COMPMID-734: CLTuner rework
Change-Id: I8f20d6ea8a09869d71003e7b08e0d33775282f6c
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/125802
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
diff --git a/arm_compute/core/CL/CLTypes.h b/arm_compute/core/CL/CLTypes.h
index a9d5fdd..14c31fa 100644
--- a/arm_compute/core/CL/CLTypes.h
+++ b/arm_compute/core/CL/CLTypes.h
@@ -60,5 +60,16 @@
CL20, /* the OpenCL 2.0 and above */
UNKNOWN /* unkown version */
};
-}
+
+/** OpenCL device options */
+struct CLDeviceOptions
+{
+ std::string name; /**< Device name */
+ std::string extensions; /**< List of supported extensions */
+ std::string ddk_version; /**< DDK version */
+ GPUTarget gpu_target; /**< GPU target architecture/instance */
+ size_t num_cores; /**< Number of cores */
+ size_t cache_size; /**< Cache size */
+};
+} // namespace arm_compute
#endif /* __ARM_COMPUTE_CL_TYPES_H__ */
diff --git a/arm_compute/core/CL/ICLKernel.h b/arm_compute/core/CL/ICLKernel.h
index c7d0c21..e6700cd 100644
--- a/arm_compute/core/CL/ICLKernel.h
+++ b/arm_compute/core/CL/ICLKernel.h
@@ -198,6 +198,15 @@
_lws_hint = lws_hint;
}
+ /** Return the Local-Workgroup-Size hint
+ *
+ * @return Current lws hint
+ */
+ cl::NDRange lws_hint() const
+ {
+ return _lws_hint;
+ }
+
/** Get the configuration ID
*
* @note The configuration ID can be used by the caller to distinguish different calls of the same OpenCL kernel
diff --git a/arm_compute/core/CL/kernels/CLDirectConvolutionLayerKernel.h b/arm_compute/core/CL/kernels/CLDirectConvolutionLayerKernel.h
index d47b7da..eb1bf58 100644
--- a/arm_compute/core/CL/kernels/CLDirectConvolutionLayerKernel.h
+++ b/arm_compute/core/CL/kernels/CLDirectConvolutionLayerKernel.h
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2017 ARM Limited.
+ * Copyright (c) 2017-2018 ARM Limited.
*
* SPDX-License-Identifier: MIT
*
@@ -88,7 +88,7 @@
void run(const Window &window, cl::CommandQueue &queue) override;
BorderSize border_size() const override;
-private:
+public:
const ICLTensor *_input;
const ICLTensor *_biases;
const ICLTensor *_weights;