1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
//! epoll support.
//!
//! This is an experiment, and it isn't yet clear whether epoll is the right
//! level of abstraction at which to introduce safety. But it works fairly well
//! in simple examples 🙂.
//!
//! # Examples
//!
//! ```no_run
//! # #![cfg_attr(io_lifetimes_use_std, feature(io_safety))]
//! # #[cfg(feature = "net")]
//! # fn main() -> std::io::Result<()> {
//! use io_lifetimes::AsFd;
//! use rustix::io::{epoll, ioctl_fionbio, read, write};
//! use rustix::net::{
//! accept, bind_v4, listen, socket, AddressFamily, Ipv4Addr, Protocol, SocketAddrV4,
//! SocketType,
//! };
//! use std::collections::HashMap;
//! use std::os::unix::io::AsRawFd;
//!
//! // Create a socket and listen on it.
//! let listen_sock = socket(AddressFamily::INET, SocketType::STREAM, Protocol::default())?;
//! bind_v4(&listen_sock, &SocketAddrV4::new(Ipv4Addr::LOCALHOST, 0))?;
//! listen(&listen_sock, 1)?;
//!
//! // Create an epoll object. Using `Owning` here means the epoll object will
//! // take ownership of the file descriptors registered with it.
//! let epoll = epoll::epoll_create(epoll::CreateFlags::CLOEXEC)?;
//!
//! // Register the socket with the epoll object.
//! epoll::epoll_add(&epoll, &listen_sock, 1, epoll::EventFlags::IN)?;
//!
//! // Keep track of the sockets we've opened.
//! let mut next_id = 2;
//! let mut sockets = HashMap::new();
//!
//! // Process events.
//! let mut event_list = epoll::EventVec::with_capacity(4);
//! loop {
//! epoll::epoll_wait(&epoll, &mut event_list, -1)?;
//! for (_event_flags, target) in &event_list {
//! if target == 1 {
//! // Accept a new connection, set it to non-blocking, and
//! // register to be notified when it's ready to write to.
//! let conn_sock = accept(&listen_sock)?;
//! ioctl_fionbio(&conn_sock, true)?;
//! epoll::epoll_add(
//! &epoll,
//! &conn_sock,
//! next_id,
//! epoll::EventFlags::OUT | epoll::EventFlags::ET,
//! )?;
//!
//! // Keep track of the socket.
//! sockets.insert(next_id, conn_sock);
//! next_id += 1;
//! } else {
//! // Write a message to the stream and then unregister it.
//! let target = sockets.remove(&target).unwrap();
//! write(&target, b"hello\n")?;
//! let _ = epoll::epoll_del(&epoll, &target)?;
//! }
//! }
//! }
//! # }
//! # #[cfg(not(feature = "net"))]
//! # fn main() {}
//! ```
#![allow(unsafe_code)]
use super::super::c;
use crate::backend::io::syscalls;
use crate::fd::{AsFd, AsRawFd, OwnedFd};
use crate::io;
use alloc::vec::Vec;
use bitflags::bitflags;
bitflags! {
/// `EPOLL_*` for use with [`Epoll::new`].
pub struct CreateFlags: c::c_uint {
/// `EPOLL_CLOEXEC`
const CLOEXEC = linux_raw_sys::general::EPOLL_CLOEXEC;
}
}
bitflags! {
/// `EPOLL*` for use with [`Epoll::add`].
#[derive(Default)]
pub struct EventFlags: u32 {
/// `EPOLLIN`
const IN = linux_raw_sys::general::EPOLLIN as u32;
/// `EPOLLOUT`
const OUT = linux_raw_sys::general::EPOLLOUT as u32;
/// `EPOLLPRI`
const PRI = linux_raw_sys::general::EPOLLPRI as u32;
/// `EPOLLERR`
const ERR = linux_raw_sys::general::EPOLLERR as u32;
/// `EPOLLHUP`
const HUP = linux_raw_sys::general::EPOLLHUP as u32;
/// `EPOLLRDNORM`
const RDNORM = linux_raw_sys::general::EPOLLRDNORM as u32;
/// `EPOLLRDBAND`
const RDBAND = linux_raw_sys::general::EPOLLRDBAND as u32;
/// `EPOLLWRNORM`
const WRNORM = linux_raw_sys::general::EPOLLWRNORM as u32;
/// `EPOLLWRBAND`
const WRBAND = linux_raw_sys::general::EPOLLWRBAND as u32;
/// `EPOLLMSG`
const MSG = linux_raw_sys::general::EPOLLMSG as u32;
/// `EPOLLRDHUP`
const RDHUP = linux_raw_sys::general::EPOLLRDHUP as u32;
/// `EPOLLET`
const ET = linux_raw_sys::general::EPOLLET as u32;
/// `EPOLLONESHOT`
const ONESHOT = linux_raw_sys::general::EPOLLONESHOT as u32;
/// `EPOLLWAKEUP`
const WAKEUP = linux_raw_sys::general::EPOLLWAKEUP as u32;
/// `EPOLLEXCLUSIVE`
const EXCLUSIVE = linux_raw_sys::general::EPOLLEXCLUSIVE as u32;
}
}
/// `epoll_create1(flags)`—Creates a new `Epoll`.
///
/// Use the [`CreateFlags::CLOEXEC`] flag to prevent the resulting file
/// descriptor from being implicitly passed across `exec` boundaries.
#[inline]
#[doc(alias = "epoll_create1")]
pub fn epoll_create(flags: CreateFlags) -> io::Result<OwnedFd> {
syscalls::epoll_create(flags)
}
/// `epoll_ctl(self, EPOLL_CTL_ADD, data, event)`—Adds an element to an
/// `Epoll`.
