Files
wehub-resource-sync 26446540fa
Lint / lint (push) Has been cancelled
CI / MacOS (push) Has been cancelled
CI / Windows (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 13:36:25 +08:00

302 lines
13 KiB
C++

/*
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you 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.
*/
#ifndef TVM_S_TIR_META_SCHEDULE_TASK_SCHEDULER_H_
#define TVM_S_TIR_META_SCHEDULE_TASK_SCHEDULER_H_
#include <tvm/ffi/container/array.h>
#include <tvm/ffi/function.h>
#include <tvm/ffi/optional.h>
#include <tvm/ffi/reflection/registry.h>
#include <tvm/runtime/base.h>
#include <tvm/s_tir/meta_schedule/builder.h>
#include <tvm/s_tir/meta_schedule/cost_model.h>
#include <tvm/s_tir/meta_schedule/measure_callback.h>
#include <tvm/s_tir/meta_schedule/runner.h>
#include <tvm/s_tir/meta_schedule/tune_context.h>
#include <tvm/s_tir/random_engine.h>
#include <string>
#include <vector>
namespace tvm {
namespace s_tir {
namespace meta_schedule {
class TaskRecordNode : public ffi::Object {
public:
/*! \brief The tune context of the task. */
TuneContext ctx{ffi::UnsafeInit()};
/*! \brief The weight of the task */
double task_weight{1.0};
/*! \brief The FLOP count of the task */
double flop{1.0};
/*! \brief Whether the tuning task has been stopped or finished. */
bool is_terminated = false;
/*! \brief Builder errors happens in the task */
int build_error_count = 0;
/*! \brief Runner errors happens in the task */
int run_error_count = 0;
/*! \brief The latency of each run, in milliseconds. */
std::vector<double> latency_ms = {};
/*! \brief The measure candidates. */
ffi::Optional<ffi::Array<MeasureCandidate>> measure_candidates = std::nullopt;
/*! \brief The building results. */
ffi::Optional<ffi::Array<BuilderResult>> builder_results = std::nullopt;
/*! \brief Packed functions to fetch the runner results asynchronously. */
ffi::Optional<ffi::Array<RunnerFuture>> runner_futures = std::nullopt;
static void RegisterReflection() {
namespace refl = tvm::ffi::reflection;
refl::ObjectDef<TaskRecordNode>()
.def_ro("ctx", &TaskRecordNode::ctx)
.def_ro("task_weight", &TaskRecordNode::task_weight)
.def_ro("flop", &TaskRecordNode::flop)
.def_ro("is_terminated", &TaskRecordNode::is_terminated)
.def_ro("build_error_count", &TaskRecordNode::build_error_count)
.def_ro("run_error_count", &TaskRecordNode::run_error_count)
.def_ro("measure_candidates", &TaskRecordNode::measure_candidates)
.def_ro("builder_results", &TaskRecordNode::builder_results)
.def_ro("runner_futures", &TaskRecordNode::runner_futures);
}
static constexpr const bool _type_mutable = true;
TVM_FFI_DECLARE_OBJECT_INFO_FINAL("s_tir.meta_schedule.TaskRecord", TaskRecordNode, ffi::Object);
};
/*!
* \brief Managed reference to TaskRecordNode.
* \sa TaskRecordNode
*/
class TaskRecord : public ffi::ObjectRef {
public:
/*! \brief Constructor */
explicit TaskRecord(TuneContext task, double task_weight);
TVM_FFI_DEFINE_OBJECT_REF_METHODS_NOTNULLABLE(TaskRecord, ffi::ObjectRef, TaskRecordNode);
};
/*!
* \brief The abstract interface of task schedulers.
