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_mult_q31.c 00009 * 00010 * Description: Q31 vector multiplication. 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 0.0.5 2010/04/26 00030 * incorporated review comments and updated with latest CMSIS layer 00031 * 00032 * Version 0.0.3 2010/03/10 00033 * Initial version 00034 * -------------------------------------------------------------------- */ 00035 00036 #include "arm_math.h" 00037 00061 void arm_mult_q31( 00062 q31_t * pSrcA, 00063 q31_t * pSrcB, 00064 q31_t * pDst, 00065 uint32_t blockSize) 00066 { 00067 uint32_t blkCnt; /* loop counters */ 00068 00069 #ifndef ARM_MATH_CM0 00070 00071 /* Run the below code for Cortex-M4 and Cortex-M3 */ 00072 /* loop Unrolling */ 00073 blkCnt = blockSize >> 2u; 00074 00075 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00076 ** a second loop below computes the remaining 1 to 3 samples. */ 00077 while(blkCnt > 0u) 00078 { 00079 /* C = A * B */ 00080 /* Multiply the inputs and then store the results in the destination buffer. */ 00081 *pDst++ = 00082 (q31_t) clip_q63_to_q31(((q63_t) (*pSrcA++) * (*pSrcB++)) >> 31); 00083 *pDst++ = 00084 (q31_t) clip_q63_to_q31(((q63_t) (*pSrcA++) * (*pSrcB++)) >> 31); 00085 *pDst++ = 00086 (q31_t) clip_q63_to_q31(((q63_t) (*pSrcA++) * (*pSrcB++)) >> 31); 00087 *pDst++ = 00088 (q31_t) clip_q63_to_q31(((q63_t) (*pSrcA++) * (*pSrcB++)) >> 31); 00089 00090 /* Decrement the blockSize loop counter */ 00091 blkCnt--; 00092 } 00093 00094 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00095 ** No loop unrolling is used. */ 00096 blkCnt = blockSize % 0x4u; 00097 00098 #else 00099 00100 /* Run the below code for Cortex-M0 */ 00101 00102 /* Initialize blkCnt with number of samples */ 00103 blkCnt = blockSize; 00104 00105 #endif /* #ifndef ARM_MATH_CM0 */ 00106 00107 while(blkCnt > 0u) 00108 { 00109 /* C = A * B */ 00110 /* Multiply the inputs and then store the results in the destination buffer. */ 00111 *pDst++ = 00112 (q31_t) clip_q63_to_q31(((q63_t) (*pSrcA++) * (*pSrcB++)) >> 31); 00113 00114 /* Decrement the blockSize loop counter */ 00115 blkCnt--; 00116 } 00117 } 00118