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。


Committer:
emilmont
Date:
Wed Nov 28 12:30:09 2012 +0000
Revision:
1:fdd22bb7aa52
Child:
2:da51fb522205
DSP library code

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* ----------------------------------------------------------------------------
emilmont 1:fdd22bb7aa52 2 * Copyright (C) 2010 ARM Limited. All rights reserved.
emilmont 1:fdd22bb7aa52 3 *
emilmont 1:fdd22bb7aa52 4 * $Date: 15. February 2012
emilmont 1:fdd22bb7aa52 5 * $Revision: V1.1.0
emilmont 1:fdd22bb7aa52 6 *
emilmont 1:fdd22bb7aa52 7 * Project: CMSIS DSP Library
emilmont 1:fdd22bb7aa52 8 * Title: arm_mat_add_f32.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 1:fdd22bb7aa52 10 * Description: Floating-point matrix addition
emilmont 1:fdd22bb7aa52 11 *
emilmont 1:fdd22bb7aa52 12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
emilmont 1:fdd22bb7aa52 13 *
emilmont 1:fdd22bb7aa52 14 * Version 1.1.0 2012/02/15
emilmont 1:fdd22bb7aa52 15 * Updated with more optimizations, bug fixes and minor API changes.
emilmont 1:fdd22bb7aa52 16 *
emilmont 1:fdd22bb7aa52 17 * Version 1.0.10 2011/7/15
emilmont 1:fdd22bb7aa52 18 * Big Endian support added and Merged M0 and M3/M4 Source code.
emilmont 1:fdd22bb7aa52 19 *
emilmont 1:fdd22bb7aa52 20 * Version 1.0.3 2010/11/29
emilmont 1:fdd22bb7aa52 21 * Re-organized the CMSIS folders and updated documentation.
emilmont 1:fdd22bb7aa52 22 *
emilmont 1:fdd22bb7aa52 23 * Version 1.0.2 2010/11/11
emilmont 1:fdd22bb7aa52 24 * Documentation updated.
emilmont 1:fdd22bb7aa52 25 *
emilmont 1:fdd22bb7aa52 26 * Version 1.0.1 2010/10/05
emilmont 1:fdd22bb7aa52 27 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 28 *
emilmont 1:fdd22bb7aa52 29 * Version 1.0.0 2010/09/20
emilmont 1:fdd22bb7aa52 30 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 31 *
emilmont 1:fdd22bb7aa52 32 * Version 0.0.5 2010/04/26
emilmont 1:fdd22bb7aa52 33 * incorporated review comments and updated with latest CMSIS layer
emilmont 1:fdd22bb7aa52 34 *
emilmont 1:fdd22bb7aa52 35 * Version 0.0.3 2010/03/10
emilmont 1:fdd22bb7aa52 36 * Initial version
emilmont 1:fdd22bb7aa52 37 * -------------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 38
emilmont 1:fdd22bb7aa52 39 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 40
emilmont 1:fdd22bb7aa52 41 /**
emilmont 1:fdd22bb7aa52 42 * @ingroup groupMatrix
emilmont 1:fdd22bb7aa52 43 */
emilmont 1:fdd22bb7aa52 44
emilmont 1:fdd22bb7aa52 45 /**
emilmont 1:fdd22bb7aa52 46 * @defgroup MatrixAdd Matrix Addition
emilmont 1:fdd22bb7aa52 47 *
emilmont 1:fdd22bb7aa52 48 * Adds two matrices.
emilmont 1:fdd22bb7aa52 49 * \image html MatrixAddition.gif "Addition of two 3 x 3 matrices"
emilmont 1:fdd22bb7aa52 50 *
emilmont 1:fdd22bb7aa52 51 * The functions check to make sure that
emilmont 1:fdd22bb7aa52 52 * <code>pSrcA</code>, <code>pSrcB</code>, and <code>pDst</code> have the same
emilmont 1:fdd22bb7aa52 53 * number of rows and columns.
emilmont 1:fdd22bb7aa52 54 */
emilmont 1:fdd22bb7aa52 55
emilmont 1:fdd22bb7aa52 56 /**
emilmont 1:fdd22bb7aa52 57 * @addtogroup MatrixAdd
emilmont 1:fdd22bb7aa52 58 * @{
emilmont 1:fdd22bb7aa52 59 */
emilmont 1:fdd22bb7aa52 60
emilmont 1:fdd22bb7aa52 61
emilmont 1:fdd22bb7aa52 62 /**
emilmont 1:fdd22bb7aa52 63 * @brief Floating-point matrix addition.
emilmont 1:fdd22bb7aa52 64 * @param[in] *pSrcA points to the first input matrix structure
emilmont 1:fdd22bb7aa52 65 * @param[in] *pSrcB points to the second input matrix structure
emilmont 1:fdd22bb7aa52 66 * @param[out] *pDst points to output matrix structure
emilmont 1:fdd22bb7aa52 67 * @return The function returns either
emilmont 1:fdd22bb7aa52 68 * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.
