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_rms_q31.c 00009 * 00010 * Description: Root Mean Square of the elements of a Q31 vector. 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 00030 #include "arm_math.h" 00031 00062 void arm_rms_q31( 00063 q31_t * pSrc, 00064 uint32_t blockSize, 00065 q31_t * pResult) 00066 { 00067 q63_t sum = 0; /* accumulator */ 00068 q31_t in; /* Temporary variable to store the input */ 00069 uint32_t blkCnt; /* loop counter */ 00070 00071 #ifndef ARM_MATH_CM0 00072 00073 /* Run the below code for Cortex-M4 and Cortex-M3 */ 00074 00075 q31_t *pIn1 = pSrc; /* SrcA pointer */ 00076 00077 /*loop Unrolling */ 00078 blkCnt = blockSize >> 2u; 00079 00080 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00081 ** a second loop below computes the remaining 1 to 3 samples. */ 00082 while(blkCnt > 0u) 00083 { 00084 /* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */ 00085 /* Compute sum of the squares and then store the result in a temporary variable, sum */ 00086 in = *pIn1++; 00087 sum += (q63_t) in *in; 00088 in = *pIn1++; 00089 sum += (q63_t) in *in; 00090 in = *pIn1++; 00091 sum += (q63_t) in *in; 00092 in = *pIn1++; 00093 sum += (q63_t) in *in; 00094 00095 /* Decrement the loop counter */ 00096 blkCnt--; 00097 } 00098 00099 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00100 ** No loop unrolling is used. */ 00101 blkCnt = blockSize % 0x4u; 00102 00103 while(blkCnt > 0u) 00104 { 00105 /* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */ 00106 /* Compute sum of the squares and then store the results in a temporary variable, sum */ 00107 in = *pIn1++; 00108 sum += (q63_t) in *in; 00109 00110 /* Decrement the loop counter */ 00111 blkCnt--; 00112 } 00113 00114 #else 00115 00116 /* Run the below code for Cortex-M0 */ 00117 00118 /* Loop over blockSize number of values */ 00119 blkCnt = blockSize; 00120 00121 while(blkCnt > 0u) 00122 { 00123 /* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */ 00124 /* Compute sum of the squares and then store the results in a temporary variable, sum */ 00125 in = *pSrc++; 00126 sum += (q63_t) in *in; 00127 00128 /* Decrement the loop counter */ 00129 blkCnt--; 00130 } 00131 00132 #endif /* #ifndef ARM_MATH_CM0 */ 00133 00134 /* Convert data in 2.62 to 1.31 by 31 right shifts */ 00135 sum = sum >> 31; 00136 00137 /* Compute Rms and store the result in the destination vector */ 00138 arm_sqrt_q31((q31_t) (sum / (int32_t) blockSize), pResult); 00139 } 00140