utils
/
stm32f407-openocd
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
stm32f407-openocd/Drivers/CMSIS/NN/Include/arm_nnsupportfunctions.h

1187 lines
49 KiB
C
Raw Permalink Normal View History

first
2024-06-12 08:32:58 +00:00
/*
* Copyright (C) 2010-2022 Arm Limited or its affiliates.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
/* ----------------------------------------------------------------------
* Project: CMSIS NN Library
* Title: arm_nnsupportfunctions.h
* Description: Public header file of support functions for CMSIS NN Library
*
* $Date: 19. April 2022
* $Revision: V.7.0.1
*
* Target Processor: Cortex-M CPUs
* -------------------------------------------------------------------- */
#ifndef _ARM_NNSUPPORTFUNCTIONS_H_
#define _ARM_NNSUPPORTFUNCTIONS_H_
#include "arm_nn_math_types.h"
#include "arm_nn_types.h"
#include <stdbool.h>
#ifdef __cplusplus
extern "C" {
#endif
#define LEFT_SHIFT(_shift) (_shift > 0 ? _shift : 0)
#define RIGHT_SHIFT(_shift) (_shift > 0 ? 0 : -_shift)
#define MASK_IF_ZERO(x) (x) == 0 ? ~0 : 0
#define MASK_IF_NON_ZERO(x) (x) != 0 ? ~0 : 0
#define SELECT_USING_MASK(mask, a, b) ((mask) & (a)) ^ (~(mask) & (b))
#define MAX(A, B) ((A) > (B) ? (A) : (B))
#define MIN(A, B) ((A) < (B) ? (A) : (B))
#define CLAMP(x, h, l) MAX(MIN((x), (h)), (l))
#define REDUCE_MULTIPLIER(_mult) ((_mult < 0x7FFF0000) ? ((_mult + (1 << 15)) >> 16) : 0x7FFF)
/**
* @brief definition to pack four 8 bit values.
*/
#define PACK_Q7x4_32x1(v0, v1, v2, v3) \
((((int32_t)(v0) << 0) & (int32_t)0x000000FF) | (((int32_t)(v1) << 8) & (int32_t)0x0000FF00) | \
(((int32_t)(v2) << 16) & (int32_t)0x00FF0000) | (((int32_t)(v3) << 24) & (int32_t)0xFF000000))
/**
* @brief Union for SIMD access of q31/q15/q7 types
*/
union arm_nnword
{
q31_t word;
/**< q31 type */
q15_t half_words[2];
/**< q15 type */
q7_t bytes[4];
/**< q7 type */
};
/**
* @brief Union for data type long long
*/
struct arm_nn_double
{
uint32_t low;
int32_t high;
};
union arm_nn_long_long
{
int64_t long_long;
struct arm_nn_double word;
};
/**
* @defgroup nndata_convert Neural Network Data Conversion Functions
*
* Perform data type conversion in-between neural network operations
*
*/
/**
* @brief Converts the elements of the q7 vector to q15 vector without left-shift
* @param[in] *pSrc points to the q7 input vector
* @param[out] *pDst points to the q15 output vector
* @param[in] blockSize length of the input vector
*
*/
void arm_q7_to_q15_no_shift(const q7_t *pSrc, q15_t *pDst, uint32_t blockSize);
/**
* @brief Non-saturating addition of elements of a q7 vector
* @param[in] *input Pointer to the q7 input vector
* @param[out] *output Pointer to the q31 output variable.
* @param[in] block_size length of the input vector
* \par Description:
*
* 2^24 samples can be added without saturating the result.
*
* The equation used for the conversion process is:
*
* <pre>
* sum = input[0] + input[1] + .. + input[block_size -1]
* </pre>
*
* */
void arm_nn_add_q7(const q7_t *input, q31_t *output, uint32_t block_size);
/**
* @brief Converts the elements of the q7 vector to reordered q15 vector without left-shift
* @param[in] *pSrc points to the q7 input vector
* @param[out] *pDst points to the q15 output vector
* @param[in] blockSize length of the input vector
* @return none.
*
*/
void arm_q7_to_q15_reordered_no_shift(const q7_t *pSrc, q15_t *pDst, uint32_t blockSize);
/**
* @brief Converts the elements from a q7 vector to a q15 vector with an added offset
* @param[in] src pointer to the q7 input vector
* @param[out] dst pointer to the q15 output vector
* @param[in] block_size length of the input vector
* @param[in] offset q7 offset to be added to each input vector element.
*
* \par Description:
*
* The equation used for the conversion process is:
*
* <pre>
* dst[n] = (q15_t) src[n] + offset; 0 <= n < block_size.
* </pre>
*
*/
void arm_q7_to_q15_with_offset(const q7_t *src, q15_t *dst, uint32_t block_size, q15_t offset);
/**
* @brief Converts the elements of the q7 vector to reordered q15 vector with an added offset
* @param[in] src pointer to the q7 input vector
* @param[out] dst pointer to the q15 output vector
* @param[in] block_size length of the input vector
* @param[in] offset offset to be added to each input vector element.
* @return none.
*
* @details This function does the q7 to q15 expansion with re-ordering of bytes. Re-ordering is a consequence of
* the sign extension intrinsic(DSP extension). The tail (i.e., last (N % 4) elements) retains its
* original order.
*
*/
void arm_q7_to_q15_reordered_with_offset(const q7_t *src, q15_t *dst, uint32_t block_size, q15_t offset);
/**
* @brief Converts the elements from a q7 vector and accumulate to a q15 vector
* @param[in] *src points to the q7 input vector
* @param[out] *dst points to the q15 output vector
* @param[in] block_size length of the input vector
*
* \par Description:
*
* The equation used for the conversion process is:
*
* <pre>
* dst[n] += (q15_t) src[n] ; 0 <= n < block_size.
