diff options
Diffstat (limited to 'src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp')
| -rw-r--r-- | src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp | 578 |
1 files changed, 331 insertions, 247 deletions
diff --git a/src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp b/src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp index 2622274587..71c247de79 100644 --- a/src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp +++ b/src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp @@ -52,7 +52,7 @@ namespace { inline bool validate_gemm_kernel(CLGEMMKernelType kernel_type) { - switch(kernel_type) + switch (kernel_type) { case CLGEMMKernelType::NATIVE: case CLGEMMKernelType::RESHAPED_ONLY_RHS: @@ -71,32 +71,41 @@ inline bool validate_gemm_kernel(CLGEMMKernelType kernel_type) inline CLGEMMKernelType auto_select_gemm_kernel(auto_heuristics::CommonQuery query, bool reshape_b_only_on_first_run) { auto gemm_kernel = auto_heuristics::select_mlgo_gemm_kernel(query, reshape_b_only_on_first_run); - if(bool(gemm_kernel)) + if (bool(gemm_kernel)) { - if(validate_gemm_kernel(gemm_kernel.gemm_type)) + if (validate_gemm_kernel(gemm_kernel.gemm_type)) { - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use gemm kernel from mlgo heuristics: %s.", to_string(gemm_kernel.gemm_type).c_str()); + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use gemm kernel from mlgo heuristics: %s.", + to_string(gemm_kernel.gemm_type).c_str()); return gemm_kernel.gemm_type; } } gemm_kernel = auto_heuristics::select_default_gemm_kernel(query, reshape_b_only_on_first_run); - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use gemm kernel from default heuristics: %s.", to_string(gemm_kernel.gemm_type).c_str()); + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use gemm kernel from default heuristics: %s.", + to_string(gemm_kernel.gemm_type).c_str()); return gemm_kernel.gemm_type; } // Validate lhs_info and rhs_info for native kernel -inline bool validate_lhs_rhs_info_native(const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const ITensorInfo *a, const ITensorInfo *b, const GEMMReshapeInfo &reshape_info) +inline bool validate_lhs_rhs_info_native(const GEMMLHSMatrixInfo &lhs_info, + const GEMMRHSMatrixInfo &rhs_info, + const ITensorInfo *a, + const ITensorInfo *b, + const GEMMReshapeInfo &reshape_info) { // Validate GEMMLHSMatrixInfo and GEMMRHSMatrixInfo for reshaped only rhs kernel TensorInfo mm_result_s32_info{}; // Output tensor auto initialization if not yet initialized - auto_init_if_empty(mm_result_s32_info, a->clone()->set_tensor_shape(compute_mm_shape(*a, *b, false, reshape_info)).set_data_type(DataType::S32)); + auto_init_if_empty( + mm_result_s32_info, + a->clone()->set_tensor_shape(compute_mm_shape(*a, *b, false, reshape_info)).set_data_type(DataType::S32)); // Validate mm kernel // NOTE: Ignore all other parameters (eg. output stage etc.) and only validate lhs and rhs info // NOTE: This assumes: // 1. lhs and rhs info's validity does not depend on these other parameters and vice versa(in CLGEMMLowpMatrixMultiplyNativeKernel.cpp validate_arguments). // 2. lhs and rhs info does not cause window and padding issues through side effects (in CLGEMMLowpMatrixMultiplyNativeKernel.cpp validate_and_configure_window). - if(!bool(ClGemmLowpMatrixMultiplyNativeKernel::validate(a, b, &mm_result_s32_info, lhs_info, rhs_info, reshape_info))) + if (!bool(ClGemmLowpMatrixMultiplyNativeKernel::validate(a, b, &mm_result_s32_info, lhs_info, rhs_info, + reshape_info))) { return false; } @@ -104,31 +113,45 @@ inline bool validate_lhs_rhs_info_native(const GEMMLHSMatrixInfo &lhs_info, cons } // Automatically select between mlgo (prioritized) and default heuristics for native kernel configs -std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> auto_select_gemm_config_native(auto_heuristics::CommonQuery query, const ITensorInfo *a, const ITensorInfo *b, const GEMMReshapeInfo &reshape_info) +std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> auto_select_gemm_config_native(auto_heuristics::CommonQuery query, + const ITensorInfo *a, + const ITensorInfo *b, + const GEMMReshapeInfo &reshape_info) { auto config = auto_heuristics::select_mlgo_gemm_config_native(query); - if(config) + if (config) { - if(validate_lhs_rhs_info_native(config.lhs_info, config.rhs_info, a, b, reshape_info)) + if (validate_lhs_rhs_info_native(config.lhs_info, config.rhs_info, a, b, reshape_info)) { - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use native config from mlgo heuristics: LHS info: %s ; RHS info: %s ", to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); - return { config.lhs_info, config.rhs_info }; + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE( + "Use native config from mlgo heuristics: LHS info: %s ; RHS info: %s ", + to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); + return {config.lhs_info, config.rhs_info}; } } config = auto_heuristics::select_default_gemm_config_native(query); - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use native config from default heuristics: LHS info: %s ; RHS info: %s ", to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); - return { config.lhs_info, config.rhs_info }; + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use native config from default heuristics: LHS info: %s ; RHS info: %s ", + to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); + return {config.lhs_info, config.rhs_info}; } // Validate lhs_info and rhs_info for reshaped only rhs kernel -inline bool validate_lhs_rhs_info_reshaped_only_rhs(const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *output, - unsigned int m, unsigned int n, unsigned int k, bool reinterpret_input_as_3d, int depth_output_gemm3d) +inline bool validate_lhs_rhs_info_reshaped_only_rhs(const GEMMLHSMatrixInfo &lhs_info, + const GEMMRHSMatrixInfo &rhs_info, + const ITensorInfo *a, + const ITensorInfo *b, + const ITensorInfo *output, + unsigned int m, + unsigned int n, + unsigned int k, + bool reinterpret_input_as_3d, + int depth_output_gemm3d) { // Validate GEMMLHSMatrixInfo and GEMMRHSMatrixInfo for reshaped only rhs kernel TensorInfo tmp_b_info{}; // Validate reshape RHS kernel