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	638x	inline epoll_scheduler::epoll_scheduler(capy::execution_context& ctx, int)
158	638x	: epoll_fd_(-1)
159	638x	, event_fd_(-1)
160	638x	, timer_fd_(-1)
161	1276x	, event_buffer_(max_events_per_poll_)
162		{
163	638x	epoll_fd_ = ::epoll_create1(EPOLL_CLOEXEC);
164	638x	if (epoll_fd_ < 0)
165	✗	detail::throw_system_error(make_err(errno), "epoll_create1");
166
167	638x	event_fd_ = ::eventfd(0, EFD_NONBLOCK \| EFD_CLOEXEC);
168	638x	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	638x	timer_fd_ = ::timerfd_create(CLOCK_MONOTONIC, TFD_NONBLOCK \| TFD_CLOEXEC);
176	638x	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	638x	epoll_event ev{};
185	638x	ev.events = EPOLLIN \| EPOLLET;
186	638x	ev.data.ptr = nullptr;
187	638x	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	638x	epoll_event timer_ev{};
197	638x	timer_ev.events = EPOLLIN \| EPOLLERR;
198	638x	timer_ev.data.ptr = &timer_fd_;
199	638x	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	638x	timer_svc_ = &get_timer_service(ctx, *this);
209	638x	timer_svc_->set_on_earliest_changed(
210	5313x	timer_service::callback(this, [](void* p) {
211	4675x	auto* self = static_cast<epoll_scheduler*>(p);
212	4675x	self->timerfd_stale_.store(true, std::memory_order_release);
213	4675x	self->interrupt_reactor();
214	4675x	}));
215
216	638x	get_resolver_service(ctx, *this);
217	638x	get_signal_service(ctx, *this);
218	638x	get_stream_file_service(ctx, *this);
219	638x	get_random_access_file_service(ctx, *this);
220
221	638x	completed_ops_.push(&task_op_);
222	638x	}
223
224	1276x	inline epoll_scheduler::~epoll_scheduler()
225		{
226	638x	if (timer_fd_ >= 0)
227	638x	::close(timer_fd_);
228	638x	if (event_fd_ >= 0)
229	638x	::close(event_fd_);
230	638x	if (epoll_fd_ >= 0)
231	638x	::close(epoll_fd_);
232	1276x	}
233
234		inline void
235	638x	epoll_scheduler::shutdown()
236		{
237	638x	shutdown_drain();
238
239	638x	if (event_fd_ >= 0)
240	638x	interrupt_reactor();
241	638x	}
242
243		inline void
244	16x	epoll_scheduler::configure_reactor(
245		unsigned max_events,
246		unsigned budget_init,
247		unsigned budget_max,
248		unsigned unassisted)
249		{
250	16x	reactor_scheduler::configure_reactor(
251		max_events, budget_init, budget_max, unassisted);
252	15x	event_buffer_.resize(max_events_per_poll_);
253	15x	}
254
255		inline void
256	8025x	epoll_scheduler::register_descriptor(int fd, reactor_descriptor_state* desc) const
257		{
258	8025x	epoll_event ev{};
259	8025x	ev.events = EPOLLIN \| EPOLLOUT \| EPOLLET \| EPOLLERR \| EPOLLHUP;
260	8025x	ev.data.ptr = desc;
261
262	8025x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, fd, &ev) < 0)
263	✗	detail::throw_system_error(make_err(errno), "epoll_ctl (register)");
264
265	8025x	desc->registered_events = ev.events;
266	8025x	desc->fd = fd;
267	8025x	desc->scheduler_ = this;
268	8025x	desc->mutex.set_enabled(reactor_io_locking_);
269	8025x	desc->ready_events_.store(0, std::memory_order_relaxed);
270
271	8025x	conditionally_enabled_mutex::scoped_lock lock(desc->mutex);
272	8025x	desc->impl_ref_.reset();
273	8025x	desc->read_ready = false;
274	8025x	desc->write_ready = false;
275	8025x	}
276
277		inline void
278	7984x	epoll_scheduler::deregister_descriptor(int fd) const
279		{
280	7984x	::epoll_ctl(epoll_fd_, EPOLL_CTL_DEL, fd, nullptr);
281	7984x	}
282
283		inline void
284	5821x	epoll_scheduler::interrupt_reactor() const
285		{
286	5821x	bool expected = false;