///
/// This registers interest in any of the events set in `events` occurring
/// on the file descriptor associated with `data`.
///
/// Note that if `epoll_del` is not called on the I/O source passed into
/// this function before the I/O source is `close`d, then the `epoll` will
/// act as if the I/O source is still registered with it. This can lead to
/// spurious events being returned from `epoll_wait`. If a file descriptor
/// is an `Arc<dyn SystemResource>`, then `epoll` can be thought to maintain
/// a `Weak<dyn SystemResource>` to the file descriptor.
#[doc(alias = "epoll_ctl")]
pub fn epoll_add(
epoll: impl AsFd,
source: impl AsFd,
data: u64,
event_flags: EventFlags,
) -> io::Result<()> {
// SAFETY: We're calling `epoll_ctl` via FFI and we know how it
// behaves.
unsafe {
syscalls::epoll_add(
epoll.as_fd(),
source.as_fd().as_raw_fd(),
&linux_raw_sys::general::epoll_event {
events: event_flags.bits(),
data,
},
)
}
}
/// `epoll_ctl(self, EPOLL_CTL_MOD, target, event)`—Modifies an element in
/// this `Epoll`.
///
/// This sets the events of interest with `target` to `events`.
#[doc(alias = "epoll_ctl")]
pub fn epoll_mod(
epoll: impl AsFd,
source: impl AsFd,
data: u64,
event_flags: EventFlags,
) -> io::Result<()> {
// SAFETY: We're calling `epoll_ctl` via FFI and we know how it
// behaves.
unsafe {
let raw_fd = source.as_fd().as_raw_fd();
syscalls::epoll_mod(
epoll.as_fd(),
raw_fd,
&linux_raw_sys::general::epoll_event {
events: event_flags.bits(),
data,
},
)
}
}
/// `epoll_ctl(self, EPOLL_CTL_DEL, target, NULL)`—Removes an element in
/// this `Epoll`.
///
/// This also returns the owning `Data`.
#[doc(alias = "epoll_ctl")]
pub fn epoll_del(epoll: impl AsFd, source: impl AsFd) -> io::Result<()> {
// SAFETY: We're calling `epoll_ctl` via FFI and we know how it
// behaves.
unsafe {
let raw_fd = source.as_fd().as_raw_fd();
syscalls::epoll_del(epoll.as_fd(), raw_fd)
}
}
/// `epoll_wait(self, events, timeout)`—Waits for registered events of
/// interest.
///
/// For each event of interest, an element is written to `events`. On
/// success, this returns the number of written elements.
pub fn epoll_wait(
epoll: impl AsFd,
event_list: &mut EventVec,
timeout: c::c_int,
) -> io::Result<()> {
// SAFETY: We're calling `epoll_wait` via FFI and we know how it
// behaves.
unsafe {
event_list.events.set_len(0);
let nfds = syscalls::epoll_wait(
epoll.as_fd(),
event_list.events[..].as_mut_ptr().cast(),
event_list.events.capacity(),
timeout,
)?;
event_list.events.set_len(nfds);
}
Ok(())
}
/// An iterator over the `Event`s in an `EventVec`.
pub struct Iter<'a> {
iter: core::slice::Iter<'a, Event>,
}
impl<'a> Iterator for Iter<'a> {
type Item = (EventFlags, u64);
fn next(&mut self) -> Option<Self::Item> {
self.iter
.next()
.map(|event| (event.event_flags, event.data))
}
}
/// A record of an event that occurred.
#[repr(C)]
#[cfg_attr(target_arch = "x86_64", repr(packed))]
struct Event {
// Match the layout of `linux_raw_sys::general::epoll_event`. We just use a
// `u64` instead of the full union.
event_flags: EventFlags,
data: u64,
}
/// A vector of `Event`s, plus context for interpreting them.
pub struct EventVec {
events: Vec<Event>,
}
impl EventVec {
/// Constructs an `EventVec` with memory for `capacity` `Event`s.
#[inline]
pub fn with_capacity(capacity: usize) -> Self {
Self {
events: Vec::with_capacity(capacity),
}
}
/// Returns the current `Event` capacity of this `EventVec`.
#[inline]
pub fn capacity(&self) -> usize {
self.events.capacity()
}
/// Reserves enough memory for at least `additional` more `Event`s.
#[inline]
pub fn reserve(&mut self, additional: usize) {
self.events.reserve(additional);
}
/// Reserves enough memory for exactly `additional` more `Event`s.
#[inline]
pub fn reserve_exact(&mut self, additional: usize) {
self.events.reserve_exact(additional);
}
/// Clears all the `Events` out of this `EventVec`.
#[inline]
pub fn clear(&mut self) {
self.events.clear();
}
/// Shrinks the capacity of this `EventVec` as much as possible.
#[inline]
pub fn shrink_to_fit(&mut self) {
self.events.shrink_to_fit();
}
/// Returns an iterator over the `Event`s in this `EventVec`.
#[inline]
pub fn iter(&self) -> Iter<'_> {
Iter {
iter: self.events.iter(),
}
}
/// Returns the number of `Event`s logically contained in this `EventVec`.
#[inline]
pub fn len(&mut self) -> usize {
self.events.len()
}
/// Tests whether this `EventVec` is logically empty.
#[inline]
pub fn is_empty(&mut self) -> bool {
self.events.is_empty()
}
}
impl<'a> IntoIterator for &'a EventVec {
type IntoIter = Iter<'a>;
type Item = (EventFlags, u64);
#[inline]
fn into_iter(self) -> Self::IntoIter {
self.iter()
}
}