/* * Copyright (c) 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. */ #ifndef ARM_COMPUTE_TEST_FIXED_POINT_NEON_TARGET #define ARM_COMPUTE_TEST_FIXED_POINT_NEON_TARGET #include "arm_compute/core/NEON/NEFixedPoint.h" #include "tests/Globals.h" #include "tests/Types.h" namespace arm_compute { namespace test { namespace validation { namespace { template void compute_target_impl(const TensorShape &shape, DataType dt, FixedPointOp op, int fixed_point_position, TensorType &src, TensorType &dst) { Window window; switch(dt) { case DataType::QS8: { constexpr unsigned int num_elems_processed_per_iteration = 16; window = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration)); AccessWindowHorizontal input_access(src.info(), 0, num_elems_processed_per_iteration); AccessWindowHorizontal output_access(dst.info(), 0, num_elems_processed_per_iteration); break; } case DataType::QS16: { constexpr unsigned int num_elems_processed_per_iteration = 8; window = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration)); AccessWindowHorizontal input_access(src.info(), 0, num_elems_processed_per_iteration); AccessWindowHorizontal output_access(dst.info(), 0, num_elems_processed_per_iteration); break; } default: ARM_COMPUTE_ERROR("Not Supported"); break; } int min; int max; switch(op) { case FixedPointOp::EXP: { // Fill tensors. Keep the range between [-1.0, 1.0) so the result won't // overflow. min = -(1 << (fixed_point_position - 1)); max = (1 << (fixed_point_position - 1)); break; } case FixedPointOp::INV_SQRT: { if(dt == DataType::QS8) { // Fill tensors. Keep the range between [1, 127). min = 1; max = 127; } else { // Fill tensors. Keep the range between [1, 0x7FFF) min = 1; max = 0x7FFF; } break; } case FixedPointOp::LOG: { if(dt == DataType::QS8) { // Fill tensors. Keep the range between [(1 << (fixed_point_position - 1), 63) so the result won't // overflow. E.g. for Q2.5 ln(0.001) = -6.9, which cannot be represented. min = (1 << (fixed_point_position - 1)); max = 0x3F; } else { // Fill tensors. Keep the range between [(1 << (fixed_point_position - 1), 0x3FFF) so the result won't // overflow. min = (1 << (fixed_point_position - 1)); max = 0x3FFF; } break; } case FixedPointOp::RECIPROCAL: { if(dt == DataType::QS8) { // Fill tensors. Keep the range between [15, 100) so the result won't // overflow. E.g. for Q2.5 reciprocal(0.001) = 1000, which cannot be represented. min = 15; max = 0x7F; } else { // Fill tensors. Keep the range between [15, 0x7FFF) so the result won't // overflow. min = 15; max = 0x7FFF; } break; } default: ARM_COMPUTE_ERROR("Not Supported"); break; } std::uniform_int_distribution<> distribution(min, max); library->fill(AccessorType(src), distribution, 0); Iterator input(&src, window); Iterator output(&dst, window); const auto loop_function = [&](const Coordinates & id) { switch(dt) { case DataType::QS8: { const qint8x16_t qs8in = vld1q_s8(reinterpret_cast(input.ptr())); switch(op) { case FixedPointOp::EXP: { // Use saturated exp vst1q_s8(reinterpret_cast(output.ptr()), vqexpq_qs8(qs8in, fixed_point_position)); break; } case FixedPointOp::INV_SQRT: { vst1q_s8(reinterpret_cast(output.ptr()), vqinvsqrtq_qs8(qs8in, fixed_point_position)); break; } case FixedPointOp::LOG: { vst1q_s8(reinterpret_cast(output.ptr()), vlogq_qs8(qs8in, fixed_point_position)); break; } case FixedPointOp::RECIPROCAL: { vst1q_s8(reinterpret_cast(output.ptr()), vrecipq_qs8(qs8in, fixed_point_position)); break; } default: ARM_COMPUTE_ERROR("Not Supported"); break; } break; } case DataType::QS16: { const qint16x8_t qs16in = vld1q_qs16(reinterpret_cast(input.ptr())); switch(op) { case FixedPointOp::EXP: { // Use saturated exp vst1q_qs16(reinterpret_cast(output.ptr()), vqexpq_qs16(qs16in, fixed_point_position)); break; } case FixedPointOp::INV_SQRT: { vst1q_qs16(reinterpret_cast(output.ptr()), vqinvsqrtq_qs16(qs16in, fixed_point_position)); break; } case FixedPointOp::LOG: { vst1q_qs16(reinterpret_cast(output.ptr()), vlogq_qs16(qs16in, fixed_point_position)); break; } case FixedPointOp::RECIPROCAL: { vst1q_qs16(reinterpret_cast(output.ptr()), vqrecipq_qs16(qs16in, fixed_point_position)); break; } default: ARM_COMPUTE_ERROR("Not Supported"); break; } break; } default: ARM_COMPUTE_ERROR("Not Supported"); break; } }; execute_window_loop(window, loop_function, input, output); } } // namespace } // namespace validation } // namespace test } // namespace arm_compute #endif /* ARM_COMPUTE_TEST_FIXED_POINT_NEON_TARGET */