include/boost/corosio/native/detail/epoll/epoll_scheduler.hpp

85.5% Lines (130/152) 100.0% List of functions (11/11)

[ ] prev/next file n p prev/next uncovered line

epoll_scheduler.hpp

f(x) Functions (11)

Function Calls Lines Blocks

Line	Hits	Source Code
1		//
2		// Copyright (c) 2026 Steve Gerbino
3		// Copyright (c) 2026 Michael Vandeberg
4		//
5		// Distributed under the Boost Software License, Version 1.0. (See accompanying
6		// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
7		//
8		// Official repository: https://github.com/cppalliance/corosio
9		//
10
11		#ifndef BOOST_COROSIO_NATIVE_DETAIL_EPOLL_EPOLL_SCHEDULER_HPP
12		#define BOOST_COROSIO_NATIVE_DETAIL_EPOLL_EPOLL_SCHEDULER_HPP
13
14		#include <boost/corosio/detail/platform.hpp>
15
16		#if BOOST_COROSIO_HAS_EPOLL
17
18		#include <boost/corosio/detail/config.hpp>
19		#include <boost/capy/ex/execution_context.hpp>
20
21		#include <boost/corosio/native/detail/reactor/reactor_scheduler.hpp>
22		#include <boost/corosio/native/detail/reactor/reactor_signal_pipe.hpp>
23
24		#include <boost/corosio/native/detail/epoll/epoll_traits.hpp>
25		#include <boost/corosio/detail/timer_service.hpp>
26		#include <boost/corosio/native/detail/make_err.hpp>
27		#include <boost/corosio/native/detail/posix/posix_resolver_service.hpp>
28		#include <boost/corosio/native/detail/posix/posix_signal_service.hpp>
29		#include <boost/corosio/native/detail/posix/posix_stream_file_service.hpp>
30		#include <boost/corosio/native/detail/posix/posix_random_access_file_service.hpp>
31
32		#include <boost/corosio/detail/except.hpp>
33
34		#include <atomic>
35		#include <chrono>
36		#include <cstdint>
37		#include <mutex>
38		#include <vector>
39
40		#include <errno.h>
41		#include <sys/epoll.h>
42		#include <sys/eventfd.h>
43		#include <sys/timerfd.h>
44		#include <unistd.h>
45
46		namespace boost::corosio::detail {
47
48		/** Linux scheduler using epoll for I/O multiplexing.
49
50		This scheduler implements the scheduler interface using Linux epoll
51		for efficient I/O event notification. It uses a single reactor model
52		where one thread runs epoll_wait while other threads
53		wait on a condition variable for handler work. This design provides:
54
55		- Handler parallelism: N posted handlers can execute on N threads
56		- No thundering herd: condition_variable wakes exactly one thread
57		- IOCP parity: Behavior matches Windows I/O completion port semantics
58
59		When threads call run(), they first try to execute queued handlers.
60		If the queue is empty and no reactor is running, one thread becomes
61		the reactor and runs epoll_wait. Other threads wait on a condition
62		variable until handlers are available.
63
64		@par Thread Safety
65		All public member functions are thread-safe.
66		*/
67		class BOOST_COROSIO_DECL epoll_scheduler final : public reactor_scheduler
68		{
69		public:
70		/** Construct the scheduler.
71
72		Creates an epoll instance, eventfd for reactor interruption,
73		and timerfd for kernel-managed timer expiry.
74
75		@param ctx Reference to the owning execution_context.
76		@param concurrency_hint Hint for expected thread count (unused).
77		*/
78		epoll_scheduler(capy::execution_context& ctx, int concurrency_hint = -1);
79
80		/// Destroy the scheduler.
81		~epoll_scheduler() override;
82
83		epoll_scheduler(epoll_scheduler const&) = delete;
84		epoll_scheduler& operator=(epoll_scheduler const&) = delete;
85
86		/// Shut down the scheduler, draining pending operations.
87		void shutdown() override;
88
89		/// Apply runtime configuration, resizing the event buffer.
