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

85.3% Lines (128/150) 100.0% List of functions (10/10)

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

epoll_scheduler.hpp

f(x) Functions (10)

Function Calls Lines Blocks

Line	Hits	Source Code
1		//
2		// Copyright (c) 2026 Steve Gerbino
3		//
4		// Distributed under the Boost Software License, Version 1.0. (See accompanying
5		// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
6		//
7		// Official repository: https://github.com/cppalliance/corosio
8		//
9
10		#ifndef BOOST_COROSIO_NATIVE_DETAIL_EPOLL_EPOLL_SCHEDULER_HPP
11		#define BOOST_COROSIO_NATIVE_DETAIL_EPOLL_EPOLL_SCHEDULER_HPP
12
13		#include <boost/corosio/detail/platform.hpp>
14
15		#if BOOST_COROSIO_HAS_EPOLL
16
17		#include <boost/corosio/detail/config.hpp>
18		#include <boost/capy/ex/execution_context.hpp>
19
20		#include <boost/corosio/native/detail/reactor/reactor_scheduler.hpp>
21
22		#include <boost/corosio/native/detail/epoll/epoll_traits.hpp>
23		#include <boost/corosio/detail/timer_service.hpp>
24		#include <boost/corosio/native/detail/make_err.hpp>
25		#include <boost/corosio/native/detail/posix/posix_resolver_service.hpp>
26		#include <boost/corosio/native/detail/posix/posix_signal_service.hpp>
27		#include <boost/corosio/native/detail/posix/posix_stream_file_service.hpp>
28		#include <boost/corosio/native/detail/posix/posix_random_access_file_service.hpp>
29
30		#include <boost/corosio/detail/except.hpp>
31
32		#include <atomic>
33		#include <chrono>
34		#include <cstdint>
35		#include <mutex>
36		#include <vector>
37
38		#include <errno.h>
39		#include <sys/epoll.h>
40		#include <sys/eventfd.h>
41		#include <sys/timerfd.h>
42		#include <unistd.h>
43
44		namespace boost::corosio::detail {
45
46		/** Linux scheduler using epoll for I/O multiplexing.
47
48		This scheduler implements the scheduler interface using Linux epoll
49		for efficient I/O event notification. It uses a single reactor model
50		where one thread runs epoll_wait while other threads
51		wait on a condition variable for handler work. This design provides:
52
53		- Handler parallelism: N posted handlers can execute on N threads
54		- No thundering herd: condition_variable wakes exactly one thread
55		- IOCP parity: Behavior matches Windows I/O completion port semantics
56
57		When threads call run(), they first try to execute queued handlers.
58		If the queue is empty and no reactor is running, one thread becomes
59		the reactor and runs epoll_wait. Other threads wait on a condition
60		variable until handlers are available.
61
62		@par Thread Safety
63		All public member functions are thread-safe.
64		*/
65		class BOOST_COROSIO_DECL epoll_scheduler final : public reactor_scheduler
66		{
67		public:
68		/** Construct the scheduler.
69
70		Creates an epoll instance, eventfd for reactor interruption,
71		and timerfd for kernel-managed timer expiry.
72
73		@param ctx Reference to the owning execution_context.
74		@param concurrency_hint Hint for expected thread count (unused).
75		*/
76		epoll_scheduler(capy::execution_context& ctx, int concurrency_hint = -1);
77
78		/// Destroy the scheduler.
79		~epoll_scheduler() override;
80
81		epoll_scheduler(epoll_scheduler const&) = delete;
82		epoll_scheduler& operator=(epoll_scheduler const&) = delete;
83
84		/// Shut down the scheduler, draining pending operations.
85		void shutdown() override;
86
87		/// Apply runtime configuration, resizing the event buffer.
88		void configure_reactor(
89		unsigned max_events,
90		unsigned budget_init,
91		unsigned budget_max,
92		unsigned unassisted) override;
93
94		/** Return the epoll file descriptor.
95
96		Used by socket services to register file descriptors
97		for I/O event notification.
98
99		@return The epoll file descriptor.
100		*/
101		int epoll_fd() const noexcept
102		{
103		return epoll_fd_;
104		}
105
106		/** Register a descriptor for persistent monitoring.
107
108		The fd is registered once and stays registered until explicitly
109		deregistered. Events are dispatched via reactor_descriptor_state which
110		tracks pending read/write/connect operations.
111
112		@param fd The file descriptor to register.
113		@param desc Pointer to descriptor data (stored in epoll_event.data.ptr).
114		*/
