Struct catalyzer::internals::crates::tokio::io::unix::AsyncFdReadyMutGuard

pub struct AsyncFdReadyMutGuard<'a, T>where
    T: AsRawFd,
{ /* private fields */ }

Expand description

Represents an IO-ready event detected on a particular file descriptor that has not yet been acknowledged. This is a must_use structure to help ensure that you do not forget to explicitly clear (or not clear) the event.

This type exposes a mutable reference to the underlying IO object.

Implementations§

§

impl<'a, Inner> AsyncFdReadyMutGuard<'a, Inner>
where Inner: AsRawFd,

pub fn clear_ready(&mut self)

Indicates to tokio that the file descriptor is no longer ready. All internal readiness flags will be cleared, and tokio will wait for the next edge-triggered readiness notification from the OS.

This function is commonly used with guards returned by AsyncFd::readable_mut and AsyncFd::writable_mut.

It is critical that this function not be called unless your code actually observes that the file descriptor is not ready. Do not call it simply because, for example, a read succeeded; it should be called when a read is observed to block.

This method only clears readiness events that happened before the creation of this guard. In other words, if the IO resource becomes ready between the creation of the guard and this call to clear_ready, then the readiness is not actually cleared.

pub fn clear_ready_matching(&mut self, ready: Ready)

Indicates to tokio that the file descriptor no longer has a specific readiness. The internal readiness flag will be cleared, and tokio will wait for the next edge-triggered readiness notification from the OS.

This function is useful in combination with the AsyncFd::ready_mut method when a combined interest like Interest::READABLE | Interest::WRITABLE is used.

It is critical that this function not be called unless your code actually observes that the file descriptor is not ready for the provided Ready. Do not call it simply because, for example, a read succeeded; it should be called when a read is observed to block. Only clear the specific readiness that is observed to block. For example when a read blocks when using a combined interest, only clear Ready::READABLE.

This method only clears readiness events that happened before the creation of this guard. In other words, if the IO resource becomes ready between the creation of the guard and this call to clear_ready, then the readiness is not actually cleared.

§Examples

Concurrently read and write to a std::net::TcpStream on the same task without splitting.

use std::error::Error;
use std::io;
use std::io::{Read, Write};
use std::net::TcpStream;
use tokio::io::unix::AsyncFd;
use tokio::io::{Interest, Ready};

#[tokio::main]
async fn main() -> Result<(), Box<dyn Error>> {
    let stream = TcpStream::connect("127.0.0.1:8080")?;
    stream.set_nonblocking(true)?;
    let mut stream = AsyncFd::new(stream)?;

    loop {
        let mut guard = stream
            .ready_mut(Interest::READABLE | Interest::WRITABLE)
            .await?;

        if guard.ready().is_readable() {
            let mut data = vec![0; 1024];
            // Try to read data, this may still fail with `WouldBlock`
            // if the readiness event is a false positive.
            match guard.get_inner_mut().read(&mut data) {
                Ok(n) => {
                    println!("read {} bytes", n);
                }
                Err(ref e) if e.kind() == io::ErrorKind::WouldBlock => {
                    // a read has blocked, but a write might still succeed.
                    // clear only the read readiness.
                    guard.clear_ready_matching(Ready::READABLE);
                    continue;
                }
                Err(e) => {
                    return Err(e.into());
                }
            }
        }

        if guard.ready().is_writable() {
            // Try to write data, this may still fail with `WouldBlock`
            // if the readiness event is a false positive.
            match guard.get_inner_mut().write(b"hello world") {
                Ok(n) => {
                    println!("write {} bytes", n);
                }
                Err(ref e) if e.kind() == io::ErrorKind::WouldBlock => {
                    // a write has blocked, but a read might still succeed.
                    // clear only the write readiness.
                    guard.clear_ready_matching(Ready::WRITABLE);
                    continue;
                }
                Err(e) => {
                    return Err(e.into());
                }
            }
        }
    }
}

pub fn retain_ready(&mut self)

This method should be invoked when you intentionally want to keep the ready flag asserted.

While this function is itself a no-op, it satisfies the #[must_use] constraint on the AsyncFdReadyGuard type.

pub fn ready(&self) -> Ready

Get the Ready value associated with this guard.

This method will return the empty readiness state if AsyncFdReadyGuard::clear_ready has been called on the guard.

pub fn try_io<R>( &mut self, f: impl FnOnce(&mut AsyncFd<Inner>) -> Result<R, Error> ) -> Result<Result<R, Error>, TryIoError>

Performs the provided IO operation.

If f returns a WouldBlock error, the readiness state associated with this file descriptor is cleared, and the method returns Err(TryIoError::WouldBlock). You will typically need to poll the AsyncFd again when this happens.

This method helps ensure that the readiness state of the underlying file descriptor remains in sync with the tokio-side readiness state, by clearing the tokio-side state only when a WouldBlock condition occurs. It is the responsibility of the caller to ensure that f returns WouldBlock only if the file descriptor that originated this AsyncFdReadyGuard no longer expresses the readiness state that was queried to create this AsyncFdReadyGuard.