287	5821x	if (eventfd_armed_.compare_exchange_strong(
288		expected, true, std::memory_order_release,
289		std::memory_order_relaxed))
290		{
291	4802x	std::uint64_t val = 1;
292	4802x	[[maybe_unused]] auto r = ::write(event_fd_, &val, sizeof(val));
293		}
294	5821x	}
295
296		inline void
297	8030x	epoll_scheduler::update_timerfd() const
298		{
299	8030x	auto nearest = timer_svc_->nearest_expiry();
300
301	8030x	itimerspec ts{};
302	8030x	int flags = 0;
303
304	8030x	if (nearest == timer_service::time_point::max())
305		{
306		// No timers — disarm by setting to 0 (relative)
307		}
308		else
309		{
310	7897x	auto now = std::chrono::steady_clock::now();
311	7897x	if (nearest <= now)
312		{
313		// Use 1ns instead of 0 — zero disarms the timerfd
314	530x	ts.it_value.tv_nsec = 1;
315		}
316		else
317		{
318	7367x	auto nsec = std::chrono::duration_cast<std::chrono::nanoseconds>(
319	7367x	nearest - now)
320	7367x	.count();
321	7367x	ts.it_value.tv_sec = nsec / 1000000000;
322	7367x	ts.it_value.tv_nsec = nsec % 1000000000;
323	7367x	if (ts.it_value.tv_sec == 0 && ts.it_value.tv_nsec == 0)
324	✗	ts.it_value.tv_nsec = 1;
325		}
326		}
327
328	8030x	if (::timerfd_settime(timer_fd_, flags, &ts, nullptr) < 0)
329	✗	detail::throw_system_error(make_err(errno), "timerfd_settime");
330	8030x	}
331
332		inline void
333	32301x	epoll_scheduler::run_task(
334		lock_type& lock, context_type* ctx, long timeout_us)
335		{
336		int timeout_ms;
337	32301x	if (task_interrupted_)
338	21376x	timeout_ms = 0;
339	10925x	else if (timeout_us < 0)
340	10921x	timeout_ms = -1;
341		else
342	4x	timeout_ms = static_cast<int>((timeout_us + 999) / 1000);
343
344	32301x	if (lock.owns_lock())
345	10927x	lock.unlock();
346
347	32301x	task_cleanup on_exit{this, &lock, ctx};
348
349		// Flush deferred timerfd programming before blocking
350	32301x	if (timerfd_stale_.exchange(false, std::memory_order_acquire))
351	4020x	update_timerfd();
352
353	32301x	int nfds = ::epoll_wait(
354		epoll_fd_, event_buffer_.data(),
355	32301x	static_cast<int>(event_buffer_.size()), timeout_ms);
356
357	32301x	if (nfds < 0 && errno != EINTR)
358	✗	detail::throw_system_error(make_err(errno), "epoll_wait");
359
360	32301x	bool check_timers = false;
361	32301x	ready_queue local_ops;
362
363	73939x	for (int i = 0; i < nfds; ++i)
364		{
365	41638x	if (event_buffer_[i].data.ptr == nullptr)
366		{
367		std::uint64_t val;
368		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
369	4164x	[[maybe_unused]] auto r = ::read(event_fd_, &val, sizeof(val));
370	4164x	eventfd_armed_.store(false, std::memory_order_relaxed);
371	4164x	continue;
372	4164x	}
373
374	37474x	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	4010x	::read(timer_fd_, &expirations, sizeof(expirations));
380	4010x	check_timers = true;
381	4010x	continue;
382	4010x	}
383
384		auto* desc =
385	33464x	static_cast<reactor_descriptor_state*>(event_buffer_[i].data.ptr);
386	33464x	desc->add_ready_events(event_buffer_[i].events);
387
388	33464x	bool expected = false;
389	33464x	if (desc->is_enqueued_.compare_exchange_strong(
390		expected, true, std::memory_order_release,
391		std::memory_order_relaxed))
392		{
393	33464x	local_ops.push(desc);
394		}
395		}
396
397	32301x	if (check_timers)
398		{
399	4010x	timer_svc_->process_expired();
400	4010x	update_timerfd();
401		}
402
403	32301x	lock.lock();
404
405	32301x	completed_ops_.splice(local_ops);
406	32301x	}
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