MIPS optimizations for the following functions:

WebRtcSpl_ComplexBitReverse, WebRtcSpl_ComplexFFT, WebRtcSpl_ComplexIFFT, WebRtcSpl_DownsampleFast and WebRtcSpl_FilterARFastQ12.
Also, moved the common table used in complex_fft functions to a separate header file (webrtc/common_audio/signal_processing/include/complex_fft_tables.h).

R=andrew@webrtc.org, kma@webrtc.org

Review URL: https://webrtc-codereview.appspot.com/1126004

Patch from Ljubomir Papuga <lpapuga@mips.com>.

git-svn-id: http://webrtc.googlecode.com/svn/trunk@4141 4adac7df-926f-26a2-2b94-8c16560cd09d
This commit is contained in:
andrew@webrtc.org 2013-05-30 16:38:36 +00:00
parent adc64a7216
commit eed919d95d
9 changed files with 980 additions and 132 deletions

View File

@ -45,6 +45,7 @@
'signal_processing/auto_corr_to_refl_coef.c',
'signal_processing/auto_correlation.c',
'signal_processing/complex_fft.c',
'signal_processing/complex_fft_tables.h',
'signal_processing/complex_bit_reverse.c',
'signal_processing/copy_set_operations.c',
'signal_processing/cross_correlation.c',
@ -115,9 +116,18 @@
}],
['target_arch=="mipsel"', {
'sources': [
'signal_processing/complex_bit_reverse_mips.c',
'signal_processing/complex_fft_mips.c',
'signal_processing/downsample_fast_mips.c',
'signal_processing/filter_ar_fast_q12_mips.c',
'signal_processing/min_max_operations_mips.c',
'signal_processing/resample_by_2_mips.c',
],
'sources!': [
'signal_processing/complex_bit_reverse.c',
'signal_processing/complex_fft.c',
'signal_processing/filter_ar_fast_q12.c',
],
}],
], # conditions
# Ignore warning on shift operator promotion.

View File

@ -0,0 +1,177 @@
/*
* Copyright (c) 2013 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.
*/
#include "webrtc/common_audio/signal_processing/include/signal_processing_library.h"
static int16_t coefTable_7[] = {
4, 256, 8, 128, 12, 384, 16, 64,
20, 320, 24, 192, 28, 448, 36, 288,
40, 160, 44, 416, 48, 96, 52, 352,
56, 224, 60, 480, 68, 272, 72, 144,
76, 400, 84, 336, 88, 208, 92, 464,
100, 304, 104, 176, 108, 432, 116, 368,
120, 240, 124, 496, 132, 264, 140, 392,
148, 328, 152, 200, 156, 456, 164, 296,
172, 424, 180, 360, 184, 232, 188, 488,
196, 280, 204, 408, 212, 344, 220, 472,
228, 312, 236, 440, 244, 376, 252, 504,
268, 388, 276, 324, 284, 452, 300, 420,
308, 356, 316, 484, 332, 404, 348, 468,
364, 436, 380, 500, 412, 460, 444, 492
};
static int16_t coefTable_8[] = {
4, 512, 8, 256, 12, 768, 16, 128,
20, 640, 24, 384, 28, 896, 32, 64,
36, 576, 40, 320, 44, 832, 48, 192,
52, 704, 56, 448, 60, 960, 68, 544,
72, 288, 76, 800, 80, 160, 84, 672,
88, 416, 92, 928, 100, 608, 104, 352,
108, 864, 112, 224, 116, 736, 120, 480,
124, 992, 132, 528, 136, 272, 140, 784,
148, 656, 152, 400, 156, 912, 164, 592,
168, 336, 172, 848, 176, 208, 180, 720,
184, 464, 188, 976, 196, 560, 200, 304,
204, 816, 212, 688, 216, 432, 220, 944,
228, 624, 232, 368, 236, 880, 244, 752,
248, 496, 252, 1008, 260, 520, 268, 776,
276, 648, 280, 392, 284, 904, 292, 584,
296, 328, 300, 840, 308, 712, 312, 456,
316, 968, 324, 552, 332, 808, 340, 680,
344, 424, 348, 936, 356, 616, 364, 872,
372, 744, 376, 488, 380, 1000, 388, 536,
396, 792, 404, 664, 412, 920, 420, 600,
428, 856, 436, 728, 440, 472, 444, 984,
452, 568, 460, 824, 468, 696, 476, 952,
484, 632, 492, 888, 500, 760, 508, 1016,
524, 772, 532, 644, 540, 900, 548, 580,
556, 836, 564, 708, 572, 964, 588, 804,
596, 676, 604, 932, 620, 868, 628, 740,
636, 996, 652, 788, 668, 916, 684, 852,
692, 724, 700, 980, 716, 820, 732, 948,
