device_context.cc 15.9 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
Q
qijun 已提交
2 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
    http://www.apache.org/licenses/LICENSE-2.0
6

Q
qijun 已提交
7 8 9 10 11
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. */
Y
Yi Wang 已提交
12
#include "paddle/fluid/platform/device_context.h"
13
#include <set>
14
#include <string>
Y
Yu Yang 已提交
15
#include <unordered_set>
16 17
#include <vector>

Y
Yi Wang 已提交
18
#include "paddle/fluid/memory/memory.h"
19 20
#ifdef PADDLE_WITH_CUDA
#include "paddle/fluid/framework/rw_lock.h"
21
#include "paddle/fluid/memory/allocation/cuda_device_context_allocator.h"
S
sneaxiy 已提交
22
#include "paddle/fluid/platform/cuda_device_guard.h"
23
#endif
24

25 26
#include "glog/logging.h"

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
namespace paddle {
namespace memory {

AllocationPtr Alloc(const platform::DeviceContext& dev_ctx, size_t size) {
  auto place = dev_ctx.GetPlace();
#ifdef PADDLE_WITH_CUDA
  if (size == 0 || !platform::is_gpu_place(place)) {
    return Alloc(place, size);
  }
  auto* default_dev_ctx = static_cast<platform::CUDADeviceContext*>(
      platform::DeviceContextPool::Instance().Get(place));
  auto& desired_dev_ctx =
      static_cast<const platform::CUDADeviceContext&>(dev_ctx);
  if (default_dev_ctx->stream() == desired_dev_ctx.stream()) {
    return Alloc(place, size);
  } else {
    return allocation::CUDADeviceContextAllocatorPool::Instance().Alloc(
        desired_dev_ctx, size);
  }
#else
  return Alloc(place, size);
#endif
}

}  // namespace memory
}  // namespace paddle

Q
qijun 已提交
54 55 56
namespace paddle {
namespace platform {

D
dzhwinter 已提交
57 58
DeviceContextPool* DeviceContextPool::pool = nullptr;

Y
Yu Yang 已提交
59
platform::DeviceContext* DeviceContextPool::Get(const platform::Place& place) {
D
dzhwinter 已提交
60 61 62
  auto it = device_contexts_.find(place);
  if (it == device_contexts_.end()) {
    PADDLE_THROW(
63 64 65 66
        "Place %s is not supported, Please check that your paddle compiles "
        "with WITH_GPU "
        "option or check that your train process hold the correct gpu_id if "
        "you use Executor",
M
minqiyang 已提交
67
        place);
D
dzhwinter 已提交
68
  }
69
  return it->second.get().get();
D
dzhwinter 已提交
70 71
}

72 73 74 75 76 77 78 79 80 81 82
template <typename DevCtx, typename PlaceType>
inline void EmplaceDeviceContext(
    std::map<Place, std::shared_future<std::unique_ptr<DeviceContext>>>*
        map_ptr,
    platform::Place p) {
  using PtrType = std::unique_ptr<DeviceContext>;
  map_ptr->emplace(p, std::async(std::launch::deferred, [=] {
                     // lazy evaluation. i.e., only create device context at
                     // first `Get`
                     return PtrType(new DevCtx(boost::get<PlaceType>(p)));
                   }));
C
chengduozh 已提交
83 84
}

