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	662x	inline epoll_scheduler::epoll_scheduler(capy::execution_context& ctx, int)
146	662x	: epoll_fd_(-1)
147	662x	, event_fd_(-1)
148	662x	, timer_fd_(-1)
149	1324x	, event_buffer_(max_events_per_poll_)
150		{
151	662x	epoll_fd_ = ::epoll_create1(EPOLL_CLOEXEC);
152	662x	if (epoll_fd_ < 0)
153	✗	detail::throw_system_error(make_err(errno), "epoll_create1");
154
155	662x	event_fd_ = ::eventfd(0, EFD_NONBLOCK \| EFD_CLOEXEC);
156	662x	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	662x	timer_fd_ = ::timerfd_create(CLOCK_MONOTONIC, TFD_NONBLOCK \| TFD_CLOEXEC);
164	662x	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	662x	epoll_event ev{};
173	662x	ev.events = EPOLLIN \| EPOLLET;
174	662x	ev.data.ptr = nullptr;
175	662x	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	662x	epoll_event timer_ev{};
185	662x	timer_ev.events = EPOLLIN \| EPOLLERR;
186	662x	timer_ev.data.ptr = &timer_fd_;
187	662x	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	662x	timer_svc_ = &get_timer_service(ctx, *this);
197	662x	timer_svc_->set_on_earliest_changed(
198	5552x	timer_service::callback(this, [](void* p) {
199	4890x	auto* self = static_cast<epoll_scheduler*>(p);
200	4890x	self->timerfd_stale_.store(true, std::memory_order_release);
201	4890x	self->interrupt_reactor();
202	4890x	}));
203
204	662x	get_resolver_service(ctx, *this);
205	662x	get_signal_service(ctx, *this);
206	662x	get_stream_file_service(ctx, *this);
207	662x	get_random_access_file_service(ctx, *this);
208
209	662x	completed_ops_.push(&task_op_);
210	662x	}
211
212	1324x	inline epoll_scheduler::~epoll_scheduler()
213		{
214	662x	if (timer_fd_ >= 0)
215	662x	::close(timer_fd_);
216	662x	if (event_fd_ >= 0)
217	662x	::close(event_fd_);
218	662x	if (epoll_fd_ >= 0)
219	662x	::close(epoll_fd_);
220	1324x	}
221
222		inline void
223	662x	epoll_scheduler::shutdown()
224		{
225	662x	shutdown_drain();
226
227	662x	if (event_fd_ >= 0)
228	662x	interrupt_reactor();
229	662x	}
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	11423x	epoll_scheduler::register_descriptor(int fd, reactor_descriptor_state* desc) const
245		{
246	11423x	epoll_event ev{};
247	11423x	ev.events = EPOLLIN \| EPOLLOUT \| EPOLLET \| EPOLLERR \| EPOLLHUP;
248	11423x	ev.data.ptr = desc;
249
250	11423x	if (::epoll_ctl(epoll_fd_, EPOLL_CTL_ADD, fd, &ev) < 0)
251	✗	detail::throw_system_error(make_err(errno), "epoll_ctl (register)");
252
253	11423x	desc->registered_events = ev.events;
254	11423x	desc->fd = fd;
255	11423x	desc->scheduler_ = this;
256	11423x	desc->mutex.set_enabled(!single_threaded_);
257	11423x	desc->ready_events_.store(0, std::memory_order_relaxed);
258
259	11423x	conditionally_enabled_mutex::scoped_lock lock(desc->mutex);
260	11423x	desc->impl_ref_.reset();
261	11423x	desc->read_ready = false;
262	11423x	desc->write_ready = false;
263	11423x	}
264
265		inline void
266	11423x	epoll_scheduler::deregister_descriptor(int fd) const
267		{
268	11423x	::epoll_ctl(epoll_fd_, EPOLL_CTL_DEL, fd, nullptr);
269	11423x	}
270
271		inline void
272	8562x	epoll_scheduler::interrupt_reactor() const
273		{
274	8562x	bool expected = false;
275	8562x	if (eventfd_armed_.compare_exchange_strong(
276		expected, true, std::memory_order_release,
277		std::memory_order_relaxed))
278		{
279	8143x	std::uint64_t val = 1;