748, 884, 764, 1012, 796, 908, 812, 844,
828, 972, 860, 940, 892, 1004, 956, 988
};
void WebRtcSpl_ComplexBitReverse(int16_t frfi[], int stages) {
int l;
int16_t tr, ti;
int32_t tmp1, tmp2, tmp3, tmp4;
int32_t* ptr_i;
int32_t* ptr_j;
if (stages == 8) {
int16_t* pcoeftable_8 = coefTable_8;
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"addiu %[l], $zero, 120 \n\t"
"1: \n\t"
"addiu %[l], %[l], -4 \n\t"
"lh %[tr], 0(%[pcoeftable_8]) \n\t"
"lh %[ti], 2(%[pcoeftable_8]) \n\t"
"lh %[tmp3], 4(%[pcoeftable_8]) \n\t"
"lh %[tmp4], 6(%[pcoeftable_8]) \n\t"
"addu %[ptr_i], %[frfi], %[tr] \n\t"
"addu %[ptr_j], %[frfi], %[ti] \n\t"
"addu %[tr], %[frfi], %[tmp3] \n\t"
"addu %[ti], %[frfi], %[tmp4] \n\t"
"ulw %[tmp1], 0(%[ptr_i]) \n\t"
"ulw %[tmp2], 0(%[ptr_j]) \n\t"
"ulw %[tmp3], 0(%[tr]) \n\t"
"ulw %[tmp4], 0(%[ti]) \n\t"
"usw %[tmp1], 0(%[ptr_j]) \n\t"
"usw %[tmp2], 0(%[ptr_i]) \n\t"
"usw %[tmp4], 0(%[tr]) \n\t"
"usw %[tmp3], 0(%[ti]) \n\t"
"lh %[tmp1], 8(%[pcoeftable_8]) \n\t"
"lh %[tmp2], 10(%[pcoeftable_8]) \n\t"
"lh %[tr], 12(%[pcoeftable_8]) \n\t"
"lh %[ti], 14(%[pcoeftable_8]) \n\t"
"addu %[ptr_i], %[frfi], %[tmp1] \n\t"
"addu %[ptr_j], %[frfi], %[tmp2] \n\t"
"addu %[tr], %[frfi], %[tr] \n\t"
"addu %[ti], %[frfi], %[ti] \n\t"
"ulw %[tmp1], 0(%[ptr_i]) \n\t"
"ulw %[tmp2], 0(%[ptr_j]) \n\t"
"ulw %[tmp3], 0(%[tr]) \n\t"
"ulw %[tmp4], 0(%[ti]) \n\t"
"usw %[tmp1], 0(%[ptr_j]) \n\t"
"usw %[tmp2], 0(%[ptr_i]) \n\t"
"usw %[tmp4], 0(%[tr]) \n\t"
"usw %[tmp3], 0(%[ti]) \n\t"
"bgtz %[l], 1b \n\t"
" addiu %[pcoeftable_8], %[pcoeftable_8], 16 \n\t"
".set pop \n\t"
: [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2), [ptr_i] "=&r" (ptr_i),
[ptr_j] "=&r" (ptr_j), [tr] "=&r" (tr), [l] "=&r" (l),
[tmp3] "=&r" (tmp3), [pcoeftable_8] "+r" (pcoeftable_8),
[ti] "=&r" (ti), [tmp4] "=&r" (tmp4)
: [frfi] "r" (frfi)
: "memory"
);
} else if (stages == 7) {
int16_t* pcoeftable_7 = coefTable_7;
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"addiu %[l], $zero, 56 \n\t"
"1: \n\t"
"addiu %[l], %[l], -4 \n\t"
"lh %[tr], 0(%[pcoeftable_7]) \n\t"
"lh %[ti], 2(%[pcoeftable_7]) \n\t"
"lh %[tmp3], 4(%[pcoeftable_7]) \n\t"
"lh %[tmp4], 6(%[pcoeftable_7]) \n\t"
"addu %[ptr_i], %[frfi], %[tr] \n\t"
"addu %[ptr_j], %[frfi], %[ti] \n\t"
"addu %[tr], %[frfi], %[tmp3] \n\t"
"addu %[ti], %[frfi], %[tmp4] \n\t"
"ulw %[tmp1], 0(%[ptr_i]) \n\t"
"ulw %[tmp2], 0(%[ptr_j]) \n\t"
"ulw %[tmp3], 0(%[tr]) \n\t"
"ulw %[tmp4], 0(%[ti]) \n\t"
"usw %[tmp1], 0(%[ptr_j]) \n\t"
"usw %[tmp2], 0(%[ptr_i]) \n\t"
"usw %[tmp4], 0(%[tr]) \n\t"
"usw %[tmp3], 0(%[ti]) \n\t"
"lh %[tmp1], 8(%[pcoeftable_7]) \n\t"
"lh %[tmp2], 10(%[pcoeftable_7]) \n\t"
"lh %[tr], 12(%[pcoeftable_7]) \n\t"
"lh %[ti], 14(%[pcoeftable_7]) \n\t"
"addu %[ptr_i], %[frfi], %[tmp1] \n\t"
"addu %[ptr_j], %[frfi], %[tmp2] \n\t"
"addu %[tr], %[frfi], %[tr] \n\t"
"addu %[ti], %[frfi], %[ti] \n\t"
"ulw %[tmp1], 0(%[ptr_i]) \n\t"
"ulw %[tmp2], 0(%[ptr_j]) \n\t"
"ulw %[tmp3], 0(%[tr]) \n\t"
"ulw %[tmp4], 0(%[ti]) \n\t"
"usw %[tmp1], 0(%[ptr_j]) \n\t"
"usw %[tmp2], 0(%[ptr_i]) \n\t"
"usw %[tmp4], 0(%[tr]) \n\t"
"usw %[tmp3], 0(%[ti]) \n\t"
"bgtz %[l], 1b \n\t"
" addiu %[pcoeftable_7], %[pcoeftable_7], 16 \n\t"
".set pop \n\t"
: [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2), [ptr_i] "=&r" (ptr_i),
[ptr_j] "=&r" (ptr_j), [ti] "=&r" (ti), [tr] "=&r" (tr),
[l] "=&r" (l), [pcoeftable_7] "+r" (pcoeftable_7),
[tmp3] "=&r" (tmp3), [tmp4] "=&r" (tmp4)
: [frfi] "r" (frfi)
: "memory"
);
}
}