D
dzhwinter 已提交
85 86 87
DeviceContextPool::DeviceContextPool(
    const std::vector<platform::Place>& places) {
  PADDLE_ENFORCE_GT(places.size(), 0);
88
  std::set<Place> set;
Y
Yu Yang 已提交
89 90 91 92 93
  for (auto& p : places) {
    set.insert(p);
  }
  for (auto& p : set) {
    if (platform::is_cpu_place(p)) {
94
#ifdef PADDLE_WITH_MKLDNN
95
      EmplaceDeviceContext<MKLDNNDeviceContext, CPUPlace>(&device_contexts_, p);
96
#else
97
      EmplaceDeviceContext<CPUDeviceContext, CPUPlace>(&device_contexts_, p);
98
#endif
Y
Yu Yang 已提交
99
    } else if (platform::is_gpu_place(p)) {
D
dzhwinter 已提交
100
#ifdef PADDLE_WITH_CUDA
101
      EmplaceDeviceContext<CUDADeviceContext, CUDAPlace>(&device_contexts_, p);
D
dzhwinter 已提交
102 103
#else
      PADDLE_THROW(
D
dzhwinter 已提交
104
          "'CUDAPlace' is not supported, Please re-compile with WITH_GPU "
D
dzhwinter 已提交
105
          "option");
C
chengduoZH 已提交
106 107 108
#endif
    } else if (platform::is_cuda_pinned_place(p)) {
#ifdef PADDLE_WITH_CUDA
109 110
      EmplaceDeviceContext<CUDAPinnedDeviceContext, CUDAPinnedPlace>(
          &device_contexts_, p);
C
chengduoZH 已提交
111 112 113 114
#else
      PADDLE_THROW(
          "'CUDAPlace' is not supported, Please re-compile with WITH_GPU "
          "option");
D
dzhwinter 已提交
115 116 117 118 119
#endif
    }
  }
}

120 121 122 123
CPUDeviceContext::CPUDeviceContext() {
  eigen_device_.reset(new Eigen::DefaultDevice());
}

D
dzhwinter 已提交
124
CPUDeviceContext::CPUDeviceContext(CPUPlace place) : place_(place) {
125 126 127 128 129 130 131
  eigen_device_.reset(new Eigen::DefaultDevice());
}

Eigen::DefaultDevice* CPUDeviceContext::eigen_device() const {
  return eigen_device_.get();
}

D
dzhwinter 已提交
132
Place CPUDeviceContext::GetPlace() const { return place_; }
133

134
#ifdef PADDLE_WITH_CUDA
135

Q
init  
qijun 已提交
136 137 138 139 140 141 142
class EigenCudaStreamDevice : public Eigen::StreamInterface {
 public:
  EigenCudaStreamDevice() : scratch_(nullptr), semaphore_(nullptr) {
    Eigen::initializeDeviceProp();
  }
  ~EigenCudaStreamDevice() override {}

D
dzhwinter 已提交
143
  void Reinitialize(const cudaStream_t* cuda_stream, CUDAPlace place) {
Q
init  
qijun 已提交
144 145 146 147 148 149 150 151 152 153 154 155
    stream_ = cuda_stream;
    place_ = place;
    device_prop_ = &Eigen::m_deviceProperties[place.device];
  }

  const cudaStream_t& stream() const override { return *stream_; }

  const cudaDeviceProp& deviceProperties() const override {
    return *device_prop_;
  }

  void* allocate(size_t num_bytes) const override {
S
sneaxiy 已提交
156 157 158
    if (UNLIKELY(num_bytes == 0)) {
      return nullptr;
    }
159 160 161
    auto buf = memory::Alloc(place_, num_bytes);
    VLOG(4) << "Eigen allocated at " << buf->ptr() << ", size" << buf->size()
            << " requested " << num_bytes;
162
    void* retv = buf->ptr();
S
sneaxiy 已提交
163 164 165 166
    {
      std::lock_guard<std::mutex> lock(mtx_);
      allocations_.emplace(retv, std::move(buf));
    }
167
    return retv;
Q
init  
qijun 已提交
168 169
  }

