init.cc 11.9 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
D
dzhwinter 已提交
2

L
Luo Tao 已提交
3 4 5
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
D
dzhwinter 已提交
6

L
Luo Tao 已提交
7
    http://www.apache.org/licenses/LICENSE-2.0
D
dzhwinter 已提交
8

L
Luo Tao 已提交
9 10 11 12 13
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. */
C
chengduo 已提交
14
#include <fstream>
D
dzhwinter 已提交
15 16
#include <string>

T
tensor-tang 已提交
17
#include "paddle/fluid/platform/cpu_helper.h"
T
tensor-tang 已提交
18
#include "paddle/fluid/platform/cpu_info.h"
19
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
Y
Yu Yang 已提交
20
#include "paddle/fluid/platform/cuda_device_guard.h"
21 22
#endif
#ifdef PADDLE_WITH_CUDA
23
#include "paddle/fluid/platform/dynload/cupti.h"
S
sneaxiy 已提交
24
#endif
Y
Yi Wang 已提交
25
#include "paddle/fluid/platform/device_context.h"
26
#include "paddle/fluid/platform/init.h"
Y
Yi Wang 已提交
27
#include "paddle/fluid/platform/place.h"
28

29 30 31 32 33
#ifdef PADDLE_WITH_XPU
#include "paddle/fluid/platform/xpu_header.h"
#include "paddle/fluid/platform/xpu_info.h"
#endif

34 35 36 37
#ifdef WITH_WIN_DUMP_DBG
#include <stdio.h>
#include <time.h>
#include <windows.h>
38

39 40 41
#include "DbgHelp.h"
#endif

42
DECLARE_int32(paddle_num_threads);
43 44 45
DEFINE_int32(multiple_of_cupti_buffer_size, 1,
             "Multiple of the CUPTI device buffer size. If the timestamps have "
             "been dropped when you are profiling, try increasing this value.");
T
tensor-tang 已提交
46

47 48 49 50
namespace paddle {
namespace platform {

void ParseCommandLineFlags(int argc, char **argv, bool remove) {
51
  ::GFLAGS_NAMESPACE::ParseCommandLineFlags(&argc, &argv, remove);
52 53 54 55 56
}

}  // namespace platform
}  // namespace paddle

D
dzhwinter 已提交
57 58 59
namespace paddle {
namespace framework {

60 61 62 63
#ifdef _WIN32
#define strdup _strdup
#endif

D
dzhwinter 已提交
64
std::once_flag gflags_init_flag;
65
std::once_flag glog_init_flag;
D
dzhwinter 已提交
66

67
bool InitGflags(std::vector<std::string> args) {
68
  bool successed = false;
D
dzhwinter 已提交
69
  std::call_once(gflags_init_flag, [&]() {
C
chengduo 已提交
70
    FLAGS_logtostderr = true;
L
Leo Chen 已提交
71 72 73 74 75
    // NOTE(zhiqiu): dummy is needed, since the function
    // ParseNewCommandLineFlags in gflags.cc starts processing
    // commandline strings from idx 1.
    // The reason is, it assumes that the first one (idx 0) is
    // the filename of executable file.
76 77
    args.insert(args.begin(), "dummy");
    std::vector<char *> argv;
D
dzhwinter 已提交
78
    std::string line;
79 80 81 82
    int argc = args.size();
    for (auto &arg : args) {
      argv.push_back(const_cast<char *>(arg.data()));
      line += arg;
D
dzhwinter 已提交
83 84
      line += ' ';
    }
L
Leo Chen 已提交
85 86
    VLOG(1) << "Before Parse: argc is " << argc
            << ", Init commandline: " << line;
87 88

    char **arr = argv.data();
89
    ::GFLAGS_NAMESPACE::ParseCommandLineFlags(&argc, &arr, true);
90
    successed = true;
91 92

    VLOG(1) << "After Parse: argc is " << argc;
D
dzhwinter 已提交
93
  });
94
  return successed;
D
dzhwinter 已提交
95 96
}

