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_q15.c 00009 * 00010 * Description: Root Mean Square of the elements of a Q15 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 00059 void arm_rms_q15( 00060 q15_t * pSrc, 00061 uint32_t blockSize, 00062 q15_t * pResult) 00063 { 00064 q63_t sum = 0; /* accumulator */ 00065 00066 #ifndef ARM_MATH_CM0 00067 00068 /* Run the below code for Cortex-M4 and Cortex-M3 */ 00069 00070 q31_t in; /* temporary variable to store the input value */ 00071 q15_t in1; /* temporary variable to store the input value */ 00072 uint32_t blkCnt; /* loop counter */ 00073 00074 /* loop Unrolling */ 00075 blkCnt = blockSize >> 2u; 00076 00077 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00078 ** a second loop below computes the remaining 1 to 3 samples. */ 00079 while(blkCnt > 0u) 00080 { 00081 /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */ 00082 /* Compute sum of the squares and then store the results in a temporary variable, sum */ 00083 in = *__SIMD32(pSrc)++; 00084 sum = __SMLALD(in, in, sum); 00085 in = *__SIMD32(pSrc)++; 00086 sum = __SMLALD(in, in, sum); 00087 00088 /* Decrement the loop counter */ 00089 blkCnt--; 00090 } 00091 00092 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00093 ** No loop unrolling is used. */ 00094 blkCnt = blockSize % 0x4u; 00095 00096 while(blkCnt > 0u) 00097 { 00098 /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */ 00099 /* Compute sum of the squares and then store the results in a temporary variable, sum */ 00100 in1 = *pSrc++; 00101 sum = __SMLALD(in1, in1, sum); 00102 00103 /* Decrement the loop counter */ 00104 blkCnt--; 00105 } 00106 00107 /* Truncating and saturating the accumulator to 1.15 format */ 00108 sum = __SSAT((q31_t) (sum >> 15), 16); 00109 00110 in1 = (q15_t) (sum / blockSize); 00111 00112 /* Store the result in the destination */ 00113 arm_sqrt_q15(in1, pResult); 00114 00115 #else 00116 00117 /* Run the below code for Cortex-M0 */ 00118 00119 q15_t in; /* temporary variable to store the input value */ 00120 uint32_t blkCnt; /* loop counter */ 00121 00122 /* Loop over blockSize number of values */ 00123 blkCnt = blockSize; 00124 00125 while(blkCnt > 0u) 00126 { 00127 /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */ 00128 /* Compute sum of the squares and then store the results in a temporary variable, sum */ 00129 in = *pSrc++; 00130 sum += ((q31_t) in * in); 00131 00132 /* Decrement the loop counter */ 00133 blkCnt--; 00134 } 00135 00136 /* Truncating and saturating the accumulator to 1.15 format */ 00137 sum = __SSAT((q31_t) (sum >> 15), 16); 00138 00139 in = (q15_t) (sum / blockSize); 00140 00141 /* Store the result in the destination */ 00142 arm_sqrt_q15(in, pResult); 00143 00144 #endif /* #ifndef ARM_MATH_CM0 */ 00145 00146 } 00147