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
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
|
/*
* Copyright 2011-2013 Blender Foundation
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifndef __SVM_RAMP_H__
#define __SVM_RAMP_H__
CCL_NAMESPACE_BEGIN
/* NOTE: svm_ramp.h, svm_ramp_util.h and node_ramp_util.h must stay consistent */
ccl_device_inline float fetch_float(const KernelGlobals *kg, int offset)
{
uint4 node = kernel_tex_fetch(__svm_nodes, offset);
return __uint_as_float(node.x);
}
ccl_device_inline float float_ramp_lookup(const KernelGlobals *kg,
int offset,
float f,
bool interpolate,
bool extrapolate,
int table_size)
{
if ((f < 0.0f || f > 1.0f) && extrapolate) {
float t0, dy;
if (f < 0.0f) {
t0 = fetch_float(kg, offset);
dy = t0 - fetch_float(kg, offset + 1);
f = -f;
}
else {
t0 = fetch_float(kg, offset + table_size - 1);
dy = t0 - fetch_float(kg, offset + table_size - 2);
f = f - 1.0f;
}
return t0 + dy * f * (table_size - 1);
}
f = saturate(f) * (table_size - 1);
/* clamp int as well in case of NaN */
int i = clamp(float_to_int(f), 0, table_size - 1);
float t = f - (float)i;
float a = fetch_float(kg, offset + i);
if (interpolate && t > 0.0f)
a = (1.0f - t) * a + t * fetch_float(kg, offset + i + 1);
return a;
}
ccl_device_inline float4 rgb_ramp_lookup(const KernelGlobals *kg,
int offset,
float f,
bool interpolate,
bool extrapolate,
int table_size)
{
if ((f < 0.0f || f > 1.0f) && extrapolate) {
float4 t0, dy;
if (f < 0.0f) {
t0 = fetch_node_float(kg, offset);
dy = t0 - fetch_node_float(kg, offset + 1);
f = -f;
}
else {
t0 = fetch_node_float(kg, offset + table_size - 1);
dy = t0 - fetch_node_float(kg, offset + table_size - 2);
f = f - 1.0f;
}
return t0 + dy * f * (table_size - 1);
}
f = saturate(f) * (table_size - 1);
/* clamp int as well in case of NaN */
int i = clamp(float_to_int(f), 0, table_size - 1);
float t = f - (float)i;
float4 a = fetch_node_float(kg, offset + i);
if (interpolate && t > 0.0f)
a = (1.0f - t) * a + t * fetch_node_float(kg, offset + i + 1);
return a;
}
ccl_device_noinline int svm_node_rgb_ramp(
const KernelGlobals *kg, ShaderData *sd, float *stack, uint4 node, int offset)
{
uint fac_offset, color_offset, alpha_offset;
uint interpolate = node.z;
svm_unpack_node_uchar3(node.y, &fac_offset, &color_offset, &alpha_offset);
uint table_size = read_node(kg, &offset).x;
float fac = stack_load_float(stack, fac_offset);
float4 color = rgb_ramp_lookup(kg, offset, fac, interpolate, false, table_size);
if (stack_valid(color_offset))
stack_store_float3(stack, color_offset, float4_to_float3(color));
if (stack_valid(alpha_offset))
stack_store_float(stack, alpha_offset, color.w);
offset += table_size;
return offset;
}
ccl_device_noinline int svm_node_curves(
const KernelGlobals *kg, ShaderData *sd, float *stack, uint4 node, int offset)
{
uint fac_offset, color_offset, out_offset;
svm_unpack_node_uchar3(node.y, &fac_offset, &color_offset, &out_offset);
uint table_size = read_node(kg, &offset).x;
float fac = stack_load_float(stack, fac_offset);
float3 color = stack_load_float3(stack, color_offset);
const float min_x = __int_as_float(node.z), max_x = __int_as_float(node.w);
const float range_x = max_x - min_x;
const float3 relpos = (color - make_float3(min_x, min_x, min_x)) / range_x;
float r = rgb_ramp_lookup(kg, offset, relpos.x, true, true, table_size).x;
float g = rgb_ramp_lookup(kg, offset, relpos.y, true, true, table_size).y;
float b = rgb_ramp_lookup(kg, offset, relpos.z, true, true, table_size).z;
color = (1.0f - fac) * color + fac * make_float3(r, g, b);
stack_store_float3(stack, out_offset, color);
offset += table_size;
return offset;
}
ccl_device_noinline int svm_node_curve(
const KernelGlobals *kg, ShaderData *sd, float *stack, uint4 node, int offset)
{
uint fac_offset, value_in_offset, out_offset;
svm_unpack_node_uchar3(node.y, &fac_offset, &value_in_offset, &out_offset);
uint table_size = read_node(kg, &offset).x;
float fac = stack_load_float(stack, fac_offset);
float in = stack_load_float(stack, value_in_offset);
const float min = __int_as_float(node.z), max = __int_as_float(node.w);
const float range = max - min;
const float relpos = (in - min) / range;
float v = float_ramp_lookup(kg, offset, relpos, true, true, table_size);
in = (1.0f - fac) * in + fac * v;
stack_store_float(stack, out_offset, in);
offset += table_size;
return offset;
}
CCL_NAMESPACE_END
#endif /* __SVM_RAMP_H__ */
|