View File

@ -16,6 +16,7 @@
*/
#include "webrtc/common_audio/signal_processing/include/signal_processing_library.h"
#include "webrtc/common_audio/signal_processing/complex_fft_tables.h"
#define CFFTSFT 14
#define CFFTRND 1
@ -24,137 +25,6 @@
#define CIFFTSFT 14
#define CIFFTRND 1
static const int16_t kSinTable1024[] = {
0, 201, 402, 603, 804, 1005, 1206, 1406,
1607, 1808, 2009, 2209, 2410, 2610, 2811, 3011,
3211, 3411, 3611, 3811, 4011, 4210, 4409, 4608,
4807, 5006, 5205, 5403, 5601, 5799, 5997, 6195,
6392, 6589, 6786, 6982, 7179, 7375, 7571, 7766,
7961, 8156, 8351, 8545, 8739, 8932, 9126, 9319,
9511, 9703, 9895, 10087, 10278, 10469, 10659, 10849,
11038, 11227, 11416, 11604, 11792, 11980, 12166, 12353,
12539, 12724, 12909, 13094, 13278, 13462, 13645, 13827,
14009, 14191, 14372, 14552, 14732, 14911, 15090, 15268,
15446, 15623, 15799, 15975, 16150, 16325, 16499, 16672,
16845, 17017, 17189, 17360, 17530, 17699, 17868, 18036,
18204, 18371, 18537, 18702, 18867, 19031, 19194, 19357,
19519, 19680, 19840, 20000, 20159, 20317, 20474, 20631,
20787, 20942, 21096, 21249, 21402, 21554, 21705, 21855,
22004, 22153, 22301, 22448, 22594, 22739, 22883, 23027,
23169, 23311, 23452, 23592, 23731, 23869, 24006, 24143,
24278, 24413, 24546, 24679, 24811, 24942, 25072, 25201,
25329, 25456, 25582, 25707, 25831, 25954, 26077, 26198,
26318, 26437, 26556, 26673, 26789, 26905, 27019, 27132,
27244, 27355, 27466, 27575, 27683, 27790, 27896, 28001,
28105, 28208, 28309, 28410, 28510, 28608, 28706, 28802,
28897, 28992, 29085, 29177, 29268, 29358, 29446, 29534,
29621, 29706, 29790, 29873, 29955, 30036, 30116, 30195,
30272, 30349, 30424, 30498, 30571, 30643, 30713, 30783,
30851, 30918, 30984, 31049,
31113, 31175, 31236, 31297,
31356, 31413, 31470, 31525, 31580, 31633, 31684, 31735,
31785, 31833, 31880, 31926, 31970, 32014, 32056, 32097,
32137, 32176, 32213, 32249, 32284, 32318, 32350, 32382,
32412, 32441, 32468, 32495, 32520, 32544, 32567, 32588,
32609, 32628, 32646, 32662, 32678, 32692, 32705, 32717,
32727, 32736, 32744, 32751, 32757, 32761, 32764, 32766,
32767, 32766, 32764, 32761, 32757, 32751, 32744, 32736,
32727, 32717, 32705, 32692, 32678, 32662, 32646, 32628,
32609, 32588, 32567, 32544, 32520, 32495, 32468, 32441,
32412, 32382, 32350, 32318, 32284, 32249, 32213, 32176,
32137, 32097, 32056, 32014, 31970, 31926, 31880, 31833,
31785, 31735, 31684, 31633, 31580, 31525, 31470, 31413,
31356, 31297, 31236, 31175, 31113, 31049, 30984, 30918,
30851, 30783, 30713, 30643, 30571, 30498, 30424, 30349,
30272, 30195, 30116, 30036, 29955, 29873, 29790, 29706,
29621, 29534, 29446, 29358, 29268, 29177, 29085, 28992,
28897, 28802, 28706, 28608, 28510, 28410, 28309, 28208,
28105, 28001, 27896, 27790, 27683, 27575, 27466, 27355,
27244, 27132, 27019, 26905, 26789, 26673, 26556, 26437,
26318, 26198, 26077, 25954, 25831, 25707, 25582, 25456,
25329, 25201, 25072, 24942, 24811, 24679, 24546, 24413,
24278, 24143, 24006, 23869, 23731, 23592, 23452, 23311,
23169, 23027, 22883, 22739, 22594, 22448, 22301, 22153,
22004, 21855, 21705, 21554, 21402, 21249, 21096, 20942,
20787, 20631, 20474, 20317, 20159, 20000, 19840, 19680,
19519, 19357, 19194, 19031, 18867, 18702, 18537, 18371,
18204, 18036, 17868, 17699, 17530, 17360, 17189, 17017,
16845, 16672, 16499, 16325, 16150, 15975, 15799, 15623,
15446, 15268, 15090, 14911, 14732, 14552, 14372, 14191,
14009, 13827, 13645, 13462, 13278, 13094, 12909, 12724,
12539, 12353, 12166, 11980, 11792, 11604, 11416, 11227,
11038, 10849, 10659, 10469, 10278, 10087, 9895, 9703,
9511, 9319, 9126, 8932, 8739, 8545, 8351, 8156,
7961, 7766, 7571, 7375, 7179, 6982, 6786, 6589,
6392, 6195, 5997, 5799, 5601, 5403, 5205, 5006,
4807, 4608, 4409, 4210, 4011, 3811, 3611, 3411,
3211, 3011, 2811, 2610, 2410, 2209, 2009, 1808,
1607, 1406, 1206, 1005, 804, 603, 402, 201,
0, -201, -402, -603, -804, -1005, -1206, -1406,
-1607, -1808, -2009, -2209, -2410, -2610, -2811, -3011,
-3211, -3411, -3611, -3811, -4011, -4210, -4409, -4608,
-4807, -5006, -5205, -5403, -5601, -5799, -5997, -6195,
-6392, -6589, -6786, -6982, -7179, -7375, -7571, -7766,
-7961, -8156, -8351, -8545, -8739, -8932, -9126, -9319,
-9511, -9703, -9895, -10087, -10278, -10469, -10659, -10849,
-11038, -11227, -11416, -11604, -11792, -11980, -12166, -12353,
-12539, -12724, -12909, -13094, -13278, -13462, -13645, -13827,
-14009, -14191, -14372, -14552, -14732, -14911, -15090, -15268,
-15446, -15623, -15799, -15975, -16150, -16325, -16499, -16672,
-16845, -17017, -17189, -17360, -17530, -17699, -17868, -18036,
-18204, -18371, -18537, -18702, -18867, -19031, -19194, -19357,
-19519, -19680, -19840, -20000, -20159, -20317, -20474, -20631,
-20787, -20942, -21096, -21249, -21402, -21554, -21705, -21855,
-22004, -22153, -22301, -22448, -22594, -22739, -22883, -23027,
-23169, -23311, -23452, -23592, -23731, -23869, -24006, -24143,
-24278, -24413, -24546, -24679, -24811, -24942, -25072, -25201,
-25329, -25456, -25582, -25707, -25831, -25954, -26077, -26198,
-26318, -26437, -26556, -26673, -26789, -26905, -27019, -27132,
-27244, -27355, -27466, -27575, -27683, -27790, -27896, -28001,
-28105, -28208, -28309, -28410, -28510, -28608, -28706, -28802,
-28897, -28992, -29085, -29177, -29268, -29358, -29446, -29534,
-29621, -29706, -29790, -29873, -29955, -30036, -30116, -30195,
-30272, -30349, -30424, -30498, -30571, -30643, -30713, -30783,
-30851, -30918, -30984, -31049, -31113, -31175, -31236, -31297,
-31356, -31413, -31470, -31525, -31580, -31633, -31684, -31735,
-31785, -31833, -31880, -31926, -31970, -32014, -32056, -32097,
-32137, -32176, -32213, -32249, -32284, -32318, -32350, -32382,
-32412, -32441, -32468, -32495, -32520, -32544, -32567, -32588,
-32609, -32628, -32646, -32662, -32678, -32692, -32705, -32717,
-32727, -32736, -32744, -32751, -32757, -32761, -32764, -32766,
-32767, -32766, -32764, -32761, -32757, -32751, -32744, -32736,
-32727, -32717, -32705, -32692, -32678, -32662, -32646, -32628,
-32609, -32588, -32567, -32544, -32520, -32495, -32468, -32441,
-32412, -32382, -32350, -32318, -32284, -32249, -32213, -32176,
-32137, -32097, -32056, -32014, -31970, -31926, -31880, -31833,
-31785, -31735, -31684, -31633, -31580, -31525, -31470, -31413,
-31356, -31297, -31236, -31175, -31113, -31049, -30984, -30918,
-30851, -30783, -30713, -30643, -30571, -30498, -30424, -30349,
-30272, -30195, -30116, -30036, -29955, -29873, -29790, -29706,
-29621, -29534, -29446, -29358, -29268, -29177, -29085, -28992,
-28897, -28802, -28706, -28608, -28510, -28410, -28309, -28208,
-28105, -28001, -27896, -27790, -27683, -27575, -27466, -27355,
-27244, -27132, -27019, -26905, -26789, -26673, -26556, -26437,
-26318, -26198, -26077, -25954, -25831, -25707, -25582, -25456,
-25329, -25201, -25072, -24942, -24811, -24679, -24546, -24413,
-24278, -24143, -24006, -23869, -23731, -23592, -23452, -23311,
-23169, -23027, -22883, -22739, -22594, -22448, -22301, -22153,
-22004, -21855, -21705, -21554, -21402, -21249, -21096, -20942,
-20787, -20631, -20474, -20317, -20159, -20000, -19840, -19680,
-19519, -19357, -19194, -19031, -18867, -18702, -18537, -18371,
-18204, -18036, -17868, -17699, -17530, -17360, -17189, -17017,
-16845, -16672, -16499, -16325, -16150, -15975, -15799, -15623,
-15446, -15268, -15090, -14911, -14732, -14552, -14372, -14191,
-14009, -13827, -13645, -13462, -13278, -13094, -12909, -12724,
-12539, -12353, -12166, -11980, -11792, -11604, -11416, -11227,
-11038, -10849, -10659, -10469, -10278, -10087, -9895, -9703,
-9511, -9319, -9126, -8932, -8739, -8545, -8351, -8156,
-7961, -7766, -7571, -7375, -7179, -6982, -6786, -6589,
-6392, -6195, -5997, -5799, -5601, -5403, -5205, -5006,
-4807, -4608, -4409, -4210, -4011, -3811, -3611, -3411,
-3211, -3011, -2811, -2610, -2410, -2209, -2009, -1808,
-1607, -1406, -1206, -1005, -804, -603, -402, -201
};
int WebRtcSpl_ComplexFFT(int16_t frfi[], int stages, int mode)
{

