/src/ffmpeg/libswscale/ops_backend.c
Line | Count | Source |
1 | | /** |
2 | | * Copyright (C) 2025 Niklas Haas |
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 "ops_backend.h" |
22 | | |
23 | | #if AV_GCC_VERSION_AT_LEAST(4, 4) |
24 | | #pragma GCC optimize ("finite-math-only") |
25 | | #endif |
26 | | |
27 | | /* Array-based reference implementation */ |
28 | | |
29 | | #ifndef SWS_BLOCK_SIZE |
30 | 0 | # define SWS_BLOCK_SIZE 32 |
31 | | #endif |
32 | | |
33 | | typedef uint8_t u8block_t[SWS_BLOCK_SIZE]; |
34 | | typedef uint16_t u16block_t[SWS_BLOCK_SIZE]; |
35 | | typedef uint32_t u32block_t[SWS_BLOCK_SIZE]; |
36 | | typedef float f32block_t[SWS_BLOCK_SIZE]; |
37 | | |
38 | | #define BIT_DEPTH 8 |
39 | | # include "ops_tmpl_int.c" |
40 | | #undef BIT_DEPTH |
41 | | |
42 | | #define BIT_DEPTH 16 |
43 | | # include "ops_tmpl_int.c" |
44 | | #undef BIT_DEPTH |
45 | | |
46 | | #define BIT_DEPTH 32 |
47 | | # include "ops_tmpl_int.c" |
48 | | # include "ops_tmpl_float.c" |
49 | | #undef BIT_DEPTH |
50 | | |
51 | | static void process(const SwsOpExec *exec, const void *priv, |
52 | | const int bx_start, const int y_start, int bx_end, int y_end) |
53 | 0 | { |
54 | 0 | const SwsOpChain *chain = priv; |
55 | 0 | const SwsOpImpl *impl = chain->impl; |
56 | 0 | SwsOpIter iter; |
57 | |
|
58 | 0 | for (iter.y = y_start; iter.y < y_end; iter.y++) { |
59 | 0 | for (int i = 0; i < 4; i++) { |
60 | 0 | iter.in[i] = exec->in[i] + (iter.y - y_start) * exec->in_stride[i]; |
61 | 0 | iter.out[i] = exec->out[i] + (iter.y - y_start) * exec->out_stride[i]; |
62 | 0 | } |
63 | |
|
64 | 0 | for (int block = bx_start; block < bx_end; block++) { |
65 | 0 | iter.x = block * SWS_BLOCK_SIZE; |
66 | 0 | ((void (*)(SwsOpIter *, const SwsOpImpl *)) impl->cont) |
67 | 0 | (&iter, &impl[1]); |
68 | 0 | } |
69 | 0 | } |
70 | 0 | } |
71 | | |
72 | | static int compile(SwsContext *ctx, SwsOpList *ops, SwsCompiledOp *out) |
73 | 0 | { |
74 | 0 | int ret; |
75 | |
|
76 | 0 | SwsOpChain *chain = ff_sws_op_chain_alloc(); |
77 | 0 | if (!chain) |
78 | 0 | return AVERROR(ENOMEM); |
79 | | |
80 | 0 | static const SwsOpTable *const tables[] = { |
81 | 0 | &bitfn(op_table_int, u8), |
82 | 0 | &bitfn(op_table_int, u16), |
83 | 0 | &bitfn(op_table_int, u32), |
84 | 0 | &bitfn(op_table_float, f32), |
85 | 0 | }; |
86 | |
|
87 | 0 | do { |
88 | 0 | ret = ff_sws_op_compile_tables(tables, FF_ARRAY_ELEMS(tables), ops, |
89 | 0 | SWS_BLOCK_SIZE, chain); |
90 | 0 | } while (ret == AVERROR(EAGAIN)); |
91 | 0 | if (ret < 0) { |
92 | 0 | ff_sws_op_chain_free(chain); |
93 | 0 | return ret; |
94 | 0 | } |
95 | | |
96 | 0 | *out = (SwsCompiledOp) { |
97 | 0 | .func = process, |
98 | 0 | .block_size = SWS_BLOCK_SIZE, |
99 | 0 | .cpu_flags = chain->cpu_flags, |
100 | 0 | .priv = chain, |
101 | 0 | .free = ff_sws_op_chain_free_cb, |
102 | 0 | }; |
103 | 0 | return 0; |
104 | 0 | } |
105 | | |
106 | | const SwsOpBackend backend_c = { |
107 | | .name = "c", |
108 | | .compile = compile, |
109 | | }; |