[GPU] Add check of multiple axis broadcasting in is_valid_fusion() wh…

…en dynamic shape (openvinotoolkit#28252) ### Details: - Add check of multiple axis broadcasting in is_valid_fusion() when dynamic shape - Current ocl fused_op jit constant doesn't support multiple axis boradcase. ![image](https://github.com/user-attachments/assets/5064faef-252c-416e-a7ba-71151d35db1e) ### Tickets: - 159939
sgolebiewski-intel · Jan 6, 2025 · 3bd87f0 · 3bd87f0
1 parent 74041ec
commit 3bd87f0
Show file tree

Hide file tree

Showing 2 changed files with 107 additions and 0 deletions.
diff --git a/src/plugins/intel_gpu/src/graph/primitive_inst.cpp b/src/plugins/intel_gpu/src/graph/primitive_inst.cpp
@@ -2658,6 +2658,18 @@ bool primitive_inst::is_valid_fusion() const {
         if (fd.is_type<eltwise>())
             can_broadcast = ov::PartialShape::broadcast_merge_into(merged_shape, outer_dep_pshape, fd.typed_desc<eltwise>()->broadcast_spec);
 
+        // Check if broadcast happens more than single axis.
+        // Current gemm_tiled_opt kernel FUSED_OP_LOAD macro cannot support broadcast on dynamic dimension.
+        if (_node->is_type<gemm>() && can_broadcast == true && merged_shape.rank().get_length() == outer_dep_pshape.rank().get_length()) {
+            uint8_t broadcast_more_than_single_axis = 0;
+            for (int64_t i = 0; i < merged_shape.rank().get_length(); i++) {
+                if (merged_shape.get_shape().at(i) != outer_dep_pshape.get_shape().at(i))
+                    broadcast_more_than_single_axis++;
+            }
+            if (broadcast_more_than_single_axis > 1)
+                can_broadcast = false;
+        }
+
 #ifdef ENABLE_ONEDNN_FOR_GPU
         // WA for OneDNN binary add fusions: we need to broadcast batch dimension to avoid situation with
         // batch dimension mismatch in OneDNN tensor descriptors as follow:

diff --git a/src/plugins/intel_gpu/tests/functional/subgraph_tests/dynamic/dynamic_unfusion.cpp b/src/plugins/intel_gpu/tests/functional/subgraph_tests/dynamic/dynamic_unfusion.cpp
@@ -0,0 +1,95 @@
+// Copyright (C) 2024 Intel Corporation
+// SPDX-License-Identifier: Apache-2.0
+//
+
+#include "common_test_utils/ov_tensor_utils.hpp"
+#include "common_test_utils/file_utils.hpp"
+#include "shared_test_classes/base/ov_subgraph.hpp"
+
+#include "openvino/op/parameter.hpp"
+#include "openvino/op/matmul.hpp"
+#include "openvino/op/multiply.hpp"
+
+namespace {
+
+using ov::test::InputShape;
+
+using DynamicUnfusionsParams = std::tuple<std::vector<InputShape>,   // input shapes
+                                          ov::element::Type>;        // input precision
+
+class DynamicUnfusions : public testing::WithParamInterface<DynamicUnfusionsParams>,
+                         virtual public ov::test::SubgraphBaseTest {
+public:
+    static std::string getTestCaseName(testing::TestParamInfo<DynamicUnfusionsParams> obj) {
+        std::vector<InputShape> input_shapes;
+        ov::element::Type input_precision;
+
+        std::tie(input_shapes, input_precision) = obj.param;
+
+        std::ostringstream result;
+        result << "IS=(";
+        for (const auto& shape : input_shapes) {
+            result << ov::test::utils::partialShape2str({shape.first}) << "_";
+        }
+        result << ")_TS=";
+        for (const auto& shape : input_shapes) {
+            result << "(";
+            if (!shape.second.empty()) {
+                auto itr = shape.second.begin();
+                do {
+                    result << ov::test::utils::vec2str(*itr);
+                } while (++itr != shape.second.end() && result << "_");
+            }
+            result << ")_";
+        }
+        result << "input_precision=" << input_precision;
+        return result.str();
+    }
+
+protected:
+    std::shared_ptr<ov::Model> init_subgraph(std::vector<ov::PartialShape>& input_shapes,
+                                             const ov::element::Type input_precision) {
+        auto input0 = std::make_shared<ov::op::v0::Parameter>(input_precision, input_shapes[0]);
+        auto input1 = std::make_shared<ov::op::v0::Parameter>(input_precision, input_shapes[1]);
+        auto input2 = std::make_shared<ov::op::v0::Parameter>(input_precision, input_shapes[2]);
+
+        auto matmul = std::make_shared<ov::op::v0::MatMul>(input0, input1);
+        auto mul = std::make_shared<ov::op::v1::Multiply>(matmul, input2);
+
+        matmul->set_friendly_name("MatMul");
+        mul->set_friendly_name("Multiply");
+
+        return std::make_shared<ov::Model>(ov::NodeVector{mul}, ov::ParameterVector{input0, input1, input2}, "DynamicUnfusions");
+    }
+
+    void SetUp() override {
+        targetDevice = ov::test::utils::DEVICE_GPU;
+
+        std::vector<InputShape> input_shapes;
+        ov::element::Type input_precision;
+
+        std::tie(input_shapes, input_precision) = GetParam();
+
+        init_input_shapes(input_shapes);
+
+        inType = outType = input_precision;
+        function = init_subgraph(inputDynamicShapes, input_precision);
+    }
+};
+
+TEST_P(DynamicUnfusions, Inference) {
+    run();
+}
+
+const std::vector<ov::element::Type> input_precisions = {ov::element::f32};
+
+const std::vector<std::vector<InputShape>> input_shapes_dyn = {
+    {{{1024, -1}, {{1024, 1024}}}, {{-1, 1024}, {{1024, 1024}}}, {{1, -1}, {{1, 1}}}},
+};
+
+INSTANTIATE_TEST_SUITE_P(DynamicUnfusions_basic,
+                         DynamicUnfusions,
+                         ::testing::Combine(::testing::ValuesIn(input_shapes_dyn),
+                                            ::testing::ValuesIn(input_precisions)),
+                         DynamicUnfusions::getTestCaseName);
+} // namespace