1 /* ----------------------------------------------------------------------
2 * Project: CMSIS DSP Library
3 * Title: arm_copy_f16.c
4 * Description: Copies the elements of a floating-point vector
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/support_functions_f16.h"
30
31 #if defined(ARM_FLOAT16_SUPPORTED)
32
33
34 /**
35 @ingroup groupSupport
36 */
37
38
39 /**
40 @addtogroup copy
41 @{
42 */
43
44 /**
45 @brief Copies the elements of a f16 vector.
46 @param[in] pSrc points to input vector
47 @param[out] pDst points to output vector
48 @param[in] blockSize number of samples in each vector
49 */
50 #if defined(ARM_MATH_MVE_FLOAT16) && !defined(ARM_MATH_AUTOVECTORIZE)
51
arm_copy_f16(const float16_t * pSrc,float16_t * pDst,uint32_t blockSize)52 void arm_copy_f16(
53 const float16_t * pSrc,
54 float16_t * pDst,
55 uint32_t blockSize)
56 {
57 do {
58 mve_pred16_t p = vctp16q(blockSize);
59
60 vstrhq_p_f16(pDst,
61 vldrhq_z_f16((float16_t const *) pSrc, p), p);
62 /*
63 * Decrement the blockSize loop counter
64 * Advance vector source and destination pointers
65 */
66 pSrc += 8;
67 pDst += 8;
68 blockSize -= 8;
69 }
70 while ((int32_t) blockSize > 0);
71 }
72
73 #else
74
arm_copy_f16(const float16_t * pSrc,float16_t * pDst,uint32_t blockSize)75 void arm_copy_f16(
76 const float16_t * pSrc,
77 float16_t * pDst,
78 uint32_t blockSize)
79 {
80 uint32_t blkCnt; /* Loop counter */
81
82 #if defined (ARM_MATH_LOOPUNROLL)
83
84 /* Loop unrolling: Compute 4 outputs at a time */
85 blkCnt = blockSize >> 2U;
86
87 while (blkCnt > 0U)
88 {
89 /* C = A */
90
91 /* Copy and store result in destination buffer */
92 *pDst++ = *pSrc++;
93 *pDst++ = *pSrc++;
94 *pDst++ = *pSrc++;
95 *pDst++ = *pSrc++;
96
97 /* Decrement loop counter */
98 blkCnt--;
99 }
100
101 /* Loop unrolling: Compute remaining outputs */
102 blkCnt = blockSize % 0x4U;
103
104 #else
105
106 /* Initialize blkCnt with number of samples */
107 blkCnt = blockSize;
108
109 #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
110
111 while (blkCnt > 0U)
112 {
113 /* C = A */
114
115 /* Copy and store result in destination buffer */
116 *pDst++ = *pSrc++;
117
118 /* Decrement loop counter */
119 blkCnt--;
120 }
121 }
122 #endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
123
124 /**
125 @} end of BasicCopy group
126 */
127
128 #endif /* #if defined(ARM_FLOAT16_SUPPORTED) */
129
130
