arm_dot_prod_q15.c

00001 /* ----------------------------------------------------------------------  
00002 * Copyright (C) 2010 ARM Limited. All rights reserved.  
00003 *  
00004 * $Date:        29. November 2010  
00005 * $Revision:    V1.0.3  
00006 *  
00007 * Project:      CMSIS DSP Library  
00008 * Title:        arm_dot_prod_q15.c  
00009 *  
00010 * Description:  Q15 dot product.  
00011 *  
00012 * Target Processor: Cortex-M4/Cortex-M3
00013 *  
00014 * Version 1.0.3 2010/11/29 
00015 *    Re-organized the CMSIS folders and updated documentation.  
00016 *   
00017 * Version 1.0.2 2010/11/11  
00018 *    Documentation updated.   
00019 *  
00020 * Version 1.0.1 2010/10/05   
00021 *    Production release and review comments incorporated.  
00022 *  
00023 * Version 1.0.0 2010/09/20   
00024 *    Production release and review comments incorporated.  
00025 *  
00026 * Version 0.0.7  2010/06/10   
00027 *    Misra-C changes done  
00028 * -------------------------------------------------------------------- */ 
00029  
00030 #include "arm_math.h" 
00031  
00032 /**  
00033  * @ingroup groupMath  
00034  */ 
00035  
00036 /**  
00037  * @addtogroup dot_prod  
00038  * @{  
00039  */ 
00040  
00041 /**  
00042  * @brief Dot product of Q15 vectors.  
00043  * @param[in]       *pSrcA points to the first input vector  
00044  * @param[in]       *pSrcB points to the second input vector  
00045  * @param[in]       blockSize number of samples in each vector  
00046  * @param[out]      *result output result returned here  
00047  * @return none.  
00048  *  
00049  * <b>Scaling and Overflow Behavior:</b>  
00050  * \par  
00051  * The intermediate multiplications are in 1.15 x 1.15 = 2.30 format and these  
00052  * results are added to a 64-bit accumulator in 34.30 format.  
00053  * Nonsaturating additions are used and given that there are 33 guard bits in the accumulator  
00054  * there is no risk of overflow.  
00055  * The return result is in 34.30 format.  
00056  */ 
00057  
00058 void arm_dot_prod_q15( 
00059   q15_t * pSrcA, 
00060   q15_t * pSrcB, 
00061   uint32_t blockSize, 
00062   q63_t * result) 
00063 { 
00064   q63_t sum = 0;                                 /* Temporary result storage */ 
00065   uint32_t blkCnt;                               /* loop counter */ 
00066  
00067  
00068   /*loop Unrolling */ 
00069   blkCnt = blockSize >> 2u; 
00070  
00071   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.  
00072    ** a second loop below computes the remaining 1 to 3 samples. */ 
00073   while(blkCnt > 0u) 
00074   { 
00075     /* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */ 
00076     /* Calculate dot product and then store the result in a temporary buffer. */ 
00077     sum = __SMLALD(*__SIMD32(pSrcA)++, *__SIMD32(pSrcB)++, sum); 
00078     sum = __SMLALD(*__SIMD32(pSrcA)++, *__SIMD32(pSrcB)++, sum); 
00079  
00080     /* Decrement the loop counter */ 
00081     blkCnt--; 
00082   } 
00083  
00084   /* If the blockSize is not a multiple of 4, compute any remaining output samples here.  
00085    ** No loop unrolling is used. */ 
00086   blkCnt = blockSize % 0x4u; 
00087  
00088   while(blkCnt > 0u) 
00089   { 
00090     /* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */ 
00091     /* Calculate dot product and then store the results in a temporary buffer. */ 
00092     sum = __SMLALD(*pSrcA++, *pSrcB++, sum); 
00093  
00094     /* Decrement the loop counter */ 
00095     blkCnt--; 
00096   } 
00097  
00098   /* Store the result in the destination buffer in 34.30 format */ 
00099   *result = sum; 
00100 } 
00101  
00102 /**  
00103  * @} end of dot_prod group  
00104  */