S
sneaxiy 已提交
170 171 172 173 174 175
  void deallocate(void* buffer) const override {
    if (LIKELY(buffer)) {
      std::lock_guard<std::mutex> lock(mtx_);
      allocations_.erase(buffer);
    }
  }
Q
init  
qijun 已提交
176 177 178 179 180 181 182 183 184 185 186 187 188

  void* scratchpad() const override {
    if (scratch_ == NULL) {
      scratch_ = allocate(Eigen::kCudaScratchSize + sizeof(unsigned int));
    }
    return scratch_;
  }

  unsigned int* semaphore() const override {
    if (semaphore_ == NULL) {
      char* scratch =
          static_cast<char*>(scratchpad()) + Eigen::kCudaScratchSize;
      semaphore_ = reinterpret_cast<unsigned int*>(scratch);
189
      PADDLE_ENFORCE_CUDA_SUCCESS(
Q
init  
qijun 已提交
190 191 192 193 194 195
          cudaMemsetAsync(semaphore_, 0, sizeof(unsigned int), *stream_));
    }
    return semaphore_;
  }

 private:
D
dzhwinter 已提交
196
  CUDAPlace place_;
Q
init  
qijun 已提交
197 198
  const cudaStream_t* stream_;         // not owned;
  const cudaDeviceProp* device_prop_;  // not owned;
Q
qijun 已提交
199
  mutable void* scratch_;
Q
init  
qijun 已提交
200
  mutable unsigned int* semaphore_;
S
sneaxiy 已提交
201
  mutable std::mutex mtx_;  // to protect allocations_
Y
Yu Yang 已提交
202
  mutable std::unordered_map<void*, memory::AllocationPtr> allocations_;
Q
init  
qijun 已提交
203 204
};

205 206 207 208 209 210 211 212 213
void CudnnWorkspaceHandle::ReallocWorkspace(size_t required_workspace_bytes) {
  if (required_workspace_bytes <= WorkspaceSize()) {
    return;
  }
  // reset allocation first before re-allocate to save memory
  allocation_.reset();
  allocation_ = memory::Alloc(device_context_, required_workspace_bytes);
}

214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241
thread_local std::unordered_map<const CUDADeviceContext*,
                                std::unique_ptr<CUDAContext>>
    CUDADeviceContext::thread_ctx_;
thread_local std::mutex CUDADeviceContext::ctx_mtx_;

void CUDAContext::InitEigenContext(const stream::CUDAStream& stream) {
  eigen_stream_.reset(new EigenCudaStreamDevice());
  eigen_stream_->Reinitialize(&stream.stream(), place_);
  eigen_device_.reset(new Eigen::GpuDevice(eigen_stream_.get()));
}

CUDAContext::CUDAContext(const CUDAPlace& place,
                         const enum stream::Priority& priority) {
  place_ = place;
  CUDADeviceGuard guard(place_.device);
  stream_.Init(place, priority);
  InitEigenContext(stream_);
  InitCuBlasContext(stream_);
  InitCuDNNContext(stream_);
  InitCallbackManager(stream_);
}

CUDAContext::~CUDAContext() {
  CUDADeviceGuard guard(place_.device);
  DestoryCuDNNContext();
  DestoryCuBlasContext();
}

242
CUDADeviceContext::CUDADeviceContext(CUDAPlace place) : place_(place) {
Y
Yu Yang 已提交
243
  CUDADeviceGuard guard(place_.device);
C
chengduo 已提交
244 245 246
  compute_capability_ = GetCUDAComputeCapability(place_.device);
  multi_process_ = GetCUDAMultiProcessors(place_.device);
  max_threads_per_mp_ = GetCUDAMaxThreadsPerMultiProcessor(place_.device);
247
  max_grid_dim_size_ = GetGpuMaxGridDimSize(place_.device);
248
  max_threads_per_block_ = GetCUDAMaxThreadsPerBlock(place_.device);
249

C
chengduo 已提交
250 251 252
  driver_version_ = GetCUDADriverVersion(place_.device);
  runtime_version_ = GetCUDARuntimeVersion(place_.device);