emilmont 1:fdd22bb7aa52 69 */
emilmont 1:fdd22bb7aa52 70
emilmont 1:fdd22bb7aa52 71 arm_status arm_mat_add_f32(
emilmont 1:fdd22bb7aa52 72 const arm_matrix_instance_f32 * pSrcA,
emilmont 1:fdd22bb7aa52 73 const arm_matrix_instance_f32 * pSrcB,
emilmont 1:fdd22bb7aa52 74 arm_matrix_instance_f32 * pDst)
emilmont 1:fdd22bb7aa52 75 {
emilmont 1:fdd22bb7aa52 76 float32_t *pIn1 = pSrcA->pData; /* input data matrix pointer A */
emilmont 1:fdd22bb7aa52 77 float32_t *pIn2 = pSrcB->pData; /* input data matrix pointer B */
emilmont 1:fdd22bb7aa52 78 float32_t *pOut = pDst->pData; /* output data matrix pointer */
emilmont 1:fdd22bb7aa52 79
emilmont 1:fdd22bb7aa52 80 #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 81
emilmont 1:fdd22bb7aa52 82 float32_t inA1, inA2, inB1, inB2, out1, out2; /* temporary variables */
emilmont 1:fdd22bb7aa52 83
emilmont 1:fdd22bb7aa52 84 #endif // #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 85
emilmont 1:fdd22bb7aa52 86 uint32_t numSamples; /* total number of elements in the matrix */
emilmont 1:fdd22bb7aa52 87 uint32_t blkCnt; /* loop counters */
emilmont 1:fdd22bb7aa52 88 arm_status status; /* status of matrix addition */
emilmont 1:fdd22bb7aa52 89
emilmont 1:fdd22bb7aa52 90 #ifdef ARM_MATH_MATRIX_CHECK
emilmont 1:fdd22bb7aa52 91 /* Check for matrix mismatch condition */
emilmont 1:fdd22bb7aa52 92 if((pSrcA->numRows != pSrcB->numRows) ||
emilmont 1:fdd22bb7aa52 93 (pSrcA->numCols != pSrcB->numCols) ||
emilmont 1:fdd22bb7aa52 94 (pSrcA->numRows != pDst->numRows) || (pSrcA->numCols != pDst->numCols))
emilmont 1:fdd22bb7aa52 95 {
emilmont 1:fdd22bb7aa52 96 /* Set status as ARM_MATH_SIZE_MISMATCH */
emilmont 1:fdd22bb7aa52 97 status = ARM_MATH_SIZE_MISMATCH;
emilmont 1:fdd22bb7aa52 98 }
emilmont 1:fdd22bb7aa52 99 else
emilmont 1:fdd22bb7aa52 100 #endif
emilmont 1:fdd22bb7aa52 101 {
emilmont 1:fdd22bb7aa52 102
emilmont 1:fdd22bb7aa52 103 /* Total number of samples in the input matrix */
emilmont 1:fdd22bb7aa52 104 numSamples = (uint32_t) pSrcA->numRows * pSrcA->numCols;
emilmont 1:fdd22bb7aa52 105
emilmont 1:fdd22bb7aa52 106 #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 107
emilmont 1:fdd22bb7aa52 108 /* Loop unrolling */
emilmont 1:fdd22bb7aa52 109 blkCnt = numSamples >> 2u;
emilmont 1:fdd22bb7aa52 110
emilmont 1:fdd22bb7aa52 111 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emilmont 1:fdd22bb7aa52 112 ** a second loop below computes the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 113 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 114 {
emilmont 1:fdd22bb7aa52 115 /* C(m,n) = A(m,n) + B(m,n) */
emilmont 1:fdd22bb7aa52 116 /* Add and then store the results in the destination buffer. */
emilmont 1:fdd22bb7aa52 117 /* Read values from source A */
emilmont 1:fdd22bb7aa52 118 inA1 = pIn1[0];
emilmont 1:fdd22bb7aa52 119
emilmont 1:fdd22bb7aa52 120 /* Read values from source B */
emilmont 1:fdd22bb7aa52 121 inB1 = pIn2[0];
emilmont 1:fdd22bb7aa52 122
emilmont 1:fdd22bb7aa52 123 /* Read values from source A */
emilmont 1:fdd22bb7aa52 124 inA2 = pIn1[1];
emilmont 1:fdd22bb7aa52 125
emilmont 1:fdd22bb7aa52 126 /* out = sourceA + sourceB */
emilmont 1:fdd22bb7aa52 127 out1 = inA1 + inB1;
emilmont 1:fdd22bb7aa52 128
emilmont 1:fdd22bb7aa52 129 /* Read values from source B */
emilmont 1:fdd22bb7aa52 130 inB2 = pIn2[1];
