diff options
author | Anthony Barbier <anthony.barbier@arm.com> | 2017-09-04 18:44:23 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-09-17 13:03:09 +0100 |
commit | 6ff3b19ee6120edf015fad8caab2991faa3070af (patch) | |
tree | a7a6dcd16dfd56d79fa1b56a313caeebcc939b68 /src/core/CL/kernels/CLCol2ImKernel.cpp | |
download | ComputeLibrary-6ff3b19ee6120edf015fad8caab2991faa3070af.tar.gz |
COMPMID-344 Updated doxygen
Change-Id: I32f7b84daa560e460b77216add529c8fa8b327ae
Diffstat (limited to 'src/core/CL/kernels/CLCol2ImKernel.cpp')
-rw-r--r-- | src/core/CL/kernels/CLCol2ImKernel.cpp | 85 |
1 files changed, 85 insertions, 0 deletions
diff --git a/src/core/CL/kernels/CLCol2ImKernel.cpp b/src/core/CL/kernels/CLCol2ImKernel.cpp new file mode 100644 index 0000000000..ad66c39483 --- /dev/null +++ b/src/core/CL/kernels/CLCol2ImKernel.cpp @@ -0,0 +1,85 @@ +/* + * Copyright (c) 2017 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. + */ +#include "arm_compute/core/CL/kernels/CLCol2ImKernel.h" + +#include "arm_compute/core/CL/CLHelpers.h" +#include "arm_compute/core/CL/CLKernelLibrary.h" +#include "arm_compute/core/CL/ICLTensor.h" +#include "arm_compute/core/CL/OpenCL.h" +#include "arm_compute/core/Error.h" +#include "arm_compute/core/Helpers.h" +#include "arm_compute/core/Types.h" +#include "arm_compute/core/Validate.h" + +#include <cmath> + +using namespace arm_compute; + +CLCol2ImKernel::CLCol2ImKernel() + : _input(nullptr), _output(nullptr), _convolved_dims() +{ +} + +void CLCol2ImKernel::configure(const ICLTensor *input, ICLTensor *output, std::pair<unsigned int, unsigned int> convolved_dims) +{ + ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input, 1, DataType::F32); + ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output, 1, DataType::F32); + ARM_COMPUTE_ERROR_ON_MISMATCHING_DATA_TYPES(input, output); + + _input = input; + _output = output; + _convolved_dims = convolved_dims; + + // Create kernel + std::set<std::string> build_opts = { ("-DDATA_TYPE=" + get_cl_type_from_data_type(input->info()->data_type())) }; + _kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel("col2im", build_opts)); + + // Set static kernel arguments + unsigned int idx = num_arguments_per_2D_tensor() + num_arguments_per_3D_tensor(); + _kernel.setArg<cl_uint>(idx++, _convolved_dims.first); + + // Configure window + Window win = calculate_max_window(*input->info(), Steps()); + // The CLCol2ImKernel doesn't need padding so update_window_and_padding() can be skipped + output->info()->set_valid_region(ValidRegion(Coordinates(), output->info()->tensor_shape())); + ICLKernel::configure(win); +} + +void CLCol2ImKernel::run(const Window &window, cl::CommandQueue &queue) +{ + ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this); + ARM_COMPUTE_ERROR_ON_MISMATCHING_WINDOWS(ICLKernel::window(), window); + + Window slice_in = window.first_slice_window_2D(); + Window slice_out = window.first_slice_window_3D(); + do + { + // Set inputs + unsigned int idx = 0; + add_2D_tensor_argument(idx, _input, slice_in); + add_3D_tensor_argument(idx, _output, slice_out); + enqueue(queue, *this, slice_in); + } + while(window.slide_window_slice_2D(slice_in) && window.slide_window_slice_3D(slice_out)); +} |