1
0
Fork 0
mirror of https://gitlab.com/niansa/libcrosscoro.git synced 2025-03-06 20:53:32 +01:00
libcrosscoro/inc/coro/thread_pool.hpp
Josh Baldwin e9b225e42f
io_scheduler inline support (#79)
* io_scheduler inline support

* add debug info for io_scheduler size issue

* move poll info into its own file

* cleanup for feature

* Fix valgrind introduced use after free with inline processing

Running the coroutines inline with event processing caused
a use after free bug with valgrind detected in the inline
tcp server/client benchmark code.  Basically if an event
and a timeout occured in the same time period because the
inline processing would resume _inline_ with the event or the
timeout -- if the timeout and event occured in the same epoll_wait()
function call then the second one's coroutine stackframe would
already be destroyed upon resuming it so the poll_info->processed
check would be reading already free'ed memory.

The solution to this was to introduce a vector of coroutine handles
which are appended into on each epoll_wait() iteration of events
and timeouts, and only then after the events and timeouts are
deduplicated are the coroutine handles resumed.

This new vector has elided a malloc in the timeout function, but
there is still a malloc to extract the poll infos from the timeout
multimap data structure.  The vector is also on the class member
list and is only ever cleared, it is possible with a monster set
of timeouts that this vector could grow extremely large, but
I think that is worth the price of not re-allocating it.
2021-04-11 15:07:01 -06:00

