Go to most recent revision | Details | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 2 | mjames | 1 | /* ---------------------------------------------------------------------- |
| 2 | * Project: CMSIS DSP Library |
||
| 3 | * Title: arm_abs_q7.c |
||
| 4 | * Description: Q7 vector absolute value |
||
| 5 | * |
||
| 6 | * $Date: 27. January 2017 |
||
| 7 | * $Revision: V.1.5.1 |
||
| 8 | * |
||
| 9 | * Target Processor: Cortex-M cores |
||
| 10 | * -------------------------------------------------------------------- */ |
||
| 11 | /* |
||
| 12 | * Copyright (C) 2010-2017 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 "arm_math.h" |
||
| 30 | |||
| 31 | /** |
||
| 32 | * @ingroup groupMath |
||
| 33 | */ |
||
| 34 | |||
| 35 | /** |
||
| 36 | * @addtogroup BasicAbs |
||
| 37 | * @{ |
||
| 38 | */ |
||
| 39 | |||
| 40 | /** |
||
| 41 | * @brief Q7 vector absolute value. |
||
| 42 | * @param[in] *pSrc points to the input buffer |
||
| 43 | * @param[out] *pDst points to the output buffer |
||
| 44 | * @param[in] blockSize number of samples in each vector |
||
| 45 | * @return none. |
||
| 46 | * |
||
| 47 | * \par Conditions for optimum performance |
||
| 48 | * Input and output buffers should be aligned by 32-bit |
||
| 49 | * |
||
| 50 | * |
||
| 51 | * <b>Scaling and Overflow Behavior:</b> |
||
| 52 | * \par |
||
| 53 | * The function uses saturating arithmetic. |
||
| 54 | * The Q7 value -1 (0x80) will be saturated to the maximum allowable positive value 0x7F. |
||
| 55 | */ |
||
| 56 | |||
| 57 | void arm_abs_q7( |
||
| 58 | q7_t * pSrc, |
||
| 59 | q7_t * pDst, |
||
| 60 | uint32_t blockSize) |
||
| 61 | { |
||
| 62 | uint32_t blkCnt; /* loop counter */ |
||
| 63 | q7_t in; /* Input value1 */ |
||
| 64 | |||
| 65 | #if defined (ARM_MATH_DSP) |
||
| 66 | |||
| 67 | /* Run the below code for Cortex-M4 and Cortex-M3 */ |
||
| 68 | q31_t in1, in2, in3, in4; /* temporary input variables */ |
||
| 69 | q31_t out1, out2, out3, out4; /* temporary output variables */ |
||
| 70 | |||
| 71 | /*loop Unrolling */ |
||
| 72 | blkCnt = blockSize >> 2U; |
||
| 73 | |||
| 74 | /* First part of the processing with loop unrolling. Compute 4 outputs at a time. |
||
| 75 | ** a second loop below computes the remaining 1 to 3 samples. */ |
||
| 76 | while (blkCnt > 0U) |
||
| 77 | { |
||
| 78 | /* C = |A| */ |
||
| 79 | /* Read inputs */ |
||
| 80 | in1 = (q31_t) * pSrc; |
||
| 81 | in2 = (q31_t) * (pSrc + 1); |
||
| 82 | in3 = (q31_t) * (pSrc + 2); |
||
| 83 | |||
| 84 | /* find absolute value */ |
||
| 85 | out1 = (in1 > 0) ? in1 : (q31_t)__QSUB8(0, in1); |
||
| 86 | |||
| 87 | /* read input */ |
||
| 88 | in4 = (q31_t) * (pSrc + 3); |
||
| 89 | |||
| 90 | /* find absolute value */ |
||
| 91 | out2 = (in2 > 0) ? in2 : (q31_t)__QSUB8(0, in2); |
||
| 92 | |||
| 93 | /* store result to destination */ |
||
| 94 | *pDst = (q7_t) out1; |
||
| 95 | |||
| 96 | /* find absolute value */ |
||
| 97 | out3 = (in3 > 0) ? in3 : (q31_t)__QSUB8(0, in3); |
||
| 98 | |||
| 99 | /* find absolute value */ |
||
| 100 | out4 = (in4 > 0) ? in4 : (q31_t)__QSUB8(0, in4); |
||
| 101 | |||
| 102 | /* store result to destination */ |
||
| 103 | *(pDst + 1) = (q7_t) out2; |
||
| 104 | |||
| 105 | /* store result to destination */ |
||
| 106 | *(pDst + 2) = (q7_t) out3; |
||
| 107 | |||
| 108 | /* store result to destination */ |
||
| 109 | *(pDst + 3) = (q7_t) out4; |
||
| 110 | |||
| 111 | /* update pointers to process next samples */ |
||
| 112 | pSrc += 4U; |
||
| 113 | pDst += 4U; |
||
| 114 | |||
| 115 | /* Decrement the loop counter */ |
||
| 116 | blkCnt--; |
||
| 117 | } |
||
| 118 | |||
| 119 | /* If the blockSize is not a multiple of 4, compute any remaining output samples here. |
||
| 120 | ** No loop unrolling is used. */ |
||
| 121 | blkCnt = blockSize % 0x4U; |
||
| 122 | #else |
||
| 123 | |||
| 124 | /* Run the below code for Cortex-M0 */ |
||
| 125 | blkCnt = blockSize; |
||
| 126 | |||
| 127 | #endif /* #define ARM_MATH_CM0_FAMILY */ |
||
| 128 | |||
| 129 | while (blkCnt > 0U) |
||
| 130 | { |
||
| 131 | /* C = |A| */ |
||
| 132 | /* Read the input */ |
||
| 133 | in = *pSrc++; |
||
| 134 | |||
| 135 | /* Store the Absolute result in the destination buffer */ |
||
| 136 | *pDst++ = (in > 0) ? in : ((in == (q7_t) 0x80) ? 0x7f : -in); |
||
| 137 | |||
| 138 | /* Decrement the loop counter */ |
||
| 139 | blkCnt--; |
||
| 140 | } |
||
| 141 | } |
||
| 142 | |||
| 143 | /** |
||
| 144 | * @} end of BasicAbs group |
||
| 145 | */ |