280	8143x	[[maybe_unused]] auto r = ::write(event_fd_, &val, sizeof(val));
281		}
282	8562x	}
283
284		inline void
285	8998x	epoll_scheduler::update_timerfd() const
286		{
287	8998x	auto nearest = timer_svc_->nearest_expiry();
288
289	8998x	itimerspec ts{};
290	8998x	int flags = 0;
291
292	8998x	if (nearest == timer_service::time_point::max())
293		{
294		// No timers — disarm by setting to 0 (relative)
295		}
296		else
297		{
298	8901x	auto now = std::chrono::steady_clock::now();
299	8901x	if (nearest <= now)
300		{
301		// Use 1ns instead of 0 — zero disarms the timerfd
302	120x	ts.it_value.tv_nsec = 1;
303		}
304		else
305		{
306	8781x	auto nsec = std::chrono::duration_cast<std::chrono::nanoseconds>(
307	8781x	nearest - now)
308	8781x	.count();
309	8781x	ts.it_value.tv_sec = nsec / 1000000000;
310	8781x	ts.it_value.tv_nsec = nsec % 1000000000;
311	8781x	if (ts.it_value.tv_sec == 0 && ts.it_value.tv_nsec == 0)
312	✗	ts.it_value.tv_nsec = 1;
313		}
314		}
315
316	8998x	if (::timerfd_settime(timer_fd_, flags, &ts, nullptr) < 0)
317	✗	detail::throw_system_error(make_err(errno), "timerfd_settime");
318	8998x	}
319
320		inline void
321	55935x	epoll_scheduler::run_task(
322		lock_type& lock, context_type* ctx, long timeout_us)
323		{
324		int timeout_ms;
325	55935x	if (task_interrupted_)
326	41913x	timeout_ms = 0;
327	14022x	else if (timeout_us < 0)
328	14018x	timeout_ms = -1;
329		else
330	4x	timeout_ms = static_cast<int>((timeout_us + 999) / 1000);
331
332	55935x	if (lock.owns_lock())
333	14022x	lock.unlock();
334
335	55935x	task_cleanup on_exit{this, &lock, ctx};
336
337		// Flush deferred timerfd programming before blocking
338	55935x	if (timerfd_stale_.exchange(false, std::memory_order_acquire))
339	4866x	update_timerfd();
340
341	55935x	int nfds = ::epoll_wait(
342		epoll_fd_, event_buffer_.data(),
343	55935x	static_cast<int>(event_buffer_.size()), timeout_ms);
344
345	55935x	if (nfds < 0 && errno != EINTR)
346	✗	detail::throw_system_error(make_err(errno), "epoll_wait");
347
348	55935x	bool check_timers = false;
349	55935x	op_queue local_ops;
350
351	120225x	for (int i = 0; i < nfds; ++i)
352		{
353	64290x	if (event_buffer_[i].data.ptr == nullptr)
354		{
355		std::uint64_t val;
356		// NOLINTNEXTLINE(clang-analyzer-unix.BlockInCriticalSection)
357	7481x	[[maybe_unused]] auto r = ::read(event_fd_, &val, sizeof(val));
358	7481x	eventfd_armed_.store(false, std::memory_order_relaxed);
359	7481x	continue;
360	7481x	}
361
362	56809x	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	4132x	::read(timer_fd_, &expirations, sizeof(expirations));
368	4132x	check_timers = true;
369	4132x	continue;
370	4132x	}
371
372		auto* desc =
373	52677x	static_cast<reactor_descriptor_state*>(event_buffer_[i].data.ptr);
374	52677x	desc->add_ready_events(event_buffer_[i].events);
375
376	52677x	bool expected = false;
377	52677x	if (desc->is_enqueued_.compare_exchange_strong(
378		expected, true, std::memory_order_release,
379		std::memory_order_relaxed))
380		{
381	52677x	local_ops.push(desc);
382		}
383		}
384
385	55935x	if (check_timers)
386		{
387	4132x	timer_svc_->process_expired();
388	4132x	update_timerfd();
389		}
390
391	55935x	lock.lock();
392
393	55935x	if (!local_ops.empty())
394	34782x	completed_ops_.splice(local_ops);
395	55935x	}
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