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	51x	void register_signal_reader(int read_fd) override
127		{
128	51x	register_descriptor(read_fd, signal_pipe_reader_.arm());
129	51x	}
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	791x	inline epoll_scheduler::epoll_scheduler(capy::execution_context& ctx, int)
158	791x	: epoll_fd_(-1)
159	791x	, event_fd_(-1)
160	791x	, timer_fd_(-1)
161	1582x	, event_buffer_(max_events_per_poll_)
162		{
163	791x	epoll_fd_ = ::epoll_create1(EPOLL_CLOEXEC);
164	791x	if (epoll_fd_ < 0)
165	✗	detail::throw_system_error(make_err(errno), "epoll_create1");
166
167	791x	event_fd_ = ::eventfd(0, EFD_NONBLOCK \| EFD_CLOEXEC);
168	791x	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	791x	timer_fd_ = ::timerfd_create(CLOCK_MONOTONIC, TFD_NONBLOCK \| TFD_CLOEXEC);
176	791x	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	791x	epoll_event ev{};
185	791x	ev.events = EPOLLIN \| EPOLLET;
186	791x	ev.data.ptr = nullptr;
187	791x	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	791x	epoll_event timer_ev{};
197	791x	timer_ev.events = EPOLLIN \| EPOLLERR;
198	791x	timer_ev.data.ptr = &timer_fd_;
199	791x	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	791x	timer_svc_ = &get_timer_service(ctx, *this);
209	791x	timer_svc_->set_on_earliest_changed(
210	3834x	timer_service::callback(this, [](void* p) {
211	3043x	auto* self = static_cast<epoll_scheduler*>(p);
212	3043x	self->timerfd_stale_.store(true, std::memory_order_release);
213	3043x	self->interrupt_reactor();
214	3043x	}));
215
216	791x	get_resolver_service(ctx, *this);
217	791x	get_signal_service(ctx, *this);
218	791x	get_stream_file_service(ctx, *this);
219	791x	get_random_access_file_service(ctx, *this);
220
221	791x	completed_ops_.push(&task_op_);
222	791x	}
223
224	1582x	inline epoll_scheduler::~epoll_scheduler()
225		{
226	791x	if (timer_fd_ >= 0)
227	791x	::close(timer_fd_);
228	791x	if (event_fd_ >= 0)
229	791x	::close(event_fd_);
230	791x	if (epoll_fd_ >= 0)
231	791x	::close(epoll_fd_);
232	1582x	}
233
234		inline void
235	791x	epoll_scheduler::shutdown()
236		{
237	791x	shutdown_drain();
238
239	791x	if (event_fd_ >= 0)
240	791x	interrupt_reactor();
241	791x	}
242
243		inline void
244	19x	epoll_scheduler::configure_reactor(
245		unsigned max_events,
246		unsigned budget_init,
247		unsigned budget_max,
248		unsigned unassisted)
249		{
250	19x	reactor_scheduler::configure_reactor(
251		max_events, budget_init, budget_max, unassisted);
252	18x	event_buffer_.resize(max_events_per_poll_);
253	18x	}
254
255		inline void
256	5267x	epoll_scheduler::register_descriptor(int fd, reactor_descriptor_state* desc) const
257		{
258	5267x	epoll_event ev{};
259	5267x	ev.events = EPOLLIN \| EPOLLOUT \| EPOLLET \| EPOLLERR \| EPOLLHUP;
260	5267x	ev.data.ptr = desc;
261
262	5267x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, fd, &ev) < 0)
263	✗	detail::throw_system_error(make_err(errno), "epoll_ctl (register)");
264
265	5267x	desc->registered_events = ev.events;
266	5267x	desc->fd = fd;
267	5267x	desc->scheduler_ = this;
268	5267x	desc->mutex.set_enabled(reactor_io_locking_);
269	5267x	desc->ready_events_.store(0, std::memory_order_relaxed);
270
271	5267x	conditionally_enabled_mutex::scoped_lock lock(desc->mutex);
272	5267x	desc->impl_ref_.reset();
273	5267x	desc->read_ready = false;
274	5267x	desc->write_ready = false;
275	5267x	}
276
277		inline void
278	5216x	epoll_scheduler::deregister_descriptor(int fd) const
279		{
280	5216x	::epoll_ctl(epoll_fd_, EPOLL_CTL_DEL, fd, nullptr);
281	5216x	}
282
283		inline void