* </pre>
*
*/
void arm_nn_accumulate_q7_to_q15(q15_t *dst, const q7_t *src, uint32_t block_size);
/**
* @brief Depthwise conv on an im2col buffer where the input channel equals output channel.
* @param[in] row pointer to row
* @param[in] col pointer to im2col buffer, always consists of 2 columns.
* @param[in] num_ch number of channels
* @param[in] out_shift pointer to per output channel requantization shift parameter.
* @param[in] out_mult pointer to per output channel requantization multiplier parameter.
* @param[in] out_offset output tensor offset.
* @param[in] activation_min minimum value to clamp the output to. Range : int8
* @param[in] activation_max maximum value to clamp the output to. Range : int8
* @param[in] kernel_size number of elements in one column.
* @param[in] output_bias per output channel bias. Range : int32
* @param[out] out pointer to output
* @return The function returns one of the two
* 1. The incremented output pointer for a successful operation or
* 2. NULL if implementation is not available.
*
* @details Supported framework: TensorFlow Lite micro.
*/
q7_t *arm_nn_depthwise_conv_s8_core(const q7_t *row,
const q15_t *col,
const uint16_t num_ch,
const int32_t *out_shift,
const int32_t *out_mult,
const int32_t out_offset,
const int32_t activation_min,
const int32_t activation_max,
const uint16_t kernel_size,
const int32_t *const output_bias,
q7_t *out);
/**
* @brief General Matrix-multiplication function with per-channel requantization.
* @param[in] input_row pointer to row operand
* @param[in] input_col pointer to col operand
* @param[in] output_ch number of rows of input_row
* @param[in] col_batches number of column batches. Range: 1 to 4
* @param[in] output_shift pointer to per output channel requantization shift parameter.
* @param[in] output_mult pointer to per output channel requantization multiplier parameter.
* @param[in] out_offset output tensor offset.
* @param[in] col_offset input tensor(col) offset.
* @param[in] row_offset kernel offset(row). Not used.
* @param[in] out_activation_min minimum value to clamp the output to. Range : int8
* @param[in] out_activation_max maximum value to clamp the output to. Range : int8
* @param[in] row_len number of elements in each row
* @param[in] bias per output channel bias. Range : int32
* @param[in,out] out pointer to output
* @return The function returns one of the two
* 1. The incremented output pointer for a successful operation or
* 2. NULL if implementation is not available.
*
* @details Supported framework: TensorFlow Lite
*/
q7_t *arm_nn_mat_mult_s8(const q7_t *input_row,
const q7_t *input_col,
const uint16_t output_ch,
const uint16_t col_batches,
const int32_t *output_shift,
const int32_t *output_mult,
const int32_t out_offset,
const int32_t col_offset,
const int32_t row_offset,
const int16_t out_activation_min,
const int16_t out_activation_max,
const uint16_t row_len,
const int32_t *const bias,
q7_t *out);
/**
* @brief Matrix-multiplication function for convolution with per-channel requantization for 16 bits convolution.
* @param[in] input_a pointer to operand A
* @param[in] input_b pointer to operand B, always consists of 2 vectors.
* @param[in] output_ch number of rows of A
* @param[in] out_shift pointer to per output channel requantization shift parameter.
* @param[in] out_mult pointer to per output channel requantization multiplier parameter.
* @param[in] activation_min minimum value to clamp the output to. Range : int16
* @param[in] activation_max maximum value to clamp the output to. Range : int16
* @param[in] num_col_a number of columns of A
* @param[in] output_bias per output channel bias. Range : int64
* @param[in,out] out_0 pointer to output
* @return The function returns one of the two
* 1. The incremented output pointer for a successful operation or
* 2. NULL if implementation is not available.
*
* @details This function does the matrix multiplication of weight matrix for all output channels
* with 2 columns from im2col and produces two elements/output_channel. The outputs are
* clamped in the range provided by activation min and max.
* Supported framework: TensorFlow Lite micro.
*/
q15_t *arm_nn_mat_mult_kernel_s16(const q7_t *input_a,
const q15_t *input_b,
const int32_t output_ch,
const int32_t *out_shift,
const int32_t *out_mult,
const int16_t activation_min,
const int16_t activation_max,
const int32_t num_col_a,
const int64_t *const output_bias,
q15_t *out_0);
/**
* @brief General Matrix-multiplication without requantization for one row & one column
* @param[in] row_elements number of row elements
* @param[in] row_base pointer to row operand
* @param[in] col_base pointer to col operand
* @param[out] sum_col pointer to store sum of column elements
* @param[out] output pointer to store result of multiply-accumulate
* @return The function returns the multiply-accumulated result of the row by column.
*
* @details Pseudo-code
* *output = 0
* sum_col = 0
* for (i = 0; i < row_elements; i++)
* *output += row_base[i] * col_base[i]
* sum_col += col_base[i]
*
*/
arm_status arm_nn_mat_mul_core_1x_s8(int32_t row_elements,
const int8_t *row_base,
const int8_t *col_base,
int32_t *const sum_col,
int32_t *const output);
/**
* @brief Matrix-multiplication with requantization & activation function for four rows and one column
* @param[in] row_elements number of row elements
* @param[in] offset offset between rows. Can be the same as row_elements.
* For e.g, in a 1x1 conv scenario with stride as 1.
* @param[in] row_base pointer to row operand
* @param[in] col_base pointer to col operand
* @param[in] out_ch Number of output channels
* @param[in] conv_params Pointer to convolution parameters like offsets and activation values
* @param[in] quant_params Pointer to per-channel quantization parameters
* @param[in] bias Pointer to per-channel bias
* @param[out] output Pointer to output where int8 results are stored.
