/* * Copyright (c) 2017-2020 Arm Limited. * * SPDX-License-Identifier: MIT * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to * deal in the Software without restriction, including without limitation the * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or * sell copies of the Software, and to permit persons to whom the Software is * furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in all * copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. */ #ifndef ARM_COMPUTE_CLTUNER_H #define ARM_COMPUTE_CLTUNER_H #include "arm_compute/core/CL/OpenCL.h" #include "arm_compute/runtime/CL/CLTunerTypes.h" #include "arm_compute/runtime/CL/ICLTuner.h" #include namespace arm_compute { class ICLKernel; /** Basic implementation of the OpenCL tuner interface */ class CLTuner : public ICLTuner { public: /** Constructor * * @param[in] tune_new_kernels Find the optimal local workgroup size for kernels which are not present in the table ? * */ CLTuner(bool tune_new_kernels = true); /** Destructor */ ~CLTuner() = default; /** Setter for tune_new_kernels option * * @param[in] tune_new_kernels Find the optimal local workgroup size for kernels which are not present in the table ? */ void set_tune_new_kernels(bool tune_new_kernels); /** Tune kernels that are not in the LWS table * * @return True if tuning of new kernels is enabled. */ bool tune_new_kernels() const; /** Set OpenCL tuner mode * * @param[in] mode Indicates how exhaustive the search for the optimal LWS should be while tuning. Default is Exhaustive mode */ void set_tuner_mode(CLTunerMode mode); /** Get the current OpenCL tuner mode * * @return tuner_mode Indicates how exhaustive the search for the optimal LWS should be while tuning */ CLTunerMode get_tuner_mode() const; /** Manually add a LWS for a kernel * * @param[in] kernel_id Unique identifiant of the kernel * @param[in] optimal_lws Optimal local workgroup size to use for the given kernel */ void add_lws_to_table(const std::string &kernel_id, cl::NDRange optimal_lws); /** Import LWS table * * @param[in] lws_table The unordered_map container to import */ void import_lws_table(const std::unordered_map &lws_table); /** Give read access to the LWS table * * @return The lws table as unordered_map container */ const std::unordered_map &lws_table() const; /** Set the OpenCL kernel event * * @note The interceptor can use this function to store the event associated to the OpenCL kernel * * @param[in] kernel_event The OpenCL kernel event */ void set_cl_kernel_event(cl_event kernel_event); /** clEnqueueNDRangeKernel symbol */ std::function real_clEnqueueNDRangeKernel; /** Load the LWS table from file * * @param[in] filename Load the LWS table from this file.(Must exist) */ void load_from_file(const std::string &filename); /** Save the content of the LWS table to file * * @param[in] filename Save the LWS table to this file. (Content will be overwritten) */ void save_to_file(const std::string &filename) const; // Inherited methods overridden: void tune_kernel_static(ICLKernel &kernel) override; void tune_kernel_dynamic(ICLKernel &kernel) override; void tune_kernel_dynamic(ICLKernel &kernel, const InputTensorMap &inputs, const OutputTensorMap &outputs) override; /** Is the kernel_event set ? * * @return true if the kernel_event is set. */ bool kernel_event_is_set() const; private: /** Find optimal LWS using brute-force approach * * @param[in] kernel OpenCL kernel to be tuned with LWS * * @return The optimal LWS to use */ cl::NDRange find_optimal_lws(ICLKernel &kernel, const InputTensorMap &inputs, const OutputTensorMap &outputs); std::unordered_map _lws_table; cl::Event _kernel_event; bool _tune_new_kernels; CLTunerMode _tuner_mode; }; } // namespace arm_compute #endif /*ARM_COMPUTE_CLTUNER_H */