Coverage Report

Created: 2025-11-16 07:20

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/src/ffmpeg/libavcodec/rectangle.h
Line
Count
Source
1
/*
2
 * rectangle filling function
3
 * Copyright (c) 2003 Michael Niedermayer <michaelni@gmx.at>
4
 *
5
 * This file is part of FFmpeg.
6
 *
7
 * FFmpeg is free software; you can redistribute it and/or
8
 * modify it under the terms of the GNU Lesser General Public
9
 * License as published by the Free Software Foundation; either
10
 * version 2.1 of the License, or (at your option) any later version.
11
 *
12
 * FFmpeg is distributed in the hope that it will be useful,
13
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
15
 * Lesser General Public License for more details.
16
 *
17
 * You should have received a copy of the GNU Lesser General Public
18
 * License along with FFmpeg; if not, write to the Free Software
19
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20
 */
21
22
/**
23
 * @file
24
 * useful rectangle filling function
25
 * @author Michael Niedermayer <michaelni@gmx.at>
26
 */
27
28
#ifndef AVCODEC_RECTANGLE_H
29
#define AVCODEC_RECTANGLE_H
30
31
#include "config.h"
32
#include "libavutil/common.h"
33
#include "libavutil/avassert.h"
34
35
/**
36
 * fill a rectangle.
37
 * @param h height of the rectangle, should be a constant
38
 * @param w width of the rectangle, should be a constant
39
 * @param size the size of val (1, 2 or 4), should be a constant
40
 */
41
451M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
451M
    uint8_t *p= (uint8_t*)vp;
43
451M
    av_assert2(size==1 || size==2 || size==4);
44
451M
    av_assert2(w<=4);
45
46
451M
    w      *= size;
47
451M
    stride *= size;
48
49
451M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
451M
    av_assert2((stride&(w-1))==0);
51
451M
    if(w==2){
52
144M
        const uint16_t v= size==4 ? val : val*0x0101;
53
144M
        *(uint16_t*)(p + 0*stride)= v;
54
144M
        if(h==1) return;
55
144M
        *(uint16_t*)(p + 1*stride)= v;
56
144M
        if(h==2) return;
57
3.21M
        *(uint16_t*)(p + 2*stride)= v;
58
3.21M
        *(uint16_t*)(p + 3*stride)= v;
59
307M
    }else if(w==4){
60
146M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
146M
        *(uint32_t*)(p + 0*stride)= v;
62
146M
        if(h==1) return;
63
145M
        *(uint32_t*)(p + 1*stride)= v;
64
145M
        if(h==2) return;
65
131M
        *(uint32_t*)(p + 2*stride)= v;
66
131M
        *(uint32_t*)(p + 3*stride)= v;
67
160M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
71.8M
#if HAVE_FAST_64BIT
70
71.8M
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
71.8M
        *(uint64_t*)(p + 0*stride)= v;
72
71.8M
        if(h==1) return;
73
71.8M
        *(uint64_t*)(p + 1*stride)= v;
74
71.8M
        if(h==2) return;
75
9.46M
        *(uint64_t*)(p + 2*stride)= v;
76
9.46M
        *(uint64_t*)(p + 3*stride)= v;
77
88.8M
    }else if(w==16){
78
88.8M
        const uint64_t v= val*0x0100000001ULL;
79
88.8M
        *(uint64_t*)(p + 0+0*stride)= v;
80
88.8M
        *(uint64_t*)(p + 8+0*stride)= v;
81
