Compute Library
 21.11
CLBitwiseKernel.cpp
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2020-2021 Arm Limited.
3  *
4  * SPDX-License-Identifier: MIT
5  *
6  * Permission is hereby granted, free of charge, to any person obtaining a copy
7  * of this software and associated documentation files (the "Software"), to
8  * deal in the Software without restriction, including without limitation the
9  * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
10  * sell copies of the Software, and to permit persons to whom the Software is
11  * furnished to do so, subject to the following conditions:
12  *
13  * The above copyright notice and this permission notice shall be included in all
14  * copies or substantial portions of the Software.
15  *
16  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
19  * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
21  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
22  * SOFTWARE.
23  */
25 
33 #include "support/StringSupport.h"
34 
35 namespace arm_compute
36 {
38  : _input1(nullptr), _input2(nullptr), _output(nullptr)
39 {
41 }
42 
43 void CLBitwiseKernel::configure(const CLCompileContext &compile_context, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, BitwiseOperation op)
44 {
47  if(op != BitwiseOperation::NOT)
48  {
51  }
54 
55  // Output auto inizialitation if not yet initialized
56  auto_init_if_empty(*(output->info()), *(input1->info()));
57  auto padding_info = get_padding_info({ input1, input2, output });
58 
59  // Configure kernel window
60  const unsigned int vec_size_x = adjust_vec_size(16 / output->info()->element_size(), output->info()->dimension(0));
61  Window win = calculate_max_window(*output->info(), Steps(vec_size_x));
62 
63  _input1 = input1;
64  _input2 = input2;
65  _output = output;
66 
67  // Create kernel
68  std::string kernel_name = "";
69  switch(op)
70  {
72  kernel_name = "bitwise_and";
73  break;
75  kernel_name = "bitwise_not";
76  break;
78  kernel_name = "bitwise_or";
79  break;
81  kernel_name = "bitwise_xor";
82  break;
83  default:
84  ARM_COMPUTE_ERROR("Bitwise operation not supported");
85  }
86 
87  CLBuildOptions build_opts;
88  const int vec_size_x_leftovers = output->info()->dimension(0) % vec_size_x;
89  build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(vec_size_x));
90  build_opts.add_option("-DVEC_SIZE_LEFTOVER=" + support::cpp11::to_string(vec_size_x_leftovers));
91  _kernel = create_kernel(compile_context, kernel_name, build_opts.options());
92 
93  ICLKernel::configure_internal(win);
95 }
96 
97 void CLBitwiseKernel::run(const Window &window, cl::CommandQueue &queue)
98 {
101 
103 
104  do
105  {
106  unsigned int idx = 0;
107  add_2D_tensor_argument(idx, _input1, slice);
108  if(_input2 != nullptr)
109  {
110  add_2D_tensor_argument(idx, _input2, slice);
111  }
112  add_2D_tensor_argument(idx, _output, slice);
113  enqueue(queue, *this, slice, lws_hint());
114  }
115  while(window.slide_window_slice_2D(slice));
116 }
117 } // namespace arm_compute
Window first_slice_window_2D() const
First 2D slice of the window.
Definition: Window.h:283
Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps, bool skip_border, BorderSize border_size)
const Window & window() const
The maximum window the kernel can be executed on.
Definition: IKernel.cpp:28
void run(const Window &window, cl::CommandQueue &queue) override
Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue...
virtual size_t dimension(size_t index) const =0
Return the size of the requested dimension.
void enqueue(cl::CommandQueue &queue, ICLKernel &kernel, const Window &window, const cl::NDRange &lws_hint=CLKernelLibrary::get().default_ndrange(), bool use_dummy_work_items=false)
Add the kernel to the command queue with the given window.
Definition: ICLKernel.cpp:32
const StringSet & options() const
Gets the current options list set.
#define ARM_COMPUTE_ERROR(msg)
Print the given message then throw an std::runtime_error.
Definition: Error.h:352
cl::NDRange lws_hint() const
Return the Local-Workgroup-Size hint.
Definition: ICLKernel.h:318
1 channel, 1 U8 per channel
std::string to_string(T &&value)
Convert integer and float values to string.
#define ARM_COMPUTE_ERROR_ON(cond)
If the condition is true then an error message is printed and an exception thrown.
Definition: Error.h:466
Manages all the OpenCL kernels compilation and caching, provides accessors for the OpenCL Context...
CLBitwiseKernel()
Default constructor.
Bitwise NOT operation.
bool slide_window_slice_2D(Window &slice) const
Slide the passed 2D window slice.
Definition: Window.h:323
Copyright (c) 2017-2021 Arm Limited.
void configure(const CLCompileContext &compile_context, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, BitwiseOperation op)
Set the inputs and output tensors.
void add_option(std::string option)
Adds option to the existing build option list.
cl::Kernel create_kernel(const CLCompileContext &ctx, const std::string &kernel_name, const std::set< std::string > &build_opts=std::set< std::string >())
Creates an opencl kernel using a compile context.
Definition: CLHelpers.cpp:391
Class to describe a number of elements in each dimension.
Definition: Steps.h:40
BitwiseOperation
Available bitwise operations.
Definition: Types.h:502
bool auto_init_if_empty(ITensorInfo &info, const TensorShape &shape, int num_channels, DataType data_type, QuantizationInfo quantization_info=QuantizationInfo())
Auto initialize the tensor info (shape, number of channels and data type) if the current assignment i...
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor&#39;s metadata.
virtual size_t element_size() const =0
Element size in bytes calculated as data_size() * num_channels()
Elementeise CL kernel type.
Definition: CLTypes.h:84
#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)
Definition: Validate.h:915
bool has_padding_changed(const std::unordered_map< const ITensorInfo *, PaddingSize > &padding_map)
Check if the previously stored padding info has changed after configuring a kernel.
Definition: Utils.cpp:533
#define ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(t, c,...)
Definition: Validate.h:786
CLCompileContext class.
void add_2D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)
Add the passed 2D tensor&#39;s parameters to the object&#39;s kernel&#39;s arguments starting from the index idx...
Definition: ICLKernel.h:190
Bitwise XOR operation.
Interface for OpenCL tensor.
Definition: ICLTensor.h:42
Bitwise AND operation.
Bitwise OR operation.
std::unordered_map< const ITensorInfo *, PaddingSize > get_padding_info(std::initializer_list< const ITensorInfo *> infos)
Stores padding information before configuring a kernel.
Definition: Utils.cpp:518
Wrapper to configure the Khronos OpenCL C++ header.
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:157
unsigned int adjust_vec_size(unsigned int vec_size, size_t dim0)
Returns the adjusted vector size in case it is less than the input&#39;s first dimension, getting rounded down to its closest valid vector size.
Definition: Utils.h:1171
std::string kernel_name
Describe a multidimensional execution window.
Definition: Window.h:39
#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)
Definition: Validate.h:201
SimpleTensor< T > slice(const SimpleTensor< T > &src, Coordinates starts, Coordinates ends)