1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
|
/*
* Copyright (c) 2016-2020 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.
*/
#include "arm_compute/core/CL/kernels/CLScharr3x3Kernel.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 "arm_compute/core/Window.h"
#include <set>
#include <string>
using namespace arm_compute;
CLScharr3x3Kernel::CLScharr3x3Kernel()
: _run_scharr_x(false), _run_scharr_y(false), _input(nullptr), _output_x(nullptr), _output_y(nullptr)
{
}
BorderSize CLScharr3x3Kernel::border_size() const
{
return BorderSize(1);
}
void CLScharr3x3Kernel::configure(const ICLTensor *input, ICLTensor *output_x, ICLTensor *output_y, bool border_undefined)
{
configure(CLKernelLibrary::get().get_compile_context(), input, output_x, output_y, border_undefined);
}
void CLScharr3x3Kernel::configure(const CLCompileContext &compile_context, const ICLTensor *input, ICLTensor *output_x, ICLTensor *output_y, bool border_undefined)
{
ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input, 1, DataType::U8);
ARM_COMPUTE_ERROR_ON((output_x == nullptr) && (output_y == nullptr));
_run_scharr_x = output_x != nullptr;
_run_scharr_y = output_y != nullptr;
if(_run_scharr_x)
{
ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output_x, 1, DataType::S16);
}
if(_run_scharr_y)
{
ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output_y, 1, DataType::S16);
}
_input = input;
_output_x = output_x;
_output_y = output_y;
// Set build options
std::set<std::string> build_opts;
if(_run_scharr_x)
{
build_opts.insert("-DGRAD_X");
}
if(_run_scharr_y)
{
build_opts.insert("-DGRAD_Y");
}
// Create kernel
_kernel = create_kernel(compile_context, "scharr3x3", build_opts);
// Configure kernel window
constexpr unsigned int num_elems_processed_per_iteration = 8;
constexpr unsigned int num_elems_read_per_iteration = 16;
constexpr unsigned int num_elems_written_per_iteration = 8;
constexpr unsigned int num_rows_read_per_iteration = 3;
Window win = calculate_max_window(*input->info(), Steps(num_elems_processed_per_iteration), border_undefined, border_size());
AccessWindowRectangle input_access(input->info(), -border_size().left, -border_size().top, num_elems_read_per_iteration, num_rows_read_per_iteration);
AccessWindowHorizontal output_x_access(output_x == nullptr ? nullptr : output_x->info(), 0, num_elems_written_per_iteration);
AccessWindowHorizontal output_y_access(output_y == nullptr ? nullptr : output_y->info(), 0, num_elems_written_per_iteration);
update_window_and_padding(win, input_access, output_x_access, output_y_access);
output_x_access.set_valid_region(win, input->info()->valid_region(), border_undefined, border_size());
output_y_access.set_valid_region(win, input->info()->valid_region(), border_undefined, border_size());
ICLKernel::configure_internal(win);
}
void CLScharr3x3Kernel::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, _input, slice);
add_2D_tensor_argument_if((_run_scharr_x), idx, _output_x, slice);
add_2D_tensor_argument_if((_run_scharr_y), idx, _output_y, slice);
enqueue(queue, *this, slice, lws_hint());
}
while(window.slide_window_slice_2D(slice));
}
|