88.8M
        *(uint64_t*)(p + 0+1*stride)= v;
82
88.8M
        *(uint64_t*)(p + 8+1*stride)= v;
83
88.8M
        if(h==2) return;
84
85.0M
        *(uint64_t*)(p + 0+2*stride)= v;
85
85.0M
        *(uint64_t*)(p + 8+2*stride)= v;
86
85.0M
        *(uint64_t*)(p + 0+3*stride)= v;
87
85.0M
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
85.0M
    }else
120
0
        av_assert2(0);
121
228M
    av_assert2(h==4);
122
228M
}
h264dec.c:fill_rectangle
Line
Count
Source
41
152M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
152M
    uint8_t *p= (uint8_t*)vp;
43
152M
    av_assert2(size==1 || size==2 || size==4);
44
152M
    av_assert2(w<=4);
45
46
152M
    w      *= size;
47
152M
    stride *= size;
48
49
152M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
152M
    av_assert2((stride&(w-1))==0);
51
152M
    if(w==2){
52
50.7M
        const uint16_t v= size==4 ? val : val*0x0101;
53
50.7M
        *(uint16_t*)(p + 0*stride)= v;
54
50.7M
        if(h==1) return;
55
50.7M
        *(uint16_t*)(p + 1*stride)= v;
56
50.7M
        if(h==2) return;
57
0
        *(uint16_t*)(p + 2*stride)= v;
58
0
        *(uint16_t*)(p + 3*stride)= v;
59
101M
    }else if(w==4){
60
50.7M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
50.7M
        *(uint32_t*)(p + 0*stride)= v;
62
50.7M
        if(h==1) return;
63
50.7M
        *(uint32_t*)(p + 1*stride)= v;
64
50.7M
        if(h==2) return;
65
50.7M
        *(uint32_t*)(p + 2*stride)= v;
66
50.7M
        *(uint32_t*)(p + 3*stride)= v;
67
50.7M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
0
#if HAVE_FAST_64BIT
70
0
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
0
        *(uint64_t*)(p + 0*stride)= v;
72
0
        if(h==1) return;
73
0
        *(uint64_t*)(p + 1*stride)= v;
74
0
        if(h==2) return;
75
0
        *(uint64_t*)(p + 2*stride)= v;
76
0
        *(uint64_t*)(p + 3*stride)= v;
77
50.7M
    }else if(w==16){
78
50.7M
        const uint64_t v= val*0x0100000001ULL;
79
50.7M
        *(uint64_t*)(p + 0+0*stride)= v;
80
50.7M
        *(uint64_t*)(p + 8+0*stride)= v;
81
50.7M
        *(uint64_t*)(p + 0+1*stride)= v;
82
50.7M
        *(uint64_t*)(p + 8+1*stride)= v;
83
50.7M
        if(h==2) return;
84
50.7M
        *(uint64_t*)(p + 0+2*stride)= v;
85
50.7M
        *(uint64_t*)(p + 8+2*stride)= v;
86
50.7M
        *(uint64_t*)(p + 0+3*stride)= v;
87
50.7M
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
50.7M
    }else
120
0
        av_assert2(0);
121
101M
    av_assert2(h==4);
122
101M
}
h264_cavlc.c:fill_rectangle
Line
Count
Source
41
14.4M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
14.4M
    uint8_t *p= (uint8_t*)vp;
43
14.4M
    av_assert2(size==1 || size==2 || size==4);
44
14.4M
    av_assert2(w<=4);
45
46
14.4M
    w      *= size;
47
14.4M
    stride *= size;
48
49
14.4M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
14.4M
    av_assert2((stride&(w-1))==0);
51
14.4M
    if(w==2){
52
910k
        const uint16_t v= size==4 ? val : val*0x0101;
53
910k
        *(uint16_t*)(p + 0*stride)= v;
54
910k
        if(h==1) return;
55
910k
