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/MatrixFunctions/arm_mat_scale_f32.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?

UserRevisionLine numberNew 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 1:fdd22bb7aa52 8 * Title: arm_mat_scale_f32.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Multiplies a floating-point matrix by a scalar.
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 * @ingroup groupMatrix
emilmont 1:fdd22bb7aa52 45 */
emilmont 1:fdd22bb7aa52 46
emilmont 1:fdd22bb7aa52 47 /**
emilmont 1:fdd22bb7aa52 48 * @defgroup MatrixScale Matrix Scale
emilmont 1:fdd22bb7aa52 49 *
emilmont 1:fdd22bb7aa52 50 * Multiplies a matrix by a scalar. This is accomplished by multiplying each element in the
emilmont 1:fdd22bb7aa52 51 * matrix by the scalar. For example:
emilmont 1:fdd22bb7aa52 52 * \image html MatrixScale.gif "Matrix Scaling of a 3 x 3 matrix"
emilmont 1:fdd22bb7aa52 53 *
emilmont 1:fdd22bb7aa52 54 * The function checks to make sure that the input and output matrices are of the same size.
emilmont 1:fdd22bb7aa52 55 *
emilmont 1:fdd22bb7aa52 56 * In the fixed-point Q15 and Q31 functions, <code>scale</code> is represented by
emilmont 1:fdd22bb7aa52 57 * a fractional multiplication <code>scaleFract</code> and an arithmetic shift <code>shift</code>.
emilmont 1:fdd22bb7aa52 58 * The shift allows the gain of the scaling operation to exceed 1.0.
emilmont 1:fdd22bb7aa52 59 * The overall scale factor applied to the fixed-point data is
emilmont 1:fdd22bb7aa52 60 * <pre>
emilmont 1:fdd22bb7aa52 61 * scale = scaleFract * 2^shift.
emilmont 1:fdd22bb7aa52 62 * </pre>
emilmont 1:fdd22bb7aa52 63 */
emilmont 1:fdd22bb7aa52 64
emilmont 1:fdd22bb7aa52 65 /**
emilmont 1:fdd22bb7aa52 66 * @addtogroup MatrixScale
emilmont 1:fdd22bb7aa52 67 * @{
emilmont 1:fdd22bb7aa52 68 */
emilmont 1:fdd22bb7aa52 69
emilmont 1:fdd22bb7aa52 70 /**
emilmont 1:fdd22bb7aa52 71 * @brief Floating-point matrix scaling.
emilmont 1:fdd22bb7aa52 72 * @param[in] *pSrc points to input matrix structure
emilmont 1:fdd22bb7aa52 73 * @param[in] scale scale factor to be applied
emilmont 1:fdd22bb7aa52 74 * @param[out] *pDst points to output matrix structure
emilmont 2:da51fb522205 75 * @return The function returns either <code>ARM_MATH_SIZE_MISMATCH</code>
emilmont 1:fdd22bb7aa52 76 * or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.
emilmont 1:fdd22bb7aa52 77 *
emilmont 1:fdd22bb7aa52 78 */
emilmont 1:fdd22bb7aa52 79
emilmont 1:fdd22bb7aa52 80 arm_status arm_mat_scale_f32(
emilmont 1:fdd22bb7aa52 81 const arm_matrix_instance_f32 * pSrc,
emilmont 1:fdd22bb7aa52 82 float32_t scale,
emilmont 1:fdd22bb7aa52 83 arm_matrix_instance_f32 * pDst)
emilmont 1:fdd22bb7aa52 84 {
emilmont 1:fdd22bb7aa52 85 float32_t *pIn = pSrc->pData; /* input data matrix pointer */
emilmont 1:fdd22bb7aa52 86 float32_t *pOut = pDst->pData; /* output data matrix pointer */
emilmont 1:fdd22bb7aa52 87 uint32_t numSamples; /* total number of elements in the matrix */
emilmont 1:fdd22bb7aa52 88 uint32_t blkCnt; /* loop counters */
emilmont 1:fdd22bb7aa52 89 arm_status status; /* status of matrix scaling */
emilmont 1:fdd22bb7aa52 90
mbed_official 3:7a284390b0ce 91 #ifndef ARM_MATH_CM0_FAMILY
emilmont 1:fdd22bb7aa52 92
emilmont 1:fdd22bb7aa52 93 float32_t in1, in2, in3, in4; /* temporary variables */
emilmont 1:fdd22bb7aa52 94 float32_t out1, out2, out3, out4; /* temporary variables */
emilmont 1:fdd22bb7aa52 95
mbed_official 3:7a284390b0ce 96 #endif // #ifndef ARM_MATH_CM0_FAMILY
emilmont 1:fdd22bb7aa52 97
emilmont 1:fdd22bb7aa52 98 #ifdef ARM_MATH_MATRIX_CHECK
emilmont 1:fdd22bb7aa52 99 /* Check for matrix mismatch condition */
emilmont 1:fdd22bb7aa52 100 if((pSrc->numRows != pDst->numRows) || (pSrc->numCols != pDst->numCols))