View File

@ -0,0 +1,323 @@
/*
* Copyright (c) 2013 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.
*/
#include "webrtc/common_audio/signal_processing/include/signal_processing_library.h"
#include "webrtc/common_audio/signal_processing/complex_fft_tables.h"
#define CFFTSFT 14
#define CFFTRND 1
#define CFFTRND2 16384
#define CIFFTSFT 14
#define CIFFTRND 1
int WebRtcSpl_ComplexFFT(int16_t frfi[], int stages, int mode) {
int i = 0;
int l = 0;
int k = 0;
int istep = 0;
int n = 0;
int m = 0;
int32_t wr = 0, wi = 0;
int32_t tmp1 = 0;
int32_t tmp2 = 0;
int32_t tmp3 = 0;
int32_t tmp4 = 0;
int32_t tmp5 = 0;
int32_t tmp6 = 0;
int32_t tmp = 0;
int16_t* ptr_j = NULL;
int16_t* ptr_i = NULL;
n = 1 << stages;
if (n > 1024) {
return -1;
}
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"addiu %[k], $zero, 10 \n\t"
"addiu %[l], $zero, 1 \n\t"
"3: \n\t"
"sll %[istep], %[l], 1 \n\t"
"move %[m], $zero \n\t"
"sll %[tmp], %[l], 2 \n\t"
"move %[i], $zero \n\t"
"2: \n\t"
#if defined(MIPS_DSP_R1_LE)
"sllv %[tmp3], %[m], %[k] \n\t"
"addiu %[tmp2], %[tmp3], 512 \n\t"
"addiu %[m], %[m], 1 \n\t"
"lhx %[wi], %[tmp3](%[kSinTable1024]) \n\t"
"lhx %[wr], %[tmp2](%[kSinTable1024]) \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"sllv %[tmp3], %[m], %[k] \n\t"
"addu %[ptr_j], %[tmp3], %[kSinTable1024] \n\t"
"addiu %[ptr_i], %[ptr_j], 512 \n\t"
"addiu %[m], %[m], 1 \n\t"
"lh %[wi], 0(%[ptr_j]) \n\t"
"lh %[wr], 0(%[ptr_i]) \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"1: \n\t"
"sll %[tmp1], %[i], 2 \n\t"
"addu %[ptr_i], %[frfi], %[tmp1] \n\t"
"addu %[ptr_j], %[ptr_i], %[tmp] \n\t"
"lh %[tmp6], 0(%[ptr_i]) \n\t"
"lh %[tmp5], 2(%[ptr_i]) \n\t"
"lh %[tmp3], 0(%[ptr_j]) \n\t"
"lh %[tmp4], 2(%[ptr_j]) \n\t"
"addu %[i], %[i], %[istep] \n\t"
#if defined(MIPS_DSP_R2_LE)
"mult %[wr], %[tmp3] \n\t"
"madd %[wi], %[tmp4] \n\t"
"mult $ac1, %[wr], %[tmp4] \n\t"
"msub $ac1, %[wi], %[tmp3] \n\t"
"mflo %[tmp1] \n\t"
"mflo %[tmp2], $ac1 \n\t"
"sll %[tmp6], %[tmp6], 14 \n\t"
"sll %[tmp5], %[tmp5], 14 \n\t"
"shra_r.w %[tmp1], %[tmp1], 1 \n\t"
"shra_r.w %[tmp2], %[tmp2], 1 \n\t"
"subu %[tmp4], %[tmp6], %[tmp1] \n\t"
"addu %[tmp1], %[tmp6], %[tmp1] \n\t"
"addu %[tmp6], %[tmp5], %[tmp2] \n\t"
"subu %[tmp5], %[tmp5], %[tmp2] \n\t"
"shra_r.w %[tmp1], %[tmp1], 15 \n\t"
"shra_r.w %[tmp6], %[tmp6], 15 \n\t"
"shra_r.w %[tmp4], %[tmp4], 15 \n\t"
"shra_r.w %[tmp5], %[tmp5], 15 \n\t"
#else // #if defined(MIPS_DSP_R2_LE)
"mul %[tmp2], %[wr], %[tmp4] \n\t"
"mul %[tmp1], %[wr], %[tmp3] \n\t"
"mul %[tmp4], %[wi], %[tmp4] \n\t"
"mul %[tmp3], %[wi], %[tmp3] \n\t"
"sll %[tmp6], %[tmp6], 14 \n\t"
"sll %[tmp5], %[tmp5], 14 \n\t"
"addiu %[tmp6], %[tmp6], 16384 \n\t"
"addiu %[tmp5], %[tmp5], 16384 \n\t"
"addu %[tmp1], %[tmp1], %[tmp4] \n\t"
"subu %[tmp2], %[tmp2], %[tmp3] \n\t"
"addiu %[tmp1], %[tmp1], 1 \n\t"
"addiu %[tmp2], %[tmp2], 1 \n\t"
"sra %[tmp1], %[tmp1], 1 \n\t"
"sra %[tmp2], %[tmp2], 1 \n\t"
"subu %[tmp4], %[tmp6], %[tmp1] \n\t"
"addu %[tmp1], %[tmp6], %[tmp1] \n\t"
"addu %[tmp6], %[tmp5], %[tmp2] \n\t"
"subu %[tmp5], %[tmp5], %[tmp2] \n\t"
"sra %[tmp4], %[tmp4], 15 \n\t"
"sra %[tmp1], %[tmp1], 15 \n\t"
"sra %[tmp6], %[tmp6], 15 \n\t"
"sra %[tmp5], %[tmp5], 15 \n\t"
#endif // #if defined(MIPS_DSP_R2_LE)
"sh %[tmp1], 0(%[ptr_i]) \n\t"
"sh %[tmp6], 2(%[ptr_i]) \n\t"
"sh %[tmp4], 0(%[ptr_j]) \n\t"
"blt %[i], %[n], 1b \n\t"
" sh %[tmp5], 2(%[ptr_j]) \n\t"
"blt %[m], %[l], 2b \n\t"
" addu %[i], $zero, %[m] \n\t"
"move %[l], %[istep] \n\t"
"blt %[l], %[n], 3b \n\t"
" addiu %[k], %[k], -1 \n\t"
".set pop \n\t"
: [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2), [tmp3] "=&r" (tmp3),
[tmp4] "=&r" (tmp4), [tmp5] "=&r" (tmp5), [tmp6] "=&r" (tmp6),
[ptr_i] "=&r" (ptr_i), [i] "=&r" (i), [wi] "=&r" (wi), [wr] "=&r" (wr),
[m] "=&r" (m), [istep] "=&r" (istep), [l] "=&r" (l), [k] "=&r" (k),
[ptr_j] "=&r" (ptr_j), [tmp] "=&r" (tmp)
: [n] "r" (n), [frfi] "r" (frfi), [kSinTable1024] "r" (kSinTable1024)
: "hi", "lo", "$ac1hi", "$ac1lo", "memory"
);
return 0;
}
int WebRtcSpl_ComplexIFFT(int16_t frfi[], int stages, int mode) {
int i = 0, l = 0, k = 0;
int istep = 0, n = 0, m = 0;
int scale = 0, shift = 0;
int32_t wr = 0, wi = 0;
int32_t tmp1 = 0, tmp2 = 0, tmp3 = 0, tmp4 = 0;
int32_t tmp5 = 0, tmp6 = 0, tmp = 0, tempMax = 0, round2 = 0;
int16_t* ptr_j = NULL;
int16_t* ptr_i = NULL;
n = 1 << stages;
if (n > 1024) {
return -1;
}
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"addiu %[k], $zero, 10 \n\t"
"addiu %[l], $zero, 1 \n\t"
"move %[scale], $zero \n\t"
"3: \n\t"
"addiu %[shift], $zero, 14 \n\t"
"addiu %[round2], $zero, 8192 \n\t"
"move %[ptr_i], %[frfi] \n\t"
"move %[tempMax], $zero \n\t"
"addu %[i], %[n], %[n] \n\t"
"5: \n\t"
"lh %[tmp1], 0(%[ptr_i]) \n\t"
"lh %[tmp2], 2(%[ptr_i]) \n\t"
"lh %[tmp3], 4(%[ptr_i]) \n\t"
"lh %[tmp4], 6(%[ptr_i]) \n\t"
#if defined(MIPS_DSP_R1_LE)
"absq_s.w %[tmp1], %[tmp1] \n\t"
