fix(mgb/core): ensure all VarNodes would be handled in init_ready_event

GitOrigin-RevId: 0b6cb24021

src/core/impl/graph/seq_comp_node_opt_impl.cpp

@@ -303,9 +303,8 @@ void SeqCompNodeOptimizerImpl::init_ready_event(
         }
 
         opr->input_waiting_spec(std::move(waiting_spec));
-        auto&& usable_output = opr->usable_output();
-        for (size_t i = 0; i < usable_output.size(); ++ i) {
-            var2step[usable_output[i]] = {cur_step, i};
+        for (size_t i = 0; i < opr->output().size(); ++ i) {
+            var2step[opr->output(i)] = {cur_step, i};
         }
         cur_step ++;
     }

src/core/test/graph/misc.cpp

@@ -1086,6 +1086,55 @@ TEST(TestGraph, DynShapeDepCrossCN) {
 }
 
 namespace {
+MGB_DEFINE_OPR_CLASS(CustomCopy, cg::SingleCNOperatorNodeBase) // {
+    std::shared_ptr<DeviceTensorND> m_data;
+
+    void scn_do_execute() override {
+        using namespace std::literals;
+        std::this_thread::sleep_for(100ms);
+        m_data->copy_from(input(0)->dev_tensor());
+    }
+
+    void init_output_static_infer_desc() override {
+        using namespace cg::static_infer;
+        owner_graph()->static_infer_manager().register_shape_infer(
+                output(0), ShapeInferDesc::make_const({}));
+    }
+
+public:
+    CustomCopy(VarNode* x, std::shared_ptr<DeviceTensorND> dv)
+            : Super{x->owner_graph(), {dv->comp_node()}, "d2h", {x}},
+                m_data(dv) {
+        add_input({x});
+        using F = VarNode::Flag;
+        add_output(None)
+                ->add_flag(F::ALLOW_EMPTY_SHAPE)
+                .add_flag(F::VOLATILE_CONTENT);
+    }
+};
+MGB_DYN_TYPE_OBJ_FINAL_IMPL(CustomCopy);
+}
+
+TEST(TestGraph, DependentOnVolatileContent) {
+    HostTensorGenerator<> gen;
+    auto cn0 = CompNode::load("xpu0"),
+         cn1 = cn0.change_stream(1);
+    auto host_x = gen({233}, cn0);
+    auto dev_y = std::make_shared<DeviceTensorND>(cn1);
+
+    auto graph = ComputingGraph::make();
+    auto x = opr::SharedDeviceTensor::make(*graph, *host_x),
+         y = x.insert_single_output_opr<CustomCopy>(x.node(), dev_y),
+         x_new = opr::AddUpdate::make(x, x.make_scalar(1));
+
+    auto func = graph->compile({{y, {}}, {x_new, {}}});
+    func->execute().wait();
+    HostTensorND host_y;
+    host_y.copy_from(*dev_y).sync();
+    MGB_ASSERT_TENSOR_EQ(*host_x, host_y);
+}
+
+namespace {
 void check_wait(SymbolVar dest, SymbolVar dep) {
     if (!dep.node()) {
         ASSERT_EQ(0u,