auto_init_if_empty(tmp_b_info, b->clone()->set_tensor_shape(compute_rhs_reshaped_shape(*b, rhs_info))); - if(!bool(ClGemmReshapeRhsMatrixKernel::validate(b, &tmp_b_info, rhs_info))) + if (!bool(ClGemmReshapeRhsMatrixKernel::validate(b, &tmp_b_info, rhs_info))) { return false; } @@ -148,7 +171,8 @@ inline bool validate_lhs_rhs_info_reshaped_only_rhs(const GEMMLHSMatrixInfo &lhs // Since we ignore the output stage, output data type has to be S32 to pass the validation TensorInfo output_info_copy(*output); output_info_copy.set_data_type(DataType::S32); - if(!bool(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate(a, &tmp_b_info, &output_info_copy, gemm_kernel_info))) + if (!bool(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate(a, &tmp_b_info, &output_info_copy, + gemm_kernel_info))) { return false; } @@ -156,14 +180,22 @@ inline bool validate_lhs_rhs_info_reshaped_only_rhs(const GEMMLHSMatrixInfo &lhs } // Validate lhs_info and rhs_info for reshaped only rhs kernel -inline bool validate_lhs_rhs_info_reshaped_only_rhs_mmul(const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *output, - unsigned int m, unsigned int n, unsigned int k, bool reinterpret_input_as_3d, int depth_output_gemm3d) +inline bool validate_lhs_rhs_info_reshaped_only_rhs_mmul(const GEMMLHSMatrixInfo &lhs_info, + const GEMMRHSMatrixInfo &rhs_info, + const ITensorInfo *a, + const ITensorInfo *b, + const ITensorInfo *output, + unsigned int m, + unsigned int n, + unsigned int k, + bool reinterpret_input_as_3d, + int depth_output_gemm3d) { // Validate GEMMLHSMatrixInfo and GEMMRHSMatrixInfo for reshaped only rhs kernel TensorInfo tmp_b_info{}; // Validate reshape RHS kernel auto_init_if_empty(tmp_b_info, b->clone()->set_tensor_shape(compute_rhs_reshaped_shape(*b, rhs_info))); - if(!bool(ClGemmReshapeRhsMatrixKernel::validate(b, &tmp_b_info, rhs_info))) + if (!bool(ClGemmReshapeRhsMatrixKernel::validate(b, &tmp_b_info, rhs_info))) { return false; } @@ -183,7 +215,8 @@ inline bool validate_lhs_rhs_info_reshaped_only_rhs_mmul(const GEMMLHSMatrixInfo // Since we ignore the output stage, output data type has to be S32 to pass the validation TensorInfo output_info_copy(*output); output_info_copy.set_data_type(DataType::S32); - if(!bool(ClGemmLowpMatrixMultiplyReshapedOnlyRhsMMULKernel::validate(a, &tmp_b_info, &output_info_copy, gemm_kernel_info))) + if (!bool(ClGemmLowpMatrixMultiplyReshapedOnlyRhsMMULKernel::validate(a, &tmp_b_info, &output_info_copy, + gemm_kernel_info))) { return false; } @@ -191,40 +224,55 @@ inline bool validate_lhs_rhs_info_reshaped_only_rhs_mmul(const GEMMLHSMatrixInfo } // Automatically select between mlgo (prioritized) and default heuristics for reshaped only rhs kernel configs -std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> auto_select_gemm_config_reshaped_only_rhs(auto_heuristics::CommonQuery query, bool reinterpret_input_as_3d, int depth_output_gemm3d, - const ITensorInfo *a, - const ITensorInfo *b, const ITensorInfo *output) +std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> +auto_select_gemm_config_reshaped_only_rhs(auto_heuristics::CommonQuery query, + bool reinterpret_input_as_3d, + int depth_output_gemm3d, + const ITensorInfo *a, + const ITensorInfo *b, + const ITensorInfo *output) { auto config = auto_heuristics::select_mlgo_gemm_config_reshaped_only_rhs(query); - if(config) + if (config) { - if(validate_lhs_rhs_info_reshaped_only_rhs(config.lhs_info, config.rhs_info, a, b, output, query.m, query.n, query.k, reinterpret_input_as_3d, depth_output_gemm3d)) + if (validate_lhs_rhs_info_reshaped_only_rhs(config.lhs_info, config.rhs_info, a, b, output, query.m, query.n, + query.k, reinterpret_input_as_3d, depth_output_gemm3d)) { - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use reshaped_only_rhs config from mlgo heuristics: LHS info: %s ; RHS info: %s ", to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); - return { config.lhs_info, config.rhs_info }; + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE( + "Use reshaped_only_rhs config from mlgo heuristics: LHS info: %s ; RHS info: %s ", + to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); + return {config.lhs_info, config.rhs_info}; } } config = auto_heuristics::select_default_gemm_config_reshaped_only_rhs(query); - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use reshaped_only_rhs config from default heuristics: LHS info: %s ; RHS info: %s ", to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); - return { config.lhs_info, config.rhs_info }; + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE( + "Use reshaped_only_rhs config from default heuristics: LHS info: %s ; RHS info: %s ", + to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); + return {config.lhs_info, config.rhs_info}; } // Automatically select between mlgo (prioritized) and default heuristics for reshaped only rhs kernel configs -std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> auto_select_gemm_config_reshaped_only_rhs_mmul(auto_heuristics::CommonQuery query, bool reinterpret_input_as_3d, int depth_output_gemm3d, - const ITensorInfo *a, - const ITensorInfo *b, const ITensorInfo *output) +std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> +auto_select_gemm_config_reshaped_only_rhs_mmul(auto_heuristics::CommonQuery query, + bool reinterpret_input_as_3d, + int depth_output_gemm3d, + const ITensorInfo *a, + const ITensorInfo *b, + const ITensorInfo *output) { ARM_COMPUTE_UNUSED(a, b, output, reinterpret_input_as_3d, depth_output_gemm3d); auto config = auto_heuristics::select_default_gemm_config_reshaped_only_rhs(query); - validate_lhs_rhs_info_reshaped_only_rhs_mmul(config.lhs_info, config.rhs_info, a, b, output, query.m, query.n, query.k, reinterpret_input_as_3d, depth_output_gemm3d); - ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE("Use reshaped_only_rhs_mmul config from default heuristics: LHS info: %s ; RHS info: %s ", to_string(config.lhs_info).c_str(), - to_string(config.rhs_info).c_str()); - return { config.lhs_info, config.rhs_info }; + validate_lhs_rhs_info_reshaped_only_rhs_mmul(config.lhs_info, config.rhs_info, a, b, output, query.m, query.n, + query.k, reinterpret_input_as_3d, depth_output_gemm3d); + ARM_COMPUTE_LOG_INFO_MSG_WITH_FORMAT_CORE( + "Use