fft_altivec.c
Go to the documentation of this file.
1 /*
2  * FFT/IFFT transforms
3  * AltiVec-enabled
4  * Copyright (c) 2009 Loren Merritt
5  *
6  * This file is part of Libav.
7  *
8  * Libav is free software; you can redistribute it and/or
9  * modify it under the terms of the GNU Lesser General Public
10  * License as published by the Free Software Foundation; either
11  * version 2.1 of the License, or (at your option) any later version.
12  *
13  * Libav is distributed in the hope that it will be useful,
14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16  * Lesser General Public License for more details.
17  *
18  * You should have received a copy of the GNU Lesser General Public
19  * License along with Libav; if not, write to the Free Software
20  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21  */
22 #include "libavcodec/fft.h"
23 #include "util_altivec.h"
24 #include "types_altivec.h"
25 
37 
38 #if HAVE_GNU_AS
39 static void ff_imdct_half_altivec(FFTContext *s, FFTSample *output, const FFTSample *input)
40 {
41  int j, k;
42  int n = 1 << s->mdct_bits;
43  int n4 = n >> 2;
44  int n8 = n >> 3;
45  int n32 = n >> 5;
46  const uint16_t *revtabj = s->revtab;
47  const uint16_t *revtabk = s->revtab+n4;
48  const vec_f *tcos = (const vec_f*)(s->tcos+n8);
49  const vec_f *tsin = (const vec_f*)(s->tsin+n8);
50  const vec_f *pin = (const vec_f*)(input+n4);
51  vec_f *pout = (vec_f*)(output+n4);
52 
53  /* pre rotation */
54  k = n32-1;
55  do {
56  vec_f cos,sin,cos0,sin0,cos1,sin1,re,im,r0,i0,r1,i1,a,b,c,d;
57 #define CMULA(p,o0,o1,o2,o3)\
58  a = pin[ k*2+p]; /* { z[k].re, z[k].im, z[k+1].re, z[k+1].im } */\
59  b = pin[-k*2-p-1]; /* { z[-k-2].re, z[-k-2].im, z[-k-1].re, z[-k-1].im } */\
60  re = vec_perm(a, b, vcprm(0,2,s0,s2)); /* { z[k].re, z[k+1].re, z[-k-2].re, z[-k-1].re } */\
61  im = vec_perm(a, b, vcprm(s3,s1,3,1)); /* { z[-k-1].im, z[-k-2].im, z[k+1].im, z[k].im } */\
62  cos = vec_perm(cos0, cos1, vcprm(o0,o1,s##o2,s##o3)); /* { cos[k], cos[k+1], cos[-k-2], cos[-k-1] } */\
63  sin = vec_perm(sin0, sin1, vcprm(o0,o1,s##o2,s##o3));\
64  r##p = im*cos - re*sin;\
65  i##p = re*cos + im*sin;
66 #define STORE2(v,dst)\
67  j = dst;\
68  vec_ste(v, 0, output+j*2);\
69  vec_ste(v, 4, output+j*2);
70 #define STORE8(p)\
71  a = vec_perm(r##p, i##p, vcprm(0,s0,0,s0));\
72  b = vec_perm(r##p, i##p, vcprm(1,s1,1,s1));\
73  c = vec_perm(r##p, i##p, vcprm(2,s2,2,s2));\
74  d = vec_perm(r##p, i##p, vcprm(3,s3,3,s3));\
75  STORE2(a, revtabk[ p*2-4]);\
76  STORE2(b, revtabk[ p*2-3]);\
77  STORE2(c, revtabj[-p*2+2]);\
78  STORE2(d, revtabj[-p*2+3]);
79 
80  cos0 = tcos[k];
81  sin0 = tsin[k];
82  cos1 = tcos[-k-1];
83  sin1 = tsin[-k-1];
84  CMULA(0, 0,1,2,3);
85  CMULA(1, 2,3,0,1);
86  STORE8(0);
87  STORE8(1);
88  revtabj += 4;
89  revtabk -= 4;
90  k--;
91  } while(k >= 0);
92 
93  ff_fft_calc_altivec(s, (FFTComplex*)output);
94 
95  /* post rotation + reordering */
96  j = -n32;
97  k = n32-1;
98  do {
99  vec_f cos,sin,re,im,a,b,c,d;
100 #define CMULB(d0,d1,o)\
101  re = pout[o*2];\
102  im = pout[o*2+1];\
103  cos = tcos[o];\
104  sin = tsin[o];\
105  d0 = im*sin - re*cos;\
106  d1 = re*sin + im*cos;
107 
108  CMULB(a,b,j);
109  CMULB(c,d,k);
110  pout[2*j] = vec_perm(a, d, vcprm(0,s3,1,s2));
111  pout[2*j+1] = vec_perm(a, d, vcprm(2,s1,3,s0));
112  pout[2*k] = vec_perm(c, b, vcprm(0,s3,1,s2));
113  pout[2*k+1] = vec_perm(c, b, vcprm(2,s1,3,s0));
114  j++;
115  k--;
116  } while(k >= 0);
117 }
118 
119 static void ff_imdct_calc_altivec(FFTContext *s, FFTSample *output, const FFTSample *input)
120 {
121  int k;
122  int n = 1 << s->mdct_bits;
123  int n4 = n >> 2;
124  int n16 = n >> 4;
125  vec_u32 sign = {1U<<31,1U<<31,1U<<31,1U<<31};
126  vec_u32 *p0 = (vec_u32*)(output+n4);
127  vec_u32 *p1 = (vec_u32*)(output+n4*3);
128 
129  ff_imdct_half_altivec(s, output+n4, input);
130 
131  for (k = 0; k < n16; k++) {
132  vec_u32 a = p0[k] ^ sign;
133  vec_u32 b = p1[-k-1];
134  p0[-k-1] = vec_perm(a, a, vcprm(3,2,1,0));
135  p1[k] = vec_perm(b, b, vcprm(3,2,1,0));
136  }
137 }
138 #endif /* HAVE_GNU_AS */
139 
141 {
142 #if HAVE_GNU_AS
144  if (s->mdct_bits >= 5) {
145  s->imdct_calc = ff_imdct_calc_altivec;
146  s->imdct_half = ff_imdct_half_altivec;
147  }
148 #endif
149 }