"absq_s.w %[tmp2], %[tmp2] \n\t"
"absq_s.w %[tmp3], %[tmp3] \n\t"
"absq_s.w %[tmp4], %[tmp4] \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"slt %[tmp5], %[tmp1], $zero \n\t"
"subu %[tmp6], $zero, %[tmp1] \n\t"
"movn %[tmp1], %[tmp6], %[tmp5] \n\t"
"slt %[tmp5], %[tmp2], $zero \n\t"
"subu %[tmp6], $zero, %[tmp2] \n\t"
"movn %[tmp2], %[tmp6], %[tmp5] \n\t"
"slt %[tmp5], %[tmp3], $zero \n\t"
"subu %[tmp6], $zero, %[tmp3] \n\t"
"movn %[tmp3], %[tmp6], %[tmp5] \n\t"
"slt %[tmp5], %[tmp4], $zero \n\t"
"subu %[tmp6], $zero, %[tmp4] \n\t"
"movn %[tmp4], %[tmp6], %[tmp5] \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"slt %[tmp5], %[tempMax], %[tmp1] \n\t"
"movn %[tempMax], %[tmp1], %[tmp5] \n\t"
"addiu %[i], %[i], -4 \n\t"
"slt %[tmp5], %[tempMax], %[tmp2] \n\t"
"movn %[tempMax], %[tmp2], %[tmp5] \n\t"
"slt %[tmp5], %[tempMax], %[tmp3] \n\t"
"movn %[tempMax], %[tmp3], %[tmp5] \n\t"
"slt %[tmp5], %[tempMax], %[tmp4] \n\t"
"movn %[tempMax], %[tmp4], %[tmp5] \n\t"
"bgtz %[i], 5b \n\t"
" addiu %[ptr_i], %[ptr_i], 8 \n\t"
"addiu %[tmp1], $zero, 13573 \n\t"
"addiu %[tmp2], $zero, 27146 \n\t"
#if !defined(MIPS32_R2_LE)
"sll %[tempMax], %[tempMax], 16 \n\t"
"sra %[tempMax], %[tempMax], 16 \n\t"
#else // #if !defined(MIPS32_R2_LE)
"seh %[tempMax] \n\t"
#endif // #if !defined(MIPS32_R2_LE)
"slt %[tmp1], %[tmp1], %[tempMax] \n\t"
"slt %[tmp2], %[tmp2], %[tempMax] \n\t"
"addu %[tmp1], %[tmp1], %[tmp2] \n\t"
"addu %[shift], %[shift], %[tmp1] \n\t"
"addu %[scale], %[scale], %[tmp1] \n\t"
"sllv %[round2], %[round2], %[tmp1] \n\t"
"sll %[istep], %[l], 1 \n\t"
"move %[m], $zero \n\t"
"sll %[tmp], %[l], 2 \n\t"
"2: \n\t"
#if defined(MIPS_DSP_R1_LE)
"sllv %[tmp3], %[m], %[k] \n\t"
"addiu %[tmp2], %[tmp3], 512 \n\t"
"addiu %[m], %[m], 1 \n\t"
"lhx %[wi], %[tmp3](%[kSinTable1024]) \n\t"
"lhx %[wr], %[tmp2](%[kSinTable1024]) \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"sllv %[tmp3], %[m], %[k] \n\t"
"addu %[ptr_j], %[tmp3], %[kSinTable1024] \n\t"
"addiu %[ptr_i], %[ptr_j], 512 \n\t"
"addiu %[m], %[m], 1 \n\t"
"lh %[wi], 0(%[ptr_j]) \n\t"
"lh %[wr], 0(%[ptr_i]) \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"1: \n\t"
"sll %[tmp1], %[i], 2 \n\t"
"addu %[ptr_i], %[frfi], %[tmp1] \n\t"
"addu %[ptr_j], %[ptr_i], %[tmp] \n\t"
"lh %[tmp3], 0(%[ptr_j]) \n\t"
"lh %[tmp4], 2(%[ptr_j]) \n\t"
"lh %[tmp6], 0(%[ptr_i]) \n\t"
"lh %[tmp5], 2(%[ptr_i]) \n\t"
"addu %[i], %[i], %[istep] \n\t"
#if defined(MIPS_DSP_R2_LE)
"mult %[wr], %[tmp3] \n\t"
"msub %[wi], %[tmp4] \n\t"
"mult $ac1, %[wr], %[tmp4] \n\t"
"madd $ac1, %[wi], %[tmp3] \n\t"
"mflo %[tmp1] \n\t"
"mflo %[tmp2], $ac1 \n\t"
"sll %[tmp6], %[tmp6], 14 \n\t"
"sll %[tmp5], %[tmp5], 14 \n\t"
"shra_r.w %[tmp1], %[tmp1], 1 \n\t"
"shra_r.w %[tmp2], %[tmp2], 1 \n\t"
"addu %[tmp6], %[tmp6], %[round2] \n\t"
"addu %[tmp5], %[tmp5], %[round2] \n\t"
"subu %[tmp4], %[tmp6], %[tmp1] \n\t"
"addu %[tmp1], %[tmp6], %[tmp1] \n\t"
"addu %[tmp6], %[tmp5], %[tmp2] \n\t"
"subu %[tmp5], %[tmp5], %[tmp2] \n\t"
"srav %[tmp4], %[tmp4], %[shift] \n\t"
"srav %[tmp1], %[tmp1], %[shift] \n\t"
"srav %[tmp6], %[tmp6], %[shift] \n\t"
"srav %[tmp5], %[tmp5], %[shift] \n\t"
#else // #if defined(MIPS_DSP_R2_LE)
"mul %[tmp1], %[wr], %[tmp3] \n\t"
"mul %[tmp2], %[wr], %[tmp4] \n\t"
"mul %[tmp4], %[wi], %[tmp4] \n\t"
"mul %[tmp3], %[wi], %[tmp3] \n\t"
"sll %[tmp6], %[tmp6], 14 \n\t"
"sll %[tmp5], %[tmp5], 14 \n\t"
"sub %[tmp1], %[tmp1], %[tmp4] \n\t"
"addu %[tmp2], %[tmp2], %[tmp3] \n\t"
"addiu %[tmp1], %[tmp1], 1 \n\t"
"addiu %[tmp2], %[tmp2], 1 \n\t"
"sra %[tmp2], %[tmp2], 1 \n\t"
"sra %[tmp1], %[tmp1], 1 \n\t"
"addu %[tmp6], %[tmp6], %[round2] \n\t"
"addu %[tmp5], %[tmp5], %[round2] \n\t"
"subu %[tmp4], %[tmp6], %[tmp1] \n\t"
"addu %[tmp1], %[tmp6], %[tmp1] \n\t"
"addu %[tmp6], %[tmp5], %[tmp2] \n\t"
"subu %[tmp5], %[tmp5], %[tmp2] \n\t"
"sra %[tmp4], %[tmp4], %[shift] \n\t"
"sra %[tmp1], %[tmp1], %[shift] \n\t"
"sra %[tmp6], %[tmp6], %[shift] \n\t"
"sra %[tmp5], %[tmp5], %[shift] \n\t"
#endif // #if defined(MIPS_DSP_R2_LE)
"sh %[tmp1], 0(%[ptr_i]) \n\t"
"sh %[tmp6], 2(%[ptr_i]) \n\t"
"sh %[tmp4], 0(%[ptr_j]) \n\t"
"blt %[i], %[n], 1b \n\t"
" sh %[tmp5], 2(%[ptr_j]) \n\t"
"blt %[m], %[l], 2b \n\t"
" addu %[i], $zero, %[m] \n\t"
"move %[l], %[istep] \n\t"
"blt %[l], %[n], 3b \n\t"
" addiu %[k], %[k], -1 \n\t"
".set pop \n\t"
: [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2), [tmp3] "=&r" (tmp3),
[tmp4] "=&r" (tmp4), [tmp5] "=&r" (tmp5), [tmp6] "=&r" (tmp6),
[ptr_i] "=&r" (ptr_i), [i] "=&r" (i), [m] "=&r" (m), [tmp] "=&r" (tmp),
[istep] "=&r" (istep), [wi] "=&r" (wi), [wr] "=&r" (wr), [l] "=&r" (l),
[k] "=&r" (k), [round2] "=&r" (round2), [ptr_j] "=&r" (ptr_j),
[shift] "=&r" (shift), [scale] "=&r" (scale), [tempMax] "=&r" (tempMax)
: [n] "r" (n), [frfi] "r" (frfi), [kSinTable1024] "r" (kSinTable1024)
: "hi", "lo", "$ac1hi", "$ac1lo", "memory"
);
return scale;
}

View File

@ -0,0 +1,148 @@
/*
* Copyright (c) 2013 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.
*/
#ifndef WEBRTC_COMMON_AUDIO_SIGNAL_PROCESSING_COMPLEX_FFT_TABLES_H_
#define WEBRTC_COMMON_AUDIO_SIGNAL_PROCESSING_COMPLEX_FFT_TABLES_H_
#include "webrtc/typedefs.h"
static const int16_t kSinTable1024[] = {
0, 201, 402, 603, 804, 1005, 1206, 1406,
1607, 1808, 2009, 2209, 2410, 2610, 2811, 3011,
3211, 3411, 3611, 3811, 4011, 4210, 4409, 4608,
4807, 5006, 5205, 5403, 5601, 5799, 5997, 6195,
6392, 6589, 6786, 6982, 7179, 7375, 7571, 7766,
7961, 8156, 8351, 8545, 8739, 8932, 9126, 9319,
9511, 9703, 9895, 10087, 10278, 10469, 10659, 10849,
11038, 11227, 11416, 11604, 11792, 11980, 12166, 12353,
12539, 12724, 12909, 13094, 13278, 13462, 13645, 13827,
14009, 14191, 14372, 14552, 14732, 14911, 15090, 15268,
15446, 15623, 15799, 15975, 16150, 16325, 16499, 16672,
16845, 17017, 17189, 17360, 17530, 17699, 17868, 18036,
18204, 18371, 18537, 18702, 18867, 19031, 19194, 19357,
19519, 19680, 19840, 20000, 20159, 20317, 20474, 20631,
20787, 20942, 21096, 21249, 21402, 21554, 21705, 21855,
22004, 22153, 22301, 22448, 22594, 22739, 22883, 23027,
23169, 23311, 23452, 23592, 23731, 23869, 24006, 24143,
24278, 24413, 24546, 24679, 24811, 24942, 25072, 25201,
25329, 25456, 25582, 25707, 25831, 25954, 26077, 26198,
26318, 26437, 26556, 26673, 26789, 26905, 27019, 27132,
27244, 27355, 27466, 27575, 27683, 27790, 27896, 28001,
28105, 28208, 28309, 28410, 28510, 28608, 28706, 28802,
28897, 28992, 29085, 29177, 29268, 29358, 29446, 29534,
29621, 29706, 29790, 29873, 29955, 30036, 30116, 30195,
30272, 30349, 30424, 30498, 30571, 30643, 30713, 30783,
30851, 30918, 30984, 31049, 31113, 31175, 31236, 31297,
31356, 31413, 31470, 31525, 31580, 31633, 31684, 31735,
31785, 31833, 31880, 31926, 31970, 32014, 32056, 32097,
32137, 32176, 32213, 32249, 32284, 32318, 32350, 32382,
32412, 32441, 32468, 32495, 32520, 32544, 32567, 32588,
32609, 32628, 32646, 32662, 32678, 32692, 32705, 32717,
32727, 32736, 32744, 32751, 32757, 32761, 32764, 32766,
32767, 32766, 32764, 32761, 32757, 32751, 32744, 32736,
32727, 32717, 32705, 32692, 32678, 32662, 32646, 32628,
32609, 32588, 32567, 32544, 32520, 32495, 32468, 32441,
32412, 32382, 32350, 32318, 32284, 32249, 32213, 32176,
32137, 32097, 32056, 32014, 31970, 31926, 31880, 31833,
31785, 31735, 31684, 31633, 31580, 31525, 31470, 31413,
31356, 31297, 31236, 31175, 31113, 31049, 30984, 30918,
30851, 30783, 30713, 30643, 30571, 30498, 30424, 30349,
30272, 30195, 30116, 30036, 29955, 29873, 29790, 29706,
29621, 29534, 29446, 29358, 29268, 29177, 29085, 28992,
28897, 28802, 28706, 28608, 28510, 28410, 28309, 28208,
28105, 28001, 27896, 27790, 27683, 27575, 27466, 27355,
27244, 27132, 27019, 26905, 26789, 26673, 26556, 26437,
26318, 26198, 26077, 25954, 25831, 25707, 25582, 25456,
25329, 25201, 25072, 24942, 24811, 24679, 24546, 24413,
24278, 24143, 24006, 23869, 23731, 23592, 23452, 23311,
23169, 23027, 22883, 22739, 22594, 22448, 22301, 22153,
22004, 21855, 21705, 21554, 21402, 21249, 21096, 20942,
20787, 20631, 20474, 20317, 20159, 20000, 19840, 19680,
19519, 19357, 19194, 19031, 18867, 18702, 18537, 18371,
18204, 18036, 17868, 17699, 17530, 17360, 17189, 17017,
16845, 16672, 16499, 16325, 16150, 15975, 15799, 15623,
15446, 15268, 15090, 14911, 14732, 14552, 14372, 14191,
14009, 13827, 13645, 13462, 13278, 13094, 12909, 12724,
12539, 12353, 12166, 11980, 11792, 11604, 11416, 11227,
11038, 10849, 10659, 10469, 10278, 10087, 9895, 9703,
9511, 9319, 9126, 8932, 8739, 8545, 8351, 8156,
7961, 7766, 7571, 7375, 7179, 6982, 6786, 6589,
6392, 6195, 5997, 5799, 5601, 5403, 5205, 5006,
4807, 4608, 4409, 4210, 4011, 3811, 3611, 3411,
3211, 3011, 2811, 2610, 2410, 2209, 2009, 1808,
1607, 1406, 1206, 1005, 804, 603, 402, 201,
0, -201, -402, -603, -804, -1005, -1206, -1406,
-1607, -1808, -2009, -2209, -2410, -2610, -2811, -3011,
-3211, -3411, -3611, -3811, -4011, -4210, -4409, -4608,
-4807, -5006, -5205, -5403, -5601, -5799, -5997, -6195,
-6392, -6589, -6786, -6982, -7179, -7375, -7571, -7766,
-7961, -8156, -8351, -8545, -8739, -8932, -9126, -9319,
-9511, -9703, -9895, -10087, -10278, -10469, -10659, -10849,
-11038, -11227, -11416, -11604, -11792, -11980, -12166, -12353,
-12539, -12724, -12909, -13094, -13278, -13462, -13645, -13827,
-14009, -14191, -14372, -14552, -14732, -14911, -15090, -15268,
-15446, -15623, -15799, -15975, -16150, -16325, -16499, -16672,
-16845, -17017, -17189, -17360, -17530, -17699, -17868, -18036,
-18204, -18371, -18537, -18702, -18867, -19031, -19194, -19357,
-19519, -19680, -19840, -20000, -20159, -20317, -20474, -20631,
-20787, -20942, -21096, -21249, -21402, -21554, -21705, -21855,
-22004, -22153, -22301, -22448, -22594, -22739, -22883, -23027,
-23169, -23311, -23452, -23592, -23731, -23869, -24006, -24143,
-24278, -24413, -24546, -24679, -24811, -24942, -25072, -25201,
-25329, -25456, -25582, -25707, -25831, -25954, -26077, -26198,
-26318, -26437, -26556, -26673, -26789, -26905, -27019, -27132,
-27244, -27355, -27466, -27575, -27683, -27790, -27896, -28001,
-28105, -28208, -28309, -28410, -28510, -28608, -28706, -28802,
-28897, -28992, -29085, -29177, -29268, -29358, -29446, -29534,
-29621, -29706, -29790, -29873, -29955, -30036, -30116, -30195,
-30272, -30349, -30424, -30498, -30571, -30643, -30713, -30783,
-30851, -30918, -30984, -31049, -31113, -31175, -31236, -31297,
-31356, -31413, -31470, -31525, -31580, -31633, -31684, -31735,
-31785, -31833, -31880, -31926, -31970, -32014, -32056, -32097,
-32137, -32176, -32213, -32249, -32284, -32318, -32350, -32382,
-32412, -32441, -32468, -32495, -32520, -32544, -32567, -32588,
-32609, -32628, -32646, -32662, -32678, -32692, -32705, -32717,
-32727, -32736, -32744, -32751, -32757, -32761, -32764, -32766,
-32767, -32766, -32764, -32761, -32757, -32751, -32744, -32736,
-32727, -32717, -32705, -32692, -32678, -32662, -32646, -32628,
-32609, -32588, -32567, -32544, -32520, -32495, -32468, -32441,
-32412, -32382, -32350, -32318, -32284, -32249, -32213, -32176,
-32137, -32097, -32056, -32014, -31970, -31926, -31880, -31833,
-31785, -31735, -31684, -31633, -31580, -31525, -31470, -31413,
-31356, -31297, -31236, -31175, -31113, -31049, -30984, -30918,
-30851, -30783, -30713, -30643, -30571, -30498, -30424, -30349,
-30272, -30195, -30116, -30036, -29955, -29873, -29790, -29706,
-29621, -29534, -29446, -29358, -29268, -29177, -29085, -28992,
-28897, -28802, -28706, -28608, -28510, -28410, -28309, -28208,
-28105, -28001, -27896, -27790, -27683, -27575, -27466, -27355,
-27244, -27132, -27019, -26905, -26789, -26673, -26556, -26437,
-26318, -26198, -26077, -25954, -25831, -25707, -25582, -25456,
-25329, -25201, -25072, -24942, -24811, -24679, -24546, -24413,
-24278, -24143, -24006, -23869, -23731, -23592, -23452, -23311,
-23169, -23027, -22883, -22739, -22594, -22448, -22301, -22153,
-22004, -21855, -21705, -21554, -21402, -21249, -21096, -20942,
-20787, -20631, -20474, -20317, -20159, -20000, -19840, -19680,
-19519, -19357, -19194, -19031, -18867, -18702, -18537, -18371,
-18204, -18036, -17868, -17699, -17530, -17360, -17189, -17017,
-16845, -16672, -16499, -16325, -16150, -15975, -15799, -15623,
-15446, -15268, -15090, -14911, -14732, -14552, -14372, -14191,
-14009, -13827, -13645, -13462, -13278, -13094, -12909, -12724,
-12539, -12353, -12166, -11980, -11792, -11604, -11416, -11227,
-11038, -10849, -10659, -10469, -10278, -10087, -9895, -9703,
-9511, -9319, -9126, -8932, -8739, -8545, -8351, -8156,
-7961, -7766, -7571, -7375, -7179, -6982, -6786, -6589,
-6392, -6195, -5997, -5799, -5601, -5403, -5205, -5006,
-4807, -4608, -4409, -4210, -4011, -3811, -3611, -3411,
-3211, -3011, -2811, -2610, -2410, -2209, -2009, -1808,
-1607, -1406, -1206, -1005, -804, -603, -402, -201
};
#endif // WEBRTC_COMMON_AUDIO_SIGNAL_PROCESSING_COMPLEX_FFT_TABLES_H_