*
* @return The function returns the updated output pointer or NULL if implementation is not available.
*
* @details Compliant to TFLM int8 specification. MVE implementation only
*/
int8_t *arm_nn_mat_mul_core_4x_s8(const int32_t row_elements,
const int32_t offset,
const int8_t *row_base,
const int8_t *col_base,
const int32_t out_ch,
const cmsis_nn_conv_params *conv_params,
const cmsis_nn_per_channel_quant_params *quant_params,
const int32_t *bias,
int8_t *output);
/**
* @brief General Matrix-multiplication function with per-channel requantization.
* This function assumes:
* - LHS input matrix NOT transposed (nt)
* - RHS input matrix transposed (t)
*
* @note This operation also performs the broadcast bias addition before the requantization
*
* @param[in] lhs Pointer to the LHS input matrix
* @param[in] rhs Pointer to the RHS input matrix
* @param[in] bias Pointer to the bias vector. The length of this vector is equal to the number of
* output columns (or RHS input rows)
* @param[out] dst Pointer to the output matrix with "m" rows and "n" columns
* @param[in] dst_multipliers Pointer to the multipliers vector needed for the per-channel requantization.
* The length of this vector is equal to the number of output columns (or RHS input
* rows)
* @param[in] dst_shifts Pointer to the shifts vector needed for the per-channel requantization. The length
* of this vector is equal to the number of output columns (or RHS input rows)
* @param[in] lhs_rows Number of LHS input rows
* @param[in] rhs_rows Number of RHS input rows
* @param[in] rhs_cols Number of LHS/RHS input columns
* @param[in] lhs_offset Offset to be applied to the LHS input value
* @param[in] dst_offset Offset to be applied the output result
* @param[in] activation_min Minimum value to clamp down the output. Range : int8
* @param[in] activation_max Maximum value to clamp up the output. Range : int8
*
* @return The function returns <code>ARM_MATH_SUCCESS</code>
*
*/
arm_status arm_nn_mat_mult_nt_t_s8(const q7_t *lhs,
const q7_t *rhs,
const q31_t *bias,
q7_t *dst,
const int32_t *dst_multipliers,
const int32_t *dst_shifts,
const int32_t lhs_rows,
const int32_t rhs_rows,
const int32_t rhs_cols,
const int32_t lhs_offset,
const int32_t dst_offset,
const int32_t activation_min,
const int32_t activation_max);
/**
* @brief s8 Vector by Matrix (transposed) multiplication
*
* @param[in] lhs Input left-hand side vector
* @param[in] rhs Input right-hand side matrix (transposed)
* @param[in] bias Input bias
* @param[out] dst Output vector
* @param[in] lhs_offset Offset to be added to the input values of the left-hand side vector.
* Range: -127 to 128
* @param[in] rhs_offset Not used
* @param[in] dst_offset Offset to be added to the output values. Range: -127 to 128
* @param[in] dst_multiplier Output multiplier
* @param[in] dst_shift Output shift
* @param[in] rhs_cols Number of columns in the right-hand side input matrix
* @param[in] rhs_rows Number of rows in the right-hand side input matrix
* @param[in] activation_min Minimum value to clamp the output to. Range: int8
* @param[in] activation_max Maximum value to clamp the output to. Range: int8
* @param[in] address_offset Memory position offset for dst. First output is stored at 'dst', the
* second at 'dst + address_offset' and so on. Default value is typically 1.
*
* @return The function returns <code>ARM_MATH_SUCCESS</code>
*
*/
arm_status arm_nn_vec_mat_mult_t_s8(const q7_t *lhs,
const q7_t *rhs,
const q31_t *bias,
q7_t *dst,
const int32_t lhs_offset,
const int32_t rhs_offset,
const int32_t dst_offset,
const int32_t dst_multiplier,
const int32_t dst_shift,
const int32_t rhs_cols,
const int32_t rhs_rows,
const int32_t activation_min,
const int32_t activation_max,
const int32_t address_offset);
/**
* @brief s16 Vector by Matrix (transposed) multiplication
*
* @param[in] lhs Input left-hand side vector
* @param[in] rhs Input right-hand side matrix (transposed)
* @param[in] bias Input bias
* @param[out] dst Output vector
* @param[in] dst_multiplier Output multiplier
* @param[in] dst_shift Output shift
* @param[in] rhs_cols Number of columns in the right-hand side input matrix
* @param[in] rhs_rows Number of rows in the right-hand side input matrix
* @param[in] activation_min Minimum value to clamp the output to. Range: int16
* @param[in] activation_max Maximum value to clamp the output to. Range: int16
*
* @return The function returns <code>ARM_MATH_SUCCESS</code>
*
*/
arm_status arm_nn_vec_mat_mult_t_s16(const q15_t *lhs,
const q7_t *rhs,
const q63_t *bias,
q15_t *dst,
const int32_t dst_multiplier,
const int32_t dst_shift,
const int32_t rhs_cols,
const int32_t rhs_rows,
const int32_t activation_min,
const int32_t activation_max);
/**
* @brief s8 Vector by Matrix (transposed) multiplication with s16 output
*
* @param[in] lhs Input left-hand side vector
* @param[in] rhs Input right-hand side matrix (transposed)
* @param[out] dst Output vector
* @param[in] lhs_offset Offset to be added to the input values of the left-hand side
* vector. Range: -127 to 128
* @param[in] rhs_offset Not used
* @param[in] scatter_offset Address offset for dst. First output is stored at 'dst', the
* second at 'dst + scatter_offset' and so on.
