2
|
1 /* ----------------------------------------------------------------------
|
|
2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
|
|
3 *
|
|
4 * $Date: 19. March 2015
|
|
5 * $Revision: V.1.4.5
|
|
6 *
|
|
7 * Project: CMSIS DSP Library
|
|
8 * Title: arm_scale_q31.c
|
|
9 *
|
|
10 * Description: Multiplies a Q31 vector by a scalar.
|
|
11 *
|
|
12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
|
|
13 *
|
|
14 * Redistribution and use in source and binary forms, with or without
|
|
15 * modification, are permitted provided that the following conditions
|
|
16 * are met:
|
|
17 * - Redistributions of source code must retain the above copyright
|
|
18 * notice, this list of conditions and the following disclaimer.
|
|
19 * - Redistributions in binary form must reproduce the above copyright
|
|
20 * notice, this list of conditions and the following disclaimer in
|
|
21 * the documentation and/or other materials provided with the
|
|
22 * distribution.
|
|
23 * - Neither the name of ARM LIMITED nor the names of its contributors
|
|
24 * may be used to endorse or promote products derived from this
|
|
25 * software without specific prior written permission.
|
|
26 *
|
|
27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
|
29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
|
|
30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
|
|
31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
|
|
32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
|
|
33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
|
|
34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
|
|
35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
|
|
37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
|
38 * POSSIBILITY OF SUCH DAMAGE.
|
|
39 * -------------------------------------------------------------------- */
|
|
40
|
|
41 #include "arm_math.h"
|
|
42
|
|
43 /**
|
|
44 * @ingroup groupMath
|
|
45 */
|
|
46
|
|
47 /**
|
|
48 * @addtogroup scale
|
|
49 * @{
|
|
50 */
|
|
51
|
|
52 /**
|
|
53 * @brief Multiplies a Q31 vector by a scalar.
|
|
54 * @param[in] *pSrc points to the input vector
|
|
55 * @param[in] scaleFract fractional portion of the scale value
|
|
56 * @param[in] shift number of bits to shift the result by
|
|
57 * @param[out] *pDst points to the output vector
|
|
58 * @param[in] blockSize number of samples in the vector
|
|
59 * @return none.
|
|
60 *
|
|
61 * <b>Scaling and Overflow Behavior:</b>
|
|
62 * \par
|
|
63 * The input data <code>*pSrc</code> and <code>scaleFract</code> are in 1.31 format.
|
|
64 * These are multiplied to yield a 2.62 intermediate result and this is shifted with saturation to 1.31 format.
|
|
65 */
|
|
66
|
|
67 void arm_scale_q31(
|
|
68 q31_t * pSrc,
|
|
69 q31_t scaleFract,
|
|
70 int8_t shift,
|
|
71 q31_t * pDst,
|
|
72 uint32_t blockSize)
|
|
73 {
|
|
74 int8_t kShift = shift + 1; /* Shift to apply after scaling */
|
|
75 int8_t sign = (kShift & 0x80);
|
|
76 uint32_t blkCnt; /* loop counter */
|
|
77 q31_t in, out;
|
|
78
|
|
79 #ifndef ARM_MATH_CM0_FAMILY
|
|
80
|
|
81 /* Run the below code for Cortex-M4 and Cortex-M3 */
|
|
82
|
|
83 q31_t in1, in2, in3, in4; /* temporary input variables */
|
|
84 q31_t out1, out2, out3, out4; /* temporary output variabels */
|
|
85
|
|
86
|
|
87 /*loop Unrolling */
|
|
88 blkCnt = blockSize >> 2u;
|
|
89
|
|
90 if(sign == 0u)
|
|
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 /* read four inputs from source */
|
|
97 in1 = *pSrc;
|
|
98 in2 = *(pSrc + 1);
|
|
99 in3 = *(pSrc + 2);
|
|
100 in4 = *(pSrc + 3);
|
|
101
|
|
