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_sub_q7.c 00009 * 00010 * Description: Q7 vector subtraction. 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_sub_q7( 00059 q7_t * pSrcA, 00060 q7_t * pSrcB, 00061 q7_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 /*loop Unrolling */ 00071 blkCnt = blockSize >> 2u; 00072 00073 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00074 ** a second loop below computes the remaining 1 to 3 samples. */ 00075 while(blkCnt > 0u) 00076 { 00077 /* C = A - B */ 00078 /* Subtract and then store the results in the destination buffer 4 samples at a time. */ 00079 *__SIMD32(pDst)++ = __QSUB8(*__SIMD32(pSrcA)++, *__SIMD32(pSrcB)++); 00080 00081 /* Decrement the loop counter */ 00082 blkCnt--; 00083 } 00084 00085 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00086 ** No loop unrolling is used. */ 00087 blkCnt = blockSize % 0x4u; 00088 00089 while(blkCnt > 0u) 00090 { 00091 /* C = A - B */ 00092 /* Subtract and then store the result in the destination buffer. */ 00093 *pDst++ = __SSAT(*pSrcA++ - *pSrcB++, 8); 00094 00095 /* Decrement the loop counter */ 00096 blkCnt--; 00097 } 00098 00099 #else 00100 00101 /* Run the below code for Cortex-M0 */ 00102 00103 /* Initialize blkCnt with number of samples */ 00104 blkCnt = blockSize; 00105 00106 while(blkCnt > 0u) 00107 { 00108 /* C = A - B */ 00109 /* Subtract and then store the result in the destination buffer. */ 00110 *pDst++ = (q7_t) __SSAT((q15_t) * pSrcA++ - *pSrcB++, 8); 00111 00112 /* Decrement the loop counter */ 00113 blkCnt--; 00114 } 00115 00116 #endif /* #ifndef ARM_MATH_CM0 */ 00117 00118 00119 } 00120