1 /*
2 * Copyright (c) 2020-2022 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 */
24 #include "src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleKernel.h"
25
26 #include "arm_compute/core/CL/CLHelpers.h"
27 #include "arm_compute/core/CL/ICLTensor.h"
28 #include "arm_compute/core/Helpers.h"
29 #include "arm_compute/core/Validate.h"
30 #include "arm_compute/core/utils/quantization/AsymmHelpers.h"
31
32 #include "src/core/helpers/AutoConfiguration.h"
33 #include "src/core/helpers/WindowHelpers.h"
34
35 #include "support/Cast.h"
36 #include "support/StringSupport.h"
37
38 namespace arm_compute
39 {
40 namespace opencl
41 {
42 namespace kernels
43 {
44 namespace
45 {
validate_arguments(const ITensorInfo * src,const ITensorInfo * bias,const ITensorInfo * dst,const GEMMLowpOutputStageInfo * output_stage)46 Status validate_arguments(const ITensorInfo *src, const ITensorInfo *bias, const ITensorInfo *dst, const GEMMLowpOutputStageInfo *output_stage)
47 {
48 ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(src, 1, DataType::S32);
49 ARM_COMPUTE_RETURN_ERROR_ON((output_stage->output_data_type != DataType::QASYMM8) && (output_stage->output_data_type != DataType::QASYMM8_SIGNED));
50 ARM_COMPUTE_RETURN_ERROR_ON(output_stage->gemmlowp_max_bound > std::get<1>(quantization::get_min_max_values_from_quantized_data_type(output_stage->output_data_type)));
51 ARM_COMPUTE_RETURN_ERROR_ON(output_stage->gemmlowp_min_bound < std::get<0>(quantization::get_min_max_values_from_quantized_data_type(output_stage->output_data_type))
52 || output_stage->gemmlowp_min_bound > output_stage->gemmlowp_max_bound);
53
54 // Check biases if exist
55 if(bias != nullptr)
56 {
57 ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(src, bias);
58 ARM_COMPUTE_RETURN_ERROR_ON(bias->num_dimensions() > 1);
59 ARM_COMPUTE_RETURN_ERROR_ON(src->dimension(0) != bias->dimension(0));
60 }
61
62 if(dst->total_size() != 0)
63 {
64 ARM_COMPUTE_RETURN_ERROR_ON_MSG(dst->data_type() != output_stage->output_data_type, "Mismatching output data type");
65 ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES(src, dst);
66 }
67
68 return Status{};
69 }
70 } //namespace
71
ClGemmLowpQuantizeDownInt32ScaleKernel()72 ClGemmLowpQuantizeDownInt32ScaleKernel::ClGemmLowpQuantizeDownInt32ScaleKernel()
73 {
74 _type = CLKernelType::ELEMENTWISE;
75 }
76
validate(const ITensorInfo * src,const ITensorInfo * bias,const ITensorInfo * dst,const GEMMLowpOutputStageInfo * output_stage)77 Status ClGemmLowpQuantizeDownInt32ScaleKernel::validate(const ITensorInfo *src, const ITensorInfo *bias, const ITensorInfo *dst, const GEMMLowpOutputStageInfo *output_stage)
78 {
79 ARM_COMPUTE_ERROR_ON_NULLPTR(src, dst);
80 ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(src, bias, dst, output_stage));
81
82 return Status{};
83 }
84
configure(const CLCompileContext & compile_context,const ITensorInfo * src,const ITensorInfo * bias,ITensorInfo * dst,const GEMMLowpOutputStageInfo * output_stage)85 void ClGemmLowpQuantizeDownInt32ScaleKernel::configure(const CLCompileContext &compile_context, const ITensorInfo *src, const ITensorInfo *bias, ITensorInfo *dst,
86 const GEMMLowpOutputStageInfo *output_stage)
87 {
88 // Perform validate step
89 ARM_COMPUTE_ERROR_ON_NULLPTR(src, dst);
90 ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(src, bias, dst, output_stage));
91
92 auto padding_info = get_padding_info({ src, bias, dst });
93
94 // Output auto inizialitation if not yet initialized