90		void configure_reactor(
91		unsigned max_events,
92		unsigned budget_init,
93		unsigned budget_max,
94		unsigned unassisted) override;
95
96		/** Return the epoll file descriptor.
97
98		Used by socket services to register file descriptors
99		for I/O event notification.
100
101		@return The epoll file descriptor.
102		*/
103		int epoll_fd() const noexcept
104		{
105		return epoll_fd_;
106		}
107
108		/** Register a descriptor for persistent monitoring.
109
110		The fd is registered once and stays registered until explicitly
111		deregistered. Events are dispatched via reactor_descriptor_state which
112		tracks pending read/write/connect operations.
113
114		@param fd The file descriptor to register.
115		@param desc Pointer to descriptor data (stored in epoll_event.data.ptr).
116		*/
117		void register_descriptor(int fd, reactor_descriptor_state* desc) const;
118
119		/** Deregister a persistently registered descriptor.
120
121		@param fd The file descriptor to deregister.
122		*/
123		void deregister_descriptor(int fd) const;
124
125		/// Watch the read end of the POSIX signal self-pipe (see scheduler.hpp).
126	41x	void register_signal_reader(int read_fd) override
127		{
128	41x	register_descriptor(read_fd, signal_pipe_reader_.arm());
129	41x	}
130
131		private:
132		void
133		run_task(lock_type& lock, context_type* ctx,
134		long timeout_us) override;
135		void interrupt_reactor() const override;
136		void update_timerfd() const;
137
138		int epoll_fd_;
139		int event_fd_;
140		int timer_fd_;
141
142		// Watches the global signal self-pipe's read end (armed lazily by
143		// register_signal_reader on the first signal registration).
144		reactor_signal_pipe_reader signal_pipe_reader_;
145
146		// Edge-triggered eventfd state
147		mutable std::atomic<bool> eventfd_armed_{false};
148
149		// Set when the earliest timer changes; flushed before epoll_wait
150		mutable std::atomic<bool> timerfd_stale_{false};
151
152		// Event buffer sized from max_events_per_poll_ (set at construction,
153		// resized by configure_reactor via io_context_options).
154		std::vector<epoll_event> event_buffer_;
155		};
156
157	749x	inline epoll_scheduler::epoll_scheduler(capy::execution_context& ctx, int)
158	749x	: epoll_fd_(-1)
159	749x	, event_fd_(-1)
160	749x	, timer_fd_(-1)
161	1498x	, event_buffer_(max_events_per_poll_)
162		{
163	749x	epoll_fd_ = ::epoll_create1(EPOLL_CLOEXEC);
164	749x	if (epoll_fd_ < 0)
165	✗	detail::throw_system_error(make_err(errno), "epoll_create1");
166
167	749x	event_fd_ = ::eventfd(0, EFD_NONBLOCK \| EFD_CLOEXEC);
168	749x	if (event_fd_ < 0)
169		{
170	✗	int errn = errno;
171	✗	::close(epoll_fd_);
172	✗	detail::throw_system_error(make_err(errn), "eventfd");
173		}
174
175	749x	timer_fd_ = ::timerfd_create(CLOCK_MONOTONIC, TFD_NONBLOCK \| TFD_CLOEXEC);
176	749x	if (timer_fd_ < 0)
177		{
178	✗	int errn = errno;
179	✗	::close(event_fd_);
180	✗	::close(epoll_fd_);
181	✗	detail::throw_system_error(make_err(errn), "timerfd_create");
182		}
183
184	749x	epoll_event ev{};
185	749x	ev.events = EPOLLIN \| EPOLLET;
186	749x	ev.data.ptr = nullptr;
187	749x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, event_fd_, &ev) < 0)
188		{
189	✗	int errn = errno;
190	✗	::close(timer_fd_);
191	✗	::close(event_fd_);
192	✗	::close(epoll_fd_);
