dengzhihao
/
CMSIS-DSP
0
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Remove q7 , q15 and q31 version of accumulate function

Browse Source
pull/42/head
Silfurion 3 years ago committed by Christophe Favergeon
parent d2b8728b22
commit 5630232a47
4 changed files with 0 additions and 493 deletions
Show Stats Download Patch File Download Diff File

35
Include/dsp/statistics_functions.h
Unescape Escape View File

@ -995,41 +995,6 @@ const float64_t * pSrc,
uint32_t blockSize,
float64_t * pResult);
/**
* @brief Accumulation value of a Q7 vector.
* @param[in] pSrc is input pointer
* @param[in] blockSize is the number of samples to process
* @param[out] pResult is output value.
*/
void arm_accumulate_q7(
const q7_t * pSrc,
uint32_t blockSize,
q31_t * pResult);
/**
* @brief Accumulation value of a Q15 vector.
* @param[in] pSrc is input pointer
* @param[in] blockSize is the number of samples to process
* @param[out] pResult is output value.
*/
void arm_accumulate_q15(
const q15_t * pSrc,
uint32_t blockSize,
q63_t * pResult);
/**
* @brief Accumulation value of a Q31 vector.
* @param[in] pSrc is input pointer
* @param[in] blockSize is the number of samples to process
* @param[out] pResult is output value.
*/
void arm_accumulate_q31(
const q31_t * pSrc,
uint32_t blockSize,
q63_t * pResult);
#ifdef __cplusplus
}

156
Source/StatisticsFunctions/arm_accumulate_q15.c
Unescape Escape View File

@ -1,156 +0,0 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_accumulate_q15.c
* Description: Accumulation value of a Q15 vector
*
* $Date: 14 July 2022
* $Revision: V1.0.0
*
* Target Processor: Cortex-M and Cortex-A cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved.
*
* 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 "dsp/statistics_functions.h"
/**
@ingroup groupStats
*/
/**
@addtogroup Accumulation
@{
*/
/**
@brief Accumulation value of a Q15 vector.
@param[in] pSrc points to the input vector
@param[in] blockSize number of samples in input vector
@param[out] pResult sum value returned here
@return none
@par Scaling and Overflow Behavior
The function is implemented using a 32-bit internal accumulator.
The input is represented in 1.15 format and is accumulated in a 32-bit
accumulator in 17.15 format.
There is no risk of internal overflow with this approach, and the
full precision of intermediate result is preserved.
Finally, the accumulator is truncated to yield a result of 1.15 format.
*/
#if defined(ARM_MATH_MVEI) && !defined(ARM_MATH_AUTOVECTORIZE)
void arm_accumulate_q15(
const q15_t * pSrc,
uint32_t blockSize,
q63_t * pResult)
{
uint32_t blkCnt; /* loop counters */
q15x8_t vecSrc;
q63_t sum = 0L;
/* Compute 8 outputs at a time */
blkCnt = blockSize >> 3U;
while (blkCnt > 0U)
{
vecSrc = vldrhq_s16(pSrc);
/*
* sum lanes
*/
sum = vaddvaq(sum, vecSrc);
blkCnt--;
pSrc += 8;
}
/* Tail */
blkCnt = blockSize & 0x7;
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
/* Store the result to the destination */
*pResult = sum;
}
#else
void arm_accumulate_q15(
const q15_t * pSrc,
uint32_t blockSize,
q63_t * pResult)
{
uint32_t blkCnt; /* Loop counter */
q63_t sum = 0; /* Temporary result storage */
#if defined (ARM_MATH_LOOPUNROLL)
q31_t in;
#endif
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
in = read_q15x2_ia (&pSrc);
sum += ((in << 16U) >> 16U);
sum += (in >> 16U);
in = read_q15x2_ia (&pSrc);
sum += ((in << 16U) >> 16U);
sum += (in >> 16U);
/* Decrement the loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
/* Store result to destination */
*pResult = sum;
}
#endif /* defined(ARM_MATH_MVEI) */
/**
@} end of Accumulation group
*/