* @param[in] dst_multiplier Output multiplier
* @param[in] dst_shift Output shift
* @param[in] rhs_cols Number of columns in the right-hand side input matrix
* @param[in] rhs_rows Number of rows in the right-hand side input matrix
* @param[in] activation_min Minimum value to clamp the output to. Range: int16
* @param[in] activation_max Maximum value to clamp the output to. Range: int16
*
* @return The function returns <code>ARM_MATH_SUCCESS</code>
*
*/
arm_status arm_nn_vec_mat_mult_t_svdf_s8(const q7_t *lhs,
const q7_t *rhs,
q15_t *dst,
const int32_t lhs_offset,
const int32_t rhs_offset,
const int32_t scatter_offset,
const int32_t dst_multiplier,
const int32_t dst_shift,
const int32_t rhs_cols,
const int32_t rhs_rows,
const int32_t activation_min,
const int32_t activation_max);
/**
* @brief Depthwise convolution of transposed rhs matrix with 4 lhs matrices. To be used in padded cases where
* the padding is -lhs_offset(Range: int8). Dimensions are the same for lhs and rhs.
*
* @param[in] lhs Input left-hand side matrix
* @param[in] rhs Input right-hand side matrix (transposed)
* @param[in] lhs_offset LHS matrix offset(input offset). Range: -127 to 128
* @param[in] num_ch Number of channels in LHS/RHS
* @param[in] out_shift Per channel output shift. Length of vector is equal to number of channels
* @param[in] out_mult Per channel output multiplier. Length of vector is equal to number of channels
* @param[in] out_offset Offset to be added to the output values. Range: -127 to 128
* @param[in] activation_min Minimum value to clamp the output to. Range: int8
* @param[in] activation_max Maximum value to clamp the output to. Range: int8
* @param[in] row_x_col (row_dimension * col_dimension) of LHS/RHS matrix
* @param[in] output_bias Per channel output bias. Length of vector is equal to number of channels
* @param[in] out Output pointer
*
* @return The function returns one of the two
* - Updated output pointer if an implementation is available
* - NULL if no implementation is available.
*
* @note If number of channels is not a multiple of 4, upto 3 elements outside the boundary will be read
* out for the following.
* - Output shift
* - Output multiplier
* - Output bias
* - rhs
*/
q7_t *arm_nn_depthwise_conv_nt_t_padded_s8(const q7_t *lhs,
const q7_t *rhs,
const int32_t lhs_offset,
const uint16_t num_ch,
const int32_t *out_shift,
const int32_t *out_mult,
const int32_t out_offset,
const int32_t activation_min,
const int32_t activation_max,
const uint16_t row_x_col,
const int32_t *const output_bias,
q7_t *out);
/**
* @brief Depthwise convolution of transposed rhs matrix with 4 lhs matrices. To be used in non-padded cases.
* Dimensions are the same for lhs and rhs.
*
* @param[in] lhs Input left-hand side matrix
* @param[in] rhs Input right-hand side matrix (transposed)
* @param[in] lhs_offset LHS matrix offset(input offset). Range: -127 to 128
* @param[in] num_ch Number of channels in LHS/RHS
* @param[in] out_shift Per channel output shift. Length of vector is equal to number of channels.
* @param[in] out_mult Per channel output multiplier. Length of vector is equal to number of channels.
* @param[in] out_offset Offset to be added to the output values. Range: -127 to 128
* @param[in] activation_min Minimum value to clamp the output to. Range: int8
* @param[in] activation_max Maximum value to clamp the output to. Range: int8
* @param[in] row_x_col (row_dimension * col_dimension) of LHS/RHS matrix
* @param[in] output_bias Per channel output bias. Length of vector is equal to number of channels.
* @param[in] out Output pointer
*
* @return The function returns one of the two
* - Updated output pointer if an implementation is available
* - NULL if no implementation is available.
*
* @note If number of channels is not a multiple of 4, upto 3 elements outside the boundary will be read
* out for the following.
* - Output shift
* - Output multiplier
* - Output bias
* - rhs
*/
q7_t *arm_nn_depthwise_conv_nt_t_s8(const q7_t *lhs,
const q7_t *rhs,
const int32_t lhs_offset,
const uint16_t num_ch,
const int32_t *out_shift,
const int32_t *out_mult,
const int32_t out_offset,
const int32_t activation_min,
const int32_t activation_max,
const uint16_t row_x_col,
const int32_t *const output_bias,
q7_t *out);
/**
*@brief Matrix-multiplication function for convolution with reordered columns
*@param[in] pA pointer to operand A
*@param[in] pInBuffer pointer to operand B, always conssists of 2 vectors
*@param[in] ch_im_out numRow of A
*@param[in] numCol_A numCol of A
*@param[in] bias_shift amount of left-shift for bias
*@param[in] out_shift amount of right-shift for output
*@param[in] bias the bias
*@param[in,out] pOut pointer to output
*@return The function returns the incremented output pointer
*
*@details This function assumes that data in pInBuffer are reordered
*/
q7_t *arm_nn_mat_mult_kernel_q7_q15_reordered(const q7_t *pA,
const q15_t *pInBuffer,
const uint16_t ch_im_out,
const uint16_t numCol_A,
const uint16_t bias_shift,
const uint16_t out_shift,
const q7_t *bias,
q7_t *pOut);
/**
@brief Read 2 q15 elements and post increment pointer.
@param[in] in_q15 Pointer to pointer that holds address of input.
