Revert elementwise add (#53745)

* modify concat_grad add sum comp rule

* delete default mul_double_grad

* delete high grad test

* recover yaml

* modify yaml

* recover add_double_grad prim

paddle/fluid/eager/auto_code_generator/generator/eager_gen.py

@@ -67,7 +67,6 @@ black_ops_list = [
 prim_white_list = [
     "matmul_double_grad",
     "tanh_double_grad",
-    "add_double_grad",
     "subtract_double_grad",
 ]
 

paddle/phi/api/yaml/legacy_backward.yaml

@@ -30,6 +30,7 @@
   kernel :
     func : add_double_grad
   optional : grad_x_grad, grad_y_grad
+  backward : add_triple_grad
   inplace : (grad_x_grad -> grad_out_grad)
   composite : add_double_grad(y, grad_out, grad_x_grad, grad_y_grad, axis, grad_out_grad)
 
@@ -47,6 +48,17 @@
   backward : add_double_grad
   inplace : (out_grad -> x_grad)
 
+- backward_op : add_triple_grad
+  forward : add_double_grad (Tensor y, Tensor grad_out, Tensor grad_grad_x, Tensor grad_grad_y, int axis = -1) -> Tensor(grad_grad_out)
+  args : (Tensor grad_grad_x, Tensor grad_grad_y, Tensor grad_grad_out_grad, int axis = -1)
+  output : Tensor(grad_grad_x_grad), Tensor(grad_grad_y_grad)
+  infer_meta :
+    func : GeneralBinaryGradInferMeta
+    param : [grad_grad_x, grad_grad_y]
+  kernel :
+    func : add_triple_grad
+  inplace : (grad_grad_out_grad -> grad_grad_x_grad)
+
 - backward_op : amax_grad
   forward: amax (Tensor x,  int64_t[] axis={},  bool keepdim=false) -> Tensor(out)
   args : (Tensor x, Tensor out, Tensor out_grad, int64_t[] axis={},  bool keepdim=false, bool reduce_all=false)

test/prim/prim/vjp/test_comp_high_grad.py

@@ -25,7 +25,7 @@ import paddle
 from paddle import fluid
 from paddle.fluid import core
 
-
+'''
 @param.parameterized_class(
     ('shape1', 'shape2'),
     [
@@ -120,6 +120,7 @@ class TestAddHighGradCheck(unittest.TestCase):
         for p in places:
             self.func_double(p)
             self.func_triple(p)
+'''
 
 
 @param.parameterized_class(