Fork of mbed-dsp. CMSIS-DSP library of supporting NEON
Dependents: mbed-os-example-cmsis_dsp_neon
Fork of mbed-dsp by
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_add_f32.c
- Committer:
- emilmont
- Date:
- 2013-05-30
- Revision:
- 2:da51fb522205
- Parent:
- 1:fdd22bb7aa52
- Child:
- 3:7a284390b0ce
File content as of revision 2:da51fb522205:
/* ---------------------------------------------------------------------------- * Copyright (C) 2010 ARM Limited. All rights reserved. * * $Date: 15. February 2012 * $Revision: V1.1.0 * * Project: CMSIS DSP Library * Title: arm_mat_add_f32.c * * Description: Floating-point matrix addition * * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0 * * Version 1.1.0 2012/02/15 * Updated with more optimizations, bug fixes and minor API changes. * * Version 1.0.10 2011/7/15 * Big Endian support added and Merged M0 and M3/M4 Source code. * * Version 1.0.3 2010/11/29 * Re-organized the CMSIS folders and updated documentation. * * Version 1.0.2 2010/11/11 * Documentation updated. * * Version 1.0.1 2010/10/05 * Production release and review comments incorporated. * * Version 1.0.0 2010/09/20 * Production release and review comments incorporated. * * Version 0.0.5 2010/04/26 * incorporated review comments and updated with latest CMSIS layer * * Version 0.0.3 2010/03/10 * Initial version * -------------------------------------------------------------------------- */ #include "arm_math.h" /** * @ingroup groupMatrix */ /** * @defgroup MatrixAdd Matrix Addition * * Adds two matrices. * \image html MatrixAddition.gif "Addition of two 3 x 3 matrices" * * The functions check to make sure that * <code>pSrcA</code>, <code>pSrcB</code>, and <code>pDst</code> have the same * number of rows and columns. */ /** * @addtogroup MatrixAdd * @{ */ /** * @brief Floating-point matrix addition. * @param[in] *pSrcA points to the first input matrix structure * @param[in] *pSrcB points to the second input matrix structure * @param[out] *pDst points to output matrix structure * @return The function returns either * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking. */ arm_status arm_mat_add_f32( const arm_matrix_instance_f32 * pSrcA, const arm_matrix_instance_f32 * pSrcB, arm_matrix_instance_f32 * pDst) { float32_t *pIn1 = pSrcA->pData; /* input data matrix pointer A */ float32_t *pIn2 = pSrcB->pData; /* input data matrix pointer B */ float32_t *pOut = pDst->pData; /* output data matrix pointer */ #ifndef ARM_MATH_CM0 float32_t inA1, inA2, inB1, inB2, out1, out2; /* temporary variables */ #endif // #ifndef ARM_MATH_CM0 uint32_t numSamples; /* total number of elements in the matrix */ uint32_t blkCnt; /* loop counters */ arm_status status; /* status of matrix addition */ #ifdef ARM_MATH_MATRIX_CHECK /* Check for matrix mismatch condition */ if((pSrcA->numRows != pSrcB->numRows) || (pSrcA->numCols != pSrcB->numCols) || (pSrcA->numRows != pDst->numRows) || (pSrcA->numCols != pDst->numCols)) { /* Set status as ARM_MATH_SIZE_MISMATCH */ status = ARM_MATH_SIZE_MISMATCH; } else #endif { /* Total number of samples in the input matrix */ numSamples = (uint32_t) pSrcA->numRows * pSrcA->numCols; #ifndef ARM_MATH_CM0 /* Loop unrolling */ blkCnt = numSamples >> 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(m,n) = A(m,n) + B(m,n) */ /* Add and then store the results in the destination buffer. */ /* Read values from source A */ inA1 = pIn1[0]; /* Read values from source B */ inB1 = pIn2[0]; /* Read values from source A */ inA2 = pIn1[1]; /* out = sourceA + sourceB */ out1 = inA1 + inB1; /* Read values from source B */ inB2 = pIn2[1]; /* Read values from source A */ inA1 = pIn1[2]; /* out = sourceA + sourceB */ out2 = inA2 + inB2; /* Read values from source B */ inB1 = pIn2[2]; /* Store result in destination */ pOut[0] = out1; pOut[1] = out2; /* Read values from source A */ inA2 = pIn1[3]; /* Read values from source B */ inB2 = pIn2[3]; /* out = sourceA + sourceB */ out1 = inA1 + inB1; /* out = sourceA + sourceB */ out2 = inA2 + inB2; /* Store result in destination */ pOut[2] = out1; /* Store result in destination */ pOut[3] = out2; /* update pointers to process next sampels */ pIn1 += 4u; pIn2 += 4u; pOut += 4u; /* Decrement the loop counter */ blkCnt--; } /* If the numSamples is not a multiple of 4, compute any remaining output samples here. ** No loop unrolling is used. */ blkCnt = numSamples % 0x4u; #else /* Run the below code for Cortex-M0 */ /* Initialize blkCnt with number of samples */ blkCnt = numSamples; #endif /* #ifndef ARM_MATH_CM0 */ while(blkCnt > 0u) { /* C(m,n) = A(m,n) + B(m,n) */ /* Add and then store the results in the destination buffer. */ *pOut++ = (*pIn1++) + (*pIn2++); /* Decrement the loop counter */ blkCnt--; } /* set status as ARM_MATH_SUCCESS */ status = ARM_MATH_SUCCESS; } /* Return to application */ return (status); } /** * @} end of MatrixAdd group */