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_power_q7.c 00009 * 00010 * Description: Sum of the squares of the elements of a Q7 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_power_q7( 00063 q7_t * pSrc, 00064 uint32_t blockSize, 00065 q31_t * pResult) 00066 { 00067 q31_t sum = 0; /* Temporary result storage */ 00068 q7_t in; /* Temporary variable to store 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 input1; /* Temporary variable to store packed input */ 00076 q15_t in1, in2; /* Temporary variables to store input */ 00077 00078 /*loop Unrolling */ 00079 blkCnt = blockSize >> 2u; 00080 00081 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00082 ** a second loop below computes the remaining 1 to 3 samples. */ 00083 while(blkCnt > 0u) 00084 { 00085 /* Reading two inputs of pSrc vector and packing */ 00086 in1 = (q15_t) * pSrc++; 00087 in2 = (q15_t) * pSrc++; 00088 input1 = ((q31_t) in1 & 0x0000FFFF) | ((q31_t) in2 << 16); 00089 00090 /* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */ 00091 /* Compute Power and then store the result in a temporary variable, sum. */ 00092 sum = __SMLAD(input1, input1, sum); 00093 00094 /* Reading two inputs of pSrc vector and packing */ 00095 in1 = (q15_t) * pSrc++; 00096 in2 = (q15_t) * pSrc++; 00097 input1 = ((q31_t) in1 & 0x0000FFFF) | ((q31_t) in2 << 16); 00098 00099 /* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */ 00100 /* Compute Power and then store the result in a temporary variable, sum. */ 00101 sum = __SMLAD(input1, input1, sum); 00102 00103 /* Decrement the loop counter */ 00104 blkCnt--; 00105 } 00106 00107 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00108 ** No loop unrolling is used. */ 00109 blkCnt = blockSize % 0x4u; 00110 00111 #else 00112 00113 /* Run the below code for Cortex-M0 */ 00114 00115 /* Loop over blockSize number of values */ 00116 blkCnt = blockSize; 00117 00118 #endif /* #ifndef ARM_MATH_CM0 */ 00119 00120 while(blkCnt > 0u) 00121 { 00122 /* C = A[0] * A[0] + A[1] * A[1] + A[2] * A[2] + ... + A[blockSize-1] * A[blockSize-1] */ 00123 /* Compute Power and then store the result in a temporary variable, sum. */ 00124 in = *pSrc++; 00125 sum += ((q15_t) in * in); 00126 00127 /* Decrement the loop counter */ 00128 blkCnt--; 00129 } 00130 00131 /* Store the result in 18.14 format */ 00132 *pResult = sum; 00133 } 00134