00001 /* ---------------------------------------------------------------------------- 00002 * Copyright (C) 2010 ARM Limited. All rights reserved. 00003 * 00004 * $Date: 15. July 2011 00005 * $Revision: V1.0.10 00006 * 00007 * Project: CMSIS DSP Library 00008 * Title: arm_mat_add_f32.c 00009 * 00010 * Description: Floating-point matrix addition 00011 * 00012 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0 00013 * 00014 * Version 1.0.10 2011/7/15 00015 * Big Endian support added and Merged M0 and M3/M4 Source code. 00016 * 00017 * Version 1.0.3 2010/11/29 00018 * Re-organized the CMSIS folders and updated documentation. 00019 * 00020 * Version 1.0.2 2010/11/11 00021 * Documentation updated. 00022 * 00023 * Version 1.0.1 2010/10/05 00024 * Production release and review comments incorporated. 00025 * 00026 * Version 1.0.0 2010/09/20 00027 * Production release and review comments incorporated. 00028 * 00029 * Version 0.0.5 2010/04/26 00030 * incorporated review comments and updated with latest CMSIS layer 00031 * 00032 * Version 0.0.3 2010/03/10 00033 * Initial version 00034 * -------------------------------------------------------------------------- */ 00035 00036 #include "arm_math.h" 00037 00068 arm_status arm_mat_add_f32( 00069 const arm_matrix_instance_f32 * pSrcA, 00070 const arm_matrix_instance_f32 * pSrcB, 00071 arm_matrix_instance_f32 * pDst) 00072 { 00073 float32_t *pIn1 = pSrcA->pData; /* input data matrix pointer A */ 00074 float32_t *pIn2 = pSrcB->pData; /* input data matrix pointer B */ 00075 float32_t *pOut = pDst->pData; /* output data matrix pointer */ 00076 uint32_t numSamples; /* total number of elements in the matrix */ 00077 uint32_t blkCnt; /* loop counters */ 00078 arm_status status; /* status of matrix addition */ 00079 00080 #ifdef ARM_MATH_MATRIX_CHECK 00081 00082 00083 /* Check for matrix mismatch condition */ 00084 if((pSrcA->numRows != pSrcB->numRows) || 00085 (pSrcA->numCols != pSrcB->numCols) || 00086 (pSrcA->numRows != pDst->numRows) || (pSrcA->numCols != pDst->numCols)) 00087 { 00088 /* Set status as ARM_MATH_SIZE_MISMATCH */ 00089 status = ARM_MATH_SIZE_MISMATCH; 00090 } 00091 else 00092 #endif /* #ifdef ARM_MATH_MATRIX_CHECK */ 00093 00094 { 00095 00096 /* Total number of samples in the input matrix */ 00097 numSamples = (uint32_t) pSrcA->numRows * pSrcA->numCols; 00098 00099 #ifndef ARM_MATH_CM0 00100 00101 /* Run the below code for Cortex-M4 and Cortex-M3 */ 00102 00103 /* Loop unrolling */ 00104 blkCnt = numSamples >> 2u; 00105 00106 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00107 ** a second loop below computes the remaining 1 to 3 samples. */ 00108 while(blkCnt > 0u) 00109 { 00110 /* C(m,n) = A(m,n) + B(m,n) */ 00111 /* Add and then store the results in the destination buffer. */ 00112 *pOut++ = (*pIn1++) + (*pIn2++); 00113 *pOut++ = (*pIn1++) + (*pIn2++); 00114 *pOut++ = (*pIn1++) + (*pIn2++); 00115 *pOut++ = (*pIn1++) + (*pIn2++); 00116 00117 /* Decrement the loop counter */ 00118 blkCnt--; 00119 } 00120 00121 /* If the numSamples is not a multiple of 4, compute any remaining output samples here. 00122 ** No loop unrolling is used. */ 00123 blkCnt = numSamples % 0x4u; 00124 00125 #else 00126 00127 /* Run the below code for Cortex-M0 */ 00128 00129 /* Initialize blkCnt with number of samples */ 00130 blkCnt = numSamples; 00131 00132 #endif /* #ifndef ARM_MATH_CM0 */ 00133 00134 while(blkCnt > 0u) 00135 { 00136 /* C(m,n) = A(m,n) + B(m,n) */ 00137 /* Add and then store the results in the destination buffer. */ 00138 *pOut++ = (*pIn1++) + (*pIn2++); 00139 00140 /* Decrement the loop counter */ 00141 blkCnt--; 00142 } 00143 /* set status as ARM_MATH_SUCCESS */ 00144 status = ARM_MATH_SUCCESS; 00145 00146 } 00147 00148 /* Return to application */ 00149 return (status); 00150 } 00151