1 /* ----------------------------------------------------------------------
2 * Project: CMSIS DSP Library
3 * Title: arm_abs_f32.c
4 * Description: Floating-point vector absolute value
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 #include <math.h>
31
32 /**
33 @ingroup groupMath
34 */
35
36 /**
37 @defgroup BasicAbs Vector Absolute Value
38
39 Computes the absolute value of a vector on an element-by-element basis.
40
41 <pre>
42 pDst[n] = abs(pSrc[n]), 0 <= n < blockSize.
43 </pre>
44
45 The functions support in-place computation allowing the source and
46 destination pointers to reference the same memory buffer.
47 There are separate functions for floating-point, Q7, Q15, and Q31 data types.
48 */
49
50 /**
51 @addtogroup BasicAbs
52 @{
53 */
54
55 /**
56 @brief Floating-point vector absolute value.
57 @param[in] pSrc points to the input vector
58 @param[out] pDst points to the output vector
59 @param[in] blockSize number of samples in each vector
60 @return none
61 */
62
63
64 #if defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE)
65
66 #include "arm_helium_utils.h"
67
arm_abs_f32(const float32_t * pSrc,float32_t * pDst,uint32_t blockSize)68 void arm_abs_f32(
69 const float32_t * pSrc,
70 float32_t * pDst,
71 uint32_t blockSize)
72 {
73 uint32_t blkCnt; /* Loop counter */
74 f32x4_t vec1;
75 f32x4_t res;
76
77
78 /* Compute 4 outputs at a time */
79 blkCnt = blockSize >> 2U;
80
81 while (blkCnt > 0U)
82 {
83 /* C = |A| */
84
85 /* Calculate absolute values and then store the results in the destination buffer. */
86 vec1 = vld1q(pSrc);
87 res = vabsq(vec1);
88 vst1q(pDst, res);
89
90 /* Increment pointers */
91 pSrc += 4;
92 pDst += 4;
93
94 /* Decrement the loop counter */
95 blkCnt--;
96 }
97
98 /* Tail */
99 blkCnt = blockSize & 0x3;
100
101
102 if (blkCnt > 0U)
103 {
104 /* C = |A| */
105 mve_pred16_t p0 = vctp32q(blkCnt);
106 vec1 = vld1q(pSrc);
107 vstrwq_p(pDst, vabsq(vec1), p0);
108 }
109
110 }
111
112 #else
arm_abs_f32(const float32_t * pSrc,float32_t * pDst,uint32_t blockSize)113 void arm_abs_f32(
114 const float32_t * pSrc,
115 float32_t * pDst,
116 uint32_t blockSize)
117 {
118 uint32_t blkCnt; /* Loop counter */
119
120 #if defined(ARM_MATH_NEON) && !defined(ARM_MATH_AUTOVECTORIZE)
121 f32x4_t vec1;
122 f32x4_t res;
123
124 /* Compute 4 outputs at a time */
125 blkCnt = blockSize >> 2U;
126
127 while (blkCnt > 0U)
128 {
129 /* C = |A| */
130
131 /* Calculate absolute values and then store the results in the destination buffer. */
132 vec1 = vld1q_f32(pSrc);
133 res = vabsq_f32(vec1);
134 vst1q_f32(pDst, res);
135
136 /* Increment pointers */
137 pSrc += 4;
138 pDst += 4;
139
140 /* Decrement the loop counter */
141 blkCnt--;
142 }
143
144 /* Tail */
145 blkCnt = blockSize & 0x3;
146
147 #else
148 #if defined (ARM_MATH_LOOPUNROLL) && !defined(ARM_MATH_AUTOVECTORIZE)
149
150 /* Loop unrolling: Compute 4 outputs at a time */
151 blkCnt = blockSize >> 2U;
152
153 while (blkCnt > 0U)
154 {
155 /* C = |A| */
156
157 /* Calculate absolute and store result in destination buffer. */
158 *pDst++ = fabsf(*pSrc++);
159
160 *pDst++ = fabsf(*pSrc++);
161
162 *pDst++ = fabsf(*pSrc++);
163
164 *pDst++ = fabsf(*pSrc++);
165
166 /* Decrement loop counter */
167 blkCnt--;
168 }
169
170 /* Loop unrolling: Compute remaining outputs */
171 blkCnt = blockSize % 0x4U;
172
173 #else
174
175 /* Initialize blkCnt with number of samples */
176 blkCnt = blockSize;
177
178 #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
179 #endif /* #if defined(ARM_MATH_NEON) */
180
181 while (blkCnt > 0U)
182 {
183 /* C = |A| */
184
185 /* Calculate absolute and store result in destination buffer. */
186 *pDst++ = fabsf(*pSrc++);
187
188 /* Decrement loop counter */
189 blkCnt--;
190 }
191
192 }
193 #endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
194 /**
195 @} end of BasicAbs group
196 */
197