        *(uint16_t*)(p + 1*stride)= v;
56
910k
        if(h==2) return;
57
437k
        *(uint16_t*)(p + 2*stride)= v;
58
437k
        *(uint16_t*)(p + 3*stride)= v;
59
13.4M
    }else if(w==4){
60
8.72M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
8.72M
        *(uint32_t*)(p + 0*stride)= v;
62
8.72M
        if(h==1) return;
63
8.72M
        *(uint32_t*)(p + 1*stride)= v;
64
8.72M
        if(h==2) return;
65
8.30M
        *(uint32_t*)(p + 2*stride)= v;
66
8.30M
        *(uint32_t*)(p + 3*stride)= v;
67
8.30M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
436k
#if HAVE_FAST_64BIT
70
436k
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
436k
        *(uint64_t*)(p + 0*stride)= v;
72
436k
        if(h==1) return;
73
436k
        *(uint64_t*)(p + 1*stride)= v;
74
436k
        if(h==2) return;
75
436k
        *(uint64_t*)(p + 2*stride)= v;
76
436k
        *(uint64_t*)(p + 3*stride)= v;
77
4.33M
    }else if(w==16){
78
4.33M
        const uint64_t v= val*0x0100000001ULL;
79
4.33M
        *(uint64_t*)(p + 0+0*stride)= v;
80
4.33M
        *(uint64_t*)(p + 8+0*stride)= v;
81
4.33M
        *(uint64_t*)(p + 0+1*stride)= v;
82
4.33M
        *(uint64_t*)(p + 8+1*stride)= v;
83
4.33M
        if(h==2) return;
84
3.91M
        *(uint64_t*)(p + 0+2*stride)= v;
85
3.91M
        *(uint64_t*)(p + 8+2*stride)= v;
86
3.91M
        *(uint64_t*)(p + 0+3*stride)= v;
87
3.91M
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
3.91M
    }else
120
0
        av_assert2(0);
121
13.0M
    av_assert2(h==4);
122
13.0M
}
h264_direct.c:fill_rectangle
Line
Count
Source
41
130M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
130M
    uint8_t *p= (uint8_t*)vp;
43
130M
    av_assert2(size==1 || size==2 || size==4);
44
130M
    av_assert2(w<=4);
45
46
130M
    w      *= size;
47
130M
    stride *= size;
48
49
130M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
130M
    av_assert2((stride&(w-1))==0);
51
130M
    if(w==2){
52
49.5M
        const uint16_t v= size==4 ? val : val*0x0101;
53
49.5M
        *(uint16_t*)(p + 0*stride)= v;
54
49.5M
        if(h==1) return;
55
49.5M
        *(uint16_t*)(p + 1*stride)= v;
56
49.5M
        if(h==2) return;
57
0
        *(uint16_t*)(p + 2*stride)= v;
58
0
        *(uint16_t*)(p + 3*stride)= v;
59
81.3M
    }else if(w==4){
60
16.3M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
16.3M
        *(uint32_t*)(p + 0*stride)= v;
62
16.3M
        if(h==1) return;
63
16.3M
        *(uint32_t*)(p + 1*stride)= v;
64
16.3M
        if(h==2) return;
65
16.3M
        *(uint32_t*)(p + 2*stride)= v;
66
16.3M
        *(uint32_t*)(p + 3*stride)= v;
67
64.9M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
48.5M
#if HAVE_FAST_64BIT
70
48.5M
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
48.5M
        *(uint64_t*)(p + 0*stride)= v;
72
48.5M
        if(h==1) return;
73
48.5M
        *(uint64_t*)(p + 1*stride)= v;
74
48.5M
        if(h==2) return;
75
0
        *(uint64_t*)(p + 2*stride)= v;
76
0
        *(uint64_t*)(p + 3*stride)= v;
77
16.3M
    }else if(w==16){
78
16.3M
        const uint64_t v= val*0x0100000001ULL;