149
Source/StatisticsFunctions/arm_accumulate_q31.c
Unescape Escape View File

@ -1,149 +0,0 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_accumulate_q31.c
* Description: Accumulation value of a Q31 vector
*
* $Date: 14 July 2022
* $Revision: V1.0.0
*
* Target Processor: Cortex-M and Cortex-A cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved.
*
* 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 "dsp/statistics_functions.h"
/**
@ingroup groupStats
*/
/**
@addtogroup Accumulation
@{
*/
/**
@brief Accumulation value of a Q31 vector.
@param[in] pSrc points to the input vector
@param[in] blockSize number of samples in input vector
@param[out] pResult sum value returned here
@return none
@par Scaling and Overflow Behavior
The function is implemented using a 64-bit internal accumulator.
The input is represented in 1.31 format and is accumulated in a 64-bit
accumulator in 33.31 format.
There is no risk of internal overflow with this approach, and the
full precision of intermediate result is preserved.
Finally, the accumulator is truncated to yield a result of 1.31 format.
*/
#if defined(ARM_MATH_MVEI) && !defined(ARM_MATH_AUTOVECTORIZE)
void arm_accumulate_q31(
const q31_t * pSrc,
uint32_t blockSize,
q63_t * pResult)
{
uint32_t blkCnt; /* loop counters */
q31x4_t vecSrc;
q63_t sum = 0LL;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
vecSrc = vldrwq_s32(pSrc);
/*
* sum lanes
*/
sum = vaddlvaq(sum, vecSrc);
blkCnt --;
pSrc += 4;
}
/* Tail */
blkCnt = blockSize & 0x3;
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
blkCnt --;
}
*pResult = sum;
}
#else
void arm_accumulate_q31(
const q31_t * pSrc,
uint32_t blockSize,
q63_t * pResult)
{
uint32_t blkCnt; /* Loop counter */
q63_t sum = 0; /* Temporary result storage */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
sum += *pSrc++;
sum += *pSrc++;
sum += *pSrc++;
/* Decrement the loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
/* Store result to destination */
*pResult = sum;
}
#endif /* defined(ARM_MATH_MVEI) */
/**
@} end of Accumulation group
*/

153
Source/StatisticsFunctions/arm_accumulate_q7.c
Unescape Escape View File

@ -1,153 +0,0 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_accumulate_q7.c
* Description: Accumulation value of a Q7 vector
*
* $Date: 14 July 2022
* $Revision: V1.0.0
*
* Target Processor: Cortex-M and Cortex-A cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved.
*
* 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 "dsp/statistics_functions.h"
/**
@ingroup groupStats
*/
/**
@addtogroup Accumulation
@{
*/
/**
@brief Accumulation value of a Q7 vector.
@param[in] pSrc points to the input vector
@param[in] blockSize number of samples in input vector
@param[out] pResult sum value returned here
@return none
@par Scaling and Overflow Behavior
The function is implemented using a 32-bit internal accumulator.
The input is represented in 1.7 format and is accumulated in a 32-bit
accumulator in 25.7 format.
There is no risk of internal overflow with this approach, and the
full precision of intermediate result is preserved.
Finally, the accumulator is truncated to yield a result of 1.7 format.
*/
#if defined(ARM_MATH_MVEI) && !defined(ARM_MATH_AUTOVECTORIZE)
void arm_accumulate_q7(
const q7_t * pSrc,
uint32_t blockSize,
q31_t * pResult)
{
uint32_t blkCnt; /* loop counters */
q7x16_t vecSrc;
q31_t sum = 0L;
blkCnt = blockSize >> 4;
while (blkCnt > 0U)
{
vecSrc = vldrbq_s8(pSrc);
/*
* sum lanes
*/
sum = vaddvaq(sum, vecSrc);
blkCnt--;
pSrc += 16;
}
blkCnt = blockSize & 0xF;
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
/* Store the result to the destination */
*pResult = sum;
}
#else
void arm_accumulate_q7(
const q7_t * pSrc,
uint32_t blockSize,
q31_t * pResult)
{
uint32_t blkCnt; /* Loop counter */
q31_t sum = 0; /* Temporary result storage */
#if defined (ARM_MATH_LOOPUNROLL)
q31_t in;
#endif
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
in = read_q7x4_ia (&pSrc);
sum += ((in << 24U) >> 24U);
sum += ((in << 16U) >> 24U);
sum += ((in << 8U) >> 24U);
sum += (in >> 24U);
/* Decrement the loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
sum += *pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
/* Store result to destination */
*pResult = sum;
}
#endif /* defined(ARM_MATH_MVEI) */
/**
@} end of Accumulation group
*/
Write Preview
Loading…
Cancel
Save