yading@10
|
1 /*
|
yading@10
|
2 * ARM NEON optimised DSP functions
|
yading@10
|
3 * Copyright (c) 2008 Mans Rullgard <mans@mansr.com>
|
yading@10
|
4 *
|
yading@10
|
5 * This file is part of FFmpeg.
|
yading@10
|
6 *
|
yading@10
|
7 * FFmpeg is free software; you can redistribute it and/or
|
yading@10
|
8 * modify it under the terms of the GNU Lesser General Public
|
yading@10
|
9 * License as published by the Free Software Foundation; either
|
yading@10
|
10 * version 2.1 of the License, or (at your option) any later version.
|
yading@10
|
11 *
|
yading@10
|
12 * FFmpeg is distributed in the hope that it will be useful,
|
yading@10
|
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
yading@10
|
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
yading@10
|
15 * Lesser General Public License for more details.
|
yading@10
|
16 *
|
yading@10
|
17 * You should have received a copy of the GNU Lesser General Public
|
yading@10
|
18 * License along with FFmpeg; if not, write to the Free Software
|
yading@10
|
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
yading@10
|
20 */
|
yading@10
|
21
|
yading@10
|
22 #include <stdint.h>
|
yading@10
|
23
|
yading@10
|
24 #include "libavutil/attributes.h"
|
yading@10
|
25 #include "libavcodec/avcodec.h"
|
yading@10
|
26 #include "dsputil_arm.h"
|
yading@10
|
27
|
yading@10
|
28 void ff_simple_idct_neon(int16_t *data);
|
yading@10
|
29 void ff_simple_idct_put_neon(uint8_t *dest, int line_size, int16_t *data);
|
yading@10
|
30 void ff_simple_idct_add_neon(uint8_t *dest, int line_size, int16_t *data);
|
yading@10
|
31
|
yading@10
|
32 void ff_clear_block_neon(int16_t *block);
|
yading@10
|
33 void ff_clear_blocks_neon(int16_t *blocks);
|
yading@10
|
34
|
yading@10
|
35 void ff_add_pixels_clamped_neon(const int16_t *, uint8_t *, int);
|
yading@10
|
36 void ff_put_pixels_clamped_neon(const int16_t *, uint8_t *, int);
|
yading@10
|
37 void ff_put_signed_pixels_clamped_neon(const int16_t *, uint8_t *, int);
|
yading@10
|
38
|
yading@10
|
39 void ff_vector_clipf_neon(float *dst, const float *src, float min, float max,
|
yading@10
|
40 int len);
|
yading@10
|
41 void ff_vector_clip_int32_neon(int32_t *dst, const int32_t *src, int32_t min,
|
yading@10
|
42 int32_t max, unsigned int len);
|
yading@10
|
43
|
yading@10
|
44 int32_t ff_scalarproduct_int16_neon(const int16_t *v1, const int16_t *v2, int len);
|
yading@10
|
45 int32_t ff_scalarproduct_and_madd_int16_neon(int16_t *v1, const int16_t *v2,
|
yading@10
|
46 const int16_t *v3, int len, int mul);
|
yading@10
|
47
|
yading@10
|
48 void ff_apply_window_int16_neon(int16_t *dst, const int16_t *src,
|
yading@10
|
49 const int16_t *window, unsigned n);
|
yading@10
|
50
|
yading@10
|
51 av_cold void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
|
yading@10
|
52 {
|
yading@10
|
53 const int high_bit_depth = avctx->bits_per_raw_sample > 8;
|
yading@10
|
54
|
yading@10
|
55 if (!avctx->lowres && avctx->bits_per_raw_sample <= 8) {
|
yading@10
|
56 if (avctx->idct_algo == FF_IDCT_AUTO ||
|
yading@10
|
57 avctx->idct_algo == FF_IDCT_SIMPLENEON) {
|
yading@10
|
58 c->idct_put = ff_simple_idct_put_neon;
|
yading@10
|
59 c->idct_add = ff_simple_idct_add_neon;
|
yading@10
|
60 c->idct = ff_simple_idct_neon;
|
yading@10
|
61 c->idct_permutation_type = FF_PARTTRANS_IDCT_PERM;
|
yading@10
|
62 }
|
yading@10
|
63 }
|
yading@10
|
64
|
yading@10
|
65 if (!high_bit_depth) {
|
yading@10
|
66 c->clear_block = ff_clear_block_neon;
|
yading@10
|
67 c->clear_blocks = ff_clear_blocks_neon;
|
yading@10
|
68 }
|
yading@10
|
69
|
yading@10
|
70 c->add_pixels_clamped = ff_add_pixels_clamped_neon;
|
yading@10
|
71 c->put_pixels_clamped = ff_put_pixels_clamped_neon;
|
yading@10
|
72 c->put_signed_pixels_clamped = ff_put_signed_pixels_clamped_neon;
|
yading@10
|
73
|
yading@10
|
74 c->vector_clipf = ff_vector_clipf_neon;
|
yading@10
|
75 c->vector_clip_int32 = ff_vector_clip_int32_neon;
|
yading@10
|
76
|
yading@10
|
77 c->scalarproduct_int16 = ff_scalarproduct_int16_neon;
|
yading@10
|
78 c->scalarproduct_and_madd_int16 = ff_scalarproduct_and_madd_int16_neon;
|
yading@10
|
79
|
yading@10
|
80 c->apply_window_int16 = ff_apply_window_int16_neon;
|
yading@10
|
81 }
|