Compute Library
 21.02
NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2017-2020 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 
26 #include "arm_compute/core/Error.h"
30 #include "arm_compute/core/Types.h"
31 #include "arm_compute/core/Utils.h"
36 #include "src/core/NEON/NEAsymm.h"
39 
40 #include <arm_neon.h>
41 #include <cstddef>
42 #include <cstdint>
43 
44 namespace arm_compute
45 {
46 namespace
47 {
48 Status validate_arguments(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min, int max)
49 {
51  ARM_COMPUTE_RETURN_ERROR_ON(min > max);
52 
53  // Check biases if exist
54  if(bias != nullptr)
55  {
57  ARM_COMPUTE_RETURN_ERROR_ON(bias->num_dimensions() > 1);
58  ARM_COMPUTE_RETURN_ERROR_ON(input->dimension(0) != bias->dimension(0));
59  }
60 
61  if(output->total_size() != 0)
62  {
65  }
66 
67  return Status{};
68 }
69 
70 std::pair<Status, Window> validate_and_configure_window(ITensorInfo *input, ITensorInfo *output)
71 {
72  // Output auto inizialitation if not yet initialized
73  auto_init_if_empty(*output, input->clone()->set_data_type(DataType::QASYMM8));
74 
75  // Configure kernel window
76  Window win = calculate_max_window(*input, Steps());
77 
78  // NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel doesn't need padding so update_window_and_padding() can be skipped
79  Coordinates coord;
80  coord.set_num_dimensions(output->num_dimensions());
81  output->set_valid_region(ValidRegion(coord, output->tensor_shape()));
82 
83  return std::make_pair(Status{}, win);
84 }
85 } // namespace
86 
87 namespace arm_compute
88 {
89 class Coordinates;
90 } // namespace arm_compute
91 
92 template <bool is_bounded_relu>
94 {
95  const int32x4_t result_offset_after_shift_s32 = vdupq_n_s32(_result_offset_after_shift);
96  const uint8x16_t min_u8 = vdupq_n_u8(static_cast<uint8_t>(_min));
97  const uint8x16_t max_u8 = vdupq_n_u8(static_cast<uint8_t>(_max));
98 
99  ARM_COMPUTE_UNUSED(min_u8);
100  ARM_COMPUTE_UNUSED(max_u8);
101 
102  const int window_step_x = 16;
103  const auto window_start_x = static_cast<int>(window.x().start());
104  const auto window_end_x = static_cast<int>(window.x().end());
105 
106  Window win_collapsed = window.collapse_if_possible(window, Window::DimZ);
107  win_collapsed.set(Window::DimX, Window::Dimension(0, 1, 1));
108 
109  Iterator in(_input, win_collapsed);
110  Iterator out(_output, win_collapsed);
111  if(_bias != nullptr)
112  {
113  Window win_biases;
114  win_biases.set(Window::DimX, Window::Dimension(0, 1, 1));
115  win_biases.set(Window::DimY, Window::Dimension(0, 1, 1));
116 
117  Iterator bias(_bias, win_biases);
118  execute_window_loop(win_collapsed, [&](const Coordinates &)
119  {
120  // Compute 16 elements per iteration
121  int x = window_start_x;
122  for(; x <= (window_end_x - window_step_x); x += window_step_x)
123  {
124  int32x4x4_t in_s32 =
125  {
126  {
127  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 0),
128  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 4),
129  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 8),
130  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 12)
131  }
132  };
133 
134  const int32x4x4_t bias_s32 =
135  {
136  {
137  vld1q_s32(reinterpret_cast<const int32_t *>(bias.ptr()) + x + 0),
138  vld1q_s32(reinterpret_cast<const int32_t *>(bias.ptr()) + x + 4),
139  vld1q_s32(reinterpret_cast<const int32_t *>(bias.ptr()) + x + 8),
140  vld1q_s32(reinterpret_cast<const int32_t *>(bias.ptr()) + x + 12)
141  }
142  };
143 
144  // Add the bias to GEMM's result
145  in_s32.val[0] = vaddq_s32(in_s32.val[0], bias_s32.val[0]);
146  in_s32.val[1] = vaddq_s32(in_s32.val[1], bias_s32.val[1]);
147  in_s32.val[2] = vaddq_s32(in_s32.val[2], bias_s32.val[2]);
148  in_s32.val[3] = vaddq_s32(in_s32.val[3], bias_s32.val[3]);
149 
150  vst1q_u8(out.ptr() + x, finalize_quantization(in_s32, _result_fixedpoint_multiplier, _result_shift, result_offset_after_shift_s32, min_u8, max_u8, is_bounded_relu));
