Last comments

src/common/snippets/include/snippets/runtime_configurator.hpp

@@ -164,6 +164,12 @@ class RuntimeConfigurator {
         void init(const ov::snippets::lowered::LinearIRCPtr& linear_ir,
                   const std::vector<snippets::lowered::PortDescriptorPtr>& io_descs,
                   size_t in_num);
+        /**
+         * @brief Checks if optimizer is enabled
+         * @todo Ticket 148891: when RuntimeConfigurator::update will be rewritten on PassPipeline, this method should be removed
+         * We will not just register ParallelWAOptimizer in case if it is not needed
+         */
+        bool enabled();
         /**
          * @brief Checks if the current master shape can be optimized, and if yes, updates all the necessary runtime information
          * @param master_shape Master shape
@@ -174,7 +180,7 @@ class RuntimeConfigurator {
          * @param in_num Number of inputs. It is needed to distinguish input and output shapes/layouts
          * @return status if the optimization is applied
          */
-        bool optimize(ov::snippets::VectorDims& master_shape,
+        void optimize(ov::snippets::VectorDims& master_shape,
                       ov::snippets::RuntimeConfigurator::LoopInfoRuntimeParamsMap& map,
                       std::vector<ov::snippets::VectorDims>& shapes,
                       std::vector<std::vector<size_t>>& layouts,

src/common/snippets/src/runtime_configurator.cpp

@@ -53,8 +53,10 @@ void RuntimeConfigurator::update(const lowered::LinearIRCPtr& linear_ir) {
     LoopInfoRuntimeParamsMap initialized_info;
     auto shapes = extract_shapes();
     auto layouts = extract_layouts();
-    if (m_optimizer.optimize(m_config->master_shape, initialized_info, shapes, layouts, m_in_num))
+    if (m_optimizer.enabled()) {
+        m_optimizer.optimize(m_config->master_shape, initialized_info, shapes, layouts, m_in_num);
         update_tensor_rank(m_config->master_shape);
+    }
 
     if (linear_ir->is_dynamic()) {
         update_loop_info(linear_ir, initialized_info);
@@ -318,6 +320,7 @@ void RuntimeConfigurator::ParallelWAOptimizer::init(const lowered::LinearIRCPtr&
         return;
 
     concurrency = linear_ir->get_config().m_min_parallel_work_amount;
+    // At the moment this optimization is Brgemm related so there must be `unsqueezed_params`
     unsqueezed_params = find_unsqueezed_params(linear_ir, brgemms);
     OPENVINO_ASSERT(!unsqueezed_params.empty(), "unsqueezed_params mustn't be empty after initialization");
     loops_to_split = find_loops_to_split(linear_ir, unsqueezed_params);
@@ -333,20 +336,23 @@ void RuntimeConfigurator::ParallelWAOptimizer::init(const lowered::LinearIRCPtr&
     }
 }
 
-bool RuntimeConfigurator::ParallelWAOptimizer::optimize(VectorDims& master_shape,
+bool RuntimeConfigurator::ParallelWAOptimizer::enabled() {
+    return !loops_to_split.empty();
+}
+
+void RuntimeConfigurator::ParallelWAOptimizer::optimize(VectorDims& master_shape,
                                                         RuntimeConfigurator::LoopInfoRuntimeParamsMap& map,
                                                         std::vector<ov::snippets::VectorDims>& shapes,
                                                         std::vector<std::vector<size_t>>& layouts,
                                                         size_t in_num) {
     size_t new_batch_dim, new_kernel_dim;
-    if (loops_to_split.empty() || !SplitDimensionM::split(master_shape, concurrency, new_batch_dim, new_kernel_dim))
-        return false;
+    if (!SplitDimensionM::split(master_shape, concurrency, new_batch_dim, new_kernel_dim))
+        return;
 
     update_master_shape(master_shape, new_batch_dim, new_kernel_dim);
     update_split_loops_info(map, new_kernel_dim);
     update_shapes(shapes, new_batch_dim, new_kernel_dim);
     update_layouts(layouts);
-    return true;
 }
 
 void RuntimeConfigurator::ParallelWAOptimizer::update_master_shape(VectorDims& master_shape, size_t new_batch_dim, size_t new_kernel_dim) {