Files
paddlepaddle--paddle/paddle/phi/core/platform/profiler.h
T
2026-07-13 12:40:42 +08:00

229 lines
6.4 KiB
C++

// Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
//
// 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.
#pragma once
#include <forward_list>
#include <list>
#include <map>
#include <memory>
#include <mutex> // NOLINT
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <utility>
#include <vector>
#include "paddle/phi/api/profiler/event.h"
#include "paddle/phi/common/place.h"
#include "paddle/phi/core/enforce.h"
#include "paddle/phi/core/platform/profiler/event_tracing.h"
#include "paddle/phi/core/platform/profiler/mem_tracing.h"
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
#include "paddle/phi/core/platform/device/gpu/gpu_info.h"
#endif
#include "paddle/phi/api/profiler/profiler.h"
#include "paddle/utils/test_macros.h"
namespace phi {
namespace proto {
class Profile;
}
} // namespace phi
namespace paddle {
namespace platform {
using phi::Event;
using phi::EventRole;
using phi::EventType;
using phi::MemEvent;
namespace proto {
class Profile;
}
const int kEnableProfiler = 1;
const int kDisableProfiler = 2;
using ProfilerState = phi::ProfilerState;
using TracerOption = phi::TracerOption;
// Candidate keys to sort the profiling report
enum class EventSortingKey {
kDefault,
kCalls,
kTotal,
kMin,
kMax,
kAve,
kCPUTime,
kGPUTime
};
struct MemoryProfilerReport {
size_t alloc_times{0};
size_t alloc_size{0};
size_t free_times{0};
size_t free_size{0};
};
// The information of each event given in the profiling report
struct EventItem {
std::string name;
int calls;
double total_time;
double max_time;
double ave_time;
double min_time;
double cpu_time;
double gpu_time;
float ratio;
EventRole role;
};
struct OverHead {
bool print_overhead = false;
bool print_explanation = false;
double elapsed_time = 0.; // the elapsed time of all events
double accumulated_time = 0.; // the accumulated time of all events
double compute_time = 0.0;
double framework_time = 0.0;
EventItem memcpy_item;
std::vector<EventItem> sub_memcpy_items;
};
struct MemEventRecorder {
public:
void PushMemRecord(const void* ptr, const Place& place, size_t size);
void PopMemRecord(const void* ptr, const Place& place);
void PushMemRecord(const void* ptr,
const Place& place,
size_t size,
phi::TracerMemEventType type,
uint64_t current_allocated,
uint64_t current_reserved,
uint64_t peak_allocated,
uint64_t peak_reserved);
void PopMemRecord(const void* ptr,
const Place& place,
size_t size,
phi::TracerMemEventType type,
uint64_t current_allocated,
uint64_t current_reserved,
uint64_t peak_allocated,
uint64_t peak_reserved);
void Flush();
static MemEventRecorder& Instance() { return recorder; }
private:
struct PADDLE_API RecordMemEvent {
RecordMemEvent(const Place& place, size_t bytes);
~RecordMemEvent();
Place place_;
size_t bytes_;
uint64_t start_ns_;
uint64_t end_ns_;
std::string alloc_in_;
std::string free_in_;
};
static MemEventRecorder recorder;
std::map<Place,
std::unordered_map<const void*, std::unique_ptr<RecordMemEvent>>>
address_memevent_;
std::mutex mtx_;
MemEventRecorder() {}
DISABLE_COPY_AND_ASSIGN(MemEventRecorder);
};
struct PADDLE_API RecordBlock {
explicit RecordBlock(int block_id);
~RecordBlock();
private:
bool is_enabled_;
std::string name_;
uint64_t start_ns_;
};
template <typename T>
using EventList = phi::EventList<T>;
void Mark(const std::string& name);
void PushMemEvent(uint64_t start_ns,
uint64_t end_ns,
size_t bytes,
const Place& place,
const std::string& annotation);
void PopMemEvent(uint64_t start_ns,
uint64_t end_ns,
size_t bytes,
const Place& place,
const std::string& annotation);
using phi::PopEvent;
using phi::PushEvent;
// Return the event list of all threads. Assumed the returned value calls
// event_lists, event_lists[i][j] represents the j-th Event of i-th thread.
PADDLE_API std::vector<std::vector<Event>> GetAllEvents();
// Enable the profiling function.
PADDLE_API void EnableProfiler(ProfilerState state);
// Clear the phi::ProfilerHelper::g_all_event_lists, which is total event lists
// of all threads.
PADDLE_API void ResetProfiler();
PADDLE_API void DisableProfiler(EventSortingKey sorted_key,
const std::string& profile_path);
// Disable profiler but return events instead of print it.
PADDLE_API void CompleteProfilerEvents(
phi::proto::Profile* tracer_profile,
std::vector<std::vector<Event>>* time_events,
std::vector<std::vector<MemEvent>>* mem_events);
// Test if the profiler is currently enabled.
PADDLE_API bool IsProfileEnabled();
// Whether the trainer should send profiling state to PS.
PADDLE_API bool ShouldSendProfileState();
PADDLE_API std::string OpName(
const std::map<std::string, std::vector<std::string>>& name_map,
const std::string& type_name);
PADDLE_API void SetTracerOption(TracerOption option);
PADDLE_API platform::TracerOption GetTracerOption();
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
void DummyKernelAndEvent();
#endif
// Mark current process as PS by assigning a lister id.
PADDLE_API void SetProfileListener();
PADDLE_API int64_t ListenerId();
PADDLE_API void NvprofEnableRecordEvent();
PADDLE_API void NvprofDisableRecordEvent();
PADDLE_API void EnableHostEventRecorder();
PADDLE_API void DisableHostEventRecorder();
PADDLE_API void EnableMemoryRecorder();
PADDLE_API void DisableMemoryRecorder();
// Defined for UT
std::string PrintHostEvents();
} // namespace platform
} // namespace paddle