* \note The relationship between SpaceGenerator and other classes are as follows:
+--------------------------------------------------------------+
+--+-----------------------------------------------------------+ |
+--+------------------ Tune Context -----------------------------+ | |
| +---------------------+ | | |
| | | Generate | | |
| | Space Generator +--------------+ | | |
| | | | | | |
| +---------------------+ v | | |
| Design Space | | |
| +---------------------+ | | | |
| Generate | | Pretuning | | | |
| +-----------+ Search Strategy |<-------------+ | | |
| | | | | +--+
| | +---------------------+ +--+
+----+----------------------------------------------------------+
|
|
+----+---------------- Managed By Task Scheduler ---------------------+
| | +-----------+ |
| | Send to | | Send to |
| v +-------------+| Builder +----------+ |
| Measure Candidate | Builder | | Runner | |
| | | +-----------+ | |
| | +------------+------------+ | |
| | | | +-----------+ | |
| +---->| Task Scheduler | | | | |
| | | | Runner |<-----+ |
| +-------------------------+ | | |
| ^ +-----+-----+ |
| | | |
| +---- Runner Future <-------+ |
+---------------------------------------------------------------------+
*/
class TaskSchedulerNode : public ffi::Object {
public:
/*! \brief The tuning task's logging function. */
ffi::Function logger;
/*! \brief Records for each task */
ffi::Array<TaskRecord> tasks_;
/*! \brief The list of measure callbacks of the scheduler. */
ffi::Array<MeasureCallback> measure_callbacks_;
/*! \brief The database used in tuning */
ffi::Optional<Database> database_;
/*! \brief The cost model used in tuning */
ffi::Optional<CostModel> cost_model_;
/*! \brief The number of remaining tasks to be tuned. */
int remaining_tasks_;
/*! \brief The default destructor. */
virtual ~TaskSchedulerNode() = default;
static void RegisterReflection() {
namespace refl = tvm::ffi::reflection;
refl::ObjectDef<TaskSchedulerNode>()
.def_ro("tasks_", &TaskSchedulerNode::tasks_)
.def_ro("measure_callbacks_", &TaskSchedulerNode::measure_callbacks_)
.def_ro("database_", &TaskSchedulerNode::database_)
.def_ro("cost_model_", &TaskSchedulerNode::cost_model_)
.def_ro("remaining_tasks_", &TaskSchedulerNode::remaining_tasks_);
}
/*!
* \brief Fetch the next task id.
* \return The next task id.
*/
virtual int NextTaskId() = 0;
/*!
* \brief Wait until the task is finished.
* \param task_id The task id to be joined.
* \return The results from the runner.
*/
virtual ffi::Array<RunnerResult> JoinRunningTask(int task_id);
/*!
* \brief Jointly tune a given list of tasks.
* \param tasks The tasks to be tuned
* \param task_weights The weight of each task
* \param max_trials_global The maximum number of trials to be performed globally
* \param max_trials_per_task The maximum number of trials to be performed for each task
* \param num_trials_per_iter The number of trials to be performed in each iteration
* \param builder The MetaSchedule builder
* \param runner The MetaSchedule runner
* \param measure_callbacks The callbacks to be called after each measurement
* \param database The database used in tuning
* \param cost_model The cost model used in tuning
*/
virtual void Tune(ffi::Array<TuneContext> tasks, //
ffi::Array<FloatImm> task_weights, //
int max_trials_global, //
int max_trials_per_task, //
int num_trials_per_iter, //
Builder builder, //
Runner runner, //
ffi::Array<MeasureCallback> measure_callbacks, //
ffi::Optional<Database> database, //
ffi::Optional<CostModel> cost_model);
/*!
* \brief Terminate a task
* \param task_id The id of the task to be terminated
*/
void TerminateTask(int task_id);
/*!
* \brief Touch the task and update its status
* \param task_id The task id to be checked.
*/
void TouchTask(int task_id);
/*! \brief Print out a human-readable format of the tuning statistics. */
void PrintTuningStatistics();
static constexpr const bool _type_mutable = true;
TVM_FFI_DECLARE_OBJECT_INFO("s_tir.meta_schedule.TaskScheduler", TaskSchedulerNode, ffi::Object);
};
class TaskScheduler;
/*! \brief The task scheduler with customized methods on the python-side. */
class PyTaskSchedulerNode : public TaskSchedulerNode {
public:
/*!