97
#ifdef PADDLE_WITH_CUDA
98 99 100 101
void InitCupti() {
#ifdef PADDLE_WITH_CUPTI
  if (FLAGS_multiple_of_cupti_buffer_size == 1) return;
  size_t attrValue = 0, attrValueSize = sizeof(size_t);
G
GaoWei8 已提交
102 103 104 105 106 107 108 109 110 111 112 113
#define MULTIPLY_ATTR_VALUE(attr)                                            \
  {                                                                          \
    PADDLE_ENFORCE_EQ(                                                       \
        !platform::dynload::cuptiActivityGetAttribute(attr, &attrValueSize,  \
                                                      &attrValue),           \
        true, platform::errors::Unavailable("Get cupti attribute failed.")); \
    attrValue *= FLAGS_multiple_of_cupti_buffer_size;                        \
    LOG(WARNING) << "Set " #attr " " << attrValue << " byte";                \
    PADDLE_ENFORCE_EQ(                                                       \
        !platform::dynload::cuptiActivitySetAttribute(attr, &attrValueSize,  \
                                                      &attrValue),           \
        true, platform::errors::Unavailable("Set cupti attribute failed.")); \
114 115 116 117 118 119 120 121 122
  }
  MULTIPLY_ATTR_VALUE(CUPTI_ACTIVITY_ATTR_DEVICE_BUFFER_SIZE);
  MULTIPLY_ATTR_VALUE(CUPTI_ACTIVITY_ATTR_DEVICE_BUFFER_SIZE_CDP);
#if CUDA_VERSION >= 9000
  MULTIPLY_ATTR_VALUE(CUPTI_ACTIVITY_ATTR_PROFILING_SEMAPHORE_POOL_SIZE);
#endif
#undef MULTIPLY_ATTR_VALUE
#endif
}
123
#endif
124

125
void InitDevices() {
126 127 128
// CUPTI attribute should be set before any CUDA context is created (see CUPTI
// documentation about CUpti_ActivityAttribute).
#ifdef PADDLE_WITH_CUDA
129
  InitCupti();
130
#endif
131
  /*Init all available devices by default */
132
  std::vector<int> devices;
133
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
D
dzhwinter 已提交
134
  try {
135 136
    // use user specified GPUs in single-node multi-process mode.
    devices = platform::GetSelectedDevices();
D
dzhwinter 已提交
137 138
  } catch (const std::exception &exp) {
    LOG(WARNING) << "Compiled with WITH_GPU, but no GPU found in runtime.";
139
  }
140 141 142 143 144 145 146 147
#endif
#ifdef PADDLE_WITH_XPU
  try {
    // use user specified XPUs in single-node multi-process mode.
    devices = platform::GetXPUSelectedDevices();
  } catch (const std::exception &exp) {
    LOG(WARNING) << "Compiled with WITH_XPU, but no XPU found in runtime.";
  }
D
dzhwinter 已提交
148
#endif
149
  InitDevices(devices);
D
dzhwinter 已提交
150 151
}

152
void InitDevices(const std::vector<int> devices) {
153 154 155
  std::vector<platform::Place> places;

  for (size_t i = 0; i < devices.size(); ++i) {
156 157 158
    // In multi process multi gpu mode, we may have gpuid = 7
    // but count = 1.
    if (devices[i] < 0) {
159 160 161
      LOG(WARNING) << "Invalid devices id.";
      continue;
    }
162

163
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
164
    places.emplace_back(platform::CUDAPlace(devices[i]));
165 166 167 168
#endif
#ifdef PADDLE_WITH_XPU
    places.emplace_back(platform::XPUPlace(devices[i]));
#endif
169 170
  }
  places.emplace_back(platform::CPUPlace());
171
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
172 173
  places.emplace_back(platform::CUDAPinnedPlace());
#endif
174
  platform::DeviceContextPool::Init(places);
Q
qingqing01 已提交
175

176
#ifndef PADDLE_WITH_MKLDNN
T
tensor-tang 已提交
177
  platform::SetNumThreads(FLAGS_paddle_num_threads);
178
#endif
T
tensor-tang 已提交
179

T
tensor-tang 已提交
180
#if !defined(_WIN32) && !defined(__APPLE__) && !defined(__OSX__)
T
tensor-tang 已提交
181
  if (platform::MayIUse(platform::avx)) {
T
tensor-tang 已提交
182 183 184 185
#ifndef __AVX__
    LOG(WARNING) << "AVX is available, Please re-compile on local machine";
#endif
  }
186 187