115		void register_descriptor(int fd, reactor_descriptor_state* desc) const;
116
117		/** Deregister a persistently registered descriptor.
118
119		@param fd The file descriptor to deregister.
120		*/
121		void deregister_descriptor(int fd) const;
122
123		private:
124		void
125		run_task(lock_type& lock, context_type* ctx,
126		long timeout_us) override;
127		void interrupt_reactor() const override;
128		void update_timerfd() const;
129
130		int epoll_fd_;
131		int event_fd_;
132		int timer_fd_;
133
134		// Edge-triggered eventfd state
135		mutable std::atomic<bool> eventfd_armed_{false};
136
137		// Set when the earliest timer changes; flushed before epoll_wait
138		mutable std::atomic<bool> timerfd_stale_{false};
139
140		// Event buffer sized from max_events_per_poll_ (set at construction,
141		// resized by configure_reactor via io_context_options).
142		std::vector<epoll_event> event_buffer_;
143		};
144
145	593x	inline epoll_scheduler::epoll_scheduler(capy::execution_context& ctx, int)
146	593x	: epoll_fd_(-1)
147	593x	, event_fd_(-1)
148	593x	, timer_fd_(-1)
149	1186x	, event_buffer_(max_events_per_poll_)
150		{
151	593x	epoll_fd_ = ::epoll_create1(EPOLL_CLOEXEC);
152	593x	if (epoll_fd_ < 0)
153	✗	detail::throw_system_error(make_err(errno), "epoll_create1");
154
155	593x	event_fd_ = ::eventfd(0, EFD_NONBLOCK \| EFD_CLOEXEC);
156	593x	if (event_fd_ < 0)
157		{
158	✗	int errn = errno;
159	✗	::close(epoll_fd_);
160	✗	detail::throw_system_error(make_err(errn), "eventfd");
161		}
162
163	593x	timer_fd_ = ::timerfd_create(CLOCK_MONOTONIC, TFD_NONBLOCK \| TFD_CLOEXEC);
164	593x	if (timer_fd_ < 0)
165		{
166	✗	int errn = errno;
167	✗	::close(event_fd_);
168	✗	::close(epoll_fd_);
169	✗	detail::throw_system_error(make_err(errn), "timerfd_create");
170		}
171
172	593x	epoll_event ev{};
173	593x	ev.events = EPOLLIN \| EPOLLET;
174	593x	ev.data.ptr = nullptr;
175	593x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, event_fd_, &ev) < 0)
176		{
177	✗	int errn = errno;
178	✗	::close(timer_fd_);
179	✗	::close(event_fd_);
180	✗	::close(epoll_fd_);
181	✗	detail::throw_system_error(make_err(errn), "epoll_ctl");
182		}
183
184	593x	epoll_event timer_ev{};
185	593x	timer_ev.events = EPOLLIN \| EPOLLERR;
186	593x	timer_ev.data.ptr = &timer_fd_;
187	593x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, timer_fd_, &timer_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 (timerfd)");
194		}
195
196	593x	timer_svc_ = &get_timer_service(ctx, *this);
197	593x	timer_svc_->set_on_earliest_changed(
198	2550x	timer_service::callback(this, [](void* p) {
199	1957x	auto* self = static_cast<epoll_scheduler*>(p);
200	1957x	self->timerfd_stale_.store(true, std::memory_order_release);
201	1957x	self->interrupt_reactor();
202	1957x	}));
203
204	593x	get_resolver_service(ctx, *this);
205	593x	get_signal_service(ctx, *this);
206	593x	get_stream_file_service(ctx, *this);
207	593x	get_random_access_file_service(ctx, *this);
208
209	593x	completed_ops_.push(&task_op_);
210	593x	}
211
212	1186x	inline epoll_scheduler::~epoll_scheduler()
213		{
214	593x	if (timer_fd_ >= 0)
215	593x	::close(timer_fd_);
216	593x	if (event_fd_ >= 0)
217	593x	::close(event_fd_);
218	593x	if (epoll_fd_ >= 0)
219	593x	::close(epoll_fd_);
220	1186x	}
221
222		inline void
223	593x	epoll_scheduler::shutdown()
224		{
225	593x	shutdown_drain();
226
227	593x	if (event_fd_ >= 0)
228	593x	interrupt_reactor();
229	593x	}
230
231		inline void
232	8x	epoll_scheduler::configure_reactor(
233		unsigned max_events,
234		unsigned budget_init,
235		unsigned budget_max,
236		unsigned unassisted)
237		{
238	8x	reactor_scheduler::configure_reactor(
239		max_events, budget_init, budget_max, unassisted);
240	7x	event_buffer_.resize(max_events_per_poll_);
241	7x	}
242
243		inline void
244	3793x	epoll_scheduler::register_descriptor(int fd, reactor_descriptor_state* desc) const
245		{
246	3793x	epoll_event ev{};
247	3793x	ev.events = EPOLLIN \| EPOLLOUT \| EPOLLET \| EPOLLERR \| EPOLLHUP;
248	3793x	ev.data.ptr = desc;
249
250	3793x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, fd, &ev) < 0)
251	✗	detail::throw_system_error(make_err(errno), "epoll_ctl (register)");
252
253	3793x	desc->registered_events = ev.events;
254	3793x	desc->fd = fd;
255	3793x	desc->scheduler_ = this;
256	3793x	desc->mutex.set_enabled(!single_threaded_);
