arm_sub_q7.c
1 /* ---------------------------------------------------------------------- 2 * Project: CMSIS DSP Library 3 * Title: arm_sub_q7.c 4 * Description: Q7 vector subtraction 5 * 6 * $Date: 23 April 2021 7 * $Revision: V1.9.0 8 * 9 * Target Processor: Cortex-M and Cortex-A cores 10 * -------------------------------------------------------------------- */ 11 /* 12 * Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved. 13 * 14 * SPDX-License-Identifier: Apache-2.0 15 * 16 * Licensed under the Apache License, Version 2.0 (the License); you may 17 * not use this file except in compliance with the License. 18 * You may obtain a copy of the License at 19 * 20 * www.apache.org/licenses/LICENSE-2.0 21 * 22 * Unless required by applicable law or agreed to in writing, software 23 * distributed under the License is distributed on an AS IS BASIS, WITHOUT 24 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 25 * See the License for the specific language governing permissions and 26 * limitations under the License. 27 */ 28 29 #include "dsp/basic_math_functions.h" 30 31 /** 32 @ingroup groupMath 33 */ 34 35 /** 36 @addtogroup BasicSub 37 @{ 38 */ 39 40 /** 41 @brief Q7 vector subtraction. 42 @param[in] pSrcA points to the first input vector 43 @param[in] pSrcB points to the second input vector 44 @param[out] pDst points to the output vector 45 @param[in] blockSize number of samples in each vector 46 @return none 47 48 @par Scaling and Overflow Behavior 49 The function uses saturating arithmetic. 50 Results outside of the allowable Q7 range [0x80 0x7F] will be saturated. 51 */ 52 #if defined(ARM_MATH_MVEI) && !defined(ARM_MATH_AUTOVECTORIZE) 53 54 #include "arm_helium_utils.h" 55 56 void arm_sub_q7( 57 const q7_t * pSrcA, 58 const q7_t * pSrcB, 59 q7_t * pDst, 60 uint32_t blockSize) 61 { 62 uint32_t blkCnt; /* loop counters */ 63 q7x16_t vecA; 64 q7x16_t vecB; 65 66 /* Compute 16 outputs at a time */ 67 blkCnt = blockSize >> 4; 68 while (blkCnt > 0U) 69 { 70 /* 71 * C = A - B 72 * Subtract and then store the results in the destination buffer. 73 */ 74 vecA = vld1q(pSrcA); 75 vecB = vld1q(pSrcB); 76 vst1q(pDst, vqsubq(vecA, vecB)); 77 /* 78 * Decrement the blockSize loop counter 79 */ 80 blkCnt--; 81 /* 82 * advance vector source and destination pointers 83 */ 84 pSrcA += 16; 85 pSrcB += 16; 86 pDst += 16; 87 } 88 /* 89 * tail 90 */ 91 blkCnt = blockSize & 0xF; 92 if (blkCnt > 0U) 93 { 94 mve_pred16_t p0 = vctp8q(blkCnt); 95 vecA = vld1q(pSrcA); 96 vecB = vld1q(pSrcB); 97 vstrbq_p(pDst, vqsubq(vecA, vecB), p0); 98 } 99 } 100 #else 101 void arm_sub_q7( 102 const q7_t * pSrcA, 103 const q7_t * pSrcB, 104 q7_t * pDst, 105 uint32_t blockSize) 106 { 107 uint32_t blkCnt; /* Loop counter */ 108 109 #if defined (ARM_MATH_LOOPUNROLL) 110 111 /* Loop unrolling: Compute 4 outputs at a time */ 112 blkCnt = blockSize >> 2U; 113 114 while (blkCnt > 0U) 115 { 116 /* C = A - B */ 117 118 #if defined (ARM_MATH_DSP) 119 /* Subtract and store result in destination buffer (4 samples at a time). */ 120 write_q7x4_ia (&pDst, __QSUB8(read_q7x4_ia (&pSrcA), read_q7x4_ia (&pSrcB))); 121 #else 122 *pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8); 123 *pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8); 124 *pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8); 125 *pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8); 126 #endif 127 128 /* Decrement loop counter */ 129 blkCnt--; 130 } 131 132 /* Loop unrolling: Compute remaining outputs */ 133 blkCnt = blockSize % 0x4U; 134 135 #else 136 137 /* Initialize blkCnt with number of samples */ 138 blkCnt = blockSize; 139 140 #endif /* #if defined (ARM_MATH_LOOPUNROLL) */ 141 142 while (blkCnt > 0U) 143 { 144 /* C = A - B */ 145 146 /* Subtract and store result in destination buffer. */ 147 *pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8); 148 149 /* Decrement loop counter */ 150 blkCnt--; 151 } 152 153 } 154 #endif /* defined(ARM_MATH_MVEI) */ 155 156 /** 157 @} end of BasicSub group 158 */