float_altivec.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2006 Luca Barbato <lu_zero@gentoo.org>
3  *
4  * This file is part of Libav.
5  *
6  * Libav is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * Libav is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with Libav; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 #include "libavcodec/dsputil.h"
22 
23 #include "dsputil_altivec.h"
24 #include "util_altivec.h"
25 
26 static void vector_fmul_altivec(float *dst, const float *src0, const float *src1, int len)
27 {
28  int i;
29  vector float d0, d1, s, zero = (vector float)vec_splat_u32(0);
30  for(i=0; i<len-7; i+=8) {
31  d0 = vec_ld(0, src0+i);
32  s = vec_ld(0, src1+i);
33  d1 = vec_ld(16, src0+i);
34  d0 = vec_madd(d0, s, zero);
35  d1 = vec_madd(d1, vec_ld(16,src1+i), zero);
36  vec_st(d0, 0, dst+i);
37  vec_st(d1, 16, dst+i);
38  }
39 }
40 
41 static void vector_fmul_reverse_altivec(float *dst, const float *src0,
42  const float *src1, int len)
43 {
44  int i;
45  vector float d, s0, s1, h0, l0,
46  s2, s3, zero = (vector float)vec_splat_u32(0);
47  src1 += len-4;
48  for(i=0; i<len-7; i+=8) {
49  s1 = vec_ld(0, src1-i); // [a,b,c,d]
50  s0 = vec_ld(0, src0+i);
51  l0 = vec_mergel(s1, s1); // [c,c,d,d]
52  s3 = vec_ld(-16, src1-i);
53  h0 = vec_mergeh(s1, s1); // [a,a,b,b]
54  s2 = vec_ld(16, src0+i);
55  s1 = vec_mergeh(vec_mergel(l0,h0), // [d,b,d,b]
56  vec_mergeh(l0,h0)); // [c,a,c,a]
57  // [d,c,b,a]
58  l0 = vec_mergel(s3, s3);
59  d = vec_madd(s0, s1, zero);
60  h0 = vec_mergeh(s3, s3);
61  vec_st(d, 0, dst+i);
62  s3 = vec_mergeh(vec_mergel(l0,h0),
63  vec_mergeh(l0,h0));
64  d = vec_madd(s2, s3, zero);
65  vec_st(d, 16, dst+i);
66  }
67 }
68 
69 static void vector_fmul_add_altivec(float *dst, const float *src0,
70  const float *src1, const float *src2,
71  int len)
72 {
73  int i;
74  vector float d, s0, s1, s2, t0, t1, edges;
75  vector unsigned char align = vec_lvsr(0,dst),
76  mask = vec_lvsl(0, dst);
77 
78  for (i=0; i<len-3; i+=4) {
79  t0 = vec_ld(0, dst+i);
80  t1 = vec_ld(15, dst+i);
81  s0 = vec_ld(0, src0+i);
82  s1 = vec_ld(0, src1+i);
83  s2 = vec_ld(0, src2+i);
84  edges = vec_perm(t1 ,t0, mask);
85  d = vec_madd(s0,s1,s2);
86  t1 = vec_perm(d, edges, align);
87  t0 = vec_perm(edges, d, align);
88  vec_st(t1, 15, dst+i);
89  vec_st(t0, 0, dst+i);
90  }
91 }
92 
93 static void vector_fmul_window_altivec(float *dst, const float *src0, const float *src1, const float *win, int len)
94 {
95  vector float zero, t0, t1, s0, s1, wi, wj;
96  const vector unsigned char reverse = vcprm(3,2,1,0);
97  int i,j;
98 
99  dst += len;
100  win += len;
101  src0+= len;
102 
103  zero = (vector float)vec_splat_u32(0);
104 
105  for(i=-len*4, j=len*4-16; i<0; i+=16, j-=16) {
106  s0 = vec_ld(i, src0);
107  s1 = vec_ld(j, src1);
108  wi = vec_ld(i, win);
109  wj = vec_ld(j, win);
110 
111  s1 = vec_perm(s1, s1, reverse);
112  wj = vec_perm(wj, wj, reverse);
113 
114  t0 = vec_madd(s0, wj, zero);
115  t0 = vec_nmsub(s1, wi, t0);
116  t1 = vec_madd(s0, wi, zero);
117  t1 = vec_madd(s1, wj, t1);
118  t1 = vec_perm(t1, t1, reverse);
119 
120  vec_st(t0, i, dst);
121  vec_st(t1, j, dst);
122  }
123 }
124 
126 {
130  if(!(avctx->flags & CODEC_FLAG_BITEXACT)) {
132  }
133 }