diff options
Diffstat (limited to 'Drivers/CMSIS/DSP/Source/StatisticsFunctions/arm_rms_q15.c')
-rw-r--r-- | Drivers/CMSIS/DSP/Source/StatisticsFunctions/arm_rms_q15.c | 139 |
1 files changed, 67 insertions, 72 deletions
diff --git a/Drivers/CMSIS/DSP/Source/StatisticsFunctions/arm_rms_q15.c b/Drivers/CMSIS/DSP/Source/StatisticsFunctions/arm_rms_q15.c index d0e61ca31..9fcd964e5 100644 --- a/Drivers/CMSIS/DSP/Source/StatisticsFunctions/arm_rms_q15.c +++ b/Drivers/CMSIS/DSP/Source/StatisticsFunctions/arm_rms_q15.c @@ -3,13 +3,13 @@ * Title: arm_rms_q15.c * Description: Root Mean Square of the elements of a Q15 vector * - * $Date: 27. January 2017 - * $Revision: V.1.5.1 + * $Date: 18. March 2019 + * $Revision: V1.6.0 * * Target Processor: Cortex-M cores * -------------------------------------------------------------------- */ /* - * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved. + * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved. * * SPDX-License-Identifier: Apache-2.0 * @@ -29,111 +29,106 @@ #include "arm_math.h" /** - * @addtogroup RMS - * @{ + @ingroup groupStats */ /** - * @brief Root Mean Square of the elements of a Q15 vector. - * @param[in] *pSrc points to the input vector - * @param[in] blockSize length of the input vector - * @param[out] *pResult rms value returned here - * @return none. - * - * @details - * <b>Scaling and Overflow Behavior:</b> - * - * \par - * The function is implemented using a 64-bit internal accumulator. - * The input is represented in 1.15 format. - * Intermediate multiplication yields a 2.30 format, and this - * result is added without saturation to a 64-bit accumulator in 34.30 format. - * With 33 guard bits in the accumulator, there is no risk of overflow, and the - * full precision of the intermediate multiplication is preserved. - * Finally, the 34.30 result is truncated to 34.15 format by discarding the lower - * 15 bits, and then saturated to yield a result in 1.15 format. - * + @addtogroup RMS + @{ + */ + +/** + @brief Root Mean Square of the elements of a Q15 vector. + @param[in] pSrc points to the input vector + @param[in] blockSize number of samples in input vector + @param[out] pResult root mean square 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 1.15 format. + Intermediate multiplication yields a 2.30 format, and this + result is added without saturation to a 64-bit accumulator in 34.30 format. + With 33 guard bits in the accumulator, there is no risk of overflow, and the + full precision of the intermediate multiplication is preserved. + Finally, the 34.30 result is truncated to 34.15 format by discarding the lower + 15 bits, and then saturated to yield a result in 1.15 format. */ void arm_rms_q15( - q15_t * pSrc, - uint32_t blockSize, - q15_t * pResult) + const q15_t * pSrc, + uint32_t blockSize, + q15_t * pResult) { - q63_t sum = 0; /* accumulator */ + uint32_t blkCnt; /* Loop counter */ + q63_t sum = 0; /* Temporary result storage */ + q15_t in; /* Temporary variable to store input value */ -#if defined (ARM_MATH_DSP) - /* Run the below code for Cortex-M4 and Cortex-M3 */ +#if defined (ARM_MATH_LOOPUNROLL) && defined (ARM_MATH_DSP) + q31_t in32; /* Temporary variable to store input value */ +#endif - q31_t in; /* temporary variable to store the input value */ - q15_t in1; /* temporary variable to store the input value */ - uint32_t blkCnt; /* loop counter */ +#if defined (ARM_MATH_LOOPUNROLL) - /* loop Unrolling */ + /* Loop unrolling: Compute 4 outputs at a time */ blkCnt = blockSize >> 2U; - /* First part of the processing with loop unrolling. Compute 4 outputs at a time. - ** a second loop below computes the remaining 1 to 3 samples. */ while (blkCnt > 0U) { - /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */ - /* Compute sum of the squares and then store the results in a temporary variable, sum */ - in = *__SIMD32(pSrc)++; - sum = __SMLALD(in, in, sum); - in = *__SIMD32(pSrc)++; - sum = __SMLALD(in, in, sum); - - /* Decrement the loop counter */ - blkCnt--; - } + /* C = A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1] */ - /* If the blockSize is not a multiple of 4, compute any remaining output samples here. - ** No loop unrolling is used. */ - blkCnt = blockSize % 0x4U; + /* Compute sum of squares and store result in a temporary variable. */ +#if defined (ARM_MATH_DSP) + in32 = read_q15x2_ia ((q15_t **) &pSrc); + sum = __SMLALD(in32, in32, sum); - while (blkCnt > 0U) - { - /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */ - /* Compute sum of the squares and then store the results in a temporary variable, sum */ - in1 = *pSrc++; - sum = __SMLALD(in1, in1, sum); + in32 = read_q15x2_ia ((q15_t **) &pSrc); + sum = __SMLALD(in32, in32, sum); +#else + in = *pSrc++; + sum += ((q31_t) in * in); - /* Decrement the loop counter */ + in = *pSrc++; + sum += ((q31_t) in * in); + + in = *pSrc++; + sum += ((q31_t) in * in); + + in = *pSrc++; + sum += ((q31_t) in * in); +#endif /* #if defined (ARM_MATH_DSP) */ + + /* Decrement loop counter */ blkCnt--; } - /* Truncating and saturating the accumulator to 1.15 format */ - /* Store the result in the destination */ - arm_sqrt_q15(__SSAT((sum / (q63_t)blockSize) >> 15, 16), pResult); + /* Loop unrolling: Compute remaining outputs */ + blkCnt = blockSize % 0x4U; #else - /* Run the below code for Cortex-M0 */ - q15_t in; /* temporary variable to store the input value */ - uint32_t blkCnt; /* loop counter */ - - /* Loop over blockSize number of values */ + /* Initialize blkCnt with number of samples */ blkCnt = blockSize; +#endif /* #if defined (ARM_MATH_LOOPUNROLL) */ + while (blkCnt > 0U) { - /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */ - /* Compute sum of the squares and then store the results in a temporary variable, sum */ + /* C = A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1] */ + in = *pSrc++; + /* Compute sum of squares and store result in a temporary variable. */ sum += ((q31_t) in * in); - /* Decrement the loop counter */ + /* Decrement loop counter */ blkCnt--; } /* Truncating and saturating the accumulator to 1.15 format */ - /* Store the result in the destination */ + /* Store result in destination */ arm_sqrt_q15(__SSAT((sum / (q63_t)blockSize) >> 15, 16), pResult); - -#endif /* #if defined (ARM_MATH_DSP) */ - } /** - * @} end of RMS group + @} end of RMS group */ |