Renesas » Code » mbed-dsp-neon

Renesas / mbed-dsp-neon

Fork of mbed-dsp. CMSIS-DSP library of supporting NEON

Dependents: mbed-os-example-cmsis_dsp_neon

Fork of mbed-dsp by mbed official

Information

Japanese version is available in lower part of this page.
このページの後半に日本語版が用意されています．

CMSIS-DSP of supporting NEON

What is this ?

A library for CMSIS-DSP of supporting NEON.
We supported the NEON to CMSIS-DSP Ver1.4.3(CMSIS V4.1) that ARM supplied, has achieved the processing speed improvement.
If you use the mbed-dsp library, you can use to replace this library.
CMSIS-DSP of supporting NEON is provied as a library.

Library Creation environment

CMSIS-DSP library of supporting NEON was created by the following environment.

Compiler
ARMCC Version 5.03
Compile option switch[C Compiler]

   -DARM_MATH_MATRIX_CHECK -DARM_MATH_ROUNDING -O3 -Otime --cpu=Cortex-A9 --littleend --arm 
   --apcs=/interwork --no_unaligned_access --fpu=vfpv3_fp16 --fpmode=fast --apcs=/hardfp 
   --vectorize --asm

Compile option switch[Assembler]

   --cpreproc --cpu=Cortex-A9 --littleend --arm --apcs=/interwork --no_unaligned_access 
   --fpu=vfpv3_fp16 --fpmode=fast --apcs=/hardfp

Effects of NEON support

In the data which passes to each function, large size will be expected more effective than small size.
Also if the data is a multiple of 16, effect will be expected in every function in the CMSIS-DSP.

NEON対応CMSIS-DSP

概要

NEON対応したCMSIS-DSPのライブラリです。
ARM社提供のCMSIS-DSP Ver1.4.3(CMSIS V4.1)をターゲットにNEON対応を行ない、処理速度向上を実現しております。
mbed-dspライブラリを使用している場合は、本ライブラリに置き換えて使用することができます。
NEON対応したCMSIS-DSPはライブラリで提供します。

ライブラリ作成環境

NEON対応CMSIS-DSPライブラリは、以下の環境で作成しています。

コンパイラ
ARMCC Version 5.03
コンパイルオプションスイッチ[C Compiler]

   -DARM_MATH_MATRIX_CHECK -DARM_MATH_ROUNDING -O3 -Otime --cpu=Cortex-A9 --littleend --arm 
   --apcs=/interwork --no_unaligned_access --fpu=vfpv3_fp16 --fpmode=fast --apcs=/hardfp 
   --vectorize --asm

コンパイルオプションスイッチ[Assembler]

   --cpreproc --cpu=Cortex-A9 --littleend --arm --apcs=/interwork --no_unaligned_access 
   --fpu=vfpv3_fp16 --fpmode=fast --apcs=/hardfp

NEON対応による効果について

CMSIS-DSP内の各関数へ渡すデータは、小さいサイズよりも大きいサイズの方が効果が見込めます。
また、16の倍数のデータであれば、CMSIS-DSP内のどの関数でも効果が見込めます。

cmsis_dsp/StatisticsFunctions/arm_rms_q31.c@3:7a284390b0ce, 2013-11-08 (annotated)

Committer:: mbed_official
Date:: Fri Nov 08 13:45:10 2013 +0000
Revision:: 3:7a284390b0ce
Parent:: 2:da51fb522205

Synchronized with git revision e69956aba2f68a2a26ac26b051f8d349deaa1ce8

Who changed what in which revision?