// Throw some informations when CPU instructions mismatch.
188 189 190 191 192 193
#define AVX_GUIDE(compiletime, runtime)                                  \
  PADDLE_THROW(platform::errors::Unavailable(                            \
      "This version is compiled on higher instruction(" #compiletime     \
      ") system, you may encounter illegal instruction error running on" \
      " your local CPU machine. Please reinstall the " #runtime          \
      " version or compile from source code."))
194 195

#ifdef __AVX512F__
T
tensor-tang 已提交
196 197
  if (!platform::MayIUse(platform::avx512f)) {
    if (platform::MayIUse(platform::avx2)) {
198
      AVX_GUIDE(AVX512, AVX2);
T
tensor-tang 已提交
199
    } else if (platform::MayIUse(platform::avx)) {
200 201 202 203 204
      AVX_GUIDE(AVX512, AVX);
    } else {
      AVX_GUIDE(AVX512, NonAVX);
    }
  }
T
tensor-tang 已提交
205
#endif
206 207

#ifdef __AVX2__
T
tensor-tang 已提交
208 209
  if (!platform::MayIUse(platform::avx2)) {
    if (platform::MayIUse(platform::avx)) {
210 211 212 213
      AVX_GUIDE(AVX2, AVX);
    } else {
      AVX_GUIDE(AVX2, NonAVX);
    }
T
tensor-tang 已提交
214 215
  }
#endif
216 217

#ifdef __AVX__
T
tensor-tang 已提交
218
  if (!platform::MayIUse(platform::avx)) {
219
    AVX_GUIDE(AVX, NonAVX);
T
tensor-tang 已提交
220
  }
221 222
#endif
#undef AVX_GUIDE
T
tensor-tang 已提交
223 224

#endif
225 226
}

C
chengduo 已提交
227
#ifndef _WIN32
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258
// Description Quoted from
// https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/signal.h.html
const struct {
  const char *name;
  const char *error_string;
} SignalErrorStrings[] = {
    {"SIGSEGV", "Segmentation fault"},
    {"SIGILL", "Illegal instruction"},
    {"SIGFPE", "Erroneous arithmetic operation"},
    {"SIGABRT", "Process abort signal"},
    {"SIGBUS", "Access to an undefined portion of a memory object"},
    {"SIGTERM", "Termination signal"},
};

bool StartsWith(const char *str, const char *prefix) {
  size_t len_prefix = strlen(prefix);
  size_t len_str = strlen(str);
  return len_str < len_prefix ? false : memcmp(prefix, str, len_prefix) == 0;
}

const char *ParseSignalErrorString(const std::string &str) {
  for (size_t i = 0;
       i < (sizeof(SignalErrorStrings) / sizeof(*(SignalErrorStrings))); ++i) {
    if (std::string::npos != str.find(SignalErrorStrings[i].name)) {
      return SignalErrorStrings[i].error_string;
    }
  }
  return "Unknown signal";
}

// Handle SIGSEGV, SIGILL, SIGFPE, SIGABRT, SIGBUS, and SIGTERM.
259
void SignalHandle(const char *data, int size) {
C
chengduo 已提交
260
  try {
261 262
    // NOTE1: The glog FailureSignalHandler dumped messages
    //   are deal with line by line
263
    auto signal_msg_dunmer_ptr = SignalMessageDumper::Instance().Get();
264 265 266
    // NOTE2: we only deal with the time info ane signal info,
    //   the stack trace will generated by paddle self
    if (StartsWith(data, "*** Aborted at")) {
267
      *signal_msg_dunmer_ptr << "\n  [TimeInfo: " << std::string(data, size - 1)
268
                             << "]\n";
269 270 271 272 273
    } else if (StartsWith(data, "***")) {
      std::string signal_info(data, size - 1);
      std::string useless_substr("; stack trace:");
      size_t start_pos = signal_info.rfind(useless_substr);
      signal_info.replace(start_pos, useless_substr.length(), "");
274
      *signal_msg_dunmer_ptr << "  [SignalInfo: " << signal_info << "]\n";
275 276 277

