1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
|
/*
* Copyright (c) 2011 The WebRTC project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
/*
* This file contains the function WebRtcSpl_CrossCorrelation().
* The description header can be found in signal_processing_library.h
*
*/
#include "signal_processing_library.h"
void WebRtcSpl_CrossCorrelation(WebRtc_Word32* cross_correlation, WebRtc_Word16* seq1,
WebRtc_Word16* seq2, WebRtc_Word16 dim_seq,
WebRtc_Word16 dim_cross_correlation,
WebRtc_Word16 right_shifts,
WebRtc_Word16 step_seq2)
{
int i, j;
WebRtc_Word16* seq1Ptr;
WebRtc_Word16* seq2Ptr;
WebRtc_Word32* CrossCorrPtr;
#ifdef _XSCALE_OPT_
#ifdef _WIN32
#pragma message("NOTE: _XSCALE_OPT_ optimizations are used (overrides _ARM_OPT_ and requires /QRxscale compiler flag)")
#endif
__int64 macc40;
int iseq1[250];
int iseq2[250];
int iseq3[250];
int * iseq1Ptr;
int * iseq2Ptr;
int * iseq3Ptr;
int len, i_len;
seq1Ptr = seq1;
iseq1Ptr = iseq1;
for(i = 0; i < ((dim_seq + 1) >> 1); i++)
{
*iseq1Ptr = (unsigned short)*seq1Ptr++;
*iseq1Ptr++ |= (WebRtc_Word32)*seq1Ptr++ << 16;
}
if(dim_seq%2)
{
*(iseq1Ptr-1) &= 0x0000ffff;
}
*iseq1Ptr = 0;
iseq1Ptr++;
*iseq1Ptr = 0;
iseq1Ptr++;
*iseq1Ptr = 0;
if(step_seq2 < 0)
{
seq2Ptr = seq2 - dim_cross_correlation + 1;
CrossCorrPtr = &cross_correlation[dim_cross_correlation - 1];
}
else
{
seq2Ptr = seq2;
CrossCorrPtr = cross_correlation;
}
len = dim_seq + dim_cross_correlation - 1;
i_len = (len + 1) >> 1;
iseq2Ptr = iseq2;
iseq3Ptr = iseq3;
for(i = 0; i < i_len; i++)
{
*iseq2Ptr = (unsigned short)*seq2Ptr++;
*iseq3Ptr = (unsigned short)*seq2Ptr;
*iseq2Ptr++ |= (WebRtc_Word32)*seq2Ptr++ << 16;
*iseq3Ptr++ |= (WebRtc_Word32)*seq2Ptr << 16;
}
if(len % 2)
{
iseq2[i_len - 1] &= 0x0000ffff;
iseq3[i_len - 1] = 0;
}
else
iseq3[i_len - 1] &= 0x0000ffff;
iseq2[i_len] = 0;
iseq3[i_len] = 0;
iseq2[i_len + 1] = 0;
iseq3[i_len + 1] = 0;
iseq2[i_len + 2] = 0;
iseq3[i_len + 2] = 0;
// Set pointer to start value
iseq2Ptr = iseq2;
iseq3Ptr = iseq3;
i_len = (dim_seq + 7) >> 3;
for (i = 0; i < dim_cross_correlation; i++)
{
iseq1Ptr = iseq1;
macc40 = 0;
_WriteCoProcessor(macc40, 0);
if((i & 1))
{
iseq3Ptr = iseq3 + (i >> 1);
for (j = i_len; j > 0; j--)
{
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq3Ptr++);
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq3Ptr++);
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq3Ptr++);
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq3Ptr++);
}
}
else
{
iseq2Ptr = iseq2 + (i >> 1);
for (j = i_len; j > 0; j--)
{
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq2Ptr++);
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq2Ptr++);
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq2Ptr++);
_SmulAddPack_2SW_ACC(*iseq1Ptr++, *iseq2Ptr++);
}
}
macc40 = _ReadCoProcessor(0);
*CrossCorrPtr = (WebRtc_Word32)(macc40 >> right_shifts);
CrossCorrPtr += step_seq2;
}
#else // #ifdef _XSCALE_OPT_
#ifdef _ARM_OPT_
WebRtc_Word16 dim_seq8 = (dim_seq >> 3) << 3;
#endif
CrossCorrPtr = cross_correlation;
for (i = 0; i < dim_cross_correlation; i++)
{
// Set the pointer to the static vector, set the pointer to the sliding vector
// and initialize cross_correlation
seq1Ptr = seq1;
seq2Ptr = seq2 + (step_seq2 * i);
(*CrossCorrPtr) = 0;
#ifndef _ARM_OPT_
#ifdef _WIN32
#pragma message("NOTE: default implementation is used")
#endif
// Perform the cross correlation
for (j = 0; j < dim_seq; j++)
{
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr), right_shifts);
seq1Ptr++;
seq2Ptr++;
}
#else
#ifdef _WIN32
#pragma message("NOTE: _ARM_OPT_ optimizations are used")
#endif
if (right_shifts == 0)
{
// Perform the optimized cross correlation
for (j = 0; j < dim_seq8; j = j + 8)
{
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
}
for (j = dim_seq8; j < dim_seq; j++)
{
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16((*seq1Ptr), (*seq2Ptr));
seq1Ptr++;
seq2Ptr++;
}
}
else // right_shifts != 0
{
// Perform the optimized cross correlation
for (j = 0; j < dim_seq8; j = j + 8)
{
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
}
for (j = dim_seq8; j < dim_seq; j++)
{
(*CrossCorrPtr) += WEBRTC_SPL_MUL_16_16_RSFT((*seq1Ptr), (*seq2Ptr),
right_shifts);
seq1Ptr++;
seq2Ptr++;
}
}
#endif
CrossCorrPtr++;
}
#endif
}
|