79
16.3M
        *(uint64_t*)(p + 0+0*stride)= v;
80
16.3M
        *(uint64_t*)(p + 8+0*stride)= v;
81
16.3M
        *(uint64_t*)(p + 0+1*stride)= v;
82
16.3M
        *(uint64_t*)(p + 8+1*stride)= v;
83
16.3M
        if(h==2) return;
84
16.3M
        *(uint64_t*)(p + 0+2*stride)= v;
85
16.3M
        *(uint64_t*)(p + 8+2*stride)= v;
86
16.3M
        *(uint64_t*)(p + 0+3*stride)= v;
87
16.3M
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
16.3M
    }else
120
0
        av_assert2(0);
121
32.7M
    av_assert2(h==4);
122
32.7M
}
h264_mb.c:fill_rectangle
Line
Count
Source
41
11.2M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
11.2M
    uint8_t *p= (uint8_t*)vp;
43
11.2M
    av_assert2(size==1 || size==2 || size==4);
44
11.2M
    av_assert2(w<=4);
45
46
11.2M
    w      *= size;
47
11.2M
    stride *= size;
48
49
11.2M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
11.2M
    av_assert2((stride&(w-1))==0);
51
11.2M
    if(w==2){
52
6.11M
        const uint16_t v= size==4 ? val : val*0x0101;
53
6.11M
        *(uint16_t*)(p + 0*stride)= v;
54
6.11M
        if(h==1) return;
55
6.11M
        *(uint16_t*)(p + 1*stride)= v;
56
6.11M
        if(h==2) return;
57
0
        *(uint16_t*)(p + 2*stride)= v;
58
0
        *(uint16_t*)(p + 3*stride)= v;
59
5.14M
    }else if(w==4){
60
5.14M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
5.14M
        *(uint32_t*)(p + 0*stride)= v;
62
5.14M
        if(h==1) return;
63
5.14M
        *(uint32_t*)(p + 1*stride)= v;
64
5.14M
        if(h==2) return;
65
5.14M
        *(uint32_t*)(p + 2*stride)= v;
66
5.14M
        *(uint32_t*)(p + 3*stride)= v;
67
5.14M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
0
#if HAVE_FAST_64BIT
70
0
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
0
        *(uint64_t*)(p + 0*stride)= v;
72
0
        if(h==1) return;
73
0
        *(uint64_t*)(p + 1*stride)= v;
74
0
        if(h==2) return;
75
0
        *(uint64_t*)(p + 2*stride)= v;
76
0
        *(uint64_t*)(p + 3*stride)= v;
77
0
    }else if(w==16){
78
0
        const uint64_t v= val*0x0100000001ULL;
79
0
        *(uint64_t*)(p + 0+0*stride)= v;
80
0
        *(uint64_t*)(p + 8+0*stride)= v;
81
0
        *(uint64_t*)(p + 0+1*stride)= v;
82
0
        *(uint64_t*)(p + 8+1*stride)= v;
83
0
        if(h==2) return;
84
0
        *(uint64_t*)(p + 0+2*stride)= v;
85
0
        *(uint64_t*)(p + 8+2*stride)= v;
86
0
        *(uint64_t*)(p + 0+3*stride)= v;
87
0
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
0
    }else
120
0
        av_assert2(0);
121
5.14M
    av_assert2(h==4);
122
5.14M
}
h264_slice.c:fill_rectangle
Line
Count
Source
41
4.95M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
4.95M
    uint8_t *p= (uint8_t*)vp;
43
4.95M
    av_assert2(size==1 || size==2 || size==4);
44
4.95M
    av_assert2(w<=4);
45
46
4.95M
    w      *= size;
47
4.95M
    stride *= size;
48
49
4.95M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
4.95M
    av_assert2((stride&(w-1))==0);
51
4.95M
    if(w==2){
52
0
        const uint16_t v= size==4 ? val : val*0x0101;
53
0
        *(uint16_t*)(p + 0*stride)= v;