View File

@ -0,0 +1,170 @@
/*
* Copyright (c) 2013 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.
*/
#include "webrtc/common_audio/signal_processing/include/signal_processing_library.h"
// Version of WebRtcSpl_DownsampleFast() for MIPS platforms.
int WebRtcSpl_DownsampleFast_mips(const int16_t* data_in,
int data_in_length,
int16_t* data_out,
int data_out_length,
const int16_t* __restrict coefficients,
int coefficients_length,
int factor,
int delay) {
int i;
int j;
int k;
int32_t out_s32 = 0;
int endpos = delay + factor * (data_out_length - 1) + 1;
int32_t tmp1, tmp2, tmp3, tmp4, factor_2;
int16_t* p_coefficients;
int16_t* p_data_in;
int16_t* p_data_in_0 = (int16_t*)&data_in[delay];
int16_t* p_coefficients_0 = (int16_t*)&coefficients[0];
#if !defined(MIPS_DSP_R1_LE)
int32_t max_16 = 0x7FFF;
int32_t min_16 = 0xFFFF8000;
#endif // #if !defined(MIPS_DSP_R1_LE)
// Return error if any of the running conditions doesn't meet.
if (data_out_length <= 0 || coefficients_length <= 0
|| data_in_length < endpos) {
return -1;
}
#if defined(MIPS_DSP_R2_LE)
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"subu %[i], %[endpos], %[delay] \n\t"
"sll %[factor_2], %[factor], 1 \n\t"
"1: \n\t"
"move %[p_data_in], %[p_data_in_0] \n\t"
"mult $zero, $zero \n\t"
"move %[p_coefs], %[p_coefs_0] \n\t"
"sra %[j], %[coef_length], 2 \n\t"
"beq %[j], $zero, 3f \n\t"
" andi %[k], %[coef_length], 3 \n\t"
"2: \n\t"
"lwl %[tmp1], 1(%[p_data_in]) \n\t"
"lwl %[tmp2], 3(%[p_coefs]) \n\t"
"lwl %[tmp3], -3(%[p_data_in]) \n\t"
"lwl %[tmp4], 7(%[p_coefs]) \n\t"
"lwr %[tmp1], -2(%[p_data_in]) \n\t"
"lwr %[tmp2], 0(%[p_coefs]) \n\t"
"lwr %[tmp3], -6(%[p_data_in]) \n\t"
"lwr %[tmp4], 4(%[p_coefs]) \n\t"
"packrl.ph %[tmp1], %[tmp1], %[tmp1] \n\t"
"packrl.ph %[tmp3], %[tmp3], %[tmp3] \n\t"
"dpa.w.ph $ac0, %[tmp1], %[tmp2] \n\t"
"dpa.w.ph $ac0, %[tmp3], %[tmp4] \n\t"
"addiu %[j], %[j], -1 \n\t"
"addiu %[p_data_in], %[p_data_in], -8 \n\t"
"bgtz %[j], 2b \n\t"
" addiu %[p_coefs], %[p_coefs], 8 \n\t"
"3: \n\t"
"beq %[k], $zero, 5f \n\t"
" nop \n\t"
"4: \n\t"
"lhu %[tmp1], 0(%[p_data_in]) \n\t"
"lhu %[tmp2], 0(%[p_coefs]) \n\t"
"addiu %[p_data_in], %[p_data_in], -2 \n\t"
"addiu %[k], %[k], -1 \n\t"
"dpa.w.ph $ac0, %[tmp1], %[tmp2] \n\t"
"bgtz %[k], 4b \n\t"
" addiu %[p_coefs], %[p_coefs], 2 \n\t"
"5: \n\t"
"extr_r.w %[out_s32], $ac0, 12 \n\t"
"addu %[p_data_in_0], %[p_data_in_0], %[factor_2] \n\t"
"subu %[i], %[i], %[factor] \n\t"
"shll_s.w %[out_s32], %[out_s32], 16 \n\t"
"sra %[out_s32], %[out_s32], 16 \n\t"
"sh %[out_s32], 0(%[data_out]) \n\t"
"bgtz %[i], 1b \n\t"
" addiu %[data_out], %[data_out], 2 \n\t"
".set pop \n\t"
: [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2), [tmp3] "=&r" (tmp3),
[tmp4] "=&r" (tmp4), [p_data_in] "=&r" (p_data_in),
[p_data_in_0] "+r" (p_data_in_0), [p_coefs] "=&r" (p_coefficients),
[j] "=&r" (j), [out_s32] "=&r" (out_s32), [factor_2] "=&r" (factor_2),
[i] "=&r" (i), [k] "=&r" (k)
: [coef_length] "r" (coefficients_length), [data_out] "r" (data_out),
[p_coefs_0] "r" (p_coefficients_0), [endpos] "r" (endpos),
[delay] "r" (delay), [factor] "r" (factor)
: "memory", "hi", "lo"
);
#else // #if defined(MIPS_DSP_R2_LE)
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"sll %[factor_2], %[factor], 1 \n\t"
"subu %[i], %[endpos], %[delay] \n\t"
"1: \n\t"
"move %[p_data_in], %[p_data_in_0] \n\t"
"addiu %[out_s32], $zero, 2048 \n\t"
"move %[p_coefs], %[p_coefs_0] \n\t"
"sra %[j], %[coef_length], 1 \n\t"
"beq %[j], $zero, 3f \n\t"
" andi %[k], %[coef_length], 1 \n\t"
"2: \n\t"
"lh %[tmp1], 0(%[p_data_in]) \n\t"
"lh %[tmp2], 0(%[p_coefs]) \n\t"
"lh %[tmp3], -2(%[p_data_in]) \n\t"
"lh %[tmp4], 2(%[p_coefs]) \n\t"
"mul %[tmp1], %[tmp1], %[tmp2] \n\t"
"addiu %[p_coefs], %[p_coefs], 4 \n\t"
"mul %[tmp3], %[tmp3], %[tmp4] \n\t"
"addiu %[j], %[j], -1 \n\t"
"addiu %[p_data_in], %[p_data_in], -4 \n\t"
"addu %[tmp1], %[tmp1], %[tmp3] \n\t"
"bgtz %[j], 2b \n\t"
" addu %[out_s32], %[out_s32], %[tmp1] \n\t"
"3: \n\t"
"beq %[k], $zero, 4f \n\t"
" nop \n\t"
"lh %[tmp1], 0(%[p_data_in]) \n\t"
"lh %[tmp2], 0(%[p_coefs]) \n\t"
"mul %[tmp1], %[tmp1], %[tmp2] \n\t"
"addu %[out_s32], %[out_s32], %[tmp1] \n\t"
"4: \n\t"
"sra %[out_s32], %[out_s32], 12 \n\t"
"addu %[p_data_in_0], %[p_data_in_0], %[factor_2] \n\t"
#if defined(MIPS_DSP_R1_LE)
"shll_s.w %[out_s32], %[out_s32], 16 \n\t"
"sra %[out_s32], %[out_s32], 16 \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"slt %[tmp1], %[max_16], %[out_s32] \n\t"
"movn %[out_s32], %[max_16], %[tmp1] \n\t"
"slt %[tmp1], %[out_s32], %[min_16] \n\t"
"movn %[out_s32], %[min_16], %[tmp1] \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"subu %[i], %[i], %[factor] \n\t"
"sh %[out_s32], 0(%[data_out]) \n\t"
"bgtz %[i], 1b \n\t"
" addiu %[data_out], %[data_out], 2 \n\t"
".set pop \n\t"
: [tmp1] "=&r" (tmp1), [tmp2] "=&r" (tmp2), [tmp3] "=&r" (tmp3),
[tmp4] "=&r" (tmp4), [p_data_in] "=&r" (p_data_in), [k] "=&r" (k),
[p_data_in_0] "+r" (p_data_in_0), [p_coefs] "=&r" (p_coefficients),
[j] "=&r" (j), [out_s32] "=&r" (out_s32), [factor_2] "=&r" (factor_2),
[i] "=&r" (i)
: [coef_length] "r" (coefficients_length), [data_out] "r" (data_out),
[p_coefs_0] "r" (p_coefficients_0), [endpos] "r" (endpos),
#if !defined(MIPS_DSP_R1_LE)
[max_16] "r" (max_16), [min_16] "r" (min_16),
#endif // #if !defined(MIPS_DSP_R1_LE)
[delay] "r" (delay), [factor] "r" (factor)
: "memory", "hi", "lo"
);
#endif // #if defined(MIPS_DSP_R2_LE)
return 0;
}