284	4544x	epoll_scheduler::interrupt_reactor() const
285		{
286	4544x	bool expected = false;
287	4544x	if (eventfd_armed_.compare_exchange_strong(
288		expected, true, std::memory_order_release,
289		std::memory_order_relaxed))
290		{
291	3543x	std::uint64_t val = 1;
292	3543x	[[maybe_unused]] auto r = ::write(event_fd_, &val, sizeof(val));
293		}
294	4544x	}
295
296		inline void
297	5066x	epoll_scheduler::update_timerfd() const
298		{
299	5066x	auto nearest = timer_svc_->nearest_expiry();
300
301	5066x	itimerspec ts{};
302	5066x	int flags = 0;
303
304	5066x	if (nearest == timer_service::time_point::max())
305		{
306		// No timers — disarm by setting to 0 (relative)
307		}
308		else
309		{
310	4916x	auto now = std::chrono::steady_clock::now();
311	4916x	if (nearest <= now)
312		{
313		// Use 1ns instead of 0 — zero disarms the timerfd
314	336x	ts.it_value.tv_nsec = 1;
315		}
316		else
317		{
318	4580x	auto nsec = std::chrono::duration_cast<std::chrono::nanoseconds>(
319	4580x	nearest - now)
320	4580x	.count();
321	4580x	ts.it_value.tv_sec = nsec / 1000000000;
322	4580x	ts.it_value.tv_nsec = nsec % 1000000000;
323	4580x	if (ts.it_value.tv_sec == 0 && ts.it_value.tv_nsec == 0)
324	✗	ts.it_value.tv_nsec = 1;
325		}
326		}
327
328	5066x	if (::timerfd_settime(timer_fd_, flags, &ts, nullptr) < 0)
329	✗	detail::throw_system_error(make_err(errno), "timerfd_settime");
330	5066x	}
331
332		inline void
333	28788x	epoll_scheduler::run_task(
334		lock_type& lock, context_type* ctx, long timeout_us)
335		{
336		int timeout_ms;
337	28788x	if (task_interrupted_)
338	22095x	timeout_ms = 0;
339	6693x	else if (timeout_us < 0)
340	6689x	timeout_ms = -1;
341		else
342	4x	timeout_ms = static_cast<int>((timeout_us + 999) / 1000);
343
344	28788x	if (lock.owns_lock())
345	6695x	lock.unlock();
346
347	28788x	task_cleanup on_exit{this, &lock, ctx};
348
349		// Flush deferred timerfd programming before blocking
350	28788x	if (timerfd_stale_.exchange(false, std::memory_order_acquire))
351	2540x	update_timerfd();
352
353	28788x	int nfds = ::epoll_wait(
354		epoll_fd_, event_buffer_.data(),
355	28788x	static_cast<int>(event_buffer_.size()), timeout_ms);
356
357	28788x	if (nfds < 0 && errno != EINTR)
358	✗	detail::throw_system_error(make_err(errno), "epoll_wait");
359
360	28788x	bool check_timers = false;
361	28788x	ready_queue local_ops;
362
363	67225x	for (int i = 0; i < nfds; ++i)
364		{
365	38437x	if (event_buffer_[i].data.ptr == nullptr)
366		{
367		std::uint64_t val;
368		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
369	2752x	[[maybe_unused]] auto r = ::read(event_fd_, &val, sizeof(val));
370	2752x	eventfd_armed_.store(false, std::memory_order_relaxed);
371	2752x	continue;
372	2752x	}
373
374	35685x	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	2526x	::read(timer_fd_, &expirations, sizeof(expirations));
380	2526x	check_timers = true;
381	2526x	continue;
382	2526x	}
383
384		auto* desc =
385	33159x	static_cast<reactor_descriptor_state*>(event_buffer_[i].data.ptr);
386	33159x	desc->add_ready_events(event_buffer_[i].events);
387
388	33159x	bool expected = false;
389	33159x	if (desc->is_enqueued_.compare_exchange_strong(
390		expected, true, std::memory_order_release,
391		std::memory_order_relaxed))
392		{
393	33159x	local_ops.push(desc);
394		}
395		}
396
397	28788x	if (check_timers)
398		{
399	2526x	timer_svc_->process_expired();
400	2526x	update_timerfd();
401		}
402
403	28788x	lock.lock();
404
405	28788x	completed_ops_.splice(local_ops);
406	28788x	}
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