@return q31 value
*/
__STATIC_FORCEINLINE q31_t arm_nn_read_q15x2_ia(const q15_t **in_q15)
{
q31_t val;
memcpy(&val, *in_q15, 4);
*in_q15 += 2;
return (val);
}
/**
@brief Read 4 q7 from q7 pointer and post increment pointer.
@param[in] in_q7 Pointer to pointer that holds address of input.
@return q31 value
*/
__STATIC_FORCEINLINE q31_t arm_nn_read_q7x4_ia(const q7_t **in_q7)
{
q31_t val;
memcpy(&val, *in_q7, 4);
*in_q7 += 4;
return (val);
}
/**
@brief Read 2 q15 from q15 pointer.
@param[in] in_q15 pointer to address of input.
@return q31 value
*/
__STATIC_FORCEINLINE q31_t arm_nn_read_q15x2(const q15_t *in_q15)
{
q31_t val;
memcpy(&val, in_q15, 4);
return (val);
}
/**
@brief Read 4 q7 values.
@param[in] in_q7 pointer to address of input.
@return q31 value
*/
__STATIC_FORCEINLINE q31_t arm_nn_read_q7x4(const q7_t *in_q7)
{
q31_t val;
memcpy(&val, in_q7, 4);
return (val);
}
/**
@brief Write four q7 to q7 pointer and increment pointer afterwards.
@param[in] in Double pointer to input value
@param[in] value Four bytes to copy
*/
__STATIC_FORCEINLINE void arm_nn_write_q7x4_ia(q7_t **in, q31_t value)
{
memcpy(*in, &value, 4);
*in += 4;
}
/**
* @brief memset optimized for MVE
* @param[in, out] dst Destination pointer
* @param[in] val Value to set
* @param[in] block_size Number of bytes to copy.
*
*/
__STATIC_FORCEINLINE void arm_memset_q7(q7_t *dst, const q7_t val, uint32_t block_size)
{
#if defined(ARM_MATH_MVEI)
__asm volatile(" vdup.8 q0, %[set_val] \n"
" wlstp.8 lr, %[cnt], 1f \n"
"2: \n"
" vstrb.8 q0, [%[in]], #16 \n"
" letp lr, 2b \n"
"1: \n"
: [ in ] "+r"(dst)
: [ cnt ] "r"(block_size), [ set_val ] "r"(val)
: "q0", "memory", "r14");
#else
memset(dst, val, block_size);
#endif
}
#if defined(ARM_MATH_DSP)
/**
* @brief read and expand one q7 word into two q15 words
*/
__STATIC_FORCEINLINE const q7_t *read_and_pad(const q7_t *source, q31_t *out1, q31_t *out2)
{
q31_t inA = arm_nn_read_q7x4_ia(&source);
q31_t inAbuf1 = __SXTB16_RORn((uint32_t)inA, 8);
q31_t inAbuf2 = __SXTB16(inA);
#ifndef ARM_MATH_BIG_ENDIAN
*out2 = (int32_t)(__PKHTB(inAbuf1, inAbuf2, 16));
*out1 = (int32_t)(__PKHBT(inAbuf2, inAbuf1, 16));
#else
*out1 = (int32_t)(__PKHTB(inAbuf1, inAbuf2, 16));
*out2 = (int32_t)(__PKHBT(inAbuf2, inAbuf1, 16));
#endif
return source;
}
/**
* @brief read and expand one q7 word into two q15 words with reordering
*/
__STATIC_FORCEINLINE const q7_t *read_and_pad_reordered(const q7_t *source, q31_t *out1, q31_t *out2)
{
q31_t inA = arm_nn_read_q7x4_ia(&source);
#ifndef ARM_MATH_BIG_ENDIAN
*out2 = __SXTB16(__ROR((uint32_t)inA, 8));
*out1 = __SXTB16(inA);
#else
*out1 = __SXTB16(__ROR((uint32_t)inA, 8));
*out2 = __SXTB16(inA);
#endif
return source;
}
/**
* @brief read and expand one q7 word into two q15 words with reordering and add an offset
*/
__STATIC_FORCEINLINE const q7_t *
read_and_pad_reordered_with_offset(const q7_t *source, q31_t *out1, q31_t *out2, q31_t offset)
{
q31_t inA = arm_nn_read_q7x4_ia(&source);
#ifndef ARM_MATH_BIG_ENDIAN
*out2 = __SXTB16(__ROR((uint32_t)inA, 8));
*out1 = __SXTB16(inA);
#else
*out1 = __SXTB16(__ROR((uint32_t)inA, 8));
*out2 = __SXTB16(inA);
#endif
*out1 = __QADD16(*out1, offset);
*out2 = __QADD16(*out2, offset);
return source;
}
#endif
/**
* @defgroup NNBasicMath Basic Math Functions for Neural Network Computation
*
* Basic Math Functions for Neural Network Computation
*
*/
/**
* @brief q7 vector multiplication with variable output shifts
* @param[in] *pSrcA pointer to the first input vector
* @param[in] *pSrcB pointer to the second input vector
* @param[out] *pDst pointer to the output vector
* @param[in] out_shift amount of right-shift for output
* @param[in] blockSize number of samples in each vector
* @return none.
*
* <b>Scaling and Overflow Behavior:</b>
* \par
* The function uses saturating arithmetic.
* Results outside of the allowable q15 range [0x8000 0x7FFF] will be saturated.
*/
void arm_nn_mult_q15(q15_t *pSrcA, q15_t *pSrcB, q15_t *pDst, const uint16_t out_shift, uint32_t blockSize);
/**
* @brief q7 vector multiplication with variable output shifts
* @param[in] *pSrcA pointer to the first input vector
* @param[in] *pSrcB pointer to the second input vector
* @param[out] *pDst pointer to the output vector
* @param[in] out_shift amount of right-shift for output
* @param[in] blockSize number of samples in each vector
* @return none.