reshaped_only_rhs_mmul config from default heuristics: LHS info: %s ; RHS info: %s ", + to_string(config.lhs_info).c_str(), to_string(config.rhs_info).c_str()); + return {config.lhs_info, config.rhs_info}; } inline bool is_gemm_reshaped(CLGEMMKernelType kernel_type) { - switch(kernel_type) + switch (kernel_type) { case CLGEMMKernelType::NATIVE: return false; @@ -254,8 +302,11 @@ ClGemmLowpMatrixMultiplyCore::ClGemmLowpMatrixMultiplyCore() ClGemmLowpMatrixMultiplyCore::~ClGemmLowpMatrixMultiplyCore() = default; void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_context, - ITensorInfo *a, ITensorInfo *b, ITensorInfo *c, ITensorInfo *output, - const GEMMInfo &gemm_info) + ITensorInfo *a, + ITensorInfo *b, + ITensorInfo *c, + ITensorInfo *output, + const GEMMInfo &gemm_info) { ARM_COMPUTE_ERROR_ON_NULLPTR(a, b, output); ARM_COMPUTE_ERROR_THROW_ON(ClGemmLowpMatrixMultiplyCore::validate(a, b, c, output, gemm_info)); @@ -263,8 +314,8 @@ void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_con _reshape_b_only_on_first_run = gemm_info.reshape_b_only_on_first_run(); _a_offset = a->quantization_info().uniform().offset; - _convert_to_qasymm8 = is_data_type_quantized_per_channel(b->data_type()) && is_data_type_quantized_symmetric(b->data_type()) - && a->data_type() == DataType::QASYMM8; + _convert_to_qasymm8 = is_data_type_quantized_per_channel(b->data_type()) && + is_data_type_quantized_symmetric(b->data_type()) && a->data_type() == DataType::QASYMM8; _b_offset = _convert_to_qasymm8 ? -128 : b->quantization_info().uniform().offset; _gemm_info = gemm_info; @@ -282,17 +333,18 @@ void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_con // Arguments used by GEMMReshapeInfo // in order to know how the matrices have been reshaped bool reinterpret_input_as_3d = gemm_info.reinterpret_input_as_3d(); - const unsigned int m = reinterpret_input_as_3d ? (a->dimension(1) * a->dimension(2)) : a->dimension(1); - const unsigned int n = b->dimension(0); - const unsigned int k = a->dimension(0); - const unsigned int batch_size = reinterpret_input_as_3d ? a->dimension(3) : a->dimension(2); - const int depth_output_gemm3d = gemm_info.depth_output_gemm3d(); + const unsigned int m = reinterpret_input_as_3d ? (a->dimension(1) * a->dimension(2)) : a->dimension(1); + const unsigned int n = b->dimension(0); + const unsigned int k = a->dimension(0); + const unsigned int batch_size = reinterpret_input_as_3d ? a->dimension(3) : a->dimension(2); + const int depth_output_gemm3d = gemm_info.depth_output_gemm3d(); const auto reshape_info = GEMMReshapeInfo(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d); - _gemm_kernel_type = auto_select_gemm_kernel(auto_heuristics::CommonQuery{ gpu_target, a->data_type(), m, n, k, batch_size }, _reshape_b_only_on_first_run); + _gemm_kernel_type = auto_select_gemm_kernel( + auto_heuristics::CommonQuery{gpu_target, a->data_type(), m, n, k, batch_size}, _reshape_b_only_on_first_run); - if(_convert_to_qasymm8) + if (_convert_to_qasymm8) { // Set data type for converted weights _qasymm8_weights = *b; @@ -301,47 +353,50 @@ void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_con } ITensorInfo *matrix_b = _convert_to_qasymm8 ? &_qasymm8_weights : b; - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) { matrix_b = &_tmp_b; // Pick up the GEMM configuration // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - std::tie(lhs_info, rhs_info) = auto_select_gemm_config_reshaped_only_rhs(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }, reinterpret_input_as_3d, - depth_output_gemm3d, - a, _convert_to_qasymm8 ? &_qasymm8_weights : b, output); + std::tie(lhs_info, rhs_info) = auto_select_gemm_config_reshaped_only_rhs( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}, reinterpret_input_as_3d, + depth_output_gemm3d, a, _convert_to_qasymm8 ? &_qasymm8_weights : b, output); // Configure reshape RHS kernel - _mtx_b_reshape_kernel->configure(compile_context, _convert_to_qasymm8 ? &_qasymm8_weights : b, &_tmp_b, rhs_info); + _mtx_b_reshape_kernel->configure(compile_context, _convert_to_qasymm8 ? &_qasymm8_weights : b, &_tmp_b, + rhs_info); } - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) { matrix_b = &_tmp_b; // Pick up the GEMM configuration // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - std::tie(lhs_info, rhs_info) = auto_select_gemm_config_reshaped_only_rhs_mmul(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }, reinterpret_input_as_3d, - depth_output_gemm3d, - a, _convert_to_qasymm8 ? &_qasymm8_weights : b, output); + std::tie(lhs_info, rhs_info) = auto_select_gemm_config_reshaped_only_rhs_mmul( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}, reinterpret_input_as_3d, + depth_output_gemm3d, a, _convert_to_qasymm8 ? &_qasymm8_weights : b, output); // Configure reshape RHS kernel - _mtx_b_reshape_kernel->configure(compile_context, _convert_to_qasymm8 ? &_qasymm8_weights : b, &_tmp_b, rhs_info); + _mtx_b_reshape_kernel->configure(compile_context, _convert_to_qasymm8 ? &_qasymm8_weights : b, &_tmp_b, + rhs_info); } // Using default reduction info - const GEMMLowpReductionKernelInfo reduction_info {}; + const GEMMLowpReductionKernelInfo reduction_info{}; // Initialize matrix B reduction kernel only if _a_offset is not equal to 0 - if(_a_offset != 0) + if (_a_offset != 0) { _vector_sum_col = TensorInfo(compute_reductionA_shape(*b), 1, DataType::S32); // Configure Matrix B reduction kernel - _mtx_b_reduction_kernel->configure(compile_context, _convert_to_qasymm8 ? &_qasymm8_weights : b, &_vector_sum_col, reduction_info); + _mtx_b_reduction_kernel->configure(compile_context, _convert_to_qasymm8 ? &_qasymm8_weights : b, + &_vector_sum_col, reduction_info); } // Initialize Matrix A reduction kernel only if _b_offset is not equal to 0 - if(_b_offset != 0) + if (_b_offset != 0) { _vector_sum_row = TensorInfo(compute_reductionB_shape(*a), 1, DataType::S32); @@ -360,17 +415,19 @@ void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_con gemm_kernel_info.a_offset = _a_offset; gemm_kernel_info.b_offset = _b_offset; // If GEMMLowpOutputStage != NONE, fuse the offset contribution with the output stage - if(gemm_info.gemmlowp_output_stage().type != GEMMLowpOutputStageType::NONE) + if (gemm_info.gemmlowp_output_stage().type != GEMMLowpOutputStageType::NONE) { // Configure offset contribution kernel - const size_t num_filters = (gemm_info.gemmlowp_output_stage().is_quantized_per_channel) ? gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.size() : 1; + const size_t num_filters = (gemm_info.gemmlowp_output_stage().is_quantized_per_channel) + ? gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.size() + : 1; _gemm_output_stage_multipliers = TensorInfo(TensorShape(num_filters), 1, DataType::S32); _gemm_output_stage_shifts = TensorInfo(TensorShape(num_filters), 1, DataType::S32); GEMMLowpOutputStageInfo gemmlowp_output_stage = gemm_info.gemmlowp_output_stage(); gemmlowp_output_stage.output_data_type = a->data_type(); - if(num_filters == 1) + if (num_filters == 1) { // Per-channel quantization with OFM == 1 is equivalent to uniform quantization. // Setting this flag to false prevents the kernel from adding useless padding to the output multipliers and shifts @@ -379,55 +436,67 @@ void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_con gemm_kernel_info.output_stage = gemmlowp_output_stage; - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS && gemmlowp_output_stage.type == GEMMLowpOutputStageType::QUANTIZE_DOWN_FIXEDPOINT) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS && + gemmlowp_output_stage.type == GEMMLowpOutputStageType::QUANTIZE_DOWN_FIXEDPOINT) { // Configure and tune matrix multiply kernel with fused output stage - _mm_reshaped_only_rhs_kernel->configure(compile_context, a, matrix_b, output, gemm_kernel_info, _a_offset == 0 ? nullptr : &_vector_sum_col, - _b_offset == 0 ? nullptr : &_vector_sum_row, c != nullptr ? c : nullptr, &_gemm_output_stage_multipliers, &_gemm_output_stage_shifts); + _mm_reshaped_only_rhs_kernel->configure( + compile_context, a, matrix_b, output, gemm_kernel_info, _a_offset == 0 ? nullptr : &_vector_sum_col, + _b_offset == 0 ? nullptr : &_vector_sum_row, c != nullptr ? c : nullptr, + &_gemm_output_stage_multipliers, &_gemm_output_stage_shifts); } - else if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL && gemmlowp_output_stage.type == GEMMLowpOutputStageType::QUANTIZE_DOWN_FIXEDPOINT) + else if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL && + gemmlowp_output_stage.type == GEMMLowpOutputStageType::QUANTIZE_DOWN_FIXEDPOINT) { // Configure and tune matrix multiply kernel with fused output stage - _mm_reshaped_only_rhs_mmul_kernel->configure(compile_context, a, matrix_b, output, gemm_kernel_info, _a_offset == 0 ? nullptr : &_vector_sum_col, - _b_offset == 0 ? nullptr : &_vector_sum_row, c != nullptr ? c : nullptr, &_gemm_output_stage_multipliers, &_gemm_output_stage_shifts); + _mm_reshaped_only_rhs_mmul_kernel->configure( + compile_context, a, matrix_b, output, gemm_kernel_info, _a_offset == 0 ? nullptr : &_vector_sum_col, + _b_offset == 0 ? nullptr : &_vector_sum_row, c != nullptr ? c : nullptr, + &_gemm_output_stage_multipliers, &_gemm_output_stage_shifts); } else { _run_output_stage = true; - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) { - _mm_reshaped_only_rhs_kernel->configure(compile_context, a, matrix_b, &_mm_result_s32, gemm_kernel_info); + _mm_reshaped_only_rhs_kernel->configure(compile_context, a, matrix_b, &_mm_result_s32, + gemm_kernel_info); } - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) { - _mm_reshaped_only_rhs_mmul_kernel->configure(compile_context, a, matrix_b, &_mm_result_s32, gemm_kernel_info); + _mm_reshaped_only_rhs_mmul_kernel->configure(compile_context, a, matrix_b, &_mm_result_s32, + gemm_kernel_info); } else { // Pick up the GEMM configuration // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - std::tie(lhs_info, rhs_info) = auto_select_gemm_config_native(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }, - a, _convert_to_qasymm8 ? &_qasymm8_weights : matrix_b, reshape_info); + std::tie(lhs_info, rhs_info) = auto_select_gemm_config_native( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}, a, + _convert_to_qasymm8 ? &_qasymm8_weights : matrix_b, reshape_info); // Configure matrix multiply kernel - _mm_native_kernel->configure(compile_context, a, matrix_b, &_mm_result_s32, lhs_info, rhs_info, reshape_info); - - _offset_contribution_output_stage_kernel->configure(compile_context, &_mm_result_s32, _a_offset == 0 ? nullptr : &_vector_sum_col, _b_offset == 0 ? nullptr : &_vector_sum_row, - c != nullptr ? c : nullptr, output, a->dimension(0), _a_offset, _b_offset, gemmlowp_output_stage, - &_gemm_output_stage_multipliers, &_gemm_output_stage_shifts); + _mm_native_kernel->configure(compile_context, a, matrix_b, &_mm_result_s32, lhs_info, rhs_info, + reshape_info); + + _offset_contribution_output_stage_kernel->configure( + compile_context, &_mm_result_s32, _a_offset == 0 ? nullptr : &_vector_sum_col, + _b_offset == 0 ? nullptr : &_vector_sum_row, c != nullptr ? c : nullptr, output, a->dimension(0), + _a_offset, _b_offset, gemmlowp_output_stage, &_gemm_output_stage_multipliers, + &_gemm_output_stage_shifts); } } } else { _run_offset_contribution = true; - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) { // Configure and tune matrix multiply kernel _mm_reshaped_only_rhs_kernel->configure(compile_context, a, matrix_b, output, gemm_kernel_info); } - else if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) + else if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) { // Configure and tune matrix multiply kernel _mm_reshaped_only_rhs_mmul_kernel->configure(compile_context, a, matrix_b, output, gemm_kernel_info); @@ -436,44 +505,65 @@ void ClGemmLowpMatrixMultiplyCore::configure(const CLCompileContext &compile_con { // Pick up the GEMM configuration // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - std::tie(lhs_info, rhs_info) = auto_select_gemm_config_native(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }, - a, _convert_to_qasymm8 ? &_qasymm8_weights : b, reshape_info); + std::tie(lhs_info, rhs_info) = auto_select_gemm_config_native( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}, a, + _convert_to_qasymm8 ? &_qasymm8_weights : b, reshape_info); // Configure matrix multiply kernel _mm_native_kernel->configure(compile_context, a, matrix_b, output, lhs_info, rhs_info, reshape_info); } // Configure