253 254
  LOG_FIRST_N(WARNING, 1) << "Please NOTE: device: " << place_.device
                          << ", CUDA Capability: " << compute_capability_
C
chengduo 已提交
255
                          << ", Driver API Version: " << driver_version_ / 1000
256
                          << "." << (driver_version_ % 100) / 10
C
chengduo 已提交
257 258 259
                          << ", Runtime API Version: "
                          << runtime_version_ / 1000 << "."
                          << (runtime_version_ % 100) / 10;
260 261 262
  size_t cudnn_dso_ver = dynload::cudnnGetVersion();
  LOG_FIRST_N(WARNING, 1) << "device: " << place_.device
                          << ", cuDNN Version: " << cudnn_dso_ver / 1000 << "."
263
                          << (cudnn_dso_ver % 1000) / 100 << ".";
S
sneaxiy 已提交
264 265 266

  {
    // Check CUDA/CUDNN version compatiblity
267 268 269 270
    auto local_cuda_version =
        (driver_version_ / 1000) * 10 + (driver_version_ % 100) / 10;
    auto compile_cuda_version =
        (CUDA_VERSION / 1000) * 10 + (CUDA_VERSION % 100) / 10;
S
sneaxiy 已提交
271 272 273 274 275 276 277 278 279 280 281 282
    if (local_cuda_version < compile_cuda_version) {
      LOG_FIRST_N(WARNING, 1)
          << "WARNING: device: " << place_.device
          << ". The installed Paddle is compiled with CUDA "
          << compile_cuda_version / 10 << "." << compile_cuda_version % 10
          << ", but CUDA runtime version in your machine is "
          << local_cuda_version / 10 << "." << local_cuda_version % 10
          << ", which may cause serious incompatible bug. "
          << "Please recompile or reinstall Paddle with compatible CUDA "
             "version.";
    }
  }
283
  default_ctx_.reset(new CUDAContext(place_));
284 285 286 287
}

CUDADeviceContext::~CUDADeviceContext() {
  SetDeviceId(place_.device);
L
liaogang 已提交
288
  Wait();
S
sneaxiy 已提交
289
  WaitStreamCallback();
290 291
}

L
liaogang 已提交
292
Place CUDADeviceContext::GetPlace() const { return place_; }
293

294
void CUDADeviceContext::Wait() const { context()->Wait(); }
295

K
Kexin Zhao 已提交
296
int CUDADeviceContext::GetComputeCapability() const {
C
chengduo 已提交
297
  return compute_capability_;
K
Kexin Zhao 已提交
298 299
}

300
int CUDADeviceContext::GetMaxPhysicalThreadCount() const {
C
chengduo 已提交
301
  return multi_process_ * max_threads_per_mp_;
302 303
}

304 305 306 307 308 309
int CUDADeviceContext::GetSMCount() const { return multi_process_; }

int CUDADeviceContext::GetMaxThreadsPerBlock() const {
  return max_threads_per_block_;
}

310
Eigen::GpuDevice* CUDADeviceContext::eigen_device() const {
311
  return context()->EigenDevice().get();
312 313
}

314
bool CUDADeviceContext::tensor_core_available() const {
315
  return context()->CublasTensorCoreHandle() != nullptr;
S
sneaxiy 已提交
316 317
}

318 319 320 321
dim3 CUDADeviceContext::GetCUDAMaxGridDimSize() const {
  return max_grid_dim_size_;
}

322 323 324
cudnnHandle_t CUDADeviceContext::cudnn_handle() const {
  return context()->CudnnHandle();
}
325

S
sneaxiy 已提交
326
CudnnWorkspaceHandle CUDADeviceContext::cudnn_workspace_handle() const {
327
  return CudnnWorkspaceHandle(*this, &cudnn_handle_mtx_);
328
}
329

330
cudaStream_t CUDADeviceContext::stream() const { return context()->Stream(); }
Q
qijun 已提交
331

C
chengduoZH 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344 345
CUDAPinnedDeviceContext::CUDAPinnedDeviceContext() {
  eigen_device_.reset(new Eigen::DefaultDevice());
}