emilmont 1:fdd22bb7aa52 131
emilmont 1:fdd22bb7aa52 132 /* Read values from source A */
emilmont 1:fdd22bb7aa52 133 inA1 = pIn1[2];
emilmont 1:fdd22bb7aa52 134
emilmont 1:fdd22bb7aa52 135 /* out = sourceA + sourceB */
emilmont 1:fdd22bb7aa52 136 out2 = inA2 + inB2;
emilmont 1:fdd22bb7aa52 137
emilmont 1:fdd22bb7aa52 138 /* Read values from source B */
emilmont 1:fdd22bb7aa52 139 inB1 = pIn2[2];
emilmont 1:fdd22bb7aa52 140
emilmont 1:fdd22bb7aa52 141 /* Store result in destination */
emilmont 1:fdd22bb7aa52 142 pOut[0] = out1;
emilmont 1:fdd22bb7aa52 143 pOut[1] = out2;
emilmont 1:fdd22bb7aa52 144
emilmont 1:fdd22bb7aa52 145 /* Read values from source A */
emilmont 1:fdd22bb7aa52 146 inA2 = pIn1[3];
emilmont 1:fdd22bb7aa52 147
emilmont 1:fdd22bb7aa52 148 /* Read values from source B */
emilmont 1:fdd22bb7aa52 149 inB2 = pIn2[3];
emilmont 1:fdd22bb7aa52 150
emilmont 1:fdd22bb7aa52 151 /* out = sourceA + sourceB */
emilmont 1:fdd22bb7aa52 152 out1 = inA1 + inB1;
emilmont 1:fdd22bb7aa52 153
emilmont 1:fdd22bb7aa52 154 /* out = sourceA + sourceB */
emilmont 1:fdd22bb7aa52 155 out2 = inA2 + inB2;
emilmont 1:fdd22bb7aa52 156
emilmont 1:fdd22bb7aa52 157 /* Store result in destination */
emilmont 1:fdd22bb7aa52 158 pOut[2] = out1;
emilmont 1:fdd22bb7aa52 159
emilmont 1:fdd22bb7aa52 160 /* Store result in destination */
emilmont 1:fdd22bb7aa52 161 pOut[3] = out2;
emilmont 1:fdd22bb7aa52 162
emilmont 1:fdd22bb7aa52 163
emilmont 1:fdd22bb7aa52 164 /* update pointers to process next sampels */
emilmont 1:fdd22bb7aa52 165 pIn1 += 4u;
emilmont 1:fdd22bb7aa52 166 pIn2 += 4u;
emilmont 1:fdd22bb7aa52 167 pOut += 4u;
emilmont 1:fdd22bb7aa52 168 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 169 blkCnt--;
emilmont 1:fdd22bb7aa52 170 }
emilmont 1:fdd22bb7aa52 171
emilmont 1:fdd22bb7aa52 172 /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 173 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 174 blkCnt = numSamples % 0x4u;
emilmont 1:fdd22bb7aa52 175
emilmont 1:fdd22bb7aa52 176 #else
emilmont 1:fdd22bb7aa52 177
emilmont 1:fdd22bb7aa52 178 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 179
emilmont 1:fdd22bb7aa52 180 /* Initialize blkCnt with number of samples */
emilmont 1:fdd22bb7aa52 181 blkCnt = numSamples;
emilmont 1:fdd22bb7aa52 182
emilmont 1:fdd22bb7aa52 183 #endif /* #ifndef ARM_MATH_CM0 */
emilmont 1:fdd22bb7aa52 184
emilmont 1:fdd22bb7aa52 185 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 186 {
emilmont 1:fdd22bb7aa52 187 /* C(m,n) = A(m,n) + B(m,n) */
emilmont 1:fdd22bb7aa52 188 /* Add and then store the results in the destination buffer. */
emilmont 1:fdd22bb7aa52 189 *pOut++ = (*pIn1++) + (*pIn2++);
emilmont 1:fdd22bb7aa52 190
emilmont 1:fdd22bb7aa52 191 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 192 blkCnt--;
emilmont 1:fdd22bb7aa52 193 }
emilmont 1:fdd22bb7aa52 194
emilmont 1:fdd22bb7aa52 195 /* set status as ARM_MATH_SUCCESS */
emilmont 1:fdd22bb7aa52 196 status = ARM_MATH_SUCCESS;
emilmont 1:fdd22bb7aa52 197
emilmont 1:fdd22bb7aa52 198 }
emilmont 1:fdd22bb7aa52 199
emilmont 1:fdd22bb7aa52 200 /* Return to application */
emilmont 1:fdd22bb7aa52 201 return (status);
emilmont 1:fdd22bb7aa52 202 }
emilmont 1:fdd22bb7aa52 203
emilmont 1:fdd22bb7aa52 204 /**
emilmont 1:fdd22bb7aa52 205 * @} end of MatrixAdd group
emilmont 1:fdd22bb7aa52 206 */