offset contribution kernel - _offset_contribution_kernel->configure(compile_context, output, _a_offset == 0 ? nullptr : &_vector_sum_col, _b_offset == 0 ? nullptr : &_vector_sum_row, - c != nullptr ? c : nullptr, a->dimension(0), _a_offset, _b_offset); + _offset_contribution_kernel->configure(compile_context, output, _a_offset == 0 ? nullptr : &_vector_sum_col, + _b_offset == 0 ? nullptr : &_vector_sum_row, c != nullptr ? c : nullptr, + a->dimension(0), _a_offset, _b_offset); } // Request memory - _aux_mem[RhsQAsymm8] = MemoryInfo(offset_int_vec(RhsQAsymm8), _reshape_b_only_on_first_run ? MemoryLifetime::Persistent : MemoryLifetime::Temporary, _qasymm8_weights.total_size()); - if(is_gemm_reshaped(_gemm_kernel_type)) + _aux_mem[RhsQAsymm8] = + MemoryInfo(offset_int_vec(RhsQAsymm8), + _reshape_b_only_on_first_run ? MemoryLifetime::Persistent : MemoryLifetime::Temporary, + _qasymm8_weights.total_size()); + if (is_gemm_reshaped(_gemm_kernel_type)) { // Overwrite Rhs as prepare if gemm is reshaped as there will be a two-step transformation - _aux_mem[RhsQAsymm8] = MemoryInfo(offset_int_vec(RhsQAsymm8), _reshape_b_only_on_first_run ? MemoryLifetime::Prepare : MemoryLifetime::Temporary, _qasymm8_weights.total_size()); - _aux_mem[RhsReshape] = MemoryInfo(offset_int_vec(RhsReshape), _reshape_b_only_on_first_run ? MemoryLifetime::Persistent : MemoryLifetime::Temporary, _tmp_b.total_size()); - } - if(_a_offset != 0) - { - _aux_mem[VecSumCol] = MemoryInfo(offset_int_vec(VecSumCol), _reshape_b_only_on_first_run ? MemoryLifetime::Persistent : MemoryLifetime::Temporary, _vector_sum_col.total_size()); - } - if(_b_offset != 0) - { - _aux_mem[VecSumRow] = MemoryInfo(offset_int_vec(VecSumRow), MemoryLifetime::Temporary, _vector_sum_row.total_size()); - } - _aux_mem[ResultS32] = MemoryInfo(offset_int_vec(ResultS32), MemoryLifetime::Temporary, _mm_result_s32.total_size()); - _aux_mem[Multipliers] = MemoryInfo(offset_int_vec(Multipliers), MemoryLifetime::Persistent, _gemm_output_stage_multipliers.total_size()); - _aux_mem[Shifts] = MemoryInfo(offset_int_vec(Shifts), MemoryLifetime::Persistent, _gemm_output_stage_shifts.total_size()); + _aux_mem[RhsQAsymm8] = + MemoryInfo(offset_int_vec(RhsQAsymm8), + _reshape_b_only_on_first_run ? MemoryLifetime::Prepare : MemoryLifetime::Temporary, + _qasymm8_weights.total_size()); + _aux_mem[RhsReshape] = MemoryInfo( + offset_int_vec(RhsReshape), + _reshape_b_only_on_first_run ? MemoryLifetime::Persistent : MemoryLifetime::Temporary, _tmp_b.total_size()); + } + if (_a_offset != 0) + { + _aux_mem[VecSumCol] = + MemoryInfo(offset_int_vec(VecSumCol), + _reshape_b_only_on_first_run ? MemoryLifetime::Persistent : MemoryLifetime::Temporary, + _vector_sum_col.total_size()); + } + if (_b_offset != 0) + { + _aux_mem[VecSumRow] = + MemoryInfo(offset_int_vec(VecSumRow), MemoryLifetime::Temporary, _vector_sum_row.total_size()); + } + _aux_mem[ResultS32] = MemoryInfo(offset_int_vec(ResultS32), MemoryLifetime::Temporary, _mm_result_s32.total_size()); + _aux_mem[Multipliers] = MemoryInfo(offset_int_vec(Multipliers), MemoryLifetime::Persistent, + _gemm_output_stage_multipliers.total_size()); + _aux_mem[Shifts] = + MemoryInfo(offset_int_vec(Shifts), MemoryLifetime::Persistent, _gemm_output_stage_shifts.total_size()); } -Status ClGemmLowpMatrixMultiplyCore::validate(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, const ITensorInfo *output, const GEMMInfo &gemm_info) +Status ClGemmLowpMatrixMultiplyCore::validate(const ITensorInfo *a, + const ITensorInfo *b, + const ITensorInfo *c, + const ITensorInfo *output, + const GEMMInfo &gemm_info) { ARM_COMPUTE_ERROR_ON_NULLPTR(a, b, output); ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(a, 1, DataType::QASYMM8, DataType::QASYMM8_SIGNED); - ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(b, 1, DataType::QASYMM8, DataType::QASYMM8_SIGNED, DataType::QSYMM8, DataType::QSYMM8_PER_CHANNEL); + ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(b, 1, DataType::QASYMM8, DataType::QASYMM8_SIGNED, + DataType::QSYMM8, DataType::QSYMM8_PER_CHANNEL); ARM_COMPUTE_RETURN_ERROR_ON(a->data_type() == DataType::QASYMM8 && b->data_type() == DataType::QASYMM8_SIGNED); ARM_COMPUTE_RETURN_ERROR_ON(a->data_type() == DataType::QASYMM8_SIGNED && b->data_type() == DataType::QASYMM8); ARM_COMPUTE_RETURN_ERROR_ON_MSG(gemm_info.is_a_reshaped(), "Matrix A already reshaped is not supported"); @@ -492,39 +582,44 @@ Status ClGemmLowpMatrixMultiplyCore::validate(const ITensorInfo *a, const ITenso const GPUTarget gpu_target = CLScheduler::get().target(); bool reinterpret_input_as_3d = gemm_info.reinterpret_input_as_3d(); - const unsigned int m = reinterpret_input_as_3d ? (a->dimension(1) * a->dimension(2)) : a->dimension(1); - const unsigned int n = b->dimension(0); - const unsigned int k = a->dimension(0); - const unsigned int batch_size = reinterpret_input_as_3d ? a->dimension(3) : a->dimension(2); - const int depth_output_gemm3d = gemm_info.depth_output_gemm3d(); + const unsigned int m = reinterpret_input_as_3d ? (a->dimension(1) * a->dimension(2)) : a->dimension(1); + const unsigned int n = b->dimension(0); + const unsigned int k = a->dimension(0); + const unsigned int batch_size = reinterpret_input_as_3d ? a->dimension(3) : a->dimension(2); + const int depth_output_gemm3d = gemm_info.depth_output_gemm3d(); - bool reshape_matrix_b = is_gemm_reshaped(auto_select_gemm_kernel(auto_heuristics::CommonQuery{ gpu_target, a->data_type(), m, n, k, batch_size }, gemm_info.reshape_b_only_on_first_run())); + bool reshape_matrix_b = is_gemm_reshaped( + auto_select_gemm_kernel(auto_heuristics::CommonQuery{gpu_target, a->data_type(), m, n, k, batch_size}, + gemm_info.reshape_b_only_on_first_run())); const GEMMReshapeInfo reshape_info = GEMMReshapeInfo(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d); - bool convert_to_qasymm8 = is_data_type_quantized_per_channel(b->data_type()) && is_data_type_quantized_symmetric(b->data_type()) - && is_data_type_quantized_asymmetric(a->data_type()); + bool convert_to_qasymm8 = is_data_type_quantized_per_channel(b->data_type()) && + is_data_type_quantized_symmetric(b->data_type()) && + is_data_type_quantized_asymmetric(a->data_type()); TensorInfo