remove allreduce before c_allgather (#55143)

* remove allreduce before c_allgather

* update reshard insert_fill_constant_op func

* insert_fill_constant_op add shape arg

python/paddle/distributed/auto_parallel/static/reshard.py

@@ -568,7 +568,7 @@ class Inserter:
         return outs
 
     @staticmethod
-    def insert_fill_constant_op(block, idx, op_role):
+    def insert_fill_constant_op(block, idx, op_role, shape):
         """Insert fill constant op into block at the given index."""
         # to avoid name conflict with framework
         helper = LayerHelper('fill_constant@RESHARD', **locals())
@@ -591,7 +591,7 @@ class Inserter:
         attrs['dtype'] = out.dtype
         attrs['op_role'] = op_role
         paddle.utils.get_shape_tensor_inputs(
-            inputs=inputs, attrs=attrs, shape=[0], op_type='fill_constant'
+            inputs=inputs, attrs=attrs, shape=shape, op_type='fill_constant'
         )
         fillconstant_op = block._insert_op(
             idx,
@@ -611,38 +611,6 @@ class Inserter:
         group = new_process_group(ranks)
         idx_offset = 0
 
-        # instant process group before insert allgather op.
-        if not group.is_instantiate():
-            # insert fill_constant op
-            fill_constant_out = Inserter.insert_fill_constant_op(
-                block, idx, op_role
-            )
-            fill_constant_out.stop_gradient = True
-
-            # insert c_allreduce_sum op
-            allreduce_op = block._insert_op(
-                idx + 1,
-                type="c_allreduce_sum",
-                inputs={'X': [fill_constant_out]},
-                outputs={'Out': [fill_constant_out]},
-                attrs={
-                    'ring_id': 0,
-                    'use_calc_stream': True,
-                    'op_role': op_role,
-                },
-            )
-            allreduce_op._set_attr('op_namescope', "/auto_parallel/reshard")
-            # insert c_sync_calc_stream op
-            sync_calc_op = block._insert_op(
-                idx + 2,
-                type="c_sync_calc_stream",
-                inputs={'X': [fill_constant_out]},
-                outputs={'Out': [fill_constant_out]},
-                attrs={'op_role': op_role},
-            )
-            sync_calc_op._set_attr('op_namescope', "/auto_parallel/reshard")
-            idx_offset = 3
-
         # insert c_allgather op
         op_type = 'c_allgather'
         # to avoid name conflict with framework