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
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
|
/*
* Copyright 2011-2016 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.
*/
#include "util/util_debug.h"
#include <stdlib.h>
#include "util/util_logging.h"
#include "util/util_string.h"
CCL_NAMESPACE_BEGIN
DebugFlags::CPU::CPU()
: avx2(true),
avx(true),
sse41(true),
sse3(true),
sse2(true),
qbvh(true),
split_kernel(false)
{
reset();
}
void DebugFlags::CPU::reset()
{
#define STRINGIFY(x) #x
#define CHECK_CPU_FLAGS(flag, env) \
do { \
flag = (getenv(env) == NULL); \
if(!flag) { \
VLOG(1) << "Disabling " << STRINGIFY(flag) << " instruction set."; \
} \
} while(0)
CHECK_CPU_FLAGS(avx2, "CYCLES_CPU_NO_AVX2");
CHECK_CPU_FLAGS(avx, "CYCLES_CPU_NO_AVX");
CHECK_CPU_FLAGS(sse41, "CYCLES_CPU_NO_SSE41");
CHECK_CPU_FLAGS(sse3, "CYCLES_CPU_NO_SSE3");
CHECK_CPU_FLAGS(sse2, "CYCLES_CPU_NO_SSE2");
#undef STRINGIFY
#undef CHECK_CPU_FLAGS
qbvh = true;
split_kernel = false;
}
DebugFlags::CUDA::CUDA()
: adaptive_compile(false),
split_kernel(false)
{
reset();
}
void DebugFlags::CUDA::reset()
{
if(getenv("CYCLES_CUDA_ADAPTIVE_COMPILE") != NULL)
adaptive_compile = true;
split_kernel = false;
}
DebugFlags::OpenCL::OpenCL()
: device_type(DebugFlags::OpenCL::DEVICE_ALL),
kernel_type(DebugFlags::OpenCL::KERNEL_DEFAULT),
debug(false),
single_program(false)
{
reset();
}
void DebugFlags::OpenCL::reset()
{
/* Initialize device type from environment variables. */
device_type = DebugFlags::OpenCL::DEVICE_ALL;
char *device = getenv("CYCLES_OPENCL_TEST");
if(device) {
if(strcmp(device, "NONE") == 0) {
device_type = DebugFlags::OpenCL::DEVICE_NONE;
}
else if(strcmp(device, "ALL") == 0) {
device_type = DebugFlags::OpenCL::DEVICE_ALL;
}
else if(strcmp(device, "DEFAULT") == 0) {
device_type = DebugFlags::OpenCL::DEVICE_DEFAULT;
}
else if(strcmp(device, "CPU") == 0) {
device_type = DebugFlags::OpenCL::DEVICE_CPU;
}
else if(strcmp(device, "GPU") == 0) {
device_type = DebugFlags::OpenCL::DEVICE_GPU;
}
else if(strcmp(device, "ACCELERATOR") == 0) {
device_type = DebugFlags::OpenCL::DEVICE_ACCELERATOR;
}
}
/* Initialize kernel type from environment variables. */
kernel_type = DebugFlags::OpenCL::KERNEL_DEFAULT;
if(getenv("CYCLES_OPENCL_MEGA_KERNEL_TEST") != NULL) {
kernel_type = DebugFlags::OpenCL::KERNEL_MEGA;
}
else if(getenv("CYCLES_OPENCL_SPLIT_KERNEL_TEST") != NULL) {
kernel_type = DebugFlags::OpenCL::KERNEL_SPLIT;
}
/* Initialize other flags from environment variables. */
debug = (getenv("CYCLES_OPENCL_DEBUG") != NULL);
single_program = (getenv("CYCLES_OPENCL_MULTI_PROGRAM") == NULL);
}
DebugFlags::DebugFlags()
: viewport_static_bvh(false)
{
/* Nothing for now. */
}
void DebugFlags::reset()
{
viewport_static_bvh = false;
cpu.reset();
cuda.reset();
opencl.reset();
}
std::ostream& operator <<(std::ostream &os,
DebugFlagsConstRef debug_flags)
{
os << "CPU flags:\n"
<< " AVX2 : " << string_from_bool(debug_flags.cpu.avx2) << "\n"
<< " AVX : " << string_from_bool(debug_flags.cpu.avx) << "\n"
<< " SSE4.1 : " << string_from_bool(debug_flags.cpu.sse41) << "\n"
<< " SSE3 : " << string_from_bool(debug_flags.cpu.sse3) << "\n"
<< " SSE2 : " << string_from_bool(debug_flags.cpu.sse2) << "\n"
<< " QBVH : " << string_from_bool(debug_flags.cpu.qbvh) << "\n"
<< " Split : " << string_from_bool(debug_flags.cpu.split_kernel) << "\n";
os << "CUDA flags:\n"
<< " Adaptive Compile: " << string_from_bool(debug_flags.cuda.adaptive_compile) << "\n";
const char *opencl_device_type,
*opencl_kernel_type;
switch(debug_flags.opencl.device_type) {
case DebugFlags::OpenCL::DEVICE_NONE:
opencl_device_type = "NONE";
break;
case DebugFlags::OpenCL::DEVICE_ALL:
opencl_device_type = "ALL";
break;
case DebugFlags::OpenCL::DEVICE_DEFAULT:
opencl_device_type = "DEFAULT";
break;
case DebugFlags::OpenCL::DEVICE_CPU:
opencl_device_type = "CPU";
break;
case DebugFlags::OpenCL::DEVICE_GPU:
opencl_device_type = "GPU";
break;
case DebugFlags::OpenCL::DEVICE_ACCELERATOR:
opencl_device_type = "ACCELERATOR";
break;
}
switch(debug_flags.opencl.kernel_type) {
case DebugFlags::OpenCL::KERNEL_DEFAULT:
opencl_kernel_type = "DEFAULT";
break;
case DebugFlags::OpenCL::KERNEL_MEGA:
opencl_kernel_type = "MEGA";
break;
case DebugFlags::OpenCL::KERNEL_SPLIT:
opencl_kernel_type = "SPLIT";
break;
}
os << "OpenCL flags:\n"
<< " Device type : " << opencl_device_type << "\n"
<< " Kernel type : " << opencl_kernel_type << "\n"
<< " Debug : " << string_from_bool(debug_flags.opencl.debug) << "\n"
<< " Single program : " << string_from_bool(debug_flags.opencl.single_program) << "\n"
<< " Memory limit : " << string_human_readable_size(debug_flags.opencl.mem_limit) << "\n";
return os;
}
CCL_NAMESPACE_END
|
