1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
|
/*
* Copyright (c) 2021 - 2023 the ThorVG project. All rights reserved.
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
* The above copyright notice and this permission notice shall be included in all
* copies or substantial portions of the Software.
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#ifdef THORVG_NEON_VECTOR_SUPPORT
#include <arm_neon.h>
static inline uint8x8_t ALPHA_BLEND(uint8x8_t c, uint8x8_t a)
{
uint16x8_t t = vmull_u8(c, a);
return vshrn_n_u16(t, 8);
}
static void neonRasterRGBA32(uint32_t *dst, uint32_t val, uint32_t offset, int32_t len)
{
uint32_t iterations = len / 4;
uint32_t neonFilled = iterations * 4;
dst += offset;
uint32x4_t vectorVal = {val, val, val, val};
for (uint32_t i = 0; i < iterations; ++i) {
vst1q_u32(dst, vectorVal);
dst += 4;
}
int32_t leftovers = len - neonFilled;
while (leftovers--) *dst++ = val;
}
static bool neonRasterTranslucentRle(SwSurface* surface, const SwRleData* rle, uint8_t r, uint8_t g, uint8_t b, uint8_t a)
{
if (surface->channelSize != sizeof(uint32_t)) {
TVGERR("SW_ENGINE", "Unsupported Channel Size = %d", surface->channelSize);
return false;
}
auto color = surface->blender.join(r, g, b, a);
auto span = rle->spans;
uint32_t src;
uint8x8_t *vDst = nullptr;
uint16_t align;
for (uint32_t i = 0; i < rle->size; ++i) {
if (span->coverage < 255) src = ALPHA_BLEND(color, span->coverage);
else src = color;
auto dst = &surface->buf32[span->y * surface->stride + span->x];
auto ialpha = 255 - _alpha(src);
if ((((uint32_t) dst) & 0x7) != 0) {
//fill not aligned byte
*dst = src + ALPHA_BLEND(*dst, ialpha);
vDst = (uint8x8_t*)(dst + 1);
align = 1;
} else {
vDst = (uint8x8_t*) dst;
align = 0;
}
uint8x8_t vSrc = (uint8x8_t) vdup_n_u32(src);
uint8x8_t vIalpha = vdup_n_u8((uint8_t) ialpha);
for (uint32_t x = 0; x < (span->len - align) / 2; ++x)
vDst[x] = vadd_u8(vSrc, ALPHA_BLEND(vDst[x], vIalpha));
auto leftovers = (span->len - align) % 2;
if (leftovers > 0) dst[span->len - 1] = src + ALPHA_BLEND(dst[span->len - 1], ialpha);
++span;
}
return true;
}
static bool neonRasterTranslucentRect(SwSurface* surface, const SwBBox& region, uint8_t r, uint8_t g, uint8_t b, uint8_t a)
{
if (surface->channelSize != sizeof(uint32_t)) {
TVGERR("SW_ENGINE", "Unsupported Channel Size = %d", surface->channelSize);
return false;
}
auto color = surface->blender.join(r, g, b, a);
auto buffer = surface->buf32 + (region.min.y * surface->stride) + region.min.x;
auto h = static_cast<uint32_t>(region.max.y - region.min.y);
auto w = static_cast<uint32_t>(region.max.x - region.min.x);
auto ialpha = 255 - _alpha(color);
auto vColor = vdup_n_u32(color);
auto vIalpha = vdup_n_u8((uint8_t) ialpha);
uint8x8_t* vDst = nullptr;
uint32_t align;
for (uint32_t y = 0; y < h; ++y) {
auto dst = &buffer[y * surface->stride];
if ((((uint32_t) dst) & 0x7) != 0) {
//fill not aligned byte
*dst = color + ALPHA_BLEND(*dst, ialpha);
vDst = (uint8x8_t*) (dst + 1);
align = 1;
} else {
vDst = (uint8x8_t*) dst;
align = 0;
}
for (uint32_t x = 0; x < (w - align) / 2; ++x)
vDst[x] = vadd_u8((uint8x8_t)vColor, ALPHA_BLEND(vDst[x], vIalpha));
auto leftovers = (w - align) % 2;
if (leftovers > 0) dst[w - 1] = color + ALPHA_BLEND(dst[w - 1], ialpha);
}
return true;
}
#endif
|