From f01201abec0a102f6e7a517971f83fef1eaffd50 Mon Sep 17 00:00:00 2001 From: Giuseppe Rossini Date: Wed, 6 Nov 2019 14:57:49 +0000 Subject: COMPMID-2305: NEDepthwiseConvolution 3x3: support for QUANT8_PER_CHANNEL_SYMM Change-Id: I9a917cff6a089ce6ae16fb4e6066a4194e2e9487 Signed-off-by: Giuseppe Rossini Reviewed-on: https://review.mlplatform.org/c/2241 Tested-by: Arm Jenkins Reviewed-by: Michele Di Giorgio Comments-Addressed: Arm Jenkins Reviewed-by: Pablo Marquez --- .../depthwise/impl_qa8_qs8_per_channel.hpp | 457 +++++++++++++++++++++ 1 file changed, 457 insertions(+) create mode 100644 src/core/NEON/kernels/convolution/depthwise/impl_qa8_qs8_per_channel.hpp (limited to 'src/core/NEON/kernels/convolution/depthwise/impl_qa8_qs8_per_channel.hpp') diff --git a/src/core/NEON/kernels/convolution/depthwise/impl_qa8_qs8_per_channel.hpp b/src/core/NEON/kernels/convolution/depthwise/impl_qa8_qs8_per_channel.hpp new file mode 100644 index 0000000000..b27430c242 --- /dev/null +++ b/src/core/NEON/kernels/convolution/depthwise/impl_qa8_qs8_per_channel.hpp @@ -0,0 +1,457 @@ +/* + * Copyright (c) 2019 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. + */ + +/* + * !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + * + * NOTE: Header to be included by implementation files only. + * + * !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! + */ + +#include + +#include "arm.hpp" +#include "impl_base.hpp" +#include "depthwise_quantized.hpp" + +#pragma once + +namespace { + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols, + typename FInput, typename FOutput +> +static inline void tilefn_hybrid( + int n_channels, + const void* packed_params, + FInput &get_input_ptr, + FOutput &get_output_ptr, + int32_t clamp_min, + int32_t clamp_max, + uint8_t input_offset, + uint8_t output_offset +) +{ + constexpr int InnerTileRows = StrideRows * (OutputTileRows - 1) + KernelRows; + constexpr int InnerTileCols = StrideCols * (OutputTileCols - 1) + KernelCols; + + // Offset into channels + int channel = 0; + + // Byte type pointer to weights and biases + const int8_t *wbptr = static_cast(packed_params); + + for (; n_channels >= 8; n_channels -= 8, channel += 8) + { + const int32x4_t biases[2] = { + vld1q_s32(reinterpret_cast(wbptr)), + vld1q_s32(reinterpret_cast(wbptr) + 4), + }; + const int32x4_t multipliers[2] = { + vld1q_s32(reinterpret_cast(wbptr) + 8), + vld1q_s32(reinterpret_cast(wbptr) + 12), + }; + const int32x4_t shifts[2] = { + vld1q_s32(reinterpret_cast(wbptr) + 16), + vld1q_s32(reinterpret_cast(wbptr) + 20), + }; + wbptr += 24*sizeof(int32_t); + + int16x8_t weights[KernelRows][KernelCols]; + for (unsigned int i = 0; i < KernelRows; i++) + { + for (unsigned int j = 0; j < KernelCols; j++) + { + const auto w = vld1_s8(wbptr); + weights[i][j] = reinterpret_cast(vmovl_s8(w)); + wbptr += 8; + } + } + + int16x8_t inputs[InnerTileRows][InnerTileCols]; + const uint8x8_t ioffset = vdup_n_u8(input_offset); + for (unsigned int i = 0; i < InnerTileRows; i++) + { + for (unsigned int j = 0; j < InnerTileCols; j++) + { + const auto x = vld1_u8(get_input_ptr(i, j, channel)); + inputs[i][j] = reinterpret_cast(vsubl_u8(x, ioffset)); + } + } + + for (unsigned int oi = 0; oi < OutputTileRows; oi++) + { + for (unsigned int oj = 0; oj < OutputTileCols; oj++) + { + int32x4_t accs[2]; + for (unsigned int i = 0; i < 2; i++) + { + accs[i] = biases[i]; + } + + for (unsigned int wi = 0; wi < KernelRows; wi++) + { + for (unsigned int wj = 0; wj < KernelCols; wj++) + { + const auto w = weights[wi][wj]; + const auto x = inputs[oi * StrideRows + wi][oj * StrideCols + wj]; + accs[0] = vmlal_s16(accs[0], vget_low_s16(w), vget_low_s16(x)); + accs[1] = vmlal_s16(accs[1], vget_high_s16(w), vget_high_s16(x)); + } + } + + int32x4_t final_accs[2]; + for (unsigned int i = 0; i < 2; i++) + { + const int32x4_t y = rounding_divide_by_exp2( + saturating_doubling_high_mul(accs[i], multipliers[i]), + shifts[i]); + const int32x4_t offset = reinterpret_cast(vdupq_n_u32(output_offset)); + final_accs[i] = vaddq_s32(y, offset); + final_accs[i] = vmaxq_s32(final_accs[i], vdupq_n_s32(clamp_min)); + final_accs[i] = vminq_s32(final_accs[i], vdupq_n_s32(clamp_max)); + } + + const auto elems_s16 = vuzpq_s16(vreinterpretq_s16_s32(final_accs[0]), + vreinterpretq_s16_s32(final_accs[1])); + const int8x16_t elems = vreinterpretq_s8_s16(elems_s16.val[0]); + const uint8x8_t output = + vget_low_u8(vreinterpretq_u8_s8(vuzpq_s8(elems, elems).val[0])); + + vst1_u8(get_output_ptr(oi, oj, channel), output); + } + } + } + + for (; n_channels; n_channels--, channel++) + { + // Load bias + const int32_t bias = *reinterpret_cast(wbptr); + const int32_t multiplier = *reinterpret_cast(wbptr + sizeof(int32_t)); + const int32_t shift = *reinterpret_cast(wbptr + 2*sizeof(int32_t)); + + wbptr += 3*sizeof(int32_t); + + // Load weights + int16_t weights[KernelRows][KernelCols]; + for (unsigned int i = 0; i < KernelRows; i++) + { + for (unsigned int j = 0; j < KernelCols; j++) + { + weights[i][j] = *(wbptr++); + } + } + + // Load the input activations + int16_t inputs[InnerTileRows][InnerTileCols]; + for (unsigned int i = 0; i < InnerTileRows; i++) + { + for (unsigned int j = 0; j < InnerTileCols; j++) + { + inputs[i][j] = *(get_input_ptr(i, j, channel)) - input_offset; + } + } + + // Perform the convolution + for (unsigned int oi = 0; oi < OutputTileRows; oi++) + { + for (unsigned int oj = 0; oj < OutputTileCols; oj++) + { + int32_t acc = bias; + + for (unsigned int wi = 0; wi < KernelRows; wi++) + { + for (unsigned int wj = 0; wj < KernelCols; wj++) + { + const auto w = weights[wi][wj], x = inputs[oi*StrideRows + wi][oj*StrideCols + wj]; + acc += w * x; + } + } + + // Requantize + acc = rounding_divide_by_exp2( + saturating_doubling_high_mul(acc, multiplier), + -shift); + acc += output_offset; + acc = std::max(acc, clamp_min); + acc = std::min(acc, clamp_max); + uint8_t output = static_cast(acc); + *(get_output_ptr(oi, oj, channel)) = output; + } + } + } +} + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols, + typename FInput, typename FOutput +> +static inline void execute_tilefn_hybrid( + int n_channels, + const void* packed_params, + const ActivationFunction actfn, + const qasymm8::QAsymm8Params &input_quant, + const qasymm8::QAsymm8Params &output_quant, + FInput &get_input_ptr, + FOutput &get_output_ptr) { + + // Compute min/max clamp values + int32_t clamp_min = std::numeric_limits::min(); + int32_t clamp_max = std::numeric_limits::max(); + + if (actfn == ActivationFunction::ReLU) { + clamp_min = output_quant.offset; + } + + // Disabling Relu6 for now + if (actfn == ActivationFunction::ReLU6) { + const int32_t top_rail = output_quant.quantize(6.0f); + clamp_max = std::min(clamp_max, top_rail); + } + + // Call the tile execution method + tilefn_hybrid(n_channels, packed_params, get_input_ptr, get_output_ptr, clamp_min, clamp_max, input_quant.offset, output_quant.offset); +} +} + + + +namespace depthwise { +using namespace qsymm8; +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols +> +QSymm8HybridPerChannelDepthwiseConvolution< + OutputTileRows, OutputTileCols, KernelRows, KernelCols, StrideRows, StrideCols +>::QSymm8HybridPerChannelDepthwiseConvolution( + int n_batches, int n_input_rows, int n_input_cols, int n_channels, + const ActivationFunction activation, + const QSymm8PerChannelParams& weight_quantisation, + const qasymm8::QAsymm8Params& input_quantisation, + const qasymm8::QAsymm8Params& output_quantisation, + unsigned int padding_top, + unsigned int padding_left, + unsigned int padding_bottom, + unsigned int padding_right +) : QSymm8HybridPerChannelDepthwiseConvolution( + n_batches, n_input_rows, n_input_cols, n_channels, + activation, weight_quantisation, input_quantisation, output_quantisation, + QSymm8PerChannelRescaleParams::make_rescale_params(weight_quantisation, input_quantisation, output_quantisation), + padding_top, padding_left, padding_bottom, padding_right + ) +{ +} + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols +> +QSymm8HybridPerChannelDepthwiseConvolution< + OutputTileRows, OutputTileCols, KernelRows, KernelCols, StrideRows, StrideCols +>::QSymm8HybridPerChannelDepthwiseConvolution( + int n_batches, int n_input_rows, int n_input_cols, int n_channels, + const ActivationFunction activation, + const QSymm8PerChannelParams& weight_quantisation, + const qasymm8::QAsymm8Params& input_quantisation, + const qasymm8::QAsymm8Params& output_quantisation, + const QSymm8PerChannelRescaleParams& rescale_params, + unsigned int padding_top, + unsigned int padding_left, + unsigned int padding_bottom, + unsigned int padding_right +) : Base( + n_batches, n_input_rows, n_input_cols, n_channels, activation, + padding_top, padding_left, padding_bottom, padding_right + ), + _weights_quant(weight_quantisation), + _input_quant(input_quantisation), + _output_quant(output_quantisation), + _rescale_parameters(rescale_params) +{ +} + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols +> +uint8_t QSymm8HybridPerChannelDepthwiseConvolution< + OutputTileRows, OutputTileCols, KernelRows, KernelCols, StrideRows, StrideCols +>::_input_padding_value(void) const +{ + return _input_quant.offset; +} + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols +> +void QSymm8HybridPerChannelDepthwiseConvolution< + OutputTileRows, OutputTileCols, KernelRows, KernelCols, StrideRows, StrideCols +>::_pack_params( + void * const buffer, + const void * const weights, + const unsigned int weight_row_stride, + const unsigned int weight_col_stride, + const void * const biases +) const +{ + const int8_t *wptr = static_cast(weights); + const int32_t *bptr = static_cast(biases); + const int32_t *mptr = static_cast(_rescale_parameters.multipliers.data()); + const int32_t *sptr = static_cast(_rescale_parameters.shifts.data()); + int8_t *outptr = static_cast(buffer); + + // We set the vector length to use doubles on both Aarch64 and Aarch32. NOTE + // For SVE set this to half the vector length. + unsigned int veclen = 8; + + // While there are channels left to process, pack a vector length of them at + // a time and reduce the size of vector used as the size of the tensor + // decreases. + for ( + unsigned int n_channels = this->n_channels(); n_channels; + n_channels -= veclen, + outptr += veclen*(3*sizeof(int32_t) + this->kernel_rows*this->kernel_cols) + ) + { + // NOTE Ignore this section if using SVE, the vector length remains the + // same and we just don't fill a full register for the tail. + while (n_channels < veclen) + { + // Reduce the vector length to either 8 or 1 (scalar) + // TODO Support more vector lengths in `execute_tile`. + veclen = (veclen == 16) ? 