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_offset_q15.c 00009 * 00010 * Description: Q15 vector offset. 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_offset_q15( 00059 q15_t * pSrc, 00060 q15_t offset, 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 q31_t offset_packed; /* Offset packed to 32 bit */ 00070 00071 00072 /*loop Unrolling */ 00073 blkCnt = blockSize >> 2u; 00074 00075 /* Offset is packed to 32 bit in order to use SIMD32 for addition */ 00076 offset_packed = __PKHBT(offset, offset, 16); 00077 00078 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. 00079 ** a second loop below computes the remaining 1 to 3 samples. */ 00080 while(blkCnt > 0u) 00081 { 00082 /* C = A + offset */ 00083 /* Add offset and then store the results in the destination buffer, 2 samples at a time. */ 00084 *__SIMD32(pDst)++ = __QADD16(*__SIMD32(pSrc)++, offset_packed); 00085 *__SIMD32(pDst)++ = __QADD16(*__SIMD32(pSrc)++, offset_packed); 00086 00087 /* Decrement the loop counter */ 00088 blkCnt--; 00089 } 00090 00091 /* If the blockSize is not a multiple of 4, compute any remaining output samples here. 00092 ** No loop unrolling is used. */ 00093 blkCnt = blockSize % 0x4u; 00094 00095 while(blkCnt > 0u) 00096 { 00097 /* C = A + offset */ 00098 /* Add offset and then store the results in the destination buffer. */ 00099 *pDst++ = (q15_t) __QADD16(*pSrc++, offset); 00100 00101 /* Decrement the loop counter */ 00102 blkCnt--; 00103 } 00104 00105 #else 00106 00107 /* Run the below code for Cortex-M0 */ 00108 00109 /* Initialize blkCnt with number of samples */ 00110 blkCnt = blockSize; 00111 00112 while(blkCnt > 0u) 00113 { 00114 /* C = A + offset */ 00115 /* Add offset and then store the results in the destination buffer. */ 00116 *pDst++ = (q15_t) __SSAT(((q31_t) * pSrc++ + offset), 16); 00117 00118 /* Decrement the loop counter */ 00119 blkCnt--; 00120 } 00121 00122 #endif /* #ifndef ARM_MATH_CM0 */ 00123 00124 } 00125