Remove duplicated concat testcase with crop primitive

Signed-off-by: yuan.xiong <[email protected]>

Author: yuanxion

src/plugins/intel_gpu/tests/unit/test_cases/concatenation_gpu_test.cpp

@@ -1780,109 +1780,6 @@ TEST(concat_gpu_onednn, b_fs_yx_fsv16_input_types) {
     }
 }
 
-TEST(concat_gpu_onednn, crop_b_fs_yx_fsv16_input_types) {
-    auto& engine = get_test_engine();
-    if (!engine.get_device_info().supports_immad)
-        return;
-
-    tests::random_generator rg(GET_SUITE_NAME);
-    const int32_t input_b = 1, input_f = 3, input_y = 416, input_x = 416;
-    uint32_t test_stride = 8;
-    int32_t test_feature = 88, test_kernel = 8;
-    int32_t test_y = input_y / test_stride, test_x = input_x / test_stride;
-
-    auto test_dt = data_types::f16;
-    auto test_format = format::b_fs_yx_fsv16;
-
-    layout input0_layout = { test_dt, format::bfyx, { input_b, input_f, input_x, input_y } };
-    auto input0 = engine.allocate_memory(input0_layout);
-
-    using test_data_type = ov::float16;
-    auto data_input = rg.generate_random_4d<test_data_type>(input_b, input_f, input_y, input_x, -1, 1);
-    auto data_input_flat = flatten_4d(format::bfyx, data_input);
-    set_values<test_data_type>(input0, data_input_flat);
-
-    auto weights0 = engine.allocate_memory({ data_types::f16, format::bfyx, { test_feature * 2, input_f, test_kernel, test_kernel } });
-    auto biases0 = engine.allocate_memory({ data_types::f16, format::bfyx, { 1, test_feature * 2, 1, 1 } });
-    std::vector<test_data_type> weights0_flat(test_feature * 2 * input_f * test_kernel * test_kernel, 1.0);
-    std::vector<test_data_type> bias0_flat(1 * test_feature * 2 * 1 * 1, 1.0);
-    set_values(weights0, weights0_flat);
-    set_values(biases0, bias0_flat);
-
-    auto weights1 = engine.allocate_memory({ data_types::f16, format::bfyx, { test_feature, test_feature, 1, 1 } });
-    auto biases1 = engine.allocate_memory({ data_types::f16, format::bfyx, { 1, test_feature, 1, 1 } });
-    std::vector<test_data_type> weights1_flat(test_feature * test_feature * 1 * 1, 1.0);
-    std::vector<test_data_type> bias1_flat(1 * test_feature * 1 * 1, 1.0);
-    set_values(weights1, weights1_flat);
-    set_values(biases1, bias1_flat);
-
-    auto output_name = "reorder";
-    layout reorder_layout = { test_dt, format::bfyx, { input_b, test_feature * 2, input_x, input_y } };
-
-    ov::intel_gpu::ImplementationDesc impl;
-    std::vector<test_data_type> outputs_ref(input_b * test_feature * 2 * test_y * test_x);
-    for (int i = 0; i < 2; ++i) {
-        if (i == 0) {
-            impl = { test_format, std::string(""), impl_types::ocl };
-        } else {
-            impl = { test_format, std::string(""), impl_types::onednn };
-        }
-
-        topology topology(
-                input_layout("input0", input0_layout),
-                data("weights0", weights0),
-                data("biases0", biases0),
-                convolution( "conv0", input_info("input0"), "weights0", "biases0", 1, {test_stride, test_stride}, {1, 1}, {0, 0}, {0, 0}, false),
-                crop("variadic_split.out0", input_info("conv0"), {input_b, test_feature, test_x, test_y }, {0, 0, 0, 0}),
-                crop("variadic_split.out1", input_info("conv0"), {input_b, test_feature, test_x, test_y }, {0, test_feature, 0, 0}),
-                data("weights1", weights1),
-                data("biases1", biases1),
-                convolution( "conv1", input_info("variadic_split.out1"), "weights1", "biases1", 1, {1, 1}, {1, 1}, {0, 0}, {0, 0}, false),
-                concatenation("concat",
-                              { input_info("variadic_split.out0"), input_info("conv1") },
-                              1,
-                              test_dt),
-                reorder("reorder", input_info("concat"), reorder_layout)
-        );
-
-        ExecutionConfig cfg = get_test_default_config(engine);
-        cfg.set_property(ov::intel_gpu::custom_outputs(std::vector<std::string>{ output_name }));
-        cfg.set_property(ov::intel_gpu::force_implementations(ov::intel_gpu::ImplForcingMap{ {"concat", impl} }));
-        network network_test(engine, topology, cfg);
-
-        network_test.set_input_data("input0", input0);
-        auto outputs = network_test.execute();
-
-        ASSERT_EQ(outputs.size(), size_t(1));
-        ASSERT_EQ(outputs.begin()->first, output_name);
-
-        auto output_memory = outputs.at(output_name).get_memory();
-        auto output_layout = output_memory->get_layout();
-        int y_size = output_layout.spatial(1);
-        int x_size = output_layout.spatial(0);
-        int f_size = output_layout.feature();
-        int b_size = output_layout.batch();
-
-        ASSERT_EQ(output_layout.format, format::bfyx);
-        ASSERT_EQ(y_size, test_y);
-        ASSERT_EQ(x_size, test_x);
-        ASSERT_EQ(f_size, test_feature * 2);
-        ASSERT_EQ(b_size, input_b);
-
-        cldnn::mem_lock<test_data_type> output_ptr(output_memory, get_test_stream());
-
-        if (i == 0) {
-            for (size_t x = 0; x < output_layout.count(); ++x) {
-                outputs_ref[x] = output_ptr[x];
-            }
-        } else {
-            for (size_t x = 0; x < output_layout.count(); ++x) {
-                ASSERT_EQ(output_ptr[x], outputs_ref[x]);
-            }
-        }
-    }
-}
-
 template <typename Type>
 struct concat_gpu_4d_implicit_onednn : public concat_gpu_implicit {
 public: