_pass.py

@@ -19,8 +19,8 @@ import mindspore.nn as nn
 from mindspore import context
 from mindspore.common.tensor import Tensor
 from mindspore.ops import operations as P
-from mindspore.graph_utils.python_pass import registe_pass, unregiste_pass, _set_renorm, gen_new_parameter,\
-    cancel_new_parameter, _set_reopt
+from mindspore.graph_utils.python_pass import registe_pass, unregiste_pass, set_renorm, gen_new_parameter,\
+    cancel_new_parameter, set_reopt
 from mindspore.common.api import _generate_pip_args
 from mindspore._c_expression import generate_key, Executor_
 from mindspore.graph_utils.graph_pattern import OneOf, Prim, Call, NoneOf, Any, NewTensor, NewParameter, Imm
@@ -157,8 +157,8 @@ def test_isnot_pattern_0():
     Test IsNot pattern which expresses the IsNot semantics.
     Case: IsNot pass failed to match
     """
-    _set_renorm(False)
-    _set_reopt(False)
+    set_renorm(False)
+    set_reopt(False)
     class ConvBN(nn.Cell):
         def __init__(self):
             super(ConvBN, self).__init__()
@@ -202,7 +202,7 @@ def test_isnot_pattern_0():
     unregiste_pass(bn_pass)
     assert "ReLU6" not in transformed_repr
     assert "Softmax" in transformed_repr
-    _set_renorm(True)
+    set_renorm(True)
 
 def test_isnot_pattern_1():
     """
@@ -234,8 +234,8 @@ def test_newtensor_pattern():
     """
     Test NewTensor pattern in the target
     """
-    _set_renorm(False)
-    _set_reopt(False)
+    set_renorm(False)
+    set_reopt(False)
     inputs = Tensor(np.ones([42]), mindspore.float16)
     softmax_model = nn.Softmax()
 
@@ -252,7 +252,7 @@ def test_newtensor_pattern():
     unregiste_pass(softmax_addn_pass)
     assert "AddN" in transformed_repr
     assert "Softmax" not in transformed_repr
-    _set_renorm(True)
+    set_renorm(True)
 
 def test_newparameter_pattern():
     """
@@ -261,8 +261,8 @@ def test_newparameter_pattern():
     inputs = Tensor(np.ones([42]), mindspore.float16)
     softmax_model = nn.Softmax()
 
-    _set_renorm(False)
-    _set_reopt(False)
+    set_renorm(False)
+    set_reopt(False)
     @registe_pass(requires_grad=False, run_only_once=True)
     def softmax_addn_pass():
         x = Any()
@@ -288,8 +288,8 @@ def test_imm_target():
     inputs = Tensor(np.ones([42]), mindspore.float16)
     softmax_model = nn.Softmax()
 
-    _set_renorm(False)
-    _set_reopt(False)
+    set_renorm(False)
+    set_reopt(False)
     @registe_pass(requires_grad=False, run_only_once=True)
     def softmax_pass():
         x = Any()
@@ -313,8 +313,8 @@ def test_gen_new_parameter():
 
     default_tensor = Tensor(np.ones((4, 4)), mindspore.float32)
     new_para = NewParameter("Merlin", default_tensor)
-    _set_renorm(False)
-    _set_reopt(False)
+    set_renorm(False)
+    set_reopt(False)
     gen_new_parameter(new_para)
     @registe_pass(requires_grad=False, run_only_once=True)
     def softmax_make_tuple_pass():