*
* <b>Scaling and Overflow Behavior:</b>
* \par
* The function uses saturating arithmetic.
* Results outside of the allowable q7 range [0x80 0x7F] will be saturated.
*/
void arm_nn_mult_q7(q7_t *pSrcA, q7_t *pSrcB, q7_t *pDst, const uint16_t out_shift, uint32_t blockSize);
/**
* @brief Matrix-multiplication function for convolution with per-channel requantization.
* @param[in] input_a pointer to operand A
* @param[in] input_b pointer to operand B, always consists of 2 vectors.
* @param[in] output_ch number of rows of A
* @param[in] out_shift pointer to per output channel requantization shift parameter.
* @param[in] out_mult pointer to per output channel requantization multiplier parameter.
* @param[in] out_offset output tensor offset.
* @param[in] activation_min minimum value to clamp the output to. Range : int8
* @param[in] activation_max maximum value to clamp the output to. Range : int8
* @param[in] num_col_a number of columns of A
* @param[in] output_bias per output channel bias. Range : int32
* @param[in,out] out_0 pointer to output
* @return The function returns one of the two
* 1. The incremented output pointer for a successful operation or
* 2. NULL if implementation is not available.
*
* @details This function does the matrix multiplication of weight matrix for all output channels
* with 2 columns from im2col and produces two elements/output_channel. The outputs are
* clamped in the range provided by activation min and max.
* Supported framework: TensorFlow Lite micro.
*/
q7_t *arm_nn_mat_mult_kernel_s8_s16(const q7_t *input_a,
const q15_t *input_b,
const uint16_t output_ch,
const int32_t *out_shift,
const int32_t *out_mult,
const int32_t out_offset,
const int16_t activation_min,
const int16_t activation_max,
const uint16_t num_col_a,
const int32_t *const output_bias,
q7_t *out_0);
/**
* @brief Common softmax function for s8 input and s8 or s16 output
* @param[in] input Pointer to the input tensor
* @param[in] num_rows Number of rows in the input tensor
* @param[in] row_size Number of elements in each input row
* @param[in] mult Input quantization multiplier
* @param[in] shift Input quantization shift within the range [0, 31]
* @param[in] diff_min Minimum difference with max in row. Used to check if
* the quantized exponential operation can be performed
* @param[in] int16_output Indicating s8 output if 0 else s16 output
* @param[out] output Pointer to the output tensor
*
* @note Supported framework: TensorFlow Lite micro (bit-accurate)
*
*/
void arm_nn_softmax_common_s8(const int8_t *input,
const int32_t num_rows,
const int32_t row_size,
const int32_t mult,
const int32_t shift,
const int32_t diff_min,
const bool int16_output,
void *output);
/**
* @brief macro for adding rounding offset
*/
#ifndef ARM_NN_TRUNCATE
#define NN_ROUND(out_shift) ((0x1 << out_shift) >> 1)
#else
#define NN_ROUND(out_shift) 0
#endif
// Macros for shortening quantization functions' names and avoid long lines
#define MUL_SAT(a, b) arm_nn_doubling_high_mult((a), (b))
#define MUL_SAT_MVE(a, b) arm_doubling_high_mult_mve_32x4((a), (b))
#define MUL_POW2(a, b) arm_nn_mult_by_power_of_two((a), (b))
#define DIV_POW2(a, b) arm_nn_divide_by_power_of_two((a), (b))
#define DIV_POW2_MVE(a, b) arm_divide_by_power_of_two_mve((a), (b))
#define EXP_ON_NEG(x) arm_nn_exp_on_negative_values((x))
#define ONE_OVER1(x) arm_nn_one_over_one_plus_x_for_x_in_0_1((x))
/**
* @brief Saturating doubling high multiply. Result matches
* NEON instruction VQRDMULH.
* @param[in] m1 Multiplicand. Range: {NN_Q31_MIN, NN_Q31_MAX}
* @param[in] m2 Multiplier. Range: {NN_Q31_MIN, NN_Q31_MAX}
* @return Result of multiplication.
*
*/
__STATIC_FORCEINLINE q31_t arm_nn_doubling_high_mult(const q31_t m1, const q31_t m2)
{
q31_t result = 0;
// Rounding offset to add for a right shift of 31
q63_t mult = 1 << 30;
if ((m1 < 0) ^ (m2 < 0))
{
mult = 1 - mult;
}
// Gets resolved as a SMLAL instruction
mult = mult + (q63_t)m1 * m2;
// Utilize all of the upper 32 bits. This is the doubling step
// as well.
result = (int32_t)(mult / (1ll << 31));
if ((m1 == m2) && (m1 == (int32_t)NN_Q31_MIN))
{
result = NN_Q31_MAX;
}
return result;
}
/**
* @brief Doubling high multiply without saturation. This is intended
* for requantization where the scale is a positive integer
*
* @param[in] m1 Multiplicand. Range: {NN_Q31_MIN, NN_Q31_MAX}
* @param[in] m2 Multiplier Range: {NN_Q31_MIN, NN_Q31_MAX}
* @return Result of multiplication.
* @note The result of this matches that of neon instruction
* VQRDMULH for m1 in range {NN_Q31_MIN, NN_Q31_MAX} and m2 in
* range {NN_Q31_MIN + 1, NN_Q31_MAX}. Saturation occurs when
* m1 equals m2 equals NN_Q31_MIN and that is not handled by
* this function.