54
0
        if(h==1) return;
55
0
        *(uint16_t*)(p + 1*stride)= v;
56
0
        if(h==2) return;
57
0
        *(uint16_t*)(p + 2*stride)= v;
58
0
        *(uint16_t*)(p + 3*stride)= v;
59
4.95M
    }else if(w==4){
60
0
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
0
        *(uint32_t*)(p + 0*stride)= v;
62
0
        if(h==1) return;
63
0
        *(uint32_t*)(p + 1*stride)= v;
64
0
        if(h==2) return;
65
0
        *(uint32_t*)(p + 2*stride)= v;
66
0
        *(uint32_t*)(p + 3*stride)= v;
67
4.95M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
0
#if HAVE_FAST_64BIT
70
0
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
0
        *(uint64_t*)(p + 0*stride)= v;
72
0
        if(h==1) return;
73
0
        *(uint64_t*)(p + 1*stride)= v;
74
0
        if(h==2) return;
75
0
        *(uint64_t*)(p + 2*stride)= v;
76
0
        *(uint64_t*)(p + 3*stride)= v;
77
4.95M
    }else if(w==16){
78
4.95M
        const uint64_t v= val*0x0100000001ULL;
79
4.95M
        *(uint64_t*)(p + 0+0*stride)= v;
80
4.95M
        *(uint64_t*)(p + 8+0*stride)= v;
81
4.95M
        *(uint64_t*)(p + 0+1*stride)= v;
82
4.95M
        *(uint64_t*)(p + 8+1*stride)= v;
83
4.95M
        if(h==2) return;
84
4.95M
        *(uint64_t*)(p + 0+2*stride)= v;
85
4.95M
        *(uint64_t*)(p + 8+2*stride)= v;
86
4.95M
        *(uint64_t*)(p + 0+3*stride)= v;
87
4.95M
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
4.95M
    }else
120
0
        av_assert2(0);
121
4.95M
    av_assert2(h==4);
122
4.95M
}
h264_cabac.c:fill_rectangle
Line
Count
Source
41
113M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
113M
    uint8_t *p= (uint8_t*)vp;
43
113M
    av_assert2(size==1 || size==2 || size==4);
44
113M
    av_assert2(w<=4);
45
46
113M
    w      *= size;
47
113M
    stride *= size;
48
49
113M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
113M
    av_assert2((stride&(w-1))==0);
51
113M
    if(w==2){
52
36.9M
        const uint16_t v= size==4 ? val : val*0x0101;
53
36.9M
        *(uint16_t*)(p + 0*stride)= v;
54
36.9M
        if(h==1) return;
55
36.9M
        *(uint16_t*)(p + 1*stride)= v;
56
36.9M
        if(h==2) return;
57
2.77M
        *(uint16_t*)(p + 2*stride)= v;
58
2.77M
        *(uint16_t*)(p + 3*stride)= v;
59
76.4M
    }else if(w==4){
60
50.9M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
50.9M
        *(uint32_t*)(p + 0*stride)= v;
62
50.9M
        if(h==1) return;
63
50.9M
        *(uint32_t*)(p + 1*stride)= v;
64
50.9M
        if(h==2) return;
65
45.5M
        *(uint32_t*)(p + 2*stride)= v;
66
45.5M
        *(uint32_t*)(p + 3*stride)= v;
67
45.5M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
13.2M
#if HAVE_FAST_64BIT
70
13.2M
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
13.2M
        *(uint64_t*)(p + 0*stride)= v;
72
13.2M
        if(h==1) return;
73
13.2M
        *(uint64_t*)(p + 1*stride)= v;
74
13.2M
        if(h==2) return;
75
8.50M
        *(uint64_t*)(p + 2*stride)= v;
76
8.50M
        *(uint64_t*)(p + 3*stride)= v;
77
12.2M
    }else if(w==16){
78
12.2M
