/github/workspace/src/MatrixFunctionsStride/mat_sub_stride/kernels/plp_mat_sub_stride_i8s_xpulpv2.c
Functions
Name | |
---|---|
void | plp_mat_sub_stride_i8s_xpulpv2(const int8_t restrict pSrcA, const int8_t restrict pSrcB, uint32_t M, uint32_t N, uint32_t strideA, uint32_t strideB, uint32_t strideY, int8_t *restrict pDst) strided matrix subtraction of 8-bit integer matrices kernel for XPULPV2 extension. |
Functions Documentation
function plp_mat_sub_stride_i8s_xpulpv2
void plp_mat_sub_stride_i8s_xpulpv2(
const int8_t *__restrict__ pSrcA,
const int8_t *__restrict__ pSrcB,
uint32_t M,
uint32_t N,
uint32_t strideA,
uint32_t strideB,
uint32_t strideY,
int8_t *__restrict__ pDst
)
strided matrix subtraction of 8-bit integer matrices kernel for XPULPV2 extension.
Parameters:
- pSrcA Points to the first input matrix
- pSrcB Points to the second input matrix
- M Height of all matrices
- N Width of all matrices
- strideA Stride of matrix A (elements between each row)
- strideB Stride of matrid B (elements between each row)
- strideY Stride of output matrix (elements between each row)
- pDst Points to the output matrix
Return: none
matrix subtraction of a 8-bit integer matrices for XPULPV2 extension.
Source code
/* =====================================================================
* Project: PULP DSP Library
* Title: plp_mat_sub_stride_i8s_xpulpv2.c
* Description: 8-bit integer strided matrix subtraction for XPULPV2
*
* $Date: 1. July 2020
* $Revision: V0
*
* Target Processor: PULP cores
* ===================================================================== */
/*
* Copyright (C) 2020 ETH Zurich and University of Bologna.
*
* Author: Tibor Schneider, ETH Zurich
*
* 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.
*/
#include "plp_math.h"
void plp_mat_sub_stride_i8s_xpulpv2(const int8_t *__restrict__ pSrcA,
const int8_t *__restrict__ pSrcB,
uint32_t M,
uint32_t N,
uint32_t strideA,
uint32_t strideB,
uint32_t strideY,
int8_t *__restrict__ pDst) {
//#define BASIC_VERSION // if used don't forget to also use the undefine at end of file
#ifdef BASIC_VERSION
uint32_t m, n; // loop counters
for (m = 0; m < M; m++) {
for (n = 0; n < N; n++) {
pDst[m * strideY + n] = pSrcA[m * strideA + n] - pSrcB[m * strideB + n];
}
}
#else
uint32_t m, n; // loop counters
const int8_t *__restrict__ pSrcA2 = pSrcA + strideA;
const int8_t *__restrict__ pSrcB2 = pSrcB + strideB;
int8_t *__restrict__ pDst2 = pDst + strideY;
unsigned int m_iter = M >> 1;
unsigned int m_rem = M & 0x1;
unsigned int n_iter = N >> 2;
unsigned int n_rem = N & 0x3;
unsigned int step_a = strideA * 2 - N;
unsigned int step_b = strideB * 2 - N;
unsigned int step_y = strideY * 2 - N;
for (m = 0; m < m_iter; m++) {
for (n = 0; n < n_iter; n++) {
v4s a1 = *((v4s *)pSrcA);
v4s b1 = *((v4s *)pSrcB);
v4s a2 = *((v4s *)pSrcA2);
v4s b2 = *((v4s *)pSrcB2);
*((v4s *)pDst) = __SUB4(a1, b1);
*((v4s *)pDst2) = __SUB4(a2, b2);
pSrcA += 4;
pSrcB += 4;
pDst += 4;
pSrcA2 += 4;
pSrcB2 += 4;
pDst2 += 4;
}
for (n = 0; n < n_rem; n++) {
*pDst++ = *pSrcA++ - *pSrcB++;
*pDst2++ = *pSrcA2++ - *pSrcB2++;
}
pSrcA += step_a;
pSrcB += step_b;
pDst += step_y;
pSrcA2 += step_a;
pSrcB2 += step_b;
pDst2 += step_y;
}
if (m_rem) {
for (n = 0; n < n_iter; n++) {
v4s a = *((v4s *)pSrcA);
v4s b = *((v4s *)pSrcB);
*((v4s *)pDst) = __SUB4(a, b);
pSrcA += 4;
pSrcB += 4;
pDst += 4;
}
for (n = 0; n < n_rem; n++) {
*pDst++ = *pSrcA++ - *pSrcB++;
}
pSrcA += step_a;
pSrcB += step_b;
pDst += step_y;
}
#endif
//#undef BASIC_VERSION
}
Updated on 2023-03-01 at 16:16:33 +0000