* \brief The function type of `NextTaskId` method.
* \return The next task id.
*/
using FNextTaskId = ffi::TypedFunction<int()>;
/*!
* \brief The function type of `JoinRunningTask` method.
* \param task_id The task id to be joined.
*/
using FJoinRunningTask = ffi::TypedFunction<ffi::Array<RunnerResult>(int)>;
/*! \brief The function type of `Tune` method. */
using FTune = ffi::TypedFunction<void(ffi::Array<TuneContext> tasks, //
ffi::Array<FloatImm> task_weights, //
int max_trials_global, //
int max_trials_per_task, //
int num_trials_per_iter, //
Builder builder, //
Runner runner, //
ffi::Array<MeasureCallback> measure_callbacks, //
ffi::Optional<Database> database, //
ffi::Optional<CostModel> cost_model)>;
/*! \brief The packed function to the `NextTaskId` function. */
FNextTaskId f_next_task_id;
/*! \brief The packed function to the `JoinRunningTask` function. */
FJoinRunningTask f_join_running_task;
/*! \brief The packed function to the `Tune` function. */
FTune f_tune;
static void RegisterReflection() {
namespace refl = tvm::ffi::reflection;
refl::ObjectDef<PyTaskSchedulerNode>();
}
int NextTaskId() final;
ffi::Array<RunnerResult> JoinRunningTask(int task_id) final;
void Tune(ffi::Array<TuneContext> tasks, ffi::Array<FloatImm> task_weights, int max_trials_global,
int max_trials_per_task, int num_trials_per_iter, Builder builder, Runner runner,
ffi::Array<MeasureCallback> measure_callbacks, ffi::Optional<Database> database,
ffi::Optional<CostModel> cost_model) final;
TVM_FFI_DECLARE_OBJECT_INFO_FINAL("s_tir.meta_schedule.PyTaskScheduler", PyTaskSchedulerNode,
TaskSchedulerNode);
};
/*!
* \brief Managed reference to TaskSchedulerNode.
* \sa TaskSchedulerNode
*/
class TaskScheduler : public ffi::ObjectRef {
public:
explicit TaskScheduler(ffi::ObjectPtr<TaskSchedulerNode> data) : ffi::ObjectRef(data) {
TVM_FFI_ICHECK(data != nullptr);
}
/*!
* \brief Create a task scheduler that fetches tasks in a round-robin fashion.
* \param logger The tuning task's logging function.
* \return The task scheduler created.
*/
TVM_DLL static TaskScheduler RoundRobin(ffi::Function logger);
/*!
* \brief Create a task scheduler that fetches tasks in a gradient based fashion.
* \param logger The tuning task's logging function.
* \param alpha The parameter alpha to control gradient computation.
* \param window_size The parameter to control backward window size.
* \param seed The random seed.
* \return The task scheduler created.
*/
TVM_DLL static TaskScheduler GradientBased(ffi::Function logger, double alpha, int window_size,
LinearCongruentialEngine::TRandState seed);
/*!
* \brief Create a task scheduler with customized methods on the python-side.
* \param logger The tuning task's logging function.
* \param f_next_task_id The packed function of `NextTaskId`.
* \param f_join_running_task The packed function of `JoinRunningTask`.
* \param f_tune The packed function of `Tune`.
* \return The task scheduler created.
*/
TVM_DLL static TaskScheduler PyTaskScheduler(
ffi::Function logger, PyTaskSchedulerNode::FNextTaskId f_next_task_id,
PyTaskSchedulerNode::FJoinRunningTask f_join_running_task, PyTaskSchedulerNode::FTune f_tune);
TVM_FFI_DEFINE_OBJECT_REF_METHODS_NOTNULLABLE(TaskScheduler, ffi::ObjectRef, TaskSchedulerNode);
};
} // namespace meta_schedule
} // namespace s_tir
} // namespace tvm
#endif // TVM_S_TIR_META_SCHEDULE_TASK_SCHEDULER_H_