1 /* ----------------------------------------------------------------------
2 * Project: CMSIS DSP Library
3 * Title: arm_cmplx_conj_f16.c
4 * Description: Floating-point complex conjugate
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/complex_math_functions_f16.h"
30
31 #if defined(ARM_FLOAT16_SUPPORTED)
32 /**
33 @ingroup groupCmplxMath
34 */
35
36
37 /**
38 @addtogroup cmplx_conj
39 @{
40 */
41
42 /**
43 @brief Floating-point complex conjugate.
44 @param[in] pSrc points to the input vector
45 @param[out] pDst points to the output vector
46 @param[in] numSamples number of samples in each vector
47 */
48
49 #if defined(ARM_MATH_MVE_FLOAT16) && !defined(ARM_MATH_AUTOVECTORIZE)
50
arm_cmplx_conj_f16(const float16_t * pSrc,float16_t * pDst,uint32_t numSamples)51 void arm_cmplx_conj_f16(
52 const float16_t * pSrc,
53 float16_t * pDst,
54 uint32_t numSamples)
55 {
56 static const float16_t cmplx_conj_sign[8] = { 1.0f, -1.0f, 1.0f, -1.0f, 1.0f, -1.0f, 1.0f, -1.0f };
57 uint32_t blockSize = numSamples * CMPLX_DIM; /* loop counters */
58 uint32_t blkCnt;
59 f16x8_t vecSrc;
60 f16x8_t vecSign;
61
62 /*
63 * load sign vector
64 */
65 vecSign = *(f16x8_t *) cmplx_conj_sign;
66
67 /* Compute 4 real samples at a time */
68 blkCnt = blockSize >> 3U;
69
70 while (blkCnt > 0U)
71 {
72 vecSrc = vld1q(pSrc);
73 vst1q(pDst,vmulq(vecSrc, vecSign));
74 /*
75 * Decrement the blkCnt loop counter
76 * Advance vector source and destination pointers
77 */
78 pSrc += 8;
79 pDst += 8;
80 blkCnt--;
81 }
82
83 /* Tail */
84 blkCnt = (blockSize & 0x7) >> 1;
85
86 while (blkCnt > 0U)
87 {
88 /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
89
90 /* Calculate Complex Conjugate and store result in destination buffer. */
91 *pDst++ = *pSrc++;
92 *pDst++ = -(_Float16)*pSrc++;
93
94 /* Decrement loop counter */
95 blkCnt--;
96 }
97
98 }
99
100 #else
arm_cmplx_conj_f16(const float16_t * pSrc,float16_t * pDst,uint32_t numSamples)101 void arm_cmplx_conj_f16(
102 const float16_t * pSrc,
103 float16_t * pDst,
104 uint32_t numSamples)
105 {
106 uint32_t blkCnt; /* Loop counter */
107
108 #if defined (ARM_MATH_LOOPUNROLL) && !defined(ARM_MATH_AUTOVECTORIZE)
109
110 /* Loop unrolling: Compute 4 outputs at a time */
111 blkCnt = numSamples >> 2U;
112
113 while (blkCnt > 0U)
114 {
115 /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
116
117 /* Calculate Complex Conjugate and store result in destination buffer. */
118 *pDst++ = *pSrc++;
119 *pDst++ = -(_Float16)*pSrc++;
120
121 *pDst++ = *pSrc++;
122 *pDst++ = -(_Float16)*pSrc++;
123
124 *pDst++ = *pSrc++;
125 *pDst++ = -(_Float16)*pSrc++;
126
127 *pDst++ = *pSrc++;
128 *pDst++ = -(_Float16)*pSrc++;
129
130 /* Decrement loop counter */
131 blkCnt--;
132 }
133
134 /* Loop unrolling: Compute remaining outputs */
135 blkCnt = numSamples % 0x4U;
136
137 #else
138
139 /* Initialize blkCnt with number of samples */
140 blkCnt = numSamples;
141
142 #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
143
144 while (blkCnt > 0U)
145 {
146 /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
147
148 /* Calculate Complex Conjugate and store result in destination buffer. */
149 *pDst++ = *pSrc++;
150 *pDst++ = -(_Float16)*pSrc++;
151
152 /* Decrement loop counter */
153 blkCnt--;
154 }
155
156 }
157 #endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
158
159 /**
160 @} end of cmplx_conj group
161 */
162 #endif /* #if defined(ARM_FLOAT16_SUPPORTED) */
163