weights_info(*b); - if(convert_to_qasymm8) + if (convert_to_qasymm8) { b_offset = -128; weights_info.set_data_type(DataType::QASYMM8); ARM_COMPUTE_RETURN_ON_ERROR(ClCastKernel::validate(b, &weights_info, ConvertPolicy::WRAP)); } const ITensorInfo *matrix_b_info = &weights_info; - if(reshape_matrix_b) + if (reshape_matrix_b) { matrix_b_info = &tmp_b_info; // Pick up the GEMM configuration // NOTE: No need to validate mlgo configurations as they automatically fall back to default heuristics if validation fails // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - const auto res = select_default_gemm_config_reshaped_only_rhs(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }); - lhs_info = res.lhs_info; - rhs_info = res.rhs_info; + const auto res = select_default_gemm_config_reshaped_only_rhs( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}); + lhs_info = res.lhs_info; + rhs_info = res.rhs_info; // Validate reshape RHS kernel - auto_init_if_empty(tmp_b_info, weights_info.clone()->set_tensor_shape(compute_rhs_reshaped_shape(weights_info, rhs_info))); + auto_init_if_empty(tmp_b_info, + weights_info.clone()->set_tensor_shape(compute_rhs_reshaped_shape(weights_info, rhs_info))); ARM_COMPUTE_RETURN_ON_ERROR(ClGemmReshapeRhsMatrixKernel::validate(&weights_info, &tmp_b_info, rhs_info)); } @@ -533,21 +628,23 @@ Status ClGemmLowpMatrixMultiplyCore::validate(const ITensorInfo *a, const ITenso const GEMMLowpReductionKernelInfo reduction_info; // Validate matrix B reduction kernel only if _a_offset is not equal to 0 - if(a_offset != 0) + if (a_offset != 0) { info_vector_sum_col = TensorInfo(compute_reductionA_shape(weights_info), 1, DataType::S32); // Configure Matrix B reduction kernel - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixBReductionKernel::validate(&weights_info, &info_vector_sum_col, reduction_info)); + ARM_COMPUTE_RETURN_ON_ERROR( + ClGemmLowpMatrixBReductionKernel::validate(&weights_info, &info_vector_sum_col, reduction_info)); } // Validate Matrix A reduction kernel only if _b_offset is not equal to 0 - if(b_offset != 0) + if (b_offset != 0) { info_vector_sum_row = TensorInfo(compute_reductionB_shape(*a), 1, DataType::S32); // Configure matrix A reduction kernel - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixAReductionKernel::validate(a, &info_vector_sum_row, reduction_info)); + ARM_COMPUTE_RETURN_ON_ERROR( + ClGemmLowpMatrixAReductionKernel::validate(a, &info_vector_sum_row, reduction_info)); } GEMMKernelInfo gemm_kernel_info; @@ -560,92 +657,99 @@ Status ClGemmLowpMatrixMultiplyCore::validate(const ITensorInfo *a, const ITenso gemm_kernel_info.rhs_info = rhs_info; gemm_kernel_info.a_offset = a_offset; gemm_kernel_info.b_offset = b_offset; - if(gemm_info.gemmlowp_output_stage().type != GEMMLowpOutputStageType::NONE) + if (gemm_info.gemmlowp_output_stage().type != GEMMLowpOutputStageType::NONE) { - const size_t num_filters = (gemm_info.gemmlowp_output_stage().is_quantized_per_channel) ? gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.size() : 1; + const size_t num_filters = (gemm_info.gemmlowp_output_stage().is_quantized_per_channel) + ? gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.size() + : 1; - const TensorInfo gemm_output_stage_multipliers_shifts_info(TensorInfo(TensorShape(num_filters), 1, DataType::S32)); + const TensorInfo gemm_output_stage_multipliers_shifts_info( + TensorInfo(TensorShape(num_filters), 1, DataType::S32)); GEMMLowpOutputStageInfo gemmlowp_output_stage = gemm_info.gemmlowp_output_stage(); gemmlowp_output_stage.output_data_type = a->data_type(); gemm_kernel_info.output_stage = gemmlowp_output_stage; - if(reshape_matrix_b && gemm_info.gemmlowp_output_stage().type == GEMMLowpOutputStageType::QUANTIZE_DOWN_FIXEDPOINT) + if (reshape_matrix_b && + gemm_info.gemmlowp_output_stage().type == GEMMLowpOutputStageType::QUANTIZE_DOWN_FIXEDPOINT) { - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate(matrix_a_info, matrix_b_info, output, gemm_kernel_info, - a_offset == 0 ? nullptr : &info_vector_sum_col, - b_offset == 0 ? nullptr : &info_vector_sum_row, - c, - &gemm_output_stage_multipliers_shifts_info, - &gemm_output_stage_multipliers_shifts_info)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate( + matrix_a_info, matrix_b_info, output, gemm_kernel_info, a_offset == 0 ? nullptr : &info_vector_sum_col, + b_offset == 0 ? nullptr : &info_vector_sum_row, c, &gemm_output_stage_multipliers_shifts_info, + &gemm_output_stage_multipliers_shifts_info)); } else { TensorInfo mm_result_s32_info{}; - if(reshape_matrix_b) + if (reshape_matrix_b) { // Output tensor auto inizialitation if not yet initialized - auto_init_if_empty(mm_result_s32_info, a->clone()->set_tensor_shape(compute_mm_shape(*matrix_a_info, *matrix_b_info, reshape_info)).set_data_type(DataType::S32)); + auto_init_if_empty(mm_result_s32_info, a->clone() + ->set_tensor_shape(compute_mm_shape( + *matrix_a_info, *matrix_b_info, reshape_info)) + .set_data_type(DataType::S32)); // Validate matrix multiply - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate(matrix_a_info, matrix_b_info, &mm_result_s32_info, gemm_kernel_info)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate( + matrix_a_info, matrix_b_info, &mm_result_s32_info, gemm_kernel_info)); } else { // Output tensor auto inizialitation if not yet initialized - auto_init_if_empty(mm_result_s32_info, a->clone()->set_tensor_shape(compute_mm_shape(*matrix_a_info, *matrix_b_info, false, reshape_info)).set_data_type(DataType::S32)); + auto_init_if_empty(mm_result_s32_info, a->clone() + ->set_tensor_shape(compute_mm_shape( + *matrix_a_info, *matrix_b_info, false, reshape_info)) + .set_data_type(DataType::S32)); // Pick up the GEMM configuration // NOTE: No need to validate mlgo configurations as they automatically fall back to default heuristics if validation fails // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - const auto res = select_default_gemm_config_native(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }); - lhs_info = res.lhs_info; - rhs_info = res.rhs_info; + const auto res = select_default_gemm_config_native( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}); + lhs_info = res.lhs_info; + rhs_info = res.rhs_info; // Validate matrix multiply - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyNativeKernel::validate(matrix_a_info, matrix_b_info, &mm_result_s32_info, lhs_info, rhs_info, reshape_info)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyNativeKernel::validate( + matrix_a_info, matrix_b_info, &mm_result_s32_info, lhs_info, rhs_info, reshape_info)); } // Validate offset contribution kernel - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpOffsetContributionOutputStageKernel::validate(&mm_result_s32_info, - a_offset == 0 ? nullptr : &info_vector_sum_col, - b_offset == 0 ? nullptr : &info_vector_sum_row, - c, - output, - a_offset, b_offset, - gemmlowp_output_stage, - &gemm_output_stage_multipliers_shifts_info, - &gemm_output_stage_multipliers_shifts_info)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpOffsetContributionOutputStageKernel::validate( + &mm_result_s32_info, a_offset == 0 ? nullptr : &info_vector_sum_col, + b_offset == 0 ? nullptr : &info_vector_sum_row, c, output, a_offset, b_offset, gemmlowp_output_stage, + &gemm_output_stage_multipliers_shifts_info, &gemm_output_stage_multipliers_shifts_info)); } } else { - if(reshape_matrix_b) + if (reshape_matrix_b) { // Validate matrix multiply - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate(matrix_a_info, matrix_b_info, output, gemm_kernel_info)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel::validate( + matrix_a_info, matrix_b_info, output, gemm_kernel_info)); } else { // Pick up the GEMM configuration // It doesn't matter whether Datatype is DataType::QASYMM8 or DataType::QASYMM8_SIGNED, since it only affect the shape configuration - const auto res = select_default_gemm_config_native(auto_heuristics::CommonQuery{ gpu_target, DataType::QASYMM8, m, n, k, batch_size }); - lhs_info = res.lhs_info; - rhs_info = res.rhs_info; + const auto res = select_default_gemm_config_native( + auto_heuristics::CommonQuery{gpu_target, DataType::QASYMM8, m, n, k, batch_size}); + lhs_info = res.lhs_info; + rhs_info = res.rhs_info; // Validate matrix multiply - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyNativeKernel::validate(matrix_a_info, matrix_b_info, output, lhs_info, rhs_info, reshape_info)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpMatrixMultiplyNativeKernel::validate( + matrix_a_info, matrix_b_info, output, lhs_info, rhs_info, reshape_info)); } - if(output->total_size() != 0) + if (output->total_size() != 0) { // Validate offset contribution kernel - ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpOffsetContributionKernel::validate(output, - a_offset == 0 ? nullptr : &info_vector_sum_col, - b_offset == 0 ? nullptr : &info_vector_sum_row, - c, - a_offset, b_offset)); + ARM_COMPUTE_RETURN_ON_ERROR(ClGemmLowpOffsetContributionKernel::validate( + output, a_offset == 0 ? nullptr : &info_vector_sum_col, b_offset == 0 ? nullptr : &info_vector_sum_row, + c, a_offset, b_offset)); } } @@ -675,73 +779,61 @@ void ClGemmLowpMatrixMultiplyCore::run(ITensorPack &tensors) const ITensor *matrix_a = a; const ITensor *matrix_b = _convert_to_qasymm8 ? rhs_qasymm8.get() : b; - if(is_gemm_reshaped(_gemm_kernel_type)) + if (is_gemm_reshaped(_gemm_kernel_type)) { matrix_b = tmp_b.get(); - if(!_reshape_b_only_on_first_run) + if (!_reshape_b_only_on_first_run) { // Run reshape matrix B - ITensorPack mtx_b_reshape_pack = - { - { TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b }, - { TensorType::ACL_DST, tmp_b.get() } - }; + ITensorPack mtx_b_reshape_pack = {{TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b}, + {TensorType::ACL_DST, tmp_b.get()}}; CLScheduler::get().enqueue_op(*_mtx_b_reshape_kernel, mtx_b_reshape_pack, false); } } // Run matrix B reduction kernel only if _a_offset is not equal to 0 - if(_a_offset != 0 && !_reshape_b_only_on_first_run) + if (_a_offset != 0 && !_reshape_b_only_on_first_run) { - ITensorPack mtx_b_red_pack = - { - { TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b }, - { TensorType::ACL_DST, vec_sum_col.get() } - }; + ITensorPack mtx_b_red_pack = {{TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b}, + {TensorType::ACL_DST, vec_sum_col.get()}}; CLScheduler::get().enqueue_op(*_mtx_b_reduction_kernel, mtx_b_red_pack, false); } // Run matrix A reduction kernel only if _b_offset is not equal to 0 - if(_b_offset != 0) + if (_b_offset != 0) { - ITensorPack mtx_a_red_pack = - { - { TensorType::ACL_SRC, matrix_a }, - { TensorType::ACL_DST, vec_sum_row.get() } - }; + ITensorPack mtx_a_red_pack = {{TensorType::ACL_SRC, matrix_a}, {TensorType::ACL_DST, vec_sum_row.get()}}; CLScheduler::get().enqueue_op(*_mtx_a_reduction_kernel, mtx_a_red_pack, false); } // Run matrix multiply - if(is_gemm_reshaped(_gemm_kernel_type)) + if (is_gemm_reshaped(_gemm_kernel_type)) { ITensorPack gemm_reshaped_pack; - if(_run_offset_contribution) + if (_run_offset_contribution) { - gemm_reshaped_pack = ITensorPack({ { TensorType::ACL_SRC_0, matrix_a }, - { TensorType::ACL_SRC_1, matrix_b }, - { TensorType::ACL_DST, _run_output_stage ? res32.get() : dst } - }); + gemm_reshaped_pack = ITensorPack({{TensorType::ACL_SRC_0, matrix_a}, + {TensorType::ACL_SRC_1, matrix_b}, + {TensorType::ACL_DST, _run_output_stage ? res32.get() : dst}}); } else { - gemm_reshaped_pack = ITensorPack( - { - { TensorType::ACL_SRC, matrix_a }, - { TensorType::ACL_SRC_1, matrix_b }, - { TensorType::ACL_BIAS, c }, - { TensorType::ACL_VEC_ROW_SUM, _b_offset == 0 ? nullptr : vec_sum_row.get() }, - { TensorType::ACL_VEC_COL_SUM, _a_offset == 0 ? nullptr : vec_sum_col.get() }, - { TensorType::ACL_SHIFTS, shifts.get() }, - { TensorType::ACL_MULTIPLIERS, multipliers.get() }, - { TensorType::ACL_DST, dst }, + gemm_reshaped_pack = ITensorPack({ + {TensorType::ACL_SRC, matrix_a}, + {TensorType::ACL_SRC_1, matrix_b}, + {TensorType::ACL_BIAS, c}, + {TensorType::ACL_VEC_ROW_SUM, _b_offset == 0 ? nullptr : vec_sum_row.get()}, + {TensorType::ACL_VEC_COL_SUM, _a_offset == 0 ? nullptr : vec_sum_col.get()}, + {TensorType::ACL_SHIFTS, shifts.get()}, + {TensorType::ACL_MULTIPLIERS, multipliers.get()}, + {TensorType::ACL_DST, dst}, }); } - if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) + if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS) { CLScheduler::get().enqueue_op(*_mm_reshaped_only_rhs_kernel, gemm_reshaped_pack, false); } - else