aboutsummaryrefslogtreecommitdiff
path: root/src/core/CL/kernels/CLBitwiseXorKernel.cpp
diff options
context:
space:
mode:
authorAnthony Barbier <anthony.barbier@arm.com>2017-09-04 18:44:23 +0100
committerAnthony Barbier <anthony.barbier@arm.com>2018-09-17 13:03:09 +0100
commit6ff3b19ee6120edf015fad8caab2991faa3070af (patch)
treea7a6dcd16dfd56d79fa1b56a313caeebcc939b68 /src/core/CL/kernels/CLBitwiseXorKernel.cpp
downloadComputeLibrary-6ff3b19ee6120edf015fad8caab2991faa3070af.tar.gz
COMPMID-344 Updated doxygen
Change-Id: I32f7b84daa560e460b77216add529c8fa8b327ae
Diffstat (limited to 'src/core/CL/kernels/CLBitwiseXorKernel.cpp')
-rw-r--r--src/core/CL/kernels/CLBitwiseXorKernel.cpp89
1 files changed, 89 insertions, 0 deletions
diff --git a/src/core/CL/kernels/CLBitwiseXorKernel.cpp b/src/core/CL/kernels/CLBitwiseXorKernel.cpp
new file mode 100644
index 0000000000..c19a78e1c4
--- /dev/null
+++ b/src/core/CL/kernels/CLBitwiseXorKernel.cpp
@@ -0,0 +1,89 @@
+/*
+ * Copyright (c) 2016, 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/CLBitwiseXorKernel.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/Helpers.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/core/Validate.h"
+#include "arm_compute/core/Window.h"
+
+using namespace arm_compute;
+
+CLBitwiseXorKernel::CLBitwiseXorKernel()
+ : _input1(nullptr), _input2(nullptr), _output(nullptr)
+{
+}
+
+void CLBitwiseXorKernel::configure(const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output)
+{
+ ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input1, 1, DataType::U8);
+ ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input2, 1, DataType::U8);
+ ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output, 1, DataType::U8);
+
+ _input1 = input1;
+ _input2 = input2;
+ _output = output;
+
+ // Create kernel
+ _kernel = static_cast<cl::Kernel>(CLKernelLibrary::get().create_kernel("bitwise_xor"));
+
+ // Configure kernel window
+ constexpr unsigned int num_elems_processed_per_iteration = 16;
+
+ Window win = calculate_max_window(*input1->info(), Steps(num_elems_processed_per_iteration));
+
+ AccessWindowHorizontal input1_access(input1->info(), 0, num_elems_processed_per_iteration);
+ AccessWindowHorizontal input2_access(input2->info(), 0, num_elems_processed_per_iteration);
+ AccessWindowHorizontal output_access(output->info(), 0, num_elems_processed_per_iteration);
+
+ update_window_and_padding(win, input1_access, input2_access, output_access);
+
+ ValidRegion valid_region = intersect_valid_regions(input1->info()->valid_region(),
+ input2->info()->valid_region());
+
+ output_access.set_valid_region(win, valid_region);
+
+ ICLKernel::configure(win);
+}
+
+void CLBitwiseXorKernel::run(const Window &window, cl::CommandQueue &queue)
+{
+ ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
+ ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
+
+ Window slice = window.first_slice_window_2D();
+
+ do
+ {
+ unsigned int idx = 0;
+ add_2D_tensor_argument(idx, _input1, slice);
+ add_2D_tensor_argument(idx, _input2, slice);
+ add_2D_tensor_argument(idx, _output, slice);
+ enqueue(queue, *this, slice);
+ }
+ while(window.slide_window_slice_2D(slice));
+}