151  }
152 
153  // Compute left-over elements
154  for(; x < window_end_x; ++x)
155  {
156  const int32_t bias_value = *(reinterpret_cast<const int32_t *>(bias.ptr()) + x);
157  int32_t in_value = *(reinterpret_cast<const int32_t *>(in.ptr()) + x);
158 
159  // Add bias
160  in_value += bias_value;
161  // Finalize and store the result
162  *(out.ptr() + x) = finalize_quantization(in_value, _result_fixedpoint_multiplier, _result_shift, _result_offset_after_shift, static_cast<uint8_t>(_min), static_cast<uint8_t>(_max), is_bounded_relu);
163  }
164  },
165  in, out, bias);
166  }
167  else
168  {
169  execute_window_loop(win_collapsed, [&](const Coordinates &)
170  {
171  // Compute 16 elements per iteration
172  int x = window_start_x;
173  for(; x <= (window_end_x - window_step_x); x += window_step_x)
174  {
175  int32x4x4_t in_s32 =
176  {
177  {
178  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 0),
179  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 4),
180  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 8),
181  vld1q_s32(reinterpret_cast<const int32_t *>(in.ptr()) + x + 12)
182  }
183  };
184 
185  vst1q_u8(out.ptr() + x, finalize_quantization(in_s32, _result_fixedpoint_multiplier, _result_shift, result_offset_after_shift_s32, min_u8, max_u8, is_bounded_relu));
186  }
187 
188  // Compute left-over elements
189  for(; x < window_end_x; ++x)
190  {
191  const int32_t in_value = *(reinterpret_cast<const int32_t *>(in.ptr()) + x);
192 
193  // Finalize and store the result
194  *(out.ptr() + x) = finalize_quantization(in_value, _result_fixedpoint_multiplier, _result_shift, _result_offset_after_shift, static_cast<uint8_t>(_min), static_cast<uint8_t>(_max), is_bounded_relu);
195  }
196  },
197  in, out);
198  }
199 }
200 
202  : _func(nullptr), _input(nullptr), _bias(nullptr), _output(nullptr), _result_fixedpoint_multiplier(0), _result_shift(0), _result_offset_after_shift(0), _min(0), _max(0)
203 {
204 }
205 
206 void NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel::configure(const ITensor *input, const ITensor *bias, ITensor *output, int result_fixedpoint_multiplier, int result_shift,
207  int result_offset_after_shift, int min, int max)
208 {
209  // Perform validate step
210  ARM_COMPUTE_ERROR_ON_NULLPTR(input, output);
211  ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input->info(), (bias != nullptr) ? bias->info() : nullptr, output->info(), min, max));
212 
213  _input = input;
214  _bias = bias;
215  _output = output;
216  _result_fixedpoint_multiplier = result_fixedpoint_multiplier;
217  _result_shift = result_shift;
218  _result_offset_after_shift = result_offset_after_shift;
219  _min = min;
220  _max = max;
221 
222  // Configure kernel window
223  auto win_config = validate_and_configure_window(input->info(), output->info());
224  ARM_COMPUTE_ERROR_THROW_ON(win_config.first);
225  INEKernel::configure(win_config.second);
226 
227  // Check if we need to clamp the result using min and max
228  const bool is_bounded_relu = !(min <= 0 && max >= 255);
229  _func = is_bounded_relu ? &NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel::run<true> : &NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel::run<false>;
230 }
231 
233 {
234  ARM_COMPUTE_ERROR_ON_NULLPTR(input, output);
235  ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input, bias, output, min, max));
236  ARM_COMPUTE_RETURN_ON_ERROR(validate_and_configure_window(input->clone().get(), output->clone().get()).first);
237 
238  return Status{};
239 }
240 
242 {
243  ARM_COMPUTE_UNUSED(info);
246 
247  (this->*_func)(window);
248 }
249 } // namespace arm_compute
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 configure(const ITensor *input, const ITensor *bias, ITensor *output, int result_fixedpoint_multiplier, int result_shift, int result_offset_after_shift, int min=0, int max=0)
Initialise the kernel&#39;s input and output.