CUDAPinnedDeviceContext::CUDAPinnedDeviceContext(CUDAPinnedPlace place)
    : place_(place) {
  eigen_device_.reset(new Eigen::DefaultDevice());
}

Eigen::DefaultDevice* CUDAPinnedDeviceContext::eigen_device() const {
  return eigen_device_.get();
}

Place CUDAPinnedDeviceContext::GetPlace() const { return place_; }
L
Luo Tao 已提交
346
#endif
Q
qijun 已提交
347

T
tensor-tang 已提交
348 349
#ifdef PADDLE_WITH_MKLDNN
MKLDNNDeviceContext::MKLDNNDeviceContext(CPUPlace place)
A
Adam 已提交
350 351 352
    : CPUDeviceContext(place),
      engine_(mkldnn::engine::kind::cpu, 0),
      p_blobmap_() {
353 354
  p_blobmap_.reset(new BlobMap());
  p_mutex_.reset(new std::mutex());
T
tensor-tang 已提交
355 356
}

S
Sylwester Fraczek 已提交
357
namespace {
358 359
// Current mkldnn session id.
thread_local size_t cur_mkldnn_session_id = kMKLDNNSessionID_Default;
360 361 362 363
// Current data input shape string.
// - For fixed-shape, it's a null string in default.
// - For dynamic-shape, it's user specific.
thread_local std::string cur_input_shape_str = "";
364 365 366
// the cache capacity of different input shapes for MKLDNN.
// Default 1 means fixed input shape, not dynamic shape.
thread_local int cur_input_shape_cache_capacity = 1;
367 368 369 370
// Recently registered data_format. This is needed to
// know for converting MKL-DNN Tensor to non MKL-DNN
thread_local paddle::framework::DataLayout cur_paddle_data_layout =
    paddle::framework::DataLayout::kNCHW;
371
}  // namespace
S
Sylwester Fraczek 已提交
372

373 374
void set_cur_mkldnn_session_id(size_t sid) { cur_mkldnn_session_id = sid; }
size_t get_cur_mkldnn_session_id(void) { return cur_mkldnn_session_id; }
375 376 377
void set_cur_input_shape_str(std::string input_shape_str) {
  cur_input_shape_str = input_shape_str;
}
378 379 380
void set_cur_input_shape_cache_capacity(int input_shape_cache_capacity) {
  cur_input_shape_cache_capacity = input_shape_cache_capacity;
}
S
Sylwester Fraczek 已提交
381

382 383 384 385 386 387 388 389
void set_cur_paddle_data_layout(framework::DataLayout dl) {
  cur_paddle_data_layout = dl;
}

framework::DataLayout get_cur_paddle_data_layout(void) {
  return cur_paddle_data_layout;
}

390 391
void MKLDNNDeviceContext::ResetBlobMap() const { p_blobmap_->clear(); }

392 393 394 395 396 397 398 399 400 401 402
size_t MKLDNNDeviceContext::GetShapeBlobSize() const {
  std::lock_guard<std::mutex> lock(*p_mutex_);
  BlobMap* pMap = p_blobmap_.get();
  auto map_it = pMap->find(cur_mkldnn_session_id);
  if (map_it == pMap->end()) {
    LOG(FATAL) << "MKLDNNDeviceContext don't find cur_mkldnn_session_id : "
               << cur_mkldnn_session_id;
  }
  return map_it->second->size();
}

403 404
void MKLDNNDeviceContext::SetBlob(const std::string& name,
                                  std::shared_ptr<void> data) const {
405
  BlobMap* pMap = p_blobmap_.get();
406
  std::shared_ptr<ShapeBlob> sBlob = nullptr;
407 408
  std::shared_ptr<KeyBlob> pBlob = nullptr;