if(_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) + else if (_gemm_kernel_type == CLGEMMKernelType::RESHAPED_ONLY_RHS_MMUL) { CLScheduler::get().enqueue_op(*_mm_reshaped_only_rhs_mmul_kernel, gemm_reshaped_pack, false); } @@ -752,46 +844,39 @@ void ClGemmLowpMatrixMultiplyCore::run(ITensorPack &tensors) } else { - ITensorPack gemm_native_pack = - { - { TensorType::ACL_SRC_0, matrix_a }, - { TensorType::ACL_SRC_1, matrix_b }, - { TensorType::ACL_DST, _run_offset_contribution ? dst : res32.get() } - }; + ITensorPack gemm_native_pack = {{TensorType::ACL_SRC_0, matrix_a}, + {TensorType::ACL_SRC_1, matrix_b}, + {TensorType::ACL_DST, _run_offset_contribution ? dst : res32.get()}}; CLScheduler::get().enqueue_op(*_mm_native_kernel, gemm_native_pack, false); } - if(_run_output_stage) + if (_run_output_stage) { // Run offset contribution/output stage kernel - ITensorPack output_stage_pack = - { - { TensorType::ACL_SRC, res32.get() }, - { TensorType::ACL_BIAS, c }, - { TensorType::ACL_VEC_ROW_SUM, _b_offset == 0 ? nullptr : vec_sum_row.get() }, - { TensorType::ACL_VEC_COL_SUM, _a_offset == 0 ? nullptr : vec_sum_col.get() }, - { TensorType::ACL_SHIFTS, shifts.get() }, - { TensorType::ACL_MULTIPLIERS, multipliers.get() }, - { TensorType::ACL_DST, dst }, + ITensorPack output_stage_pack = { + {TensorType::ACL_SRC, res32.get()}, + {TensorType::ACL_BIAS, c}, + {TensorType::ACL_VEC_ROW_SUM, _b_offset == 0 ? nullptr : vec_sum_row.get()}, + {TensorType::ACL_VEC_COL_SUM, _a_offset == 0 ? nullptr : vec_sum_col.get()}, + {TensorType::ACL_SHIFTS, shifts.get()}, + {TensorType::ACL_MULTIPLIERS, multipliers.get()}, + {TensorType::ACL_DST, dst}, }; CLScheduler::get().enqueue_op(*_offset_contribution_output_stage_kernel, output_stage_pack, true); } - if(_run_offset_contribution) + if (_run_offset_contribution) { // Run offset contribution kernel - ITensorPack offset_contrib_pack = - { - { TensorType::ACL_SRC_DST, dst }, - { TensorType::ACL_BIAS, c }, - { TensorType::ACL_VEC_ROW_SUM, _b_offset == 0 ? nullptr : vec_sum_row.get() }, - { TensorType::ACL_VEC_COL_SUM, _a_offset == 0 ? nullptr : vec_sum_col.get() } - }; + ITensorPack offset_contrib_pack = {{TensorType::ACL_SRC_DST, dst}, + {TensorType::ACL_BIAS, c}, + {TensorType::ACL_VEC_ROW_SUM, _b_offset == 0 ? nullptr : vec_sum_row.get()}, + {TensorType::ACL_VEC_COL_SUM, _a_offset == 0 ? nullptr : vec_sum_col.get()}}; CLScheduler::get().enqueue_op(*_offset_contribution_kernel, offset_contrib_pack, true); } } void ClGemmLowpMatrixMultiplyCore::prepare(ITensorPack &tensors) { - if(!_is_prepared) + if (!_is_prepared) { auto b = tensors.get_const_tensor(TensorType::ACL_SRC_1); CLAuxTensorHandler tmp_b(offset_int_vec(RhsReshape), _tmp_b, tensors, true); @@ -800,56 +885,55 @@ void ClGemmLowpMatrixMultiplyCore::prepare(ITensorPack &tensors) ARM_COMPUTE_ERROR_ON_NULLPTR(b); - if(_convert_to_qasymm8) + if (_convert_to_qasymm8) { - ITensorPack convert_to_qs8_pack = { { ACL_SRC, b }, { ACL_DST, rhs_qasymm8.get() } }; + ITensorPack convert_to_qs8_pack = {{ACL_SRC, b}, {ACL_DST, rhs_qasymm8.get()}}; CLScheduler::get().enqueue_op(*_weights_to_qasymm8, convert_to_qs8_pack, false); b->mark_as_unused(); } - if(is_gemm_reshaped(_gemm_kernel_type) && _reshape_b_only_on_first_run) + if (is_gemm_reshaped(_gemm_kernel_type) && _reshape_b_only_on_first_run) { // Run reshape kernel and mark original weights tensor as unused - ITensorPack mtx_b_pack = - { - { TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b }, - { TensorType::ACL_DST, tmp_b.get() } - }; + ITensorPack mtx_b_pack = {{TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b}, + {TensorType::ACL_DST, tmp_b.get()}}; CLScheduler::get().enqueue_op(*_mtx_b_reshape_kernel, mtx_b_pack, false); b->mark_as_unused(); } // Run matrix B reduction kernel only if _a_offset is not equal to 0 - if(_a_offset != 0 && _reshape_b_only_on_first_run) + if (_a_offset != 0 && _reshape_b_only_on_first_run) { - ITensorPack mtx_b_red_pack = - { - { TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b }, - { TensorType::ACL_DST, vec_sum_col.get() } - }; + ITensorPack mtx_b_red_pack = {{TensorType::ACL_SRC, _convert_to_qasymm8 ? rhs_qasymm8.get() : b}, + {TensorType::ACL_DST, vec_sum_col.get()}}; CLScheduler::get().enqueue_op(*_mtx_b_reduction_kernel, mtx_b_red_pack, false); } // Compute GEMM output multipliers and shifts for output stage { - const size_t num_filters = (_gemm_info.gemmlowp_output_stage().is_quantized_per_channel) ? _gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.size() : 1; + const size_t num_filters = (_gemm_info.gemmlowp_output_stage().is_quantized_per_channel) + ? _gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.size() + : 1; CLAuxTensorHandler multipliers(offset_int_vec(Multipliers), _gemm_output_stage_multipliers, tensors, false); CLAuxTensorHandler shifts(offset_int_vec(Shifts), _gemm_output_stage_shifts, tensors, false); ICLTensor *multiplier_tensor = multipliers.get(); - if(multiplier_tensor != nullptr && multiplier_tensor->info()->total_size() > 0) + if (multiplier_tensor != nullptr && multiplier_tensor->info()->total_size() > 0) { multiplier_tensor->map(CLScheduler::get().queue(), true); - std::memcpy(multiplier_tensor->ptr_to_element(Coordinates(0)), _gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.data(), num_filters * sizeof(int32_t)); + std::memcpy(multiplier_tensor->ptr_to_element(Coordinates(0)), + _gemm_info.gemmlowp_output_stage().gemmlowp_multipliers.data(), + num_filters * sizeof(int32_t)); multiplier_tensor->unmap(CLScheduler::get().queue()); } ICLTensor *shifts_tensor = shifts.get(); - if(shifts.get() != nullptr && shifts_tensor->info()->total_size() > 0) + if (shifts.get() != nullptr && shifts_tensor->info()->total_size() > 0) { shifts_tensor->map(CLScheduler::get().queue(), true); - std::memcpy(shifts_tensor->ptr_to_element(Coordinates(0)), _gemm_info.gemmlowp_output_stage().gemmlowp_shifts.data(), num_filters * sizeof(int32_t)); + std::memcpy(shifts_tensor->ptr_to_element(Coordinates(0)), + _gemm_info.gemmlowp_output_stage().gemmlowp_shifts.data(), num_filters * sizeof(int32_t)); shifts_tensor->unmap(CLScheduler::get().queue()); } } |