193	✗	detail::throw_system_error(make_err(errn), "epoll_ctl");
194		}
195
196	749x	epoll_event timer_ev{};
197	749x	timer_ev.events = EPOLLIN \| EPOLLERR;
198	749x	timer_ev.data.ptr = &timer_fd_;
199	749x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, timer_fd_, &timer_ev) < 0)
200		{
201	✗	int errn = errno;
202	✗	::close(timer_fd_);
203	✗	::close(event_fd_);
204	✗	::close(epoll_fd_);
205	✗	detail::throw_system_error(make_err(errn), "epoll_ctl (timerfd)");
206		}
207
208	749x	timer_svc_ = &get_timer_service(ctx, *this);
209	749x	timer_svc_->set_on_earliest_changed(
210	5648x	timer_service::callback(this, [](void* p) {
211	4899x	auto* self = static_cast<epoll_scheduler*>(p);
212	4899x	self->timerfd_stale_.store(true, std::memory_order_release);
213	4899x	self->interrupt_reactor();
214	4899x	}));
215
216	749x	get_resolver_service(ctx, *this);
217	749x	get_signal_service(ctx, *this);
218	749x	get_stream_file_service(ctx, *this);
219	749x	get_random_access_file_service(ctx, *this);
220
221	749x	completed_ops_.push(&task_op_);
222	749x	}
223
224	1498x	inline epoll_scheduler::~epoll_scheduler()
225		{
226	749x	if (timer_fd_ >= 0)
227	749x	::close(timer_fd_);
228	749x	if (event_fd_ >= 0)
229	749x	::close(event_fd_);
230	749x	if (epoll_fd_ >= 0)
231	749x	::close(epoll_fd_);
232	1498x	}
233
234		inline void
235	749x	epoll_scheduler::shutdown()
236		{
237	749x	shutdown_drain();
238
239	749x	if (event_fd_ >= 0)
240	749x	interrupt_reactor();
241	749x	}
242
243		inline void
244	18x	epoll_scheduler::configure_reactor(
245		unsigned max_events,
246		unsigned budget_init,
247		unsigned budget_max,
248		unsigned unassisted)
249		{
250	18x	reactor_scheduler::configure_reactor(
251		max_events, budget_init, budget_max, unassisted);
252	17x	event_buffer_.resize(max_events_per_poll_);
253	17x	}
254
255		inline void
256	10431x	epoll_scheduler::register_descriptor(int fd, reactor_descriptor_state* desc) const
257		{
258	10431x	epoll_event ev{};
259	10431x	ev.events = EPOLLIN \| EPOLLOUT \| EPOLLET \| EPOLLERR \| EPOLLHUP;
260	10431x	ev.data.ptr = desc;
261
262	10431x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, fd, &ev) < 0)
263	✗	detail::throw_system_error(make_err(errno), "epoll_ctl (register)");
264
265	10431x	desc->registered_events = ev.events;
266	10431x	desc->fd = fd;
267	10431x	desc->scheduler_ = this;
268	10431x	desc->mutex.set_enabled(reactor_io_locking_);
269	10431x	desc->ready_events_.store(0, std::memory_order_relaxed);
270
271	10431x	conditionally_enabled_mutex::scoped_lock lock(desc->mutex);
272	10431x	desc->impl_ref_.reset();
273	10431x	desc->read_ready = false;
274	10431x	desc->write_ready = false;
275	10431x	}
276
277		inline void
278	10390x	epoll_scheduler::deregister_descriptor(int fd) const
279		{
280	10390x	::epoll_ctl(epoll_fd_, EPOLL_CTL_DEL, fd, nullptr);
281	10390x	}
282
283		inline void
284	8575x	epoll_scheduler::interrupt_reactor() const
285		{
286	8575x	bool expected = false;
287	8575x	if (eventfd_armed_.compare_exchange_strong(
288		expected, true, std::memory_order_release,
289		std::memory_order_relaxed))
290		{
291	7563x	std::uint64_t val = 1;
292	7563x	[[maybe_unused]] auto r = ::write(event_fd_, &val, sizeof(val));
293		}
294	8575x	}
295
296		inline void
297	8041x	epoll_scheduler::update_timerfd() const
298		{
299	8041x	auto nearest = timer_svc_->nearest_expiry();
