1 /*
2 * Copyright (C) 2012 Michael Niedermayer (michaelni@gmx.at)
3 *
4 * This file is part of libswresample
5 *
6 * libswresample 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 * libswresample 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 libswresample; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
23
24 #define PROTO(pre, in, out, cap) void ff ## pre ## _ ##in## _to_ ##out## _a_ ##cap(uint8_t **dst, const uint8_t **src, int len);
25 #define PROTO2(pre, out, cap) PROTO(pre, int16, out, cap) PROTO(pre, int32, out, cap) PROTO(pre, float, out, cap)
26 #define PROTO3(pre, cap) PROTO2(pre, int16, cap) PROTO2(pre, int32, cap) PROTO2(pre, float, cap)
27 #define PROTO4(pre) PROTO3(pre, mmx) PROTO3(pre, sse) PROTO3(pre, sse2) PROTO3(pre, ssse3) PROTO3(pre, sse4) PROTO3(pre, avx)
32
36 int channels){
38
39 ac->simd_f= NULL;
40
41 //FIXME add memcpy case
42
43 #define MULTI_CAPS_FUNC(flag, cap) \
44 if (mm_flags & flag) {\
45 if( out_fmt == AV_SAMPLE_FMT_S32 && in_fmt == AV_SAMPLE_FMT_S16 || out_fmt == AV_SAMPLE_FMT_S32P && in_fmt == AV_SAMPLE_FMT_S16P)\
46 ac->simd_f = ff_int16_to_int32_a_ ## cap;\
47 if( out_fmt == AV_SAMPLE_FMT_S16 && in_fmt == AV_SAMPLE_FMT_S32 || out_fmt == AV_SAMPLE_FMT_S16P && in_fmt == AV_SAMPLE_FMT_S32P)\
48 ac->simd_f = ff_int32_to_int16_a_ ## cap;\
49 }
50
53
55 if(channels == 6) {
57 ac->simd_f = ff_pack_6ch_float_to_float_a_mmx;
58 }
59 }
60
63 ac->simd_f = ff_int32_to_float_a_sse2;
65 ac->simd_f = ff_int16_to_float_a_sse2;
67 ac->simd_f = ff_float_to_int32_a_sse2;
69 ac->simd_f = ff_float_to_int16_a_sse2;
70
71 if(channels == 2) {
73 ac->simd_f = ff_pack_2ch_int32_to_int32_a_sse2;
75 ac->simd_f = ff_pack_2ch_int16_to_int16_a_sse2;
77 ac->simd_f = ff_pack_2ch_int16_to_int32_a_sse2;
79 ac->simd_f = ff_pack_2ch_int32_to_int16_a_sse2;
80
82 ac->simd_f = ff_unpack_2ch_int32_to_int32_a_sse2;
84 ac->simd_f = ff_unpack_2ch_int16_to_int16_a_sse2;
86 ac->simd_f = ff_unpack_2ch_int16_to_int32_a_sse2;
88 ac->simd_f = ff_unpack_2ch_int32_to_int16_a_sse2;
89
91 ac->simd_f = ff_pack_2ch_int32_to_float_a_sse2;
93 ac->simd_f = ff_pack_2ch_float_to_int32_a_sse2;
95 ac->simd_f = ff_pack_2ch_int16_to_float_a_sse2;
97 ac->simd_f = ff_pack_2ch_float_to_int16_a_sse2;
99 ac->simd_f = ff_unpack_2ch_int32_to_float_a_sse2;
101 ac->simd_f = ff_unpack_2ch_float_to_int32_a_sse2;
103 ac->simd_f = ff_unpack_2ch_int16_to_float_a_sse2;
105 ac->simd_f = ff_unpack_2ch_float_to_int16_a_sse2;
106 }
107 }
109 if(channels == 2) {
111 ac->simd_f = ff_unpack_2ch_int16_to_int16_a_ssse3;
113 ac->simd_f = ff_unpack_2ch_int16_to_int32_a_ssse3;
115 ac->simd_f = ff_unpack_2ch_int16_to_float_a_ssse3;
116 }
117 }
119 if(channels == 6) {
121 ac->simd_f = ff_pack_6ch_float_to_float_a_sse4;
123 ac->simd_f = ff_pack_6ch_int32_to_float_a_sse4;
125 ac->simd_f = ff_pack_6ch_float_to_int32_a_sse4;
126 }
127 }
130 ac->simd_f = ff_int32_to_float_a_avx;
131 if(channels == 6) {
133 ac->simd_f = ff_pack_6ch_float_to_float_a_avx;
135 ac->simd_f = ff_pack_6ch_int32_to_float_a_avx;
137 ac->simd_f = ff_pack_6ch_float_to_int32_a_avx;
138 }
139 }
140 }
141
142 #define D(type, simd) \
143 mix_1_1_func_type ff_mix_1_1_a_## type ## _ ## simd;\
144 mix_2_1_func_type ff_mix_2_1_a_## type ## _ ## simd;
145
150
151
156 int num = nb_in * nb_out;
157 int i,j;
158
159 s->mix_1_1_simd = NULL;
160 s->mix_2_1_simd = NULL;
161
164 s->mix_1_1_simd = ff_mix_1_1_a_int16_mmx;
165 s->mix_2_1_simd = ff_mix_2_1_a_int16_mmx;
166 }
168 s->mix_1_1_simd = ff_mix_1_1_a_int16_sse2;
169 s->mix_2_1_simd = ff_mix_2_1_a_int16_sse2;
170 }
171 s->native_simd_matrix =
av_mallocz(2 * num *
sizeof(int16_t));
172 s->native_simd_one =
av_mallocz(2 *
sizeof(int16_t));
173 for(i=0; i<nb_out; i++){
174 int sh = 0;
175 for(j=0; j<nb_in; j++)
176 sh =
FFMAX(sh,
FFABS(((
int*)s->native_matrix)[i * nb_in + j]));
178 for(j=0; j<nb_in; j++) {
179 ((int16_t*)s->native_simd_matrix)[2*(i * nb_in + j)+1] = 15 - sh;
180 ((int16_t*)s->native_simd_matrix)[2*(i * nb_in + j)] =
181 ((((int*)s->native_matrix)[i * nb_in + j]) + (1<<sh>>1)) >> sh;
182 }
183 }
184 ((int16_t*)s->native_simd_one)[1] = 14;
185 ((int16_t*)s->native_simd_one)[0] = 16384;
188 s->mix_1_1_simd = ff_mix_1_1_a_float_sse;
189 s->mix_2_1_simd = ff_mix_2_1_a_float_sse;
190 }
192 s->mix_1_1_simd = ff_mix_1_1_a_float_avx;
193 s->mix_2_1_simd = ff_mix_2_1_a_float_avx;
194 }
195 s->native_simd_matrix =
av_mallocz(num *
sizeof(
float));
196 memcpy(s->native_simd_matrix, s->native_matrix, num * sizeof(float));
197 s->native_simd_one =
av_mallocz(
sizeof(
float));
198 memcpy(s->native_simd_one, s->native_one, sizeof(float));
199 }
200 }