102 /* multiply input with scaler value */
|
|
103 in1 = ((q63_t) in1 * scaleFract) >> 32;
|
|
104 in2 = ((q63_t) in2 * scaleFract) >> 32;
|
|
105 in3 = ((q63_t) in3 * scaleFract) >> 32;
|
|
106 in4 = ((q63_t) in4 * scaleFract) >> 32;
|
|
107
|
|
108 /* apply shifting */
|
|
109 out1 = in1 << kShift;
|
|
110 out2 = in2 << kShift;
|
|
111
|
|
112 /* saturate the results. */
|
|
113 if(in1 != (out1 >> kShift))
|
|
114 out1 = 0x7FFFFFFF ^ (in1 >> 31);
|
|
115
|
|
116 if(in2 != (out2 >> kShift))
|
|
117 out2 = 0x7FFFFFFF ^ (in2 >> 31);
|
|
118
|
|
119 out3 = in3 << kShift;
|
|
120 out4 = in4 << kShift;
|
|
121
|
|
122 *pDst = out1;
|
|
123 *(pDst + 1) = out2;
|
|
124
|
|
125 if(in3 != (out3 >> kShift))
|
|
126 out3 = 0x7FFFFFFF ^ (in3 >> 31);
|
|
127
|
|
128 if(in4 != (out4 >> kShift))
|
|
129 out4 = 0x7FFFFFFF ^ (in4 >> 31);
|
|
130
|
|
131 /* Store result destination */
|
|
132 *(pDst + 2) = out3;
|
|
133 *(pDst + 3) = out4;
|
|
134
|
|
135 /* Update pointers to process next sampels */
|
|
136 pSrc += 4u;
|
|
137 pDst += 4u;
|
|
138
|
|
139 /* Decrement the loop counter */
|
|
140 blkCnt--;
|
|
141 }
|
|
142
|
|
143 }
|
|
144 else
|
|
145 {
|
|
146 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
|
|
147 ** a second loop below computes the remaining 1 to 3 samples. */
|
|
148 while(blkCnt > 0u)
|
|
149 {
|
|
150 /* read four inputs from source */
|
|
151 in1 = *pSrc;
|
|
152 in2 = *(pSrc + 1);
|
|
153 in3 = *(pSrc + 2);
|
|
154 in4 = *(pSrc + 3);
|
|
155
|
|
156 /* multiply input with scaler value */
|
|
157 in1 = ((q63_t) in1 * scaleFract) >> 32;
|
|
158 in2 = ((q63_t) in2 * scaleFract) >> 32;
|
|
159 in3 = ((q63_t) in3 * scaleFract) >> 32;
|
|
160 in4 = ((q63_t) in4 * scaleFract) >> 32;
|
|
161
|
|
162 /* apply shifting */
|
|
163 out1 = in1 >> -kShift;
|
|
164 out2 = in2 >> -kShift;
|
|
165
|
|
166 out3 = in3 >> -kShift;
|
|
167 out4 = in4 >> -kShift;
|
|
168
|
|
169 /* Store result destination */
|
|
170 *pDst = out1;
|
|
171 *(pDst + 1) = out2;
|
|
172
|
|
173 *(pDst + 2) = out3;
|
|
174 *(pDst + 3) = out4;
|
|
175
|
|
176 /* Update pointers to process next sampels */
|
|
177 pSrc += 4u;
|
|
178 pDst += 4u;
|
|
179
|
|
180 /* Decrement the loop counter */
|
|
181 blkCnt--;
|
|
182 }
|
|
183 }
|
|
184 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
|
|
185 ** No loop unrolling is used. */
|
|
186 blkCnt = blockSize % 0x4u;
|
|
187
|
|
188 #else
|
|
189
|
|
190 /* Run the below code for Cortex-M0 */
|
|
191
|
|
192 /* Initialize blkCnt with number of samples */
|
|
193 blkCnt = blockSize;
|
|
194
|
|
195 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
|
|
196
|
|
197 if(sign == 0)
|
|
198 {
|
|
199 while(blkCnt > 0u)
|
|
200 {
|
|
201 /* C = A * scale */
|
|
202 /* Scale the input and then store the result in the destination buffer. */
|
|
203 in = *pSrc++;
|
|
204 in = ((q63_t) in * scaleFract) >> 32;
|
|
205
|
|
206 out = in << kShift;
|
|
207
|
|
208 if(in != (out >> kShift))
|
|
209 out = 0x7FFFFFFF ^ (in >> 31);
|
|
210
|
|
211 *pDst++ = out;
|
|
212
|
|
213 /* Decrement the loop counter */
|
|
214 blkCnt--;
|
|
215 }
|
|
216 }
|
|
217 else
|
|
218 {
|
|
219 while(blkCnt > 0u)
|
|
220 {
|
|
221 /* C = A * scale */
|
|
222 /* Scale the input and then store the result in the destination buffer. */
|
|
223 in = *pSrc++;
|
|
224 in = ((q63_t) in * scaleFract) >> 32;
|
|
225
|
|
226 out = in >> -kShift;
|
|
227
|
|
228 *pDst++ = out;
|
|
229
|
|
230 /* Decrement the loop counter */
|
|
231 blkCnt--;
|
|
232 }
|
|
233
|
|
234 }
|
|
235 }
|
|
236
|
|
237 /**
|
|
238 * @} end of scale group
|
|
239 */
|