        const uint64_t v= val*0x0100000001ULL;
79
12.2M
        *(uint64_t*)(p + 0+0*stride)= v;
80
12.2M
        *(uint64_t*)(p + 8+0*stride)= v;
81
12.2M
        *(uint64_t*)(p + 0+1*stride)= v;
82
12.2M
        *(uint64_t*)(p + 8+1*stride)= v;
83
12.2M
        if(h==2) return;
84
8.95M
        *(uint64_t*)(p + 0+2*stride)= v;
85
8.95M
        *(uint64_t*)(p + 8+2*stride)= v;
86
8.95M
        *(uint64_t*)(p + 0+3*stride)= v;
87
8.95M
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
8.95M
    }else
120
0
        av_assert2(0);
121
65.7M
    av_assert2(h==4);
122
65.7M
}
rv34.c:fill_rectangle
Line
Count
Source
41
22.0M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
22.0M
    uint8_t *p= (uint8_t*)vp;
43
22.0M
    av_assert2(size==1 || size==2 || size==4);
44
22.0M
    av_assert2(w<=4);
45
46
22.0M
    w      *= size;
47
22.0M
    stride *= size;
48
49
22.0M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
22.0M
    av_assert2((stride&(w-1))==0);
51
22.0M
    if(w==2){
52
0
        const uint16_t v= size==4 ? val : val*0x0101;
53
0
        *(uint16_t*)(p + 0*stride)= v;
54
0
        if(h==1) return;
55
0
        *(uint16_t*)(p + 1*stride)= v;
56
0
        if(h==2) return;
57
0
        *(uint16_t*)(p + 2*stride)= v;
58
0
        *(uint16_t*)(p + 3*stride)= v;
59
22.0M
    }else if(w==4){
60
12.9M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
12.9M
        *(uint32_t*)(p + 0*stride)= v;
62
12.9M
        if(h==1) return;
63
12.9M
        *(uint32_t*)(p + 1*stride)= v;
64
12.9M
        if(h==2) return;
65
4.98M
        *(uint32_t*)(p + 2*stride)= v;
66
4.98M
        *(uint32_t*)(p + 3*stride)= v;
67
9.03M
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
9.03M
#if HAVE_FAST_64BIT
70
9.03M
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
9.03M
        *(uint64_t*)(p + 0*stride)= v;
72
9.03M
        if(h==1) return;
73
9.03M
        *(uint64_t*)(p + 1*stride)= v;
74
9.03M
        if(h==2) return;
75
0
        *(uint64_t*)(p + 2*stride)= v;
76
0
        *(uint64_t*)(p + 3*stride)= v;
77
0
    }else if(w==16){
78
0
        const uint64_t v= val*0x0100000001ULL;
79
0
        *(uint64_t*)(p + 0+0*stride)= v;
80
0
        *(uint64_t*)(p + 8+0*stride)= v;
81
0
        *(uint64_t*)(p + 0+1*stride)= v;
82
0
        *(uint64_t*)(p + 8+1*stride)= v;
83
0
        if(h==2) return;
84
0
        *(uint64_t*)(p + 0+2*stride)= v;
85
0
        *(uint64_t*)(p + 8+2*stride)= v;
86
0
        *(uint64_t*)(p + 0+3*stride)= v;
87
0
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
0
    }else
120
0
        av_assert2(0);
121
4.98M
    av_assert2(h==4);
122
4.98M
}
svq3.c:fill_rectangle
Line
Count
Source
41
2.08M
static av_always_inline void fill_rectangle(void *vp, int w, int h, int stride, uint32_t val, int size){
42
2.08M
    uint8_t *p= (uint8_t*)vp;
43
2.08M
    av_assert2(size==1 || size==2 || size==4);
44
2.08M
    av_assert2(w<=4);
45
46
2.08M
    w      *= size;
47
2.08M
    stride *= size;
48
49
2.08M
    av_assert2((((uintptr_t)vp)&(FFMIN(w, 8<<(HAVE_NEON|ARCH_PPC|HAVE_MMX))-1)) == 0);
50
2.08M
    av_assert2((stride&(w-1))==0);
51