User	Revision	Line number	New contents of line
emilmont	1:fdd22bb7aa52	1	/* ----------------------------------------------------------------------
mbed_official	3:7a284390b0ce	2	* Copyright (C) 2010-2013 ARM Limited. All rights reserved.
emilmont	1:fdd22bb7aa52	3	*
mbed_official	3:7a284390b0ce	4	* $Date: 17. January 2013
mbed_official	3:7a284390b0ce	5	* $Revision: V1.4.1
emilmont	1:fdd22bb7aa52	6	*
emilmont	2:da51fb522205	7	* Project: CMSIS DSP Library
emilmont	2:da51fb522205	8	* Title: arm_rms_q31.c
emilmont	1:fdd22bb7aa52	9	*
emilmont	2:da51fb522205	10	* Description: Root Mean Square of the elements of a Q31 vector.
emilmont	1:fdd22bb7aa52	11	*
emilmont	1:fdd22bb7aa52	12	* Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
emilmont	1:fdd22bb7aa52	13	*
mbed_official	3:7a284390b0ce	14	* Redistribution and use in source and binary forms, with or without
mbed_official	3:7a284390b0ce	15	* modification, are permitted provided that the following conditions
mbed_official	3:7a284390b0ce	16	* are met:
mbed_official	3:7a284390b0ce	17	* - Redistributions of source code must retain the above copyright
mbed_official	3:7a284390b0ce	18	* notice, this list of conditions and the following disclaimer.
mbed_official	3:7a284390b0ce	19	* - Redistributions in binary form must reproduce the above copyright
mbed_official	3:7a284390b0ce	20	* notice, this list of conditions and the following disclaimer in
mbed_official	3:7a284390b0ce	21	* the documentation and/or other materials provided with the
mbed_official	3:7a284390b0ce	22	* distribution.
mbed_official	3:7a284390b0ce	23	* - Neither the name of ARM LIMITED nor the names of its contributors
mbed_official	3:7a284390b0ce	24	* may be used to endorse or promote products derived from this
mbed_official	3:7a284390b0ce	25	* software without specific prior written permission.
mbed_official	3:7a284390b0ce	26	*
mbed_official	3:7a284390b0ce	27	* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
mbed_official	3:7a284390b0ce	28	* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
mbed_official	3:7a284390b0ce	29	* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
mbed_official	3:7a284390b0ce	30	* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
mbed_official	3:7a284390b0ce	31	* COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
mbed_official	3:7a284390b0ce	32	* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
mbed_official	3:7a284390b0ce	33	* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
mbed_official	3:7a284390b0ce	34	* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
mbed_official	3:7a284390b0ce	35	* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
mbed_official	3:7a284390b0ce	36	* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
mbed_official	3:7a284390b0ce	37	* ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
mbed_official	3:7a284390b0ce	38	* POSSIBILITY OF SUCH DAMAGE.
emilmont	1:fdd22bb7aa52	39	* ---------------------------------------------------------------------------- */
emilmont	1:fdd22bb7aa52	40
emilmont	1:fdd22bb7aa52	41	#include "arm_math.h"
emilmont	1:fdd22bb7aa52	42
