feat(dnn): add bool type cvt on gpu

GitOrigin-RevId: ab0fecf368

dnn/src/cuda/type_cvt/opr_impl.cpp

@@ -73,6 +73,7 @@ void exec_src_normal(const TensorND& dst, const TensorND& src,
         return;                                                    \
     }
             MEGDNN_FOREACH_COMPUTING_DTYPE(cb);
+            cb(::megdnn::dtype::Bool);
 #undef cb
             default:
                 megdnn_assert_internal(0);

src/opr/test/basic_arith/others.cpp

@@ -546,6 +546,28 @@ TEST(TestOprBasicArith, TypeCvt) {
     ASSERT_EQ(TensorShape({3, 0}), host_y.shape());
 }
 
+TEST(TestOprBasicArith, TypeCvtBool) {
+    auto graph = ComputingGraph::make();
+    HostTensorGenerator<dtype::Int32> gen;
+    auto host_x = gen({3});
+    auto px = host_x->ptr<int>();
+    px[0] = -1;
+    px[1] = 0;
+    px[2] = 1;
+
+    auto x = opr::Host2DeviceCopy::make(*graph, host_x),
+         y = opr::TypeCvt::make(x, dtype::Bool{});
+    HostTensorND host_y;
+    auto func = graph->compile({make_callback_copy(y, host_y)});
+    func->execute();
+
+    auto py = host_y.ptr<bool>();
+    for (size_t i = 0;i < 3;i ++) {
+        ASSERT_EQ(static_cast<bool>(px[i]), py[i]);
+    }
+    ASSERT_EQ(TensorShape({3}), host_y.shape());
+}
+
 TEST(TestOprBasicArith, ElemwiseMemFwd) {
     auto graph = ComputingGraph::make();
     graph->options().graph_opt_level = 0;