2.08M
    if(w==2){
52
0
        const uint16_t v= size==4 ? val : val*0x0101;
53
0
        *(uint16_t*)(p + 0*stride)= v;
54
0
        if(h==1) return;
55
0
        *(uint16_t*)(p + 1*stride)= v;
56
0
        if(h==2) return;
57
0
        *(uint16_t*)(p + 2*stride)= v;
58
0
        *(uint16_t*)(p + 3*stride)= v;
59
2.08M
    }else if(w==4){
60
1.37M
        const uint32_t v= size==4 ? val : size==2 ? val*0x00010001 : val*0x01010101;
61
1.37M
        *(uint32_t*)(p + 0*stride)= v;
62
1.37M
        if(h==1) return;
63
74.1k
        *(uint32_t*)(p + 1*stride)= v;
64
74.1k
        if(h==2) return;
65
0
        *(uint32_t*)(p + 2*stride)= v;
66
0
        *(uint32_t*)(p + 3*stride)= v;
67
713k
    }else if(w==8){
68
    // gcc cannot optimize 64-bit math on x86_32
69
557k
#if HAVE_FAST_64BIT
70
557k
        const uint64_t v=  size==2 ? val*0x0001000100010001ULL : val*0x0100000001ULL;
71
557k
        *(uint64_t*)(p + 0*stride)= v;
72
557k
        if(h==1) return;
73
532k
        *(uint64_t*)(p + 1*stride)= v;
74
532k
        if(h==2) return;
75
522k
        *(uint64_t*)(p + 2*stride)= v;
76
522k
        *(uint64_t*)(p + 3*stride)= v;
77
522k
    }else if(w==16){
78
156k
        const uint64_t v= val*0x0100000001ULL;
79
156k
        *(uint64_t*)(p + 0+0*stride)= v;
80
156k
        *(uint64_t*)(p + 8+0*stride)= v;
81
156k
        *(uint64_t*)(p + 0+1*stride)= v;
82
156k
        *(uint64_t*)(p + 8+1*stride)= v;
83
156k
        if(h==2) return;
84
149k
        *(uint64_t*)(p + 0+2*stride)= v;
85
149k
        *(uint64_t*)(p + 8+2*stride)= v;
86
149k
        *(uint64_t*)(p + 0+3*stride)= v;
87
149k
        *(uint64_t*)(p + 8+3*stride)= v;
88
#else
89
        const uint32_t v= size==2 ? val*0x00010001 : val;
90
        *(uint32_t*)(p + 0+0*stride)= v;
91
        *(uint32_t*)(p + 4+0*stride)= v;
92
        if(h==1) return;
93
        *(uint32_t*)(p + 0+1*stride)= v;
94
        *(uint32_t*)(p + 4+1*stride)= v;
95
        if(h==2) return;
96
        *(uint32_t*)(p + 0+2*stride)= v;
97
        *(uint32_t*)(p + 4+2*stride)= v;
98
        *(uint32_t*)(p + 0+3*stride)= v;
99
        *(uint32_t*)(p + 4+3*stride)= v;
100
    }else if(w==16){
101
        *(uint32_t*)(p + 0+0*stride)= val;
102
        *(uint32_t*)(p + 4+0*stride)= val;
103
        *(uint32_t*)(p + 8+0*stride)= val;
104
        *(uint32_t*)(p +12+0*stride)= val;
105
        *(uint32_t*)(p + 0+1*stride)= val;
106
        *(uint32_t*)(p + 4+1*stride)= val;
107
        *(uint32_t*)(p + 8+1*stride)= val;
108
        *(uint32_t*)(p +12+1*stride)= val;
109
        if(h==2) return;
110
        *(uint32_t*)(p + 0+2*stride)= val;
111
        *(uint32_t*)(p + 4+2*stride)= val;
112
        *(uint32_t*)(p + 8+2*stride)= val;
113
        *(uint32_t*)(p +12+2*stride)= val;
114
        *(uint32_t*)(p + 0+3*stride)= val;
115
        *(uint32_t*)(p + 4+3*stride)= val;
116
        *(uint32_t*)(p + 8+3*stride)= val;
117
        *(uint32_t*)(p +12+3*stride)= val;
118
#endif
119
149k
    }else
120
0
        av_assert2(0);
121
671k
    av_assert2(h==4);
122
671k
}
123
124
#endif /* AVCODEC_RECTANGLE_H */