signal: Touch up more impls and comments

This commit is contained in:
Alex Crichton
2018-09-10 11:29:58 -07:00
committed by Carl Lerche
parent 699b9ab89e
commit 70e4ed67ad
+57 -35
View File
@@ -85,7 +85,16 @@ fn globals() -> &'static Globals {
} }
} }
// Flag the relevant signal and wake up through a self-pipe /// Our global signal handler for all signals registered by this module.
///
/// The purpose of this signal handler is to primarily:
///
/// 1. Flag that our specific signal was received (e.g. store an atomic flag)
/// 2. Wake up driver tasks by writing a byte to a pipe
///
/// Those two operations shoudl both be async-signal safe. After that's done we
/// just try to call a previous signal handler, if any, to be "good denizens of
/// the internet"
extern fn handler(signum: c_int, extern fn handler(signum: c_int,
info: *mut libc::siginfo_t, info: *mut libc::siginfo_t,
ptr: *mut libc::c_void) { ptr: *mut libc::c_void) {
@@ -116,10 +125,18 @@ extern fn handler(signum: c_int,
} }
} }
// Make sure we listen to the given signal and provide the recipient end of the /// Enable this module to receive signal notifications for the `signal`
// self-pipe /// provided.
///
/// This will register the signal handler if it hasn't already been registered,
/// returning any error along the way if that fails.
fn signal_enable(signal: c_int) -> io::Result<()> { fn signal_enable(signal: c_int) -> io::Result<()> {
let siginfo = &globals().signals[signal as usize]; let siginfo = match globals().signals.get(signal as usize) {
Some(slot) => slot,
None => {
return Err(io::Error::new(io::ErrorKind::Other, "signal too large"))
}
};
unsafe { unsafe {
let mut err = None; let mut err = None;
siginfo.init.call_once(|| { siginfo.init.call_once(|| {
@@ -146,6 +163,12 @@ fn signal_enable(signal: c_int) -> io::Result<()> {
} }
} }
/// A helper struct to register our global receiving end of the signal pipe on
/// multiple event loops.
///
/// This structure represents registering the receiving end on all event loops,
/// and uses `EventedFd` in mio to do so. It's stored in each driver task and is
/// used to read data and register interest in new signals coming in.
struct EventedReceiver; struct EventedReceiver;
impl Evented for EventedReceiver { impl Evented for EventedReceiver {
@@ -163,6 +186,12 @@ impl Evented for EventedReceiver {
} }
} }
impl Read for EventedReceiver {
fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> {
(&globals().receiver).read(buf)
}
}
struct Driver { struct Driver {
id: CoreId, id: CoreId,
wakeup: PollEvented<EventedReceiver>, wakeup: PollEvented<EventedReceiver>,
@@ -198,30 +227,30 @@ impl Driver {
}) })
} }
// Drain all data in the pipe and maintain an interest in read-ready /// Drain all data in the global receiver, returning whether data was to be
fn drain(&self) -> bool { /// had.
// Inform tokio we're interested in reading. It also hints on ///
// if we may be readable. /// If this function returns `true` then some signal has been received since
if let Async::NotReady = self.wakeup.poll_read() { /// we last checked, otherwise `false` indicates that no signal has been
return false; /// received.
} fn drain(&mut self) -> bool {
// Read all available data (until EAGAIN)
let mut received = false; let mut received = false;
loop { loop {
match (&globals().receiver).read(&mut [0; 128]) { match self.wakeup.read(&mut [0; 128]) {
Ok(0) => panic!("EOF on self-pipe"), Ok(0) => panic!("EOF on self-pipe"),
Ok(_) => received = true, Ok(_) => received = true,
Err(ref e) if e.kind() == io::ErrorKind::WouldBlock => break, Err(ref e) if e.kind() == io::ErrorKind::WouldBlock => break,
Err(e) => panic!("Bad read on self-pipe: {}", e), Err(e) => panic!("Bad read on self-pipe: {}", e),
} }
} }
// If we got here, it's because we got EAGAIN above. Ask for more data.
self.wakeup.need_read();
received received
} }
// Go through all the signals and broadcast everything
/// Go through all the signals and broadcast everything.
///
/// Driver tasks wake up for *any* signal and simply process all globally
/// registered signal streams, so each task is sort of cooperatively working
/// for all the rest as well.
fn broadcast(&self) { fn broadcast(&self) {
for (sig, slot) in globals().signals.iter().enumerate() { for (sig, slot) in globals().signals.iter().enumerate() {
// Any signal of this kind arrived since we checked last? // Any signal of this kind arrived since we checked last?