95 auto_init_if_empty(*dst, src->clone()->set_data_type(output_stage->output_data_type));
96
97 const unsigned int num_elems_processed_per_iteration = adjust_vec_size(4, src->dimension(0));
98
99 // Set the arguments to pass at compile time
100 auto min = output_stage->gemmlowp_min_bound;
101 auto max = output_stage->gemmlowp_max_bound;
102 CLBuildOptions build_opts;
103 build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(num_elems_processed_per_iteration));
104 build_opts.add_option("-DVEC_SIZE_LEFTOVER=" + support::cpp11::to_string(src->dimension(0) % num_elems_processed_per_iteration));
105 build_opts.add_option("-DRESULT_OFFSET=" + support::cpp11::to_string(output_stage->gemmlowp_offset));
106 build_opts.add_option("-DRESULT_MULT_INT=" + support::cpp11::to_string(output_stage->gemmlowp_multiplier));
107 build_opts.add_option("-DRESULT_SHIFT=" + support::cpp11::to_string(output_stage->gemmlowp_shift));
108 build_opts.add_option_if((min > std::get<0>(quantization::get_min_max_values_from_quantized_data_type(output_stage->output_data_type))) && (min != max),
109 "-DMIN_BOUND=" + support::cpp11::to_string(min));
110 build_opts.add_option_if((max < std::get<1>(quantization::get_min_max_values_from_quantized_data_type(output_stage->output_data_type))) && (min != max),
111 "-DMAX_BOUND=" + support::cpp11::to_string(max));
112 build_opts.add_option("-DOUTPUT_DATA_TYPE=" + get_cl_type_from_data_type(dst->data_type()));
113 build_opts.add_option_if(bias != nullptr, "-DADD_BIAS");
114
115 const std::string kernel_name = "gemmlowp_output_stage_quantize_down";
116
117 // A macro guard to compile ONLY the kernel of interest
118 build_opts.add_option("-D" + upper_string(kernel_name));
119
120 // Create kernel
121 _kernel = create_kernel(compile_context, kernel_name, build_opts.options());
122
123 // Configure kernel window
124 Window win = calculate_max_window(*src, Steps(num_elems_processed_per_iteration));
125 ICLKernel::configure_internal(win);
126
127 ARM_COMPUTE_ERROR_ON(has_padding_changed(padding_info));
128 }
129
run_op(ITensorPack & tensors,const Window & window,cl::CommandQueue & queue)130 void ClGemmLowpQuantizeDownInt32ScaleKernel::run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue)
131 {
132 ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
133 ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
134
135 const auto src = utils::cast::polymorphic_downcast<const ICLTensor *>(tensors.get_const_tensor(TensorType::ACL_SRC));
136 const auto bias = utils::cast::polymorphic_downcast<const ICLTensor *>(tensors.get_const_tensor(TensorType::ACL_BIAS));
137 auto dst = utils::cast::polymorphic_downcast<ICLTensor *>(tensors.get_tensor(TensorType::ACL_DST));
138
139 Window collapsed = window.collapse_if_possible(ICLKernel::window(), Window::DimZ);
140 Window slice = collapsed.first_slice_window_3D();
141
142 unsigned int idx1 = num_arguments_per_3D_tensor();
143 if(bias != nullptr)
144 {
145 Window biases_slice(slice);
146 biases_slice.set(Window::DimY, Window::Dimension(0, 1, 1));
147 biases_slice.set(Window::DimZ, Window::Dimension(0, 1, 1));
148 add_1D_tensor_argument(idx1, bias, biases_slice);
149 }
150
151 do
152 {
153 unsigned int idx = 0;
154 add_3D_tensor_argument(idx, src, slice);
155 add_3D_tensor_argument(idx1, dst, slice);
156 enqueue(queue, *this, slice, lws_hint());
157 }
158 while(collapsed.slide_window_slice_3D(slice));
159 }
160 } // namespace kernels
161 } // namespace opencl
162 } // namespace arm_compute