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_add_q15.c 00009 * 00010 * Description: Q15 vector 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 0.0.7 2010/06/10 00030 * Misra-C changes done 00031 * -------------------------------------------------------------------- */ 00032 00033 #include "arm_math.h" 00034 00058 void arm_add_q15( 00059 q15_t * pSrcA, 00060 q15_t * pSrcB, 00061 q15_t * pDst, 00062 uint32_t blockSize) 00063 { 00064 uint32_t blkCnt; /* loop counter */ 00065 00066 #ifndef ARM_MATH_CM0 00067 00068 /* Run the below code for Cortex-M4 and Cortex-M3 */ 00069 00070 00071 /*loop Unrolling */ 00072 blkCnt = blockSize >> 2u; 00073 00074 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00075 ** a second loop below computes the remaining 1 to 3 samples. */ 00076 while(blkCnt > 0u) 00077 { 00078 /* C = A + B */ 00079 /* Add and then store the results in the destination buffer. */ 00080 *__SIMD32(pDst)++ = __QADD16(*__SIMD32(pSrcA)++, *__SIMD32(pSrcB)++); 00081 *__SIMD32(pDst)++ = __QADD16(*__SIMD32(pSrcA)++, *__SIMD32(pSrcB)++); 00082 00083 /* Decrement the loop counter */ 00084 blkCnt--; 00085 } 00086 00087 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00088 ** No loop unrolling is used. */ 00089 blkCnt = blockSize % 0x4u; 00090 00091 while(blkCnt > 0u) 00092 { 00093 /* C = A + B */ 00094 /* Add and then store the results in the destination buffer. */ 00095 *pDst++ = (q15_t) __QADD16(*pSrcA++, *pSrcB++); 00096 00097 /* Decrement the loop counter */ 00098 blkCnt--; 00099 } 00100 00101 #else 00102 00103 /* Run the below code for Cortex-M0 */ 00104 00105 00106 00107 /* Initialize blkCnt with number of samples */ 00108 blkCnt = blockSize; 00109 00110 while(blkCnt > 0u) 00111 { 00112 /* C = A + B */ 00113 /* Add and then store the results in the destination buffer. */ 00114 *pDst++ = (q15_t) __SSAT(((q31_t) * pSrcA++ + *pSrcB++), 16); 00115 00116 /* Decrement the loop counter */ 00117 blkCnt--; 00118 } 00119 00120 #endif /* #ifndef ARM_MATH_CM0 */ 00121 00122 00123 } 00124