1 /* ----------------------------------------------------------------------
2 * Project: CMSIS DSP Library
3 * Title: arm_fill_f32.c
4 * Description: Fills a constant value into 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.h"
30
31 /**
32 @ingroup groupSupport
33 */
34
35 /**
36 @defgroup Fill Vector Fill
37
38 Fills the destination vector with a constant value.
39
40 <pre>
41 pDst[n] = value; 0 <= n < blockSize.
42 </pre>
43
44 There are separate functions for floating point, Q31, Q15, and Q7 data types.
45 */
46
47 /**
48 @addtogroup Fill
49 @{
50 */
51
52 /**
53 @brief Fills a constant value into a floating-point vector.
54 @param[in] value input value to be filled
55 @param[out] pDst points to output vector
56 @param[in] blockSize number of samples in each vector
57 */
58 #if defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE)
arm_fill_f32(float32_t value,float32_t * pDst,uint32_t blockSize)59 ARM_DSP_ATTRIBUTE void arm_fill_f32(
60 float32_t value,
61 float32_t * pDst,
62 uint32_t blockSize)
63 {
64 uint32_t blkCnt;
65 blkCnt = blockSize >> 2U;
66
67 /* Compute 4 outputs at a time */
68 while (blkCnt > 0U)
69 {
70
71 vstrwq_f32(pDst,vdupq_n_f32(value));
72 /*
73 * Decrement the blockSize loop counter
74 * Advance vector source and destination pointers
75 */
76 pDst += 4;
77 blkCnt --;
78 }
79
80 blkCnt = blockSize & 3;
81
82 while (blkCnt > 0U)
83 {
84 /* C = value */
85
86 /* Fill value in destination buffer */
87 *pDst++ = value;
88
89 /* Decrement loop counter */
90 blkCnt--;
91 }
92
93 }
94 #else
95 #if defined(ARM_MATH_NEON_EXPERIMENTAL)
arm_fill_f32(float32_t value,float32_t * pDst,uint32_t blockSize)96 ARM_DSP_ATTRIBUTE void arm_fill_f32(
97 float32_t value,
98 float32_t * pDst,
99 uint32_t blockSize)
100 {
101 uint32_t blkCnt; /* loop counter */
102
103
104 float32x4_t inV = vdupq_n_f32(value);
105
106 blkCnt = blockSize >> 2U;
107
108 /* Compute 4 outputs at a time.
109 ** a second loop below computes the remaining 1 to 3 samples. */
110 while (blkCnt > 0U)
111 {
112 /* C = value */
113 /* Fill the value in the destination buffer */
114 vst1q_f32(pDst, inV);
115 pDst += 4;
116
117 /* Decrement the loop counter */
118 blkCnt--;
119 }
120
121 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
122 ** No loop unrolling is used. */
123 blkCnt = blockSize & 3;
124
125 while (blkCnt > 0U)
126 {
127 /* C = value */
128 /* Fill the value in the destination buffer */
129 *pDst++ = value;
130
131 /* Decrement the loop counter */
132 blkCnt--;
133 }
134 }
135 #else
arm_fill_f32(float32_t value,float32_t * pDst,uint32_t blockSize)136 ARM_DSP_ATTRIBUTE void arm_fill_f32(
137 float32_t value,
138 float32_t * pDst,
139 uint32_t blockSize)
140 {
141 uint32_t blkCnt; /* Loop counter */
142
143 #if defined (ARM_MATH_LOOPUNROLL)
144
145 /* Loop unrolling: Compute 4 outputs at a time */
146 blkCnt = blockSize >> 2U;
147
148 while (blkCnt > 0U)
149 {
150 /* C = value */
151
152 /* Fill value in destination buffer */
153 *pDst++ = value;
154 *pDst++ = value;
155 *pDst++ = value;
156 *pDst++ = value;
157
158 /* Decrement loop counter */
159 blkCnt--;
160 }
161
162 /* Loop unrolling: Compute remaining outputs */
163 blkCnt = blockSize % 0x4U;
164
165 #else
166
167 /* Initialize blkCnt with number of samples */
168 blkCnt = blockSize;
169
170 #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
171
172 while (blkCnt > 0U)
173 {
174 /* C = value */
175
176 /* Fill value in destination buffer */
177 *pDst++ = value;
178
179 /* Decrement loop counter */
180 blkCnt--;
181 }
182 }
183 #endif /* #if defined(ARM_MATH_NEON) */
184 #endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
185
186 /**
187 @} end of Fill group
188 */
189