300
301	8041x	itimerspec ts{};
302	8041x	int flags = 0;
303
304	8041x	if (nearest == timer_service::time_point::max())
305		{
306		// No timers — disarm by setting to 0 (relative)
307		}
308		else
309		{
310	7908x	auto now = std::chrono::steady_clock::now();
311	7908x	if (nearest <= now)
312		{
313		// Use 1ns instead of 0 — zero disarms the timerfd
314	183x	ts.it_value.tv_nsec = 1;
315		}
316		else
317		{
318	7725x	auto nsec = std::chrono::duration_cast<std::chrono::nanoseconds>(
319	7725x	nearest - now)
320	7725x	.count();
321	7725x	ts.it_value.tv_sec = nsec / 1000000000;
322	7725x	ts.it_value.tv_nsec = nsec % 1000000000;
323	7725x	if (ts.it_value.tv_sec == 0 && ts.it_value.tv_nsec == 0)
324	✗	ts.it_value.tv_nsec = 1;
325		}
326		}
327
328	8041x	if (::timerfd_settime(timer_fd_, flags, &ts, nullptr) < 0)
329	✗	detail::throw_system_error(make_err(errno), "timerfd_settime");
330	8041x	}
331
332		inline void
333	54298x	epoll_scheduler::run_task(
334		lock_type& lock, context_type* ctx, long timeout_us)
335		{
336		int timeout_ms;
337	54298x	if (task_interrupted_)
338	41323x	timeout_ms = 0;
339	12975x	else if (timeout_us < 0)
340	12971x	timeout_ms = -1;
341		else
342	4x	timeout_ms = static_cast<int>((timeout_us + 999) / 1000);
343
344	54298x	if (lock.owns_lock())
345	12977x	lock.unlock();
346
347	54298x	task_cleanup on_exit{this, &lock, ctx};
348
349		// Flush deferred timerfd programming before blocking
350	54298x	if (timerfd_stale_.exchange(false, std::memory_order_acquire))
351	4355x	update_timerfd();
352
353	54298x	int nfds = ::epoll_wait(
354		epoll_fd_, event_buffer_.data(),
355	54298x	static_cast<int>(event_buffer_.size()), timeout_ms);
356
357	54298x	if (nfds < 0 && errno != EINTR)
358	✗	detail::throw_system_error(make_err(errno), "epoll_wait");
359
360	54298x	bool check_timers = false;
361	54298x	ready_queue local_ops;
362
363	112725x	for (int i = 0; i < nfds; ++i)
364		{
365	58427x	if (event_buffer_[i].data.ptr == nullptr)
366		{
367		std::uint64_t val;
368		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
369	6814x	[[maybe_unused]] auto r = ::read(event_fd_, &val, sizeof(val));
370	6814x	eventfd_armed_.store(false, std::memory_order_relaxed);
371	6814x	continue;
372	6814x	}
373
374	51613x	if (event_buffer_[i].data.ptr == &timer_fd_)
375		{
376		std::uint64_t expirations;
377		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
378		[[maybe_unused]] auto r =
379	3686x	::read(timer_fd_, &expirations, sizeof(expirations));
380	3686x	check_timers = true;
381	3686x	continue;
382	3686x	}
383
384		auto* desc =
385	47927x	static_cast<reactor_descriptor_state*>(event_buffer_[i].data.ptr);
386	47927x	desc->add_ready_events(event_buffer_[i].events);
387
388	47927x	bool expected = false;
389	47927x	if (desc->is_enqueued_.compare_exchange_strong(
390		expected, true, std::memory_order_release,
391		std::memory_order_relaxed))
392		{
393	47927x	local_ops.push(desc);
394		}
395		}
396
397	54298x	if (check_timers)
398		{
399	3686x	timer_svc_->process_expired();
400	3686x	update_timerfd();
401		}
402
403	54298x	lock.lock();
404
405	54298x	completed_ops_.splice(local_ops);
406	54298x	}
407
408		} // namespace boost::corosio::detail
409
410		#endif // BOOST_COROSIO_HAS_EPOLL
411
412		#endif // BOOST_COROSIO_NATIVE_DETAIL_EPOLL_EPOLL_SCHEDULER_HPP
413