8 : 1; + } + + // Get pointers to bias and weight portions of the output structure. + int32_t *out_bptr = reinterpret_cast(outptr); + int32_t *out_mptr = reinterpret_cast(outptr + veclen*sizeof(int32_t)); + int32_t *out_sptr = reinterpret_cast(outptr + 2*veclen*sizeof(int32_t)); + int8_t *out_wptr = outptr + 3*veclen*sizeof(int32_t); + + // Copy a vector length of elements + for (unsigned int n = 0; n < veclen && n < n_channels; n++) + { + const int32_t bias = (bptr != nullptr) ? *(bptr++) : 0; + const int32_t multiplier = (mptr != nullptr) ? *(mptr++) : 0; + const int32_t shift = (sptr != nullptr) ? *(sptr++) : 0; + + out_bptr[n] = bias; + out_mptr[n] = multiplier; + out_sptr[n] = -shift; + + for (unsigned int i = 0; i < KernelRows; i++) + { + int8_t *row_outptr = out_wptr + i*KernelCols*veclen; + for (unsigned int j = 0; j < KernelCols; j++) + { + int8_t w = *(wptr + i*weight_row_stride + j*weight_col_stride); + row_outptr[j*veclen + n] = w; + } + } + wptr++; + } + } +} + + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols +> +template +void QSymm8HybridPerChannelDepthwiseConvolution< + OutputTileRows, OutputTileCols, KernelRows, KernelCols, StrideRows, StrideCols +>::execute_tile( + int n_channels, + const void* packed_params, + const uint8_t* inptr, + unsigned int in_row_stride, + unsigned int in_col_stride, + uint8_t* outptr, + unsigned int out_row_stride, + unsigned int out_col_stride +) { + + // Construct methods to get pointers + const auto get_input_ptr = [inptr, in_row_stride, in_col_stride]( + const int i, const int j, const int channel) { + return inptr + i * in_row_stride + j * in_col_stride + channel; + }; + + const auto get_output_ptr = [outptr, out_row_stride, out_col_stride]( + const int i, const int j, const int channel) { + return outptr + i * out_row_stride + j * out_col_stride + channel; + }; + + execute_tilefn_hybrid( + n_channels, packed_params, Activation, _input_quant, _output_quant, get_input_ptr, get_output_ptr); +} + +template < + unsigned int OutputTileRows, unsigned int OutputTileCols, + unsigned int KernelRows, unsigned int KernelCols, + unsigned int StrideRows, unsigned int StrideCols +> +template +void QSymm8HybridPerChannelDepthwiseConvolution< + OutputTileRows, OutputTileCols, KernelRows, KernelCols, StrideRows, StrideCols +>::execute_tile( + int n_channels, + const void* packed_params, + const uint8_t* inptrs[Base::inner_tile_rows][Base::inner_tile_cols], + uint8_t* outptrs[Base::output_tile_rows][Base::output_tile_cols] +) { + // Construct methods to get pointers + const auto get_input_ptr = [inptrs](const int i, const int j, + const int channel) { + return inptrs[i][j] + channel; + }; + + const auto get_output_ptr = [outptrs](const int i, const int j, + const int channel) { + return outptrs[i][j] + channel; + }; + + // Call the tile execution method + execute_tilefn_hybrid( + n_channels, packed_params, Activation, _input_quant, _output_quant, get_input_ptr, get_output_ptr); +} + +} // namespace depthwise -- cgit v1.2.1