View File

@ -0,0 +1,140 @@
/*
* Copyright (c) 2013 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.
*/
#include <assert.h>
#include "signal_processing_library.h"
void WebRtcSpl_FilterARFastQ12(const int16_t* data_in,
int16_t* data_out,
const int16_t* __restrict coefficients,
int coefficients_length,
int data_length) {
int r0, r1, r2, r3;
int coef0, offset;
int i, j, k;
int coefptr, outptr, tmpout, inptr;
#if !defined(MIPS_DSP_R1_LE)
int max16 = 0x7FFF;
int min16 = 0xFFFF8000;
#endif // #if !defined(MIPS_DSP_R1_LE)
assert(data_length > 0);
assert(coefficients_length > 1);
__asm __volatile (
".set push \n\t"
".set noreorder \n\t"
"addiu %[i], %[data_length], 0 \n\t"
"lh %[coef0], 0(%[coefficients]) \n\t"
"addiu %[j], %[coefficients_length], -1 \n\t"
"andi %[k], %[j], 1 \n\t"
"sll %[offset], %[j], 1 \n\t"
"subu %[outptr], %[data_out], %[offset] \n\t"
"addiu %[inptr], %[data_in], 0 \n\t"
"bgtz %[k], 3f \n\t"
" addu %[coefptr], %[coefficients], %[offset] \n\t"
"1: \n\t"
"lh %[r0], 0(%[inptr]) \n\t"
"addiu %[i], %[i], -1 \n\t"
"addiu %[tmpout], %[outptr], 0 \n\t"
"mult %[r0], %[coef0] \n\t"
"2: \n\t"
"lh %[r0], 0(%[tmpout]) \n\t"
"lh %[r1], 0(%[coefptr]) \n\t"
"lh %[r2], 2(%[tmpout]) \n\t"
"lh %[r3], -2(%[coefptr]) \n\t"
"addiu %[tmpout], %[tmpout], 4 \n\t"
"msub %[r0], %[r1] \n\t"
"msub %[r2], %[r3] \n\t"
"addiu %[j], %[j], -2 \n\t"
"bgtz %[j], 2b \n\t"
" addiu %[coefptr], %[coefptr], -4 \n\t"
#if defined(MIPS_DSP_R1_LE)
"extr_r.w %[r0], $ac0, 12 \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"mflo %[r0] \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"addu %[coefptr], %[coefficients], %[offset] \n\t"
"addiu %[inptr], %[inptr], 2 \n\t"
"addiu %[j], %[coefficients_length], -1 \n\t"
#if defined(MIPS_DSP_R1_LE)
"shll_s.w %[r0], %[r0], 16 \n\t"
"sra %[r0], %[r0], 16 \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"addiu %[r0], %[r0], 2048 \n\t"
"sra %[r0], %[r0], 12 \n\t"
"slt %[r1], %[max16], %[r0] \n\t"
"movn %[r0], %[max16], %[r1] \n\t"
"slt %[r1], %[r0], %[min16] \n\t"
"movn %[r0], %[min16], %[r1] \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"sh %[r0], 0(%[tmpout]) \n\t"
"bgtz %[i], 1b \n\t"
" addiu %[outptr], %[outptr], 2 \n\t"
"b 5f \n\t"
" nop \n\t"
"3: \n\t"
"lh %[r0], 0(%[inptr]) \n\t"
"addiu %[i], %[i], -1 \n\t"
"addiu %[tmpout], %[outptr], 0 \n\t"
"mult %[r0], %[coef0] \n\t"
"4: \n\t"
"lh %[r0], 0(%[tmpout]) \n\t"
"lh %[r1], 0(%[coefptr]) \n\t"
"lh %[r2], 2(%[tmpout]) \n\t"
"lh %[r3], -2(%[coefptr]) \n\t"
"addiu %[tmpout], %[tmpout], 4 \n\t"
"msub %[r0], %[r1] \n\t"
"msub %[r2], %[r3] \n\t"
"addiu %[j], %[j], -2 \n\t"
"bgtz %[j], 4b \n\t"
" addiu %[coefptr], %[coefptr], -4 \n\t"
"lh %[r0], 0(%[tmpout]) \n\t"
"lh %[r1], 0(%[coefptr]) \n\t"
"msub %[r0], %[r1] \n\t"
#if defined(MIPS_DSP_R1_LE)
"extr_r.w %[r0], $ac0, 12 \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"mflo %[r0] \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"addu %[coefptr], %[coefficients], %[offset] \n\t"
"addiu %[inptr], %[inptr], 2 \n\t"
"addiu %[j], %[coefficients_length], -1 \n\t"
#if defined(MIPS_DSP_R1_LE)
"shll_s.w %[r0], %[r0], 16 \n\t"
"sra %[r0], %[r0], 16 \n\t"
#else // #if defined(MIPS_DSP_R1_LE)
"addiu %[r0], %[r0], 2048 \n\t"
"sra %[r0], %[r0], 12 \n\t"
"slt %[r1], %[max16], %[r0] \n\t"
"movn %[r0], %[max16], %[r1] \n\t"
"slt %[r1], %[r0], %[min16] \n\t"
"movn %[r0], %[min16], %[r1] \n\t"
#endif // #if defined(MIPS_DSP_R1_LE)
"sh %[r0], 2(%[tmpout]) \n\t"
"bgtz %[i], 3b \n\t"
" addiu %[outptr], %[outptr], 2 \n\t"
"5: \n\t"
".set pop \n\t"
: [i] "=&r" (i), [j] "=&r" (j), [k] "=&r" (k), [r0] "=&r" (r0),
[r1] "=&r" (r1), [r2] "=&r" (r2), [r3] "=&r" (r3),
[coef0] "=&r" (coef0), [offset] "=&r" (offset),
[outptr] "=&r" (outptr), [inptr] "=&r" (inptr),
[coefptr] "=&r" (coefptr), [tmpout] "=&r" (tmpout)
: [coefficients] "r" (coefficients), [data_length] "r" (data_length),
[coefficients_length] "r" (coefficients_length),
#if !defined(MIPS_DSP_R1_LE)
[max16] "r" (max16), [min16] "r" (min16),
#endif
[data_out] "r" (data_out), [data_in] "r" (data_in)
: "hi", "lo", "memory"
);
}

View File

@ -771,6 +771,16 @@ int WebRtcSpl_DownsampleFastNeon(const int16_t* data_in,
int factor,
int delay);
#endif
#if defined(MIPS32_LE)
int WebRtcSpl_DownsampleFast_mips(const int16_t* data_in,
int data_in_length,
int16_t* data_out,
int data_out_length,
const int16_t* __restrict coefficients,
int coefficients_length,
int factor,
int delay);
#endif
// End: Filter operations.

View File

@ -77,7 +77,7 @@ static void InitPointersToMIPS() {
WebRtcSpl_MinValueW16 = WebRtcSpl_MinValueW16_mips;
WebRtcSpl_MinValueW32 = WebRtcSpl_MinValueW32_mips;
WebRtcSpl_CrossCorrelation = WebRtcSpl_CrossCorrelationC;
WebRtcSpl_DownsampleFast = WebRtcSpl_DownsampleFastC;
WebRtcSpl_DownsampleFast = WebRtcSpl_DownsampleFast_mips;
WebRtcSpl_ScaleAndAddVectorsWithRound =
WebRtcSpl_ScaleAndAddVectorsWithRoundC;
WebRtcSpl_RealForwardFFT = WebRtcSpl_RealForwardFFTC;