244 lines
8.6 KiB
C++

#pragma once
#include "coro/concepts/range_of.hpp"
#include "coro/event.hpp"
#include "coro/task.hpp"
#include <atomic>
#include <condition_variable>
#include <coroutine>
#include <deque>
#include <functional>
#include <mutex>
#include <optional>
#include <ranges>
#include <thread>
#include <variant>
#include <vector>
namespace coro
{
/**
* Creates a thread pool that executes arbitrary coroutine tasks in a FIFO scheduler policy.
* The thread pool by default will create an execution thread per available core on the system.
*
* When shutting down, either by the thread pool destructing or by manually calling shutdown()
* the thread pool will stop accepting new tasks but will complete all tasks that were scheduled
* prior to the shutdown request.
*/
class thread_pool
{
public:
/**
* An operation is an awaitable type with a coroutine to resume the task scheduled on one of
* the executor threads.
*/
class operation
{
friend class thread_pool;
/**
* Only thread_pools can create operations when a task is being scheduled.
* @param tp The thread pool that created this operation.
*/
explicit operation(thread_pool& tp) noexcept;
public:
/**
* Operations always pause so the executing thread can be switched.
*/
auto await_ready() noexcept -> bool { return false; }
/**
* Suspending always returns to the caller (using void return of await_suspend()) and
* stores the coroutine internally for the executing thread to resume from.
*/
auto await_suspend(std::coroutine_handle<> awaiting_coroutine) noexcept -> void;
/**
* no-op as this is the function called first by the thread pool's executing thread.
*/
auto await_resume() noexcept -> void {}
private:
/// The thread pool that this operation will execute on.
thread_pool& m_thread_pool;
/// The coroutine awaiting execution.
std::coroutine_handle<> m_awaiting_coroutine{nullptr};
};
struct options
{
/// The number of executor threads for this thread pool. Uses the hardware concurrency
/// value by default.
uint32_t thread_count = std::thread::hardware_concurrency();
/// Functor to call on each executor thread upon starting execution. The parameter is the
/// thread's ID assigned to it by the thread pool.
std::function<void(std::size_t)> on_thread_start_functor = nullptr;
/// Functor to call on each executor thread upon stopping execution. The parameter is the
/// thread's ID assigned to it by the thread pool.
std::function<void(std::size_t)> on_thread_stop_functor = nullptr;
};
/**
* @param opts Thread pool configuration options.
*/
explicit thread_pool(
options opts = options{
.thread_count = std::thread::hardware_concurrency(),
.on_thread_start_functor = nullptr,
.on_thread_stop_functor = nullptr});
thread_pool(const thread_pool&) = delete;
thread_pool(thread_pool&&) = delete;
auto operator=(const thread_pool&) -> thread_pool& = delete;
auto operator=(thread_pool&&) -> thread_pool& = delete;
virtual ~thread_pool();
/**
* @return The number of executor threads for processing tasks.
*/
auto thread_count() const noexcept -> uint32_t { return m_threads.size(); }
/**
* Schedules the currently executing coroutine to be run on this thread pool. This must be
* called from within the coroutines function body to schedule the coroutine on the thread pool.
* @throw std::runtime_error If the thread pool is `shutdown()` scheduling new tasks is not permitted.
* @return The operation to switch from the calling scheduling thread to the executor thread
* pool thread.
*/
[[nodiscard]] auto schedule() -> operation;
/**
* @throw std::runtime_error If the thread pool is `shutdown()` scheduling new tasks is not permitted.
* @param f The function to execute on the thread pool.
* @param args The arguments to call the functor with.
* @return A task that wraps the given functor to be executed on the thread pool.
*/
template<typename functor, typename... arguments>
[[nodiscard]] auto schedule(functor&& f, arguments... args) -> task<decltype(f(std::forward<arguments>(args)...))>
{
co_await schedule();
if constexpr (std::is_same_v<void, decltype(f(std::forward<arguments>(args)...))>)
{
f(std::forward<arguments>(args)...);
co_return;
}
else
{
co_return f(std::forward<arguments>(args)...);
}
}
/**
* Schedules any coroutine handle that is ready to be resumed.
* @param handle The coroutine handle to schedule.
*/
auto resume(std::coroutine_handle<> handle) noexcept -> void;
/**
* Schedules the set of coroutine handles that are ready to be resumed.
* @param handles The coroutine handles to schedule.
*/
template<coro::concepts::range_of<std::coroutine_handle<>> range_type>
auto resume(const range_type& handles) noexcept -> void
{
m_size.fetch_add(std::size(handles), std::memory_order::release);
size_t null_handles{0};
{
std::scoped_lock lk{m_wait_mutex};
for (const auto& handle : handles)
{
if (handle != nullptr) [[likely]]
{
m_queue.emplace_back(handle);
}
else
{
++null_handles;
}
}
}
if (null_handles > 0)
{
m_size.fetch_sub(null_handles, std::memory_order::release);
}
m_wait_cv.notify_one();
}
/**
* Immediately yields the current task and places it at the end of the queue of tasks waiting
* to be processed. This will immediately be picked up again once it naturally goes through the
* FIFO task queue. This function is useful to yielding long processing tasks to let other tasks
* get processing time.
*/
[[nodiscard]] auto yield() -> operation { return schedule(); }
/**
* Shutsdown the thread pool. This will finish any tasks scheduled prior to calling this
* function but will prevent the thread pool from scheduling any new tasks. This call is
* blocking and will wait until all inflight tasks are completed before returnin.
*/
auto shutdown() noexcept -> void;
/**
* @return The number of tasks waiting in the task queue + the executing tasks.
*/
auto size() const noexcept -> std::size_t { return m_size.load(std::memory_order::acquire); }
/**
* @return True if the task queue is empty and zero tasks are currently executing.
*/
auto empty() const noexcept -> bool { return size() == 0; }
/**
* @return The number of tasks waiting in the task queue to be executed.
*/
auto queue_size() const noexcept -> std::size_t
{
// Might not be totally perfect but good enough, avoids acquiring the lock for now.
std::atomic_thread_fence(std::memory_order::acquire);
return m_queue.size();
}
/**
* @return True if the task queue is currently empty.
*/
auto queue_empty() const noexcept -> bool { return queue_size() == 0; }
private:
/// The configuration options.
options m_opts;
/// The background executor threads.
std::vector<std::jthread> m_threads;
/// Mutex for executor threads to sleep on the condition variable.
std::mutex m_wait_mutex;
/// Condition variable for each executor thread to wait on when no tasks are available.
std::condition_variable_any m_wait_cv;
/// FIFO queue of tasks waiting to be executed.
std::deque<std::coroutine_handle<>> m_queue;
/**
* Each background thread runs from this function.
* @param stop_token Token which signals when shutdown() has been called.
* @param idx The executor's idx for internal data structure accesses.
*/
auto executor(std::stop_token stop_token, std::size_t idx) -> void;
/**
* @param handle Schedules the given coroutine to be executed upon the first available thread.
*/
auto schedule_impl(std::coroutine_handle<> handle) noexcept -> void;
/// The number of tasks in the queue + currently executing.
std::atomic<std::size_t> m_size{0};
/// Has the thread pool been requested to shut down?
std::atomic<bool> m_shutdown_requested{false};
};
} // namespace coro