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。


dsp/FIR_f32.h@5:a912b042151f, 2015-06-23 (annotated)

Committer:
1050186
Date:
Tue Jun 23 06:23:42 2015 +0000
Revision:
5:a912b042151f
Parent:
1:fdd22bb7aa52 
first commit

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* mbed Microcontroller Library
emilmont 1:fdd22bb7aa52 2 * Copyright (c) 2006-2012 ARM Limited
emilmont 1:fdd22bb7aa52 3 *
emilmont 1:fdd22bb7aa52 4 * Permission is hereby granted, free of charge, to any person obtaining a copy
emilmont 1:fdd22bb7aa52 5 * of this software and associated documentation files (the "Software"), to deal
emilmont 1:fdd22bb7aa52 6 * in the Software without restriction, including without limitation the rights
emilmont 1:fdd22bb7aa52 7 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
emilmont 1:fdd22bb7aa52 8 * copies of the Software, and to permit persons to whom the Software is
emilmont 1:fdd22bb7aa52 9 * furnished to do so, subject to the following conditions:
emilmont 1:fdd22bb7aa52 10 *
emilmont 1:fdd22bb7aa52 11 * The above copyright notice and this permission notice shall be included in
emilmont 1:fdd22bb7aa52 12 * all copies or substantial portions of the Software.
emilmont 1:fdd22bb7aa52 13 *
emilmont 1:fdd22bb7aa52 14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
emilmont 1:fdd22bb7aa52 15 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
emilmont 1:fdd22bb7aa52 16 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
emilmont 1:fdd22bb7aa52 17 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
emilmont 1:fdd22bb7aa52 18 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
emilmont 1:fdd22bb7aa52 19 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
emilmont 1:fdd22bb7aa52 20 * SOFTWARE.
emilmont 1:fdd22bb7aa52 21 */
emilmont 1:fdd22bb7aa52 22 #ifndef FIR_F32_H
emilmont 1:fdd22bb7aa52 23 #define FIR_F32_H
emilmont 1:fdd22bb7aa52 24
emilmont 1:fdd22bb7aa52 25 #include <stdint.h>
emilmont 1:fdd22bb7aa52 26 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 27
emilmont 1:fdd22bb7aa52 28 namespace dsp {
emilmont 1:fdd22bb7aa52 29
emilmont 1:fdd22bb7aa52 30 template<uint16_t num_taps, uint32_t block_size=32>
emilmont 1:fdd22bb7aa52 31 class FIR_f32 {
emilmont 1:fdd22bb7aa52 32 public:
emilmont 1:fdd22bb7aa52 33 FIR_f32(const float32_t *coeff) {
emilmont 1:fdd22bb7aa52 34 arm_fir_init_f32(&fir, num_taps, (float32_t*)coeff, fir_state, block_size);
emilmont 1:fdd22bb7aa52 35 }
emilmont 1:fdd22bb7aa52 36
emilmont 1:fdd22bb7aa52 37 void process(float32_t *sgn_in, float32_t *sgn_out) {
emilmont 1:fdd22bb7aa52 38 arm_fir_f32(&fir, sgn_in, sgn_out, block_size);
emilmont 1:fdd22bb7aa52 39 }
emilmont 1:fdd22bb7aa52 40
emilmont 1:fdd22bb7aa52 41 void reset(void) {
emilmont 1:fdd22bb7aa52 42 memset(fir_state, 0, sizeof(fir_state));
emilmont 1:fdd22bb7aa52 43 }
emilmont 1:fdd22bb7aa52 44
emilmont 1:fdd22bb7aa52 45 private:
emilmont 1:fdd22bb7aa52 46 arm_fir_instance_f32 fir;
emilmont 1:fdd22bb7aa52 47 float32_t fir_state[block_size + num_taps - 1];
emilmont 1:fdd22bb7aa52 48 };
emilmont 1:fdd22bb7aa52 49
emilmont 1:fdd22bb7aa52 50 }
emilmont 1:fdd22bb7aa52 51 #endif