*
*/
__STATIC_FORCEINLINE q31_t arm_nn_doubling_high_mult_no_sat(const q31_t m1, const q31_t m2)
{
q31_t result = 0;
union arm_nn_long_long mult;
// Rounding offset to add for a right shift of 31
mult.word.low = 1 << 30;
mult.word.high = 0;
// Gets resolved as a SMLAL instruction
mult.long_long = mult.long_long + (q63_t)m1 * m2;
// Utilize all of the upper 32 bits. This is the doubling step
// as well.
result = (int32_t)(mult.long_long >> 31);
return result;
}
/**
* @brief Rounding divide by power of two.
* @param[in] dividend - Dividend
* @param[in] exponent - Divisor = power(2, exponent)
* Range: [0, 31]
* @return Rounded result of division. Midpoint is rounded away from zero.
*
*/
__STATIC_FORCEINLINE q31_t arm_nn_divide_by_power_of_two(const q31_t dividend, const q31_t exponent)
{
q31_t result = 0;
const q31_t remainder_mask = (1 << exponent) - 1;
int32_t remainder = remainder_mask & dividend;
// Basic division
result = dividend >> exponent;
// Adjust 'result' for rounding (mid point away from zero)
q31_t threshold = remainder_mask >> 1;
if (result < 0)
{
threshold++;
}
if (remainder > threshold)
{
result++;
}
return result;
}
/**
* @brief Requantize a given value.
* @param[in] val Value to be requantized
* @param[in] multiplier multiplier. Range {NN_Q31_MIN + 1, Q32_MAX}
* @param[in] shift left or right shift for 'val * multiplier'
*
* @return Returns (val * multiplier)/(2 ^ shift)
*
*/
__STATIC_FORCEINLINE q31_t arm_nn_requantize(const q31_t val, const q31_t multiplier, const q31_t shift)
{
#ifdef CMSIS_NN_USE_SINGLE_ROUNDING
const int64_t total_shift = 31 - shift;
const int64_t new_val = val * (int64_t)multiplier;
int32_t result = new_val >> (total_shift - 1);
result = (result + 1) >> 1;
return result;
#else
return arm_nn_divide_by_power_of_two(arm_nn_doubling_high_mult_no_sat(val * (1 << LEFT_SHIFT(shift)), multiplier),
RIGHT_SHIFT(shift));
#endif
}
/**
* @brief Requantize a given 64 bit value.
* @param[in] val Value to be requantized in the range {-(1<<47)} to {(1<<47) - 1}
* @param[in] reduced_multiplier Reduced multiplier in the range {NN_Q31_MIN + 1, Q32_MAX} to {Q16_MIN + 1,
* Q16_MAX}
* @param[in] shift Left or right shift for 'val * multiplier' in the range {-31} to {7}
*
* @return Returns (val * multiplier)/(2 ^ shift)
*
*/
__STATIC_FORCEINLINE q31_t arm_nn_requantize_s64(const q63_t val, const q31_t reduced_multiplier, const q31_t shift)
{
const q63_t new_val = val * reduced_multiplier;
q31_t result = new_val >> (14 - shift); // 64->32 bit reduction
result = (result + 1) >> 1; // Last shift position and insert round
return result;
}
/**
* @brief memcpy optimized for MVE
* @param[in, out] dst Destination pointer
* @param[in] src Source pointer.
* @param[in] block_size Number of bytes to copy.
*
*/
__STATIC_FORCEINLINE void arm_memcpy_q7(q7_t *__RESTRICT dst, const q7_t *__RESTRICT src, uint32_t block_size)
{
#if defined(ARM_MATH_MVEI)
__asm volatile(" wlstp.8 lr, %[cnt], 1f \n"
"2: \n"
" vldrb.8 q0, [%[in]], #16 \n"
" vstrb.8 q0, [%[out]], #16 \n"
" letp lr, 2b \n"
"1: \n"
: [ in ] "+r"(src), [ out ] "+r"(dst)
: [ cnt ] "r"(block_size)
: "q0", "memory", "r14");
#else
memcpy(dst, src, block_size);
#endif
}
#if defined(ARM_MATH_MVEI)
/**
* @brief Vector saturating doubling high multiply returning high half.
* @param[in] m1 Multiplicand
* @param[in] m2 Multiplier
* @return Result of multiplication.
*
*/
__STATIC_FORCEINLINE int32x4_t arm_doubling_high_mult_mve(const int32x4_t m1, const q31_t m2)
{
return vqrdmulhq_n_s32(m1, m2);
}
/**
* @brief Vector rounding divide by power of two.
* @param[in] dividend - Dividend vector
* @param[in] exponent - Divisor = power(2, exponent)
* Range: [0, 31]
* @return Rounded result of division. Midpoint is rounded away from zero.
*
*/
__STATIC_FORCEINLINE int32x4_t arm_divide_by_power_of_two_mve(const int32x4_t dividend, const q31_t exponent)
{
const int32x4_t shift = vdupq_n_s32(-exponent);
const int32x4_t fixup = vshrq_n_s32(vandq_s32(dividend, shift), 31);
const int32x4_t fixed_up_dividend = vqaddq_s32(dividend, fixup);
return vrshlq_s32(fixed_up_dividend, shift);
}
/**
* @brief Requantize a given vector.