emilmont 1:fdd22bb7aa52 101 {
emilmont 1:fdd22bb7aa52 102 /* Set status as ARM_MATH_SIZE_MISMATCH */
emilmont 1:fdd22bb7aa52 103 status = ARM_MATH_SIZE_MISMATCH;
emilmont 1:fdd22bb7aa52 104 }
emilmont 1:fdd22bb7aa52 105 else
emilmont 1:fdd22bb7aa52 106 #endif /* #ifdef ARM_MATH_MATRIX_CHECK */
emilmont 1:fdd22bb7aa52 107 {
emilmont 1:fdd22bb7aa52 108 /* Total number of samples in the input matrix */
emilmont 1:fdd22bb7aa52 109 numSamples = (uint32_t) pSrc->numRows * pSrc->numCols;
emilmont 1:fdd22bb7aa52 110
mbed_official 3:7a284390b0ce 111 #ifndef ARM_MATH_CM0_FAMILY
emilmont 1:fdd22bb7aa52 112
emilmont 1:fdd22bb7aa52 113 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 114
emilmont 1:fdd22bb7aa52 115 /* Loop Unrolling */
emilmont 1:fdd22bb7aa52 116 blkCnt = numSamples >> 2;
emilmont 1:fdd22bb7aa52 117
emilmont 1:fdd22bb7aa52 118 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emilmont 1:fdd22bb7aa52 119 ** a second loop below computes the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 120 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 121 {
emilmont 1:fdd22bb7aa52 122 /* C(m,n) = A(m,n) * scale */
emilmont 1:fdd22bb7aa52 123 /* Scaling and results are stored in the destination buffer. */
emilmont 1:fdd22bb7aa52 124 in1 = pIn[0];
emilmont 1:fdd22bb7aa52 125 in2 = pIn[1];
emilmont 1:fdd22bb7aa52 126 in3 = pIn[2];
emilmont 1:fdd22bb7aa52 127 in4 = pIn[3];
emilmont 1:fdd22bb7aa52 128
emilmont 1:fdd22bb7aa52 129 out1 = in1 * scale;
emilmont 1:fdd22bb7aa52 130 out2 = in2 * scale;
emilmont 1:fdd22bb7aa52 131 out3 = in3 * scale;
emilmont 1:fdd22bb7aa52 132 out4 = in4 * scale;
emilmont 1:fdd22bb7aa52 133
emilmont 1:fdd22bb7aa52 134
emilmont 1:fdd22bb7aa52 135 pOut[0] = out1;
emilmont 1:fdd22bb7aa52 136 pOut[1] = out2;
emilmont 1:fdd22bb7aa52 137 pOut[2] = out3;
emilmont 1:fdd22bb7aa52 138 pOut[3] = out4;
emilmont 1:fdd22bb7aa52 139
emilmont 1:fdd22bb7aa52 140 /* update pointers to process next sampels */
emilmont 1:fdd22bb7aa52 141 pIn += 4u;
emilmont 1:fdd22bb7aa52 142 pOut += 4u;
emilmont 1:fdd22bb7aa52 143
emilmont 1:fdd22bb7aa52 144 /* Decrement the numSamples loop counter */
emilmont 1:fdd22bb7aa52 145 blkCnt--;
emilmont 1:fdd22bb7aa52 146 }
emilmont 1:fdd22bb7aa52 147
emilmont 1:fdd22bb7aa52 148 /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 149 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 150 blkCnt = numSamples % 0x4u;
emilmont 1:fdd22bb7aa52 151
emilmont 1:fdd22bb7aa52 152 #else
emilmont 1:fdd22bb7aa52 153
emilmont 1:fdd22bb7aa52 154 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 155
emilmont 1:fdd22bb7aa52 156 /* Initialize blkCnt with number of samples */
emilmont 1:fdd22bb7aa52 157 blkCnt = numSamples;
emilmont 1:fdd22bb7aa52 158
mbed_official 3:7a284390b0ce 159 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
emilmont 1:fdd22bb7aa52 160
emilmont 1:fdd22bb7aa52 161 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 162 {
emilmont 1:fdd22bb7aa52 163 /* C(m,n) = A(m,n) * scale */
emilmont 1:fdd22bb7aa52 164 /* The results are stored in the destination buffer. */
emilmont 1:fdd22bb7aa52 165 *pOut++ = (*pIn++) * scale;
emilmont 1:fdd22bb7aa52 166
emilmont 1:fdd22bb7aa52 167 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 168 blkCnt--;
emilmont 1:fdd22bb7aa52 169 }
emilmont 1:fdd22bb7aa52 170
emilmont 1:fdd22bb7aa52 171 /* Set status as ARM_MATH_SUCCESS */
emilmont 1:fdd22bb7aa52 172 status = ARM_MATH_SUCCESS;
emilmont 1:fdd22bb7aa52 173 }
emilmont 1:fdd22bb7aa52 174
emilmont 1:fdd22bb7aa52 175 /* Return to application */
emilmont 1:fdd22bb7aa52 176 return (status);
emilmont 1:fdd22bb7aa52 177 }
emilmont 1:fdd22bb7aa52 178
emilmont 1:fdd22bb7aa52 179 /**
emilmont 1:fdd22bb7aa52 180 * @} end of MatrixScale group
emilmont 1:fdd22bb7aa52 181 */