      // NOTE3: Final singal error message print.
      // Here does not throw an exception,
278
      // otherwise it will casue "terminate called recursively"
279 280 281 282 283 284 285 286 287
      std::ostringstream sout;
      sout << platform::GetCurrentTraceBackString();
      sout << "\n----------------------\nError Message "
              "Summary:\n----------------------\n";
      sout << platform::errors::Fatal(
                  "`%s` is detected by the operating system.",
                  ParseSignalErrorString(signal_info))
                  .to_string();
      std::cout << sout.str() << (*signal_msg_dunmer_ptr).str() << std::endl;
288
    }
C
chengduo 已提交
289
  } catch (...) {
290 291 292
    // Since the program has already triggered a system error,
    // no further processing is required here, glog FailureSignalHandler
    // will Kill program by the default signal handler
C
chengduo 已提交
293 294 295 296
  }
}
#endif

297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335
#ifdef WITH_WIN_DUMP_DBG
typedef BOOL(WINAPI *MINIDUMP_WRITE_DUMP)(
    IN HANDLE hProcess, IN DWORD ProcessId, IN HANDLE hFile,
    IN MINIDUMP_TYPE DumpType,
    IN CONST PMINIDUMP_EXCEPTION_INFORMATION ExceptionParam,
    OPTIONAL IN PMINIDUMP_USER_STREAM_INFORMATION UserStreamParam,
    OPTIONAL IN PMINIDUMP_CALLBACK_INFORMATION CallbackParam OPTIONAL);
void CreateDumpFile(LPCSTR lpstrDumpFilePathName,
                    EXCEPTION_POINTERS *pException) {
  HANDLE hDumpFile = CreateFile(lpstrDumpFilePathName, GENERIC_WRITE, 0, NULL,
                                CREATE_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL);
  MINIDUMP_EXCEPTION_INFORMATION dumpInfo;
  dumpInfo.ExceptionPointers = pException;
  dumpInfo.ThreadId = GetCurrentThreadId();
  dumpInfo.ClientPointers = TRUE;
  MINIDUMP_WRITE_DUMP MiniDumpWriteDump_;
  HMODULE hDbgHelp = LoadLibrary("DBGHELP.DLL");
  MiniDumpWriteDump_ =
      (MINIDUMP_WRITE_DUMP)GetProcAddress(hDbgHelp, "MiniDumpWriteDump");
  MiniDumpWriteDump_(GetCurrentProcess(), GetCurrentProcessId(), hDumpFile,
                     MiniDumpWithPrivateReadWriteMemory, &dumpInfo, NULL, NULL);
  CloseHandle(hDumpFile);
}

LONG ApplicationCrashHandler(EXCEPTION_POINTERS *pException) {
  time_t time_seconds = time(0);
  struct tm now_time;
  localtime_s(&now_time, &time_seconds);

  char buf[1024];
  sprintf_s(buf, "C:\\Paddle%04d%02d%02d-%02d%02d%02d.dmp",
            1900 + now_time.tm_year, 1 + now_time.tm_mon, now_time.tm_mday,
            now_time.tm_hour, now_time.tm_min, now_time.tm_sec);

  CreateDumpFile(buf, pException);
  return EXCEPTION_EXECUTE_HANDLER;
}
#endif

Y
Yang Yu 已提交
336
void InitGLOG(const std::string &prog_name) {
337
  std::call_once(glog_init_flag, [&]() {
338 339 340 341 342 343
// glog will not hold the ARGV[0] inside.
// Use strdup to alloc a new string.
#ifdef WITH_WIN_DUMP_DBG
    SetUnhandledExceptionFilter(
        (LPTOP_LEVEL_EXCEPTION_FILTER)ApplicationCrashHandler);
#endif
344
    google::InitGoogleLogging(strdup(prog_name.c_str()));
C
chengduo 已提交
345
#ifndef _WIN32
346 347
    google::InstallFailureSignalHandler();
    google::InstallFailureWriter(&SignalHandle);
C
chengduo 已提交
348
#endif
349
  });
Y
Yang Yu 已提交
350 351
}

D
dzhwinter 已提交
352 353
}  // namespace framework
}  // namespace paddle