#define ARM_COMPUTE_RETURN_ON_ERROR(status)
Checks if a status contains an error and returns it.
Definition: Error.h:204
Store the tensor&#39;s metadata.
Definition: ITensorInfo.h:40
#define ARM_COMPUTE_ERROR_THROW_ON(status)
Definition: Error.h:455
Describe one of the image&#39;s dimensions with a start, end and step.
Definition: Window.h:77
void run(const Window &window, const ThreadInfo &info) override
Execute the kernel on the passed window.
Status class.
Definition: Error.h:52
#define ARM_COMPUTE_RETURN_ERROR_ON(cond)
If the condition is true, an error is returned.
Definition: Error.h:296
Interface for Neon tensor.
Definition: ITensor.h:36
Copyright (c) 2017-2021 Arm Limited.
1 channel, 1 S32 per channel
static constexpr size_t DimX
Alias for dimension 0 also known as X dimension.
Definition: Window.h:43
#define ARM_COMPUTE_UNUSED(...)
To avoid unused variables warnings.
Definition: Error.h:152
Window collapse_if_possible(const Window &full_window, size_t first, size_t last, bool *has_collapsed=nullptr) const
Collapse the dimensions between first and last if possible.
Definition: Window.inl:68
quantized, asymmetric fixed-point 8-bit number unsigned
Coordinates of an item.
Definition: Coordinates.h:37
static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min=0, int max=0)
Static function to check if given info will lead to a valid configuration of NEGEMMLowpQuantizeDownIn...
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 std::unique_ptr< T > clone() const =0
Provide a clone of the current object of class T.
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor&#39;s metadata.
constexpr uint8_t * ptr() const
Return a pointer to the current pixel.
Definition: Helpers.inl:139
void set(size_t dimension, const Dimension &dim)
Set the values of a given dimension.
Definition: Window.inl:49
#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)
Definition: Validate.h:941
static constexpr size_t DimY
Alias for dimension 1 also known as Y dimension.
Definition: Window.h:45
ScaleKernelInfo info(interpolation_policy, default_border_mode, PixelValue(), sampling_policy, false)
Information about executing thread and CPU.
Definition: CPPTypes.h:235
#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES(...)
Definition: Validate.h:443
static constexpr size_t DimZ
Alias for dimension 2 also known as Z dimension.
Definition: Window.h:47
#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(...)
Definition: Validate.h:545
#define ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(t, c,...)
Definition: Validate.h:792
Status validate_arguments(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, const GEMMLowpOutputStageInfo *output_stage)
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
Definition: Validate.h:161
void execute_window_loop(const Window &w, L &&lambda_function, Ts &&... iterators)
Iterate through the passed window, automatically adjusting the iterators and calling the lambda_funct...
Definition: Helpers.inl:77
constexpr int end() const
Return the end of the dimension.
Definition: Window.h:99
Iterator updated by execute_window_loop for each window element.
Definition: Helpers.h:46
constexpr int start() const
Return the start of the dimension.
Definition: Window.h:94
Describe a multidimensional execution window.
Definition: Window.h:39
wrapper::traits::neon_vector< T, 16 >::type finalize_quantization(int32x4x4_t &in_s32, int32x4_t result_shift_s32, typename wrapper::traits::neon_vector< T, 16 >::type min, typename wrapper::traits::neon_vector< T, 16 >::type max)
#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)
Definition: Validate.h:205
constexpr const Dimension & x() const
Alias to access the first dimension of the window.
Definition: Window.h:145