* @param[in] val Vector to be requantized
* @param[in] multiplier multiplier
* @param[in] shift shift
*
* @return Returns (val * multiplier)/(2 ^ shift)
*
*/
__STATIC_FORCEINLINE int32x4_t arm_requantize_mve(const int32x4_t val, const q31_t multiplier, const q31_t shift)
{
#ifdef CMSIS_NN_USE_SINGLE_ROUNDING
const int right_shift = MIN(-1, shift);
const int left_shift = shift - right_shift;
const int32x4_t left_shift_dup = vdupq_n_s32(left_shift);
const int32x4_t right_shift_dup = vdupq_n_s32(right_shift);
int32x4_t result = vqdmulhq_n_s32(vshlq_s32(val, left_shift_dup), multiplier);
result = vrshlq_s32(result, right_shift_dup);
return result;
#else
return arm_divide_by_power_of_two_mve(
arm_doubling_high_mult_mve(vshlq_s32(val, vdupq_n_s32(LEFT_SHIFT(shift))), multiplier), RIGHT_SHIFT(shift));
#endif
}
__STATIC_FORCEINLINE int32x4_t arm_doubling_high_mult_mve_32x4(const int32x4_t m1, const int32x4_t m2)
{
return vqrdmulhq_s32(m1, m2);
}
__STATIC_FORCEINLINE int32x4_t arm_divide_by_power_of_two_mve_32x4(const int32x4_t dividend, const int32x4_t exponent)
{
const int32x4_t shift = -exponent;
const int32x4_t fixup = vshrq_n_s32(vandq_s32(dividend, shift), 31);
const int32x4_t fixed_up_dividend = vqaddq_s32(dividend, fixup);
return vrshlq_s32(fixed_up_dividend, shift);
}
__STATIC_FORCEINLINE int32x4_t arm_requantize_mve_32x4(const int32x4_t val,
const int32x4_t multiplier,
const int32x4_t shift)
{
#ifdef CMSIS_NN_USE_SINGLE_ROUNDING
const int32x4_t right_shift = vminq_s32(vdupq_n_s32(-1), shift);
const int32x4_t left_shift = vqsubq_s32(shift, right_shift);
int32x4_t result = vqdmulhq_s32(vshlq_s32(val, left_shift), multiplier);
result = vrshlq_s32(result, right_shift);
return result;
#else
const int32x4_t zz = vdupq_n_s32(0);
const mve_pred16_t p = vcmpgtq_n_s32(shift, 0);
const int32x4_t left_shift = vpselq_s32(shift, zz, p);
const int32x4_t right_shift = -vpselq_s32(zz, shift, p);
return arm_divide_by_power_of_two_mve_32x4(arm_doubling_high_mult_mve_32x4(vshlq_s32(val, left_shift), multiplier),
right_shift);
#endif
}
#endif
// @note The following functions are used only for softmax layer, scaled bits = 5 assumed
__STATIC_FORCEINLINE int32_t arm_nn_exp_on_negative_values(int32_t val)
{
int32_t mask = 0;
int32_t shift = 24;
const int32_t val_mod_minus_quarter = (val & ((1 << shift) - 1)) - (1 << shift);
const int32_t remainder = val_mod_minus_quarter - val;
const int32_t x = (val_mod_minus_quarter << 5) + (1 << 28);
const int32_t x2 = MUL_SAT(x, x);
int32_t result = 1895147668 +
MUL_SAT(1895147668, x + DIV_POW2(MUL_SAT(DIV_POW2(MUL_SAT(x2, x2), 2) + MUL_SAT(x2, x), 715827883) + x2, 1));
#define SELECT_IF_NON_ZERO(x) \
{ \
mask = MASK_IF_NON_ZERO(remainder & (1 << shift++)); \
result = SELECT_USING_MASK(mask, MUL_SAT(result, x), result); \
}
SELECT_IF_NON_ZERO(1672461947)
SELECT_IF_NON_ZERO(1302514674)
SELECT_IF_NON_ZERO(790015084)
SELECT_IF_NON_ZERO(290630308)
SELECT_IF_NON_ZERO(39332535)
SELECT_IF_NON_ZERO(720401)
SELECT_IF_NON_ZERO(242)
#undef SELECT_IF_NON_ZERO
mask = MASK_IF_ZERO(val);
return SELECT_USING_MASK(mask, NN_Q31_MAX, result);
}
__STATIC_FORCEINLINE q31_t arm_nn_mult_by_power_of_two(const int32_t val, const int32_t exp)
{
const int32_t thresh = ((1 << (31 - exp)) - 1);
int32_t result = val << exp;
result = SELECT_USING_MASK(MASK_IF_NON_ZERO(val > thresh), NN_Q31_MAX, result);
result = SELECT_USING_MASK(MASK_IF_NON_ZERO(val < -thresh), NN_Q31_MIN, result);
return result;
}
__STATIC_FORCEINLINE int32_t arm_nn_one_over_one_plus_x_for_x_in_0_1(int32_t val)
{
const int64_t sum = (int64_t)val + (int64_t)NN_Q31_MAX;
const int32_t half_denominator = (int32_t)((sum + (sum >= 0 ? 1 : -1)) / 2L);
int32_t x = 1515870810 + MUL_SAT(half_denominator, -1010580540);
const int32_t shift = (1 << 29);
x += MUL_POW2(MUL_SAT(x, shift - MUL_SAT(half_denominator, x)), 2);
x += MUL_POW2(MUL_SAT(x, shift - MUL_SAT(half_denominator, x)), 2);
x += MUL_POW2(MUL_SAT(x, shift - MUL_SAT(half_denominator, x)), 2);
return MUL_POW2(x, 1);
}
/**
@brief Write 2 q15 elements and post increment pointer.
@param[in] dest_q15 Pointer to pointer that holds address of destination.
@param[in] src_q31 Input value to be written.
*/
__STATIC_FORCEINLINE void arm_nn_write_q15x2_ia(q15_t **dest_q15, q31_t src_q31)
{
q31_t val = src_q31;
memcpy(*dest_q15, &val, 4);
*dest_q15 += 2;
}
#ifdef __cplusplus
}
#endif
#endif