emilmont	1:fdd22bb7aa52	43	/**
emilmont	1:fdd22bb7aa52	44	* @addtogroup RMS
emilmont	1:fdd22bb7aa52	45	* @{
emilmont	1:fdd22bb7aa52	46	*/
emilmont	1:fdd22bb7aa52	47
emilmont	1:fdd22bb7aa52	48
emilmont	1:fdd22bb7aa52	49	/**
emilmont	1:fdd22bb7aa52	50	* @brief Root Mean Square of the elements of a Q31 vector.
emilmont	1:fdd22bb7aa52	51	* @param[in] *pSrc points to the input vector
emilmont	1:fdd22bb7aa52	52	* @param[in] blockSize length of the input vector
emilmont	1:fdd22bb7aa52	53	* @param[out] *pResult rms value returned here
emilmont	1:fdd22bb7aa52	54	* @return none.
emilmont	1:fdd22bb7aa52	55	*
emilmont	1:fdd22bb7aa52	56	* @details
emilmont	1:fdd22bb7aa52	57	* <b>Scaling and Overflow Behavior:</b>
emilmont	1:fdd22bb7aa52	58	*
emilmont	1:fdd22bb7aa52	59	*\par
emilmont	1:fdd22bb7aa52	60	* The function is implemented using an internal 64-bit accumulator.
emilmont	1:fdd22bb7aa52	61	* The input is represented in 1.31 format, and intermediate multiplication
emilmont	1:fdd22bb7aa52	62	* yields a 2.62 format.
emilmont	1:fdd22bb7aa52	63	* The accumulator maintains full precision of the intermediate multiplication results,
emilmont	1:fdd22bb7aa52	64	* but provides only a single guard bit.
emilmont	1:fdd22bb7aa52	65	* There is no saturation on intermediate additions.
emilmont	1:fdd22bb7aa52	66	* If the accumulator overflows, it wraps around and distorts the result.
emilmont	1:fdd22bb7aa52	67	* In order to avoid overflows completely, the input signal must be scaled down by
emilmont	1:fdd22bb7aa52	68	* log2(blockSize) bits, as a total of blockSize additions are performed internally.
emilmont	1:fdd22bb7aa52	69	* Finally, the 2.62 accumulator is right shifted by 31 bits to yield a 1.31 format value.
emilmont	1:fdd22bb7aa52	70	*
emilmont	1:fdd22bb7aa52	71	*/
emilmont	1:fdd22bb7aa52	72
emilmont	1:fdd22bb7aa52	73	void arm_rms_q31(
emilmont	1:fdd22bb7aa52	74	q31_t * pSrc,
emilmont	1:fdd22bb7aa52	75	uint32_t blockSize,
emilmont	1:fdd22bb7aa52	76	q31_t * pResult)
emilmont	1:fdd22bb7aa52	77	{
emilmont	1:fdd22bb7aa52	78	q63_t sum = 0; /* accumulator */
emilmont	1:fdd22bb7aa52	79	q31_t in; /* Temporary variable to store the input */
emilmont	1:fdd22bb7aa52	80	uint32_t blkCnt; /* loop counter */
emilmont	1:fdd22bb7aa52	81
mbed_official	3:7a284390b0ce	82	#ifndef ARM_MATH_CM0_FAMILY
emilmont	1:fdd22bb7aa52	83
emilmont	1:fdd22bb7aa52	84	/* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont	1:fdd22bb7aa52	85
emilmont	1:fdd22bb7aa52	86	q31_t in1, in2, in3, in4; /* Temporary input variables */
emilmont	1:fdd22bb7aa52	87
emilmont	1:fdd22bb7aa52	88	/loop Unrolling /
emilmont	1:fdd22bb7aa52	89	blkCnt = blockSize >> 2u;
emilmont	1:fdd22bb7aa52	90
emilmont	1:fdd22bb7aa52	91	/* First part of the processing with loop unrolling. Compute 8 outputs at a time.
emilmont	1:fdd22bb7aa52	92	** a second loop below computes the remaining 1 to 7 samples. */
emilmont	1:fdd22bb7aa52	93	while(blkCnt > 0u)
emilmont	1:fdd22bb7aa52	94	{
emilmont	1:fdd22bb7aa52	95	/* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */
emilmont	1:fdd22bb7aa52	96	/* Compute sum of the squares and then store the result in a temporary variable, sum */
emilmont	1:fdd22bb7aa52	97	/* read two samples from source buffer */
emilmont	1:fdd22bb7aa52	98	in1 = pSrc[0];
emilmont	1:fdd22bb7aa52	99	in2 = pSrc[1];
emilmont	1:fdd22bb7aa52	100
emilmont	1:fdd22bb7aa52	101	/* calculate power and accumulate to accumulator */
emilmont	1:fdd22bb7aa52	102	sum += (q63_t) in1 *in1;
emilmont	1:fdd22bb7aa52	103	sum += (q63_t) in2 *in2;
emilmont	1:fdd22bb7aa52	104
emilmont	1:fdd22bb7aa52	105	/* read two samples from source buffer */
emilmont	1:fdd22bb7aa52	106	in3 = pSrc[2];
emilmont	1:fdd22bb7aa52	107	in4 = pSrc[3];
emilmont	1:fdd22bb7aa52	108
emilmont	1:fdd22bb7aa52	109	/* calculate power and accumulate to accumulator */
emilmont	1:fdd22bb7aa52	110	sum += (q63_t) in3 *in3;
emilmont	1:fdd22bb7aa52	111	sum += (q63_t) in4 *in4;
emilmont	1:fdd22bb7aa52	112
emilmont	1:fdd22bb7aa52	113
emilmont	1:fdd22bb7aa52	114	/* update source buffer to process next samples */
emilmont	1:fdd22bb7aa52	115	pSrc += 4u;
emilmont	1:fdd22bb7aa52	116
emilmont	1:fdd22bb7aa52	117	/* Decrement the loop counter */
emilmont	1:fdd22bb7aa52	118	blkCnt--;
emilmont	1:fdd22bb7aa52	119	}
emilmont	1:fdd22bb7aa52	120
emilmont	1:fdd22bb7aa52	121	/* If the blockSize is not a multiple of 8, compute any remaining output samples here.
emilmont	1:fdd22bb7aa52	122	** No loop unrolling is used. */
emilmont	1:fdd22bb7aa52	123	blkCnt = blockSize % 0x4u;
emilmont	1:fdd22bb7aa52	124
emilmont	1:fdd22bb7aa52	125	#else
emilmont	1:fdd22bb7aa52	126
emilmont	1:fdd22bb7aa52	127	/* Run the below code for Cortex-M0 */
emilmont	1:fdd22bb7aa52	128	blkCnt = blockSize;
emilmont	1:fdd22bb7aa52	129
mbed_official	3:7a284390b0ce	130	#endif /* #ifndef ARM_MATH_CM0_FAMILY */
emilmont	1:fdd22bb7aa52	131
emilmont	1:fdd22bb7aa52	132	while(blkCnt > 0u)
emilmont	1:fdd22bb7aa52	133	{
emilmont	1:fdd22bb7aa52	134	/* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */
emilmont	1:fdd22bb7aa52	135	/* Compute sum of the squares and then store the results in a temporary variable, sum */
emilmont	1:fdd22bb7aa52	136	in = *pSrc++;
emilmont	1:fdd22bb7aa52	137	sum += (q63_t) in *in;
emilmont	1:fdd22bb7aa52	138
emilmont	1:fdd22bb7aa52	139	/* Decrement the loop counter */
emilmont	1:fdd22bb7aa52	140	blkCnt--;
emilmont	1:fdd22bb7aa52	141	}
emilmont	1:fdd22bb7aa52	142
emilmont	1:fdd22bb7aa52	143	/* Convert data in 2.62 to 1.31 by 31 right shifts and saturate */
emilmont	1:fdd22bb7aa52	144
emilmont	1:fdd22bb7aa52	145	sum = __SSAT(sum >> 31, 31);
emilmont	1:fdd22bb7aa52	146
emilmont	1:fdd22bb7aa52	147
emilmont	1:fdd22bb7aa52	148	/* Compute Rms and store the result in the destination vector */
emilmont	1:fdd22bb7aa52	149	arm_sqrt_q31((q31_t) ((q31_t) sum / (int32_t) blockSize), pResult);
emilmont	1:fdd22bb7aa52	150	}
emilmont	1:fdd22bb7aa52	151
emilmont	1:fdd22bb7aa52	152	/**
emilmont	1:fdd22bb7aa52	153	* @} end of RMS group
emilmont	1:fdd22bb7aa52	154	*/

Repository toolbox

Export to desktop IDE

Repository details

Type:	Library
Created:	23 Jun 2015
Imports:	13
Forks:	0
Commits:	6
Dependents:	1
Dependencies:	0
Followers:	58