257	3793x	desc->ready_events_.store(0, std::memory_order_relaxed);
258
259	3793x	conditionally_enabled_mutex::scoped_lock lock(desc->mutex);
260	3793x	desc->impl_ref_.reset();
261	3793x	desc->read_ready = false;
262	3793x	desc->write_ready = false;
263	3793x	}
264
265		inline void
266	3793x	epoll_scheduler::deregister_descriptor(int fd) const
267		{
268	3793x	::epoll_ctl(epoll_fd_, EPOLL_CTL_DEL, fd, nullptr);
269	3793x	}
270
271		inline void
272	3007x	epoll_scheduler::interrupt_reactor() const
273		{
274	3007x	bool expected = false;
275	3007x	if (eventfd_armed_.compare_exchange_strong(
276		expected, true, std::memory_order_release,
277		std::memory_order_relaxed))
278		{
279	2660x	std::uint64_t val = 1;
280	2660x	[[maybe_unused]] auto r = ::write(event_fd_, &val, sizeof(val));
281		}
282	3007x	}
283
284		inline void
285	3870x	epoll_scheduler::update_timerfd() const
286		{
287	3870x	auto nearest = timer_svc_->nearest_expiry();
288
289	3870x	itimerspec ts{};
290	3870x	int flags = 0;
291
292	3870x	if (nearest == timer_service::time_point::max())
293		{
294		// No timers — disarm by setting to 0 (relative)
295		}
296		else
297		{
298	3773x	auto now = std::chrono::steady_clock::now();
299	3773x	if (nearest <= now)
300		{
301		// Use 1ns instead of 0 — zero disarms the timerfd
302	613x	ts.it_value.tv_nsec = 1;
303		}
304		else
305		{
306	3160x	auto nsec = std::chrono::duration_cast<std::chrono::nanoseconds>(
307	3160x	nearest - now)
308	3160x	.count();
309	3160x	ts.it_value.tv_sec = nsec / 1000000000;
310	3160x	ts.it_value.tv_nsec = nsec % 1000000000;
311	3160x	if (ts.it_value.tv_sec == 0 && ts.it_value.tv_nsec == 0)
312	✗	ts.it_value.tv_nsec = 1;
313		}
314		}
315
316	3870x	if (::timerfd_settime(timer_fd_, flags, &ts, nullptr) < 0)
317	✗	detail::throw_system_error(make_err(errno), "timerfd_settime");
318	3870x	}
319
320		inline void
321	30490x	epoll_scheduler::run_task(
322		lock_type& lock, context_type* ctx, long timeout_us)
323		{
324		int timeout_ms;
325	30490x	if (task_interrupted_)
326	25717x	timeout_ms = 0;
327	4773x	else if (timeout_us < 0)
328	4770x	timeout_ms = -1;
329		else
330	3x	timeout_ms = static_cast<int>((timeout_us + 999) / 1000);
331
332	30490x	if (lock.owns_lock())
333	4773x	lock.unlock();
334
335	30490x	task_cleanup on_exit{this, &lock, ctx};
336
337		// Flush deferred timerfd programming before blocking
338	30490x	if (timerfd_stale_.exchange(false, std::memory_order_acquire))
339	1933x	update_timerfd();
340
341	30490x	int nfds = ::epoll_wait(
342		epoll_fd_, event_buffer_.data(),
343	30490x	static_cast<int>(event_buffer_.size()), timeout_ms);
344
345	30490x	if (nfds < 0 && errno != EINTR)
346	✗	detail::throw_system_error(make_err(errno), "epoll_wait");
347
348	30490x	bool check_timers = false;
349	30490x	op_queue local_ops;
350
351	72264x	for (int i = 0; i < nfds; ++i)
352		{
353	41774x	if (event_buffer_[i].data.ptr == nullptr)
354		{
355		std::uint64_t val;
356		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
357	2067x	[[maybe_unused]] auto r = ::read(event_fd_, &val, sizeof(val));
358	2067x	eventfd_armed_.store(false, std::memory_order_relaxed);
359	2067x	continue;
360	2067x	}
361
362	39707x	if (event_buffer_[i].data.ptr == &timer_fd_)
363		{
364		std::uint64_t expirations;
365		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
366		[[maybe_unused]] auto r =
367	1937x	::read(timer_fd_, &expirations, sizeof(expirations));
368	1937x	check_timers = true;
369	1937x	continue;
370	1937x	}
371
372		auto* desc =
373	37770x	static_cast<reactor_descriptor_state*>(event_buffer_[i].data.ptr);
374	37770x	desc->add_ready_events(event_buffer_[i].events);
375
376	37770x	bool expected = false;
377	37770x	if (desc->is_enqueued_.compare_exchange_strong(
378		expected, true, std::memory_order_release,
379		std::memory_order_relaxed))
380		{
381	37770x	local_ops.push(desc);
382		}
383		}
384
385	30490x	if (check_timers)
386		{
387	1937x	timer_svc_->process_expired();
388	1937x	update_timerfd();
389		}
390
391	30490x	lock.lock();
392
393	30490x	if (!local_ops.empty())
394	25148x	completed_ops_.splice(local_ops);
395	30490x	}
396
397		} // namespace boost::corosio::detail
398
399		#endif // BOOST_COROSIO_HAS_EPOLL
400
401		#endif // BOOST_COROSIO_NATIVE_DETAIL_EPOLL_EPOLL_SCHEDULER_HPP
402