Fix shape checking for scatterop in dynamic shape scene

mindspore/core/abstract/prim_arrays.cc

@@ -169,8 +169,8 @@ AbstractBasePtr InferImplUnique(const AnalysisEnginePtr &, const PrimitivePtr &p
     std::make_shared<AbstractTensor>(input->element(), std::make_shared<Shape>(ids_shape, min_shape, max_shape));
   // Currently we choose the same data type as input for the idx.
   TypePtr ids_idx_type = kInt32;
-  if (input->element() != nullptr && input->element()->GetTypeTrack() != nullptr) {
-    ids_idx_type = input->element()->GetTypeTrack();
+  if (input->element() != nullptr && input->element()->GetTypeTrack() == kInt64) {
+    ids_idx_type = kInt64;
   }
   auto ids_idx = std::make_shared<AbstractTensor>(ids_idx_type, shape->shape());
   // outputs: ids, ids_idx

mindspore/ops/operations/array_ops.py

@@ -47,7 +47,7 @@ class _ScatterOp(PrimitiveWithInfer):
     )
 
     def _check_scatter_shape(self, x_shape, indices_shape, updates_shape, prim_name):
-        if updates_shape and updates_shape != indices_shape + x_shape[1:]:
+        if indices_shape != [-1] and updates_shape and updates_shape != indices_shape + x_shape[1:]:
             raise ValueError(f"For '{prim_name}', "
                              f"updates_shape = indices_shape + x_shape[1:], but got x_shape: {x_shape}, "
                              f"indices_shape: {indices_shape}, updates_shape: {updates_shape}.")