409
  int sid = platform::get_cur_mkldnn_session_id();
T
tensor-tang 已提交
410

411
  std::lock_guard<std::mutex> lock(*p_mutex_);
T
tensor-tang 已提交
412

413 414
  // Find ShapeBlob for current mkldnn session id.
  auto map_it = pMap->find(sid);
415 416 417

  if (map_it == pMap->end()) {
    // 1st time to set blob in current thread
418
    sBlob = std::shared_ptr<ShapeBlob>(new ShapeBlob());
419 420
    (*pMap)[sid] = sBlob;
    VLOG(2) << "SetBlob: sid=" << sid << ", add new sid\n";
421
  } else {
422
    sBlob = map_it->second;
423
  }
T
tensor-tang 已提交
424

425 426
  // Find KeyBlob for current input shape
  auto key_it = sBlob->find(cur_input_shape_str);
427

428
  if (key_it == sBlob->end()) {
429 430
    // In cache clearing mode, cur_input_shape_cache_capacity defines
    // max pblob capacity
431 432
    if ((static_cast<size_t>(sid) == kMKLDNNSessionID_CacheClearing) &&
        sBlob->size() &&
433 434 435 436 437 438
        (sBlob->size() >=
         static_cast<size_t>(cur_input_shape_cache_capacity))) {
      VLOG(2) << "sid=" << sid
              << ", remove all blobs of shape: " << sBlob->begin()->first;
      sBlob->erase(sBlob->begin()->first);
    }
439 440
    pBlob = std::shared_ptr<KeyBlob>(new KeyBlob());
    (*sBlob)[cur_input_shape_str] = pBlob;
441
  } else {
442
    pBlob = key_it->second;
443 444
  }

445 446 447 448 449 450 451
  // Find Blob via name
  auto blob_it = pBlob->find(name);
  if (blob_it == pBlob->end()) {
    (*pBlob)[name] = data;
  } else {
    blob_it->second = data;  // set data to existing blob
  }
452
  VLOG(2) << "SetBlob: sid=" << sid << ", add blob=" << name << "\n";
453
  // lock will be automatically released when out of scope
454
  return;
T
tensor-tang 已提交
455 456
}

457 458
std::shared_ptr<void> MKLDNNDeviceContext::GetBlob(
    const std::string& name) const {
459
  BlobMap* pMap = p_blobmap_.get();
460
  std::shared_ptr<ShapeBlob> sBlob = nullptr;
461
  std::shared_ptr<KeyBlob> pBlob = nullptr;
T
tensor-tang 已提交
462

463
  int sid = platform::get_cur_mkldnn_session_id();
T
tensor-tang 已提交
464

465
  std::lock_guard<std::mutex> lock(*p_mutex_);
466

467 468
  // Find ShapeBlob for current mkldnn session id firstly
  auto map_it = pMap->find(sid);
469
  if (map_it == pMap->end()) {
470
    VLOG(2) << "GetBlob: sid=" << sid << ", miss sid\n";
471 472 473 474 475 476 477
    return nullptr;
  }
  sBlob = map_it->second;

  // Find KeyBlob for current input shape secondly
  auto sBlob_it = sBlob->find(cur_input_shape_str);
  if (sBlob_it == sBlob->end()) {
478
    VLOG(2) << "GetBlob: sid=" << cur_input_shape_str
479 480 481 482
            << ", miss input_shape_str\n";
    return nullptr;
  }
  pBlob = sBlob_it->second;
483 484 485 486

  // Find Blob via name
  auto key_it = pBlob->find(name);

487
  if (key_it == pBlob->end()) {
488
    VLOG(2) << "GetBlob sid=" << sid << ", miss blob=" << name << "\n";
489 490
    return nullptr;
  }
491

492
  VLOG(2) << "GetBlob sid=" << sid << ", get blob=" << name << "\n";
493 494
  // lock will be automatically released when out of scope
  return key_it->second;
T
tensor-tang 已提交
495 496 497 498
}

#endif

Q
qijun 已提交
499
}  // namespace platform
Q
qijun 已提交
500
}  // namespace paddle