1 | /* ----------------------------------------------------------------------
|
---|
2 | * Project: CMSIS DSP Library
|
---|
3 | * Title: arm_scale_f32.c
|
---|
4 | * Description: Multiplies a floating-point vector by a scalar
|
---|
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 | * @defgroup scale Vector Scale
|
---|
37 | *
|
---|
38 | * Multiply a vector by a scalar value. For floating-point data, the algorithm used is:
|
---|
39 | *
|
---|
40 | * <pre>
|
---|
41 | * pDst[n] = pSrc[n] * scale, 0 <= n < blockSize.
|
---|
42 | * </pre>
|
---|
43 | *
|
---|
44 | * In the fixed-point Q7, Q15, and Q31 functions, <code>scale</code> is represented by
|
---|
45 | * a fractional multiplication <code>scaleFract</code> and an arithmetic shift <code>shift</code>.
|
---|
46 | * The shift allows the gain of the scaling operation to exceed 1.0.
|
---|
47 | * The algorithm used with fixed-point data is:
|
---|
48 | *
|
---|
49 | * <pre>
|
---|
50 | * pDst[n] = (pSrc[n] * scaleFract) << shift, 0 <= n < blockSize.
|
---|
51 | * </pre>
|
---|
52 | *
|
---|
53 | * The overall scale factor applied to the fixed-point data is
|
---|
54 | * <pre>
|
---|
55 | * scale = scaleFract * 2^shift.
|
---|
56 | * </pre>
|
---|
57 | *
|
---|
58 | * The functions support in-place computation allowing the source and destination
|
---|
59 | * pointers to reference the same memory buffer.
|
---|
60 | */
|
---|
61 |
|
---|
62 | /**
|
---|
63 | * @addtogroup scale
|
---|
64 | * @{
|
---|
65 | */
|
---|
66 |
|
---|
67 | /**
|
---|
68 | * @brief Multiplies a floating-point vector by a scalar.
|
---|
69 | * @param[in] *pSrc points to the input vector
|
---|
70 | * @param[in] scale scale factor to be applied
|
---|
71 | * @param[out] *pDst points to the output vector
|
---|
72 | * @param[in] blockSize number of samples in the vector
|
---|
73 | * @return none.
|
---|
74 | */
|
---|
75 |
|
---|
76 |
|
---|
77 | void arm_scale_f32(
|
---|
78 | float32_t * pSrc,
|
---|
79 | float32_t scale,
|
---|
80 | float32_t * pDst,
|
---|
81 | uint32_t blockSize)
|
---|
82 | {
|
---|
83 | uint32_t blkCnt; /* loop counter */
|
---|
84 | #if defined (ARM_MATH_DSP)
|
---|
85 |
|
---|
86 | /* Run the below code for Cortex-M4 and Cortex-M3 */
|
---|
87 | float32_t in1, in2, in3, in4; /* temporary variabels */
|
---|
88 |
|
---|
89 | /*loop Unrolling */
|
---|
90 | blkCnt = blockSize >> 2U;
|
---|
91 |
|
---|
92 | /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
|
---|
93 | ** a second loop below computes the remaining 1 to 3 samples. */
|
---|
94 | while (blkCnt > 0U)
|
---|
95 | {
|
---|
96 | /* C = A * scale */
|
---|
97 | /* Scale the input and then store the results in the destination buffer. */
|
---|
98 | /* read input samples from source */
|
---|
99 | in1 = *pSrc;
|
---|
100 | in2 = *(pSrc + 1);
|
---|
101 |
|
---|
102 | /* multiply with scaling factor */
|
---|
103 | in1 = in1 * scale;
|
---|
104 |
|
---|
105 | /* read input sample from source */
|
---|
106 | in3 = *(pSrc + 2);
|
---|
107 |
|
---|
108 | /* multiply with scaling factor */
|
---|
109 | in2 = in2 * scale;
|
---|
110 |
|
---|
111 | /* read input sample from source */
|
---|
112 | in4 = *(pSrc + 3);
|
---|
113 |
|
---|
114 | /* multiply with scaling factor */
|
---|
115 | in3 = in3 * scale;
|
---|
116 | in4 = in4 * scale;
|
---|
117 | /* store the result to destination */
|
---|
118 | *pDst = in1;
|
---|
119 | *(pDst + 1) = in2;
|
---|
120 | *(pDst + 2) = in3;
|
---|
121 | *(pDst + 3) = in4;
|
---|
122 |
|
---|
123 | /* update pointers to process next samples */
|
---|
124 | pSrc += 4U;
|
---|
125 | pDst += 4U;
|
---|
126 |
|
---|
127 | /* Decrement the loop counter */
|
---|
128 | blkCnt--;
|
---|
129 | }
|
---|
130 |
|
---|
131 | /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
|
---|
132 | ** No loop unrolling is used. */
|
---|
133 | blkCnt = blockSize % 0x4U;
|
---|
134 |
|
---|
135 | #else
|
---|
136 |
|
---|
137 | /* Run the below code for Cortex-M0 */
|
---|
138 |
|
---|
139 | /* Initialize blkCnt with number of samples */
|
---|
140 | blkCnt = blockSize;
|
---|
141 |
|
---|
142 | #endif /* #if defined (ARM_MATH_DSP) */
|
---|
143 |
|
---|
144 | while (blkCnt > 0U)
|
---|
145 | {
|
---|
146 | /* C = A * scale */
|
---|
147 | /* Scale the input and then store the result in the destination buffer. */
|
---|
148 | *pDst++ = (*pSrc++) * scale;
|
---|
149 |
|
---|
150 | /* Decrement the loop counter */
|
---|
151 | blkCnt--;
|
---|
152 | }
|
---|
153 | }
|
---|
154 |
|
---|
155 | /**
|
---|
156 | * @} end of scale group
|
---|
157 | */
|
---|