@@ -234,7 +263,8 @@ impl Driver {
// Notify all waiters on this signal that the signal has been // Notify all waiters on this signal that the signal has been
// received. If we can't push a message into the queue then we don't // received. If we can't push a message into the queue then we don't
// worry about it as everything is coalesced anyway. // worry about it as everything is coalesced anyway. If the channel
// has gone away then we can remove that slot.
for i in (0..recipients.len()).rev() { for i in (0..recipients.len()).rev() {
// TODO: This thing probably generates unnecessary wakups of // TODO: This thing probably generates unnecessary wakups of
// this task. But let's optimise it later on, when we // this task. But let's optimise it later on, when we
@@ -249,7 +279,6 @@ impl Driver {
} }
} }
// TODO: Go through the docs, they are a copy-paste from the previous version
/// An implementation of `Stream` for receiving a particular type of signal. /// An implementation of `Stream` for receiving a particular type of signal.
/// ///
/// This structure implements the `Stream` trait and represents notifications /// This structure implements the `Stream` trait and represents notifications
@@ -261,14 +290,6 @@ impl Driver {
/// structure is no exception! There are some important limitations to keep in /// structure is no exception! There are some important limitations to keep in
/// mind when using `Signal` streams: /// mind when using `Signal` streams:
/// ///
/// * While multiple event loops are supported, the *first* event loop to
/// register a signal handler is required to be active to ensure that signals
/// for other event loops are delivered. In other words, once an event loop
/// registers a signal, it's best to keep it around and running. This is
/// normally just a problem for tests, and the "workaround" is to spawn a
/// thread in the background at the beginning of the test suite which is
/// running an event loop (and listening for a signal).
///
/// * Signals handling in Unix already necessitates coalescing signals /// * Signals handling in Unix already necessitates coalescing signals
/// together sometimes. This `Signal` stream is also no exception here in /// together sometimes. This `Signal` stream is also no exception here in
/// that it will also coalesce signals. That is, even if the signal handler /// that it will also coalesce signals. That is, even if the signal handler
@@ -278,11 +299,16 @@ impl Driver {
/// Once `poll` has been called, however, a further signal is guaranteed to /// Once `poll` has been called, however, a further signal is guaranteed to
/// be yielded as an item. /// be yielded as an item.
/// ///
/// Put another way, any element pulled off the returned stream corresponds to
/// *at least one* signal, but possibly more.
///
/// * Signal handling in general is relatively inefficient. Although some /// * Signal handling in general is relatively inefficient. Although some
/// improvements are possible in this crate, it's recommended to not plan on /// improvements are possible in this crate, it's recommended to not plan on
/// having millions of signal channels open. /// having millions of signal channels open.
/// ///
/// * Currently the "driver task" to process incoming signals never exits. /// * Currently the "driver task" to process incoming signals never exits. This
/// driver task runs in the background of the event loop provided, and
/// in general you shouldn't need to worry about it.
/// ///
/// If you've got any questions about this feel free to open an issue on the /// If you've got any questions about this feel free to open an issue on the
/// repo, though, as I'd love to chat about this! In other words, I'd love to /// repo, though, as I'd love to chat about this! In other words, I'd love to
@@ -290,9 +316,8 @@ impl Driver {
pub struct Signal(Receiver<c_int>); pub struct Signal(Receiver<c_int>);
impl Signal { impl Signal {
// TODO: Revisit the docs, they are from the previous version
/// Creates a new stream which will receive notifications when the current /// Creates a new stream which will receive notifications when the current
/// process receives the signal `signum`. /// process receives the signal `signal`.
/// ///
/// This function will create a new stream which may be based on the /// This function will create a new stream which may be based on the
/// event loop handle provided. This function returns a future which will /// event loop handle provided. This function returns a future which will
@@ -303,10 +328,7 @@ impl Signal {
/// found on `Signal` itself, but to reiterate: /// found on `Signal` itself, but to reiterate:
/// ///
/// * Signals may be coalesced beyond what the kernel already does. /// * Signals may be coalesced beyond what the kernel already does.
/// * While multiple event loops are supported, the first event loop to /// * Once a signal handler is registered with the process the underlying
/// register a signal handler must be active to deliver signal
/// notifications
/// * Once a signal handle is registered with the process the underlying
/// libc signal handler is never unregistered. /// libc signal handler is never unregistered.
/// ///
/// A `Signal` stream can be created for a particular signal number /// A `Signal` stream can be created for a particular signal number