1 /*
2 * Copyright (c) 2007 Luca Barbato <lu_zero@gentoo.org>
3 *
4 * This file is part of FFmpeg.
5 *
6 * FFmpeg 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 * FFmpeg 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 FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
21 #include "config.h"
22
23 #include "libavutil/attributes.h"
24 #include "libavutil/cpu.h"
25 #include "libavutil/ppc/cpu.h"
26 #include "libavutil/ppc/util_altivec.h"
27
28 #include "libavcodec/lossless_audiodsp.h"
29
30 #if HAVE_BIGENDIAN
31 #define GET_T(tt0,tt1,src,a,b){ \
32 a = vec_ld(16, src); \
33 tt0 = vec_perm(b, a, align); \
34 b = vec_ld(32, src); \
35 tt1 = vec_perm(a, b, align); \
36 }
37 #else
38 #define GET_T(tt0,tt1,src,a,b){ \
39 tt0 = vec_vsx_ld(0, src); \
40 tt1 = vec_vsx_ld(16, src); \
41 }
42 #endif
43
44 #if HAVE_ALTIVEC
scalarproduct_and_madd_int16_altivec(int16_t * v1,const int16_t * v2,const int16_t * v3,int order,int mul)45 static int32_t scalarproduct_and_madd_int16_altivec(int16_t *v1,
46 const int16_t *v2,
47 const int16_t *v3,
48 int order, int mul)
49 {
50 LOAD_ZERO;
51 vec_s16 *pv1 = (vec_s16 *) v1;
52 register vec_s16 muls = { mul, mul, mul, mul, mul, mul, mul, mul };
53 register vec_s16 t0, t1, i0, i1, i4, i2, i3;
54 register vec_s32 res = zero_s32v;
55 #if HAVE_BIGENDIAN
56 register vec_u8 align = vec_lvsl(0, v2);
57 i2 = vec_ld(0, v2);
58 i3 = vec_ld(0, v3);
59 #endif
60 int32_t ires;
61
62 order >>= 4;
63 do {
64 GET_T(t0,t1,v2,i1,i2);
65 i0 = pv1[0];
66 i1 = pv1[1];
67 res = vec_msum(t0, i0, res);
68 res = vec_msum(t1, i1, res);
69 GET_T(t0,t1,v3,i4,i3);
70 pv1[0] = vec_mladd(t0, muls, i0);
71 pv1[1] = vec_mladd(t1, muls, i1);
72 pv1 += 2;
73 v2 += 16;
74 v3 += 16;
75 } while (--order);
76 res = vec_splat(vec_sums(res, zero_s32v), 3);
77 vec_ste(res, 0, &ires);
78
79 return ires;
80 }
81 #endif /* HAVE_ALTIVEC */
82
ff_llauddsp_init_ppc(LLAudDSPContext * c)83 av_cold void ff_llauddsp_init_ppc(LLAudDSPContext *c)
84 {
85 #if HAVE_ALTIVEC
86 if (!PPC_ALTIVEC(av_get_cpu_flags()))
87 return;
88
89 c->scalarproduct_and_madd_int16 = scalarproduct_and_madd_int16_altivec;
90 #endif /* HAVE_ALTIVEC */
91 }
92