2003-01-07 19:15:48 +01:00
|
|
|
/*
|
|
|
|
* FFT/IFFT transforms
|
|
|
|
* AltiVec-enabled
|
2003-01-08 19:47:49 +01:00
|
|
|
* Copyright (c) 2003 Romain Dolbeau <romain@dolbeau.org>
|
2009-01-19 16:46:40 +01:00
|
|
|
* Based on code Copyright (c) 2002 Fabrice Bellard
|
2003-01-07 19:15:48 +01:00
|
|
|
*
|
2006-10-07 17:30:46 +02:00
|
|
|
* This file is part of FFmpeg.
|
|
|
|
*
|
|
|
|
* FFmpeg is free software; you can redistribute it and/or
|
2003-01-07 19:15:48 +01:00
|
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
|
|
* License as published by the Free Software Foundation; either
|
2006-10-07 17:30:46 +02:00
|
|
|
* version 2.1 of the License, or (at your option) any later version.
|
2003-01-07 19:15:48 +01:00
|
|
|
*
|
2006-10-07 17:30:46 +02:00
|
|
|
* FFmpeg is distributed in the hope that it will be useful,
|
2003-01-07 19:15:48 +01:00
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
* Lesser General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU Lesser General Public
|
2006-10-07 17:30:46 +02:00
|
|
|
* License along with FFmpeg; if not, write to the Free Software
|
2006-01-12 23:43:26 +01:00
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
2003-01-07 19:15:48 +01:00
|
|
|
*/
|
2008-05-09 13:56:36 +02:00
|
|
|
#include "libavcodec/dsputil.h"
|
2007-10-02 13:39:32 +02:00
|
|
|
#include "dsputil_ppc.h"
|
|
|
|
#include "util_altivec.h"
|
2003-01-07 19:15:48 +01:00
|
|
|
/**
|
2004-03-13 22:43:24 +01:00
|
|
|
* Do a complex FFT with the parameters defined in ff_fft_init(). The
|
2003-01-07 19:15:48 +01:00
|
|
|
* input data must be permuted before with s->revtab table. No
|
|
|
|
* 1.0/sqrt(n) normalization is done.
|
|
|
|
* AltiVec-enabled
|
|
|
|
* This code assumes that the 'z' pointer is 16 bytes-aligned
|
|
|
|
* It also assumes all FFTComplex are 8 bytes-aligned pair of float
|
|
|
|
* The code is exactly the same as the SSE version, except
|
2003-01-08 19:47:49 +01:00
|
|
|
* that successive MUL + ADD/SUB have been merged into
|
2003-01-07 19:15:48 +01:00
|
|
|
* fused multiply-add ('vec_madd' in altivec)
|
|
|
|
*/
|
2004-03-13 22:43:24 +01:00
|
|
|
void ff_fft_calc_altivec(FFTContext *s, FFTComplex *z)
|
2003-01-07 19:15:48 +01:00
|
|
|
{
|
2003-07-09 22:18:13 +02:00
|
|
|
POWERPC_PERF_DECLARE(altivec_fft_num, s->nbits >= 6);
|
2006-11-07 13:45:29 +01:00
|
|
|
register const vector float vczero = (const vector float)vec_splat_u32(0.);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
int ln = s->nbits;
|
2005-12-22 02:10:11 +01:00
|
|
|
int j, np, np2;
|
|
|
|
int nblocks, nloops;
|
2003-01-07 19:15:48 +01:00
|
|
|
register FFTComplex *p, *q;
|
|
|
|
FFTComplex *cptr, *cptr1;
|
|
|
|
int k;
|
|
|
|
|
2003-07-09 22:18:13 +02:00
|
|
|
POWERPC_PERF_START_COUNT(altivec_fft_num, s->nbits >= 6);
|
2003-01-16 22:54:55 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
np = 1 << ln;
|
|
|
|
|
|
|
|
{
|
|
|
|
vector float *r, a, b, a1, c1, c2;
|
|
|
|
|
|
|
|
r = (vector float *)&z[0];
|
|
|
|
|
|
|
|
c1 = vcii(p,p,n,n);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2008-07-20 20:58:30 +02:00
|
|
|
if (s->inverse) {
|
|
|
|
c2 = vcii(p,p,n,p);
|
|
|
|
} else {
|
|
|
|
c2 = vcii(p,p,p,n);
|
|
|
|
}
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
j = (np >> 2);
|
|
|
|
do {
|
|
|
|
a = vec_ld(0, r);
|
|
|
|
a1 = vec_ld(sizeof(vector float), r);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
b = vec_perm(a,a,vcprmle(1,0,3,2));
|
|
|
|
a = vec_madd(a,c1,b);
|
|
|
|
/* do the pass 0 butterfly */
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
b = vec_perm(a1,a1,vcprmle(1,0,3,2));
|
|
|
|
b = vec_madd(a1,c1,b);
|
|
|
|
/* do the pass 0 butterfly */
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
/* multiply third by -i */
|
|
|
|
b = vec_perm(b,b,vcprmle(2,3,1,0));
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
/* do the pass 1 butterfly */
|
|
|
|
vec_st(vec_madd(b,c2,a), 0, r);
|
|
|
|
vec_st(vec_nmsub(b,c2,a), sizeof(vector float), r);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
r += 2;
|
|
|
|
} while (--j != 0);
|
|
|
|
}
|
|
|
|
/* pass 2 .. ln-1 */
|
|
|
|
|
|
|
|
nblocks = np >> 3;
|
|
|
|
nloops = 1 << 2;
|
|
|
|
np2 = np >> 1;
|
|
|
|
|
|
|
|
cptr1 = s->exptab1;
|
|
|
|
do {
|
|
|
|
p = z;
|
|
|
|
q = z + nloops;
|
|
|
|
j = nblocks;
|
|
|
|
do {
|
|
|
|
cptr = cptr1;
|
|
|
|
k = nloops >> 1;
|
|
|
|
do {
|
|
|
|
vector float a,b,c,t1;
|
|
|
|
|
|
|
|
a = vec_ld(0, (float*)p);
|
|
|
|
b = vec_ld(0, (float*)q);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
/* complex mul */
|
|
|
|
c = vec_ld(0, (float*)cptr);
|
|
|
|
/* cre*re cim*re */
|
|
|
|
t1 = vec_madd(c, vec_perm(b,b,vcprmle(2,2,0,0)),vczero);
|
|
|
|
c = vec_ld(sizeof(vector float), (float*)cptr);
|
|
|
|
/* -cim*im cre*im */
|
|
|
|
b = vec_madd(c, vec_perm(b,b,vcprmle(3,3,1,1)),t1);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
/* butterfly */
|
|
|
|
vec_st(vec_add(a,b), 0, (float*)p);
|
|
|
|
vec_st(vec_sub(a,b), 0, (float*)q);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
p += 2;
|
|
|
|
q += 2;
|
|
|
|
cptr += 4;
|
|
|
|
} while (--k);
|
2005-12-17 19:14:38 +01:00
|
|
|
|
2003-01-07 19:15:48 +01:00
|
|
|
p += nloops;
|
|
|
|
q += nloops;
|
|
|
|
} while (--j);
|
|
|
|
cptr1 += nloops * 2;
|
|
|
|
nblocks = nblocks >> 1;
|
|
|
|
nloops = nloops << 1;
|
|
|
|
} while (nblocks != 0);
|
|
|
|
|
2003-07-09 22:18:13 +02:00
|
|
|
POWERPC_PERF_STOP_COUNT(altivec_fft_num, s->nbits >= 6);
|
2003-01-16 22:54:55 +01:00
|
|
|
}
|