-
Notifications
You must be signed in to change notification settings - Fork 12
/
Ray.cpp
149 lines (134 loc) · 5.43 KB
/
Ray.cpp
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
#include "Ray.h"
#include <regex>
#include "internal/Core.h"
#ifdef ENABLE_REF_IMPL
#include "internal/RendererRef.h"
#else // ENABLE_REF_IMPL
#pragma message("Ray: Compiling without reference backend")
#endif // ENABLE_REF_IMPL
#ifdef ENABLE_SIMD_IMPL
#if !defined(__arm__) && !defined(__aarch64__) && !defined(_M_ARM) && !defined(_M_ARM64)
#include "internal/RendererAVX.h"
#include "internal/RendererAVX2.h"
#include "internal/RendererAVX512.h"
#include "internal/RendererSSE2.h"
#include "internal/RendererSSE41.h"
#elif defined(__arm__) || defined(__aarch64__) || defined(_M_ARM) || defined(_M_ARM64)
#include "internal/RendererNEON.h"
#elif defined(__i386__) || defined(__x86_64__)
#include "internal/RendererSSE2.h"
#endif
#else // ENABLE_SIMD_IMPL
#pragma message("Ray: Compiling without SIMD support")
#endif // #ifdef ENABLE_SIMD_IMPL
#if defined(ENABLE_DX_IMPL) && defined(_WIN32)
#include "internal/RendererDX.h"
#endif // defined(ENABLE_DX_IMPL) && defined(_WIN32)
#ifdef ENABLE_VK_IMPL
#include "internal/RendererVK.h"
#endif // ENABLE_VK_IMPL
#if !(defined(ENABLE_DX_IMPL) && defined(_WIN32)) && !defined(ENABLE_VK_IMPL)
#pragma message("Ray: Compiling without GPU support")
#endif
#include "internal/simd/detect.h"
#include "third-party/renderdoc/renderdoc_app.h"
namespace Ray {
LogNull g_null_log;
LogStdout g_stdout_log;
RENDERDOC_DevicePointer g_rdoc_device = {};
extern const std::pair<uint32_t, const char *> KnownGPUVendors[] = {
{0x1002, "AMD"}, {0x10DE, "NVIDIA"}, {0x8086, "INTEL"}, {0x13B5, "ARM"}};
extern const int KnownGPUVendorsCount = 4;
} // namespace Ray
Ray::RendererBase *Ray::CreateRenderer(const settings_t &s, ILog *log,
const std::function<void(int, int, ParallelForFunction &&)> ¶llel_for,
const Bitmask<eRendererType> enabled_types) {
#if defined(ENABLE_VK_IMPL)
if (enabled_types & eRendererType::Vulkan) {
log->Info("Ray: Creating Vulkan renderer %ix%i", s.w, s.h);
try {
return Vk::CreateRenderer(s, log, parallel_for);
} catch (std::exception &e) {
log->Info("Ray: Failed to create Vulkan renderer, %s", e.what());
}
}
#endif // ENABLE_VK_IMPL
#if defined(ENABLE_DX_IMPL) && defined(_WIN32)
if (enabled_types & eRendererType::DirectX12) {
log->Info("Ray: Creating DirectX12 renderer %ix%i", s.w, s.h);
try {
return Dx::CreateRenderer(s, log, parallel_for);
} catch (std::exception &e) {
log->Info("Ray: Failed to create DirectX12 renderer, %s", e.what());
}
}
#endif // defined(ENABLE_DX_IMPL) && defined(_WIN32)
#if !defined(__arm__) && !defined(__aarch64__) && !defined(_M_ARM) && !defined(_M_ARM64)
#ifdef ENABLE_SIMD_IMPL
const CpuFeatures features = GetCpuFeatures();
if ((enabled_types & eRendererType::SIMD_AVX512) && features.avx512_supported) {
log->Info("Ray: Creating AVX512 renderer %ix%i", s.w, s.h);
return Avx512::CreateRenderer(s, log);
}
if ((enabled_types & eRendererType::SIMD_AVX2) && features.avx2_supported) {
log->Info("Ray: Creating AVX2 renderer %ix%i", s.w, s.h);
return Avx2::CreateRenderer(s, log);
}
if ((enabled_types & eRendererType::SIMD_AVX) && features.avx_supported) {
log->Info("Ray: Creating AVX renderer %ix%i", s.w, s.h);
return Avx::CreateRenderer(s, log);
}
if ((enabled_types & eRendererType::SIMD_SSE41) && features.sse41_supported) {
log->Info("Ray: Creating SSE41 renderer %ix%i", s.w, s.h);
return Sse41::CreateRenderer(s, log);
}
if ((enabled_types & eRendererType::SIMD_SSE2) && features.sse2_supported) {
log->Info("Ray: Creating SSE2 renderer %ix%i", s.w, s.h);
return Sse2::CreateRenderer(s, log);
}
#endif
#ifdef ENABLE_REF_IMPL
if (enabled_types & eRendererType::Reference) {
log->Info("Ray: Creating Reference renderer %ix%i", s.w, s.h);
return Ref::CreateRenderer(s, log);
}
#endif
#elif defined(__arm__) || defined(__aarch64__) || defined(_M_ARM) || defined(_M_ARM64)
#ifdef ENABLE_SIMD_IMPL
if (enabled_types & eRendererType::SIMD_NEON) {
log->Info("Ray: Creating NEON renderer %ix%i", s.w, s.h);
return Neon::CreateRenderer(s, log);
}
#endif
#ifdef ENABLE_REF_IMPL
if (enabled_types & eRendererType::Reference) {
log->Info("Ray: Creating Reference renderer %ix%i", s.w, s.h);
return Ref::CreateRenderer(s, log);
}
#endif
#endif
#ifdef ENABLE_REF_IMPL
log->Info("Ray: Creating Reference renderer %ix%i", s.w, s.h);
return Ref::CreateRenderer(s, log);
#else
return nullptr;
#endif
}
int Ray::QueryAvailableGPUDevices(ILog *log, gpu_device_t out_devices[], const int capacity) {
int count = 0;
#ifdef ENABLE_VK_IMPL
count = Vk::Context::QueryAvailableDevices(log, out_devices, capacity);
#endif
#if defined(ENABLE_DX_IMPL) && defined(_WIN32)
count = Dx::Context::QueryAvailableDevices(log, out_devices, capacity);
#endif
return count;
}
bool Ray::MatchDeviceNames(const char *name, const char *pattern) {
std::regex match_name(pattern);
return std::regex_search(name, match_name) || strcmp(name, pattern) == 0;
}
const char *Ray::Version() { return "v0.3.0-unknown-commit"; }
// Workaround for identical template instantiations merging causing illegal instruction error with Windows Clang
template class std::vector<float, Ray::aligned_allocator<float, 64>>;
template class std::vector<uint16_t>;