1 /* Copyright (C) 2007-2008 Jean-Marc Valin
2 * Copyright (C) 2008 Thorvald Natvig
3 */
4 /**
5 @file resample_sse.h
6 @brief Resampler functions (SSE version)
7 */
8 /*
9 Redistribution and use in source and binary forms, with or without
10 modification, are permitted provided that the following conditions
11 are met:
12
13 - Redistributions of source code must retain the above copyright
14 notice, this list of conditions and the following disclaimer.
15
16 - Redistributions in binary form must reproduce the above copyright
17 notice, this list of conditions and the following disclaimer in the
18 documentation and/or other materials provided with the distribution.
19
20 - Neither the name of the Xiph.org Foundation nor the names of its
21 contributors may be used to endorse or promote products derived from
22 this software without specific prior written permission.
23
24 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
25 ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
26 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
27 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR
28 CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
29 EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
30 PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
31 PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
32 LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
33 NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
34 SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
35 */
36
37 #include "simd_detect.h"
38
39 #include <xmmintrin.h>
40
41 #define OVERRIDE_INNER_PRODUCT_SINGLE
inner_product_single(const float * a,const float * b,unsigned int len)42 float inner_product_single(const float *a, const float *b, unsigned int len)
43 {
44 int i;
45 float ret;
46 __m128 sum = _mm_setzero_ps();
47 for (i=0;i<len;i+=8)
48 {
49 sum = _mm_add_ps(sum, _mm_mul_ps(_mm_loadu_ps(a+i), _mm_loadu_ps(b+i)));
50 sum = _mm_add_ps(sum, _mm_mul_ps(_mm_loadu_ps(a+i+4), _mm_loadu_ps(b+i+4)));
51 }
52 sum = _mm_add_ps(sum, _mm_movehl_ps(sum, sum));
53 sum = _mm_add_ss(sum, _mm_shuffle_ps(sum, sum, 0x55));
54 _mm_store_ss(&ret, sum);
55 return ret;
56 }
57
58 #define OVERRIDE_INTERPOLATE_PRODUCT_SINGLE
interpolate_product_single(const float * a,const float * b,unsigned int len,const spx_uint32_t oversample,float * frac)59 float interpolate_product_single(const float *a, const float *b, unsigned int len, const spx_uint32_t oversample, float *frac) {
60 int i;
61 float ret;
62 __m128 sum = _mm_setzero_ps();
63 __m128 f = _mm_loadu_ps(frac);
64 for(i=0;i<len;i+=2)
65 {
66 sum = _mm_add_ps(sum, _mm_mul_ps(_mm_load1_ps(a+i), _mm_loadu_ps(b+i*oversample)));
67 sum = _mm_add_ps(sum, _mm_mul_ps(_mm_load1_ps(a+i+1), _mm_loadu_ps(b+(i+1)*oversample)));
68 }
69 sum = _mm_mul_ps(f, sum);
70 sum = _mm_add_ps(sum, _mm_movehl_ps(sum, sum));
71 sum = _mm_add_ss(sum, _mm_shuffle_ps(sum, sum, 0x55));
72 _mm_store_ss(&ret, sum);
73 return ret;
74 }
75
76 #ifdef _USE_SSE2
77 #include <emmintrin.h>
78 #define OVERRIDE_INNER_PRODUCT_DOUBLE
79
inner_product_double(const float * a,const float * b,unsigned int len)80 double inner_product_double(const float *a, const float *b, unsigned int len)
81 {
82 int i;
83 double ret;
84 __m128d sum = _mm_setzero_pd();
85 __m128 t;
86 for (i=0;i<len;i+=8)
87 {
88 t = _mm_mul_ps(_mm_loadu_ps(a+i), _mm_loadu_ps(b+i));
89 sum = _mm_add_pd(sum, _mm_cvtps_pd(t));
90 sum = _mm_add_pd(sum, _mm_cvtps_pd(_mm_movehl_ps(t, t)));
91
92 t = _mm_mul_ps(_mm_loadu_ps(a+i+4), _mm_loadu_ps(b+i+4));
93 sum = _mm_add_pd(sum, _mm_cvtps_pd(t));
94 sum = _mm_add_pd(sum, _mm_cvtps_pd(_mm_movehl_ps(t, t)));
95 }
96 sum = _mm_add_sd(sum, _mm_unpackhi_pd(sum, sum));
97 _mm_store_sd(&ret, sum);
98 return ret;
99 }
100
101 #define OVERRIDE_INTERPOLATE_PRODUCT_DOUBLE
interpolate_product_double(const float * a,const float * b,unsigned int len,const spx_uint32_t oversample,float * frac)102 double interpolate_product_double(const float *a, const float *b, unsigned int len, const spx_uint32_t oversample, float *frac) {
103 int i;
104 double ret;
105 __m128d sum;
106 __m128d sum1 = _mm_setzero_pd();
107 __m128d sum2 = _mm_setzero_pd();
108 __m128 f = _mm_loadu_ps(frac);
109 __m128d f1 = _mm_cvtps_pd(f);
110 __m128d f2 = _mm_cvtps_pd(_mm_movehl_ps(f,f));
111 __m128 t;
112 for(i=0;i<len;i+=2)
113 {
114 t = _mm_mul_ps(_mm_load1_ps(a+i), _mm_loadu_ps(b+i*oversample));
115 sum1 = _mm_add_pd(sum1, _mm_cvtps_pd(t));
116 sum2 = _mm_add_pd(sum2, _mm_cvtps_pd(_mm_movehl_ps(t, t)));
117
118 t = _mm_mul_ps(_mm_load1_ps(a+i+1), _mm_loadu_ps(b+(i+1)*oversample));
119 sum1 = _mm_add_pd(sum1, _mm_cvtps_pd(t));
120 sum2 = _mm_add_pd(sum2, _mm_cvtps_pd(_mm_movehl_ps(t, t)));
121 }
122 sum1 = _mm_mul_pd(f1, sum1);
123 sum2 = _mm_mul_pd(f2, sum2);
124 sum = _mm_add_pd(sum1, sum2);
125 sum = _mm_add_sd(sum, _mm_unpackhi_pd(sum, sum));
126 _mm_store_sd(&ret, sum);
127 return ret;
128 }
129
130 #endif
131