rust/src/libstd/io/timer.rs

369 lines
11 KiB
Rust
Raw Normal View History

2013-07-19 16:24:07 -07:00
// Copyright 2013 The Rust Project Developers. See the COPYRIGHT
2013-07-19 16:03:02 -07:00
// file at the top-level directory of this distribution and at
2013-07-19 16:24:07 -07:00
// http://rust-lang.org/COPYRIGHT.
2013-07-19 16:03:02 -07:00
//
2013-07-19 16:24:07 -07:00
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
2013-07-19 16:03:02 -07:00
// except according to those terms.
2013-07-19 16:24:07 -07:00
/*!
Synchronous Timers
This module exposes the functionality to create timers, block the current task,
and create receivers which will receive notifications after a period of time.
*/
2014-06-03 20:09:39 -07:00
use comm::{Receiver, Sender, channel};
use io::{IoResult, IoError};
use kinds::Send;
use owned::Box;
2014-06-03 20:09:39 -07:00
use rt::rtio::{IoFactory, LocalIo, RtioTimer, Callback};
2013-07-19 16:03:02 -07:00
/// A synchronous timer object
///
/// Values of this type can be used to put the current task to sleep for a
/// period of time. Handles to this timer can also be created in the form of
/// receivers which will receive notifications over time.
///
/// # Example
///
/// ```
/// # fn main() {}
/// # fn foo() {
/// use std::io::Timer;
///
/// let mut timer = Timer::new().unwrap();
/// timer.sleep(10); // block the task for awhile
///
/// let timeout = timer.oneshot(10);
/// // do some work
/// timeout.recv(); // wait for the timeout to expire
///
/// let periodic = timer.periodic(10);
/// loop {
/// periodic.recv();
/// // this loop is only executed once every 10ms
/// }
/// # }
/// ```
///
2014-05-22 22:50:31 +10:00
/// If only sleeping is necessary, then a convenience API is provided through
/// the `io::timer` module.
///
/// ```
/// # fn main() {}
/// # fn foo() {
/// use std::io::timer;
///
/// // Put this task to sleep for 5 seconds
/// timer::sleep(5000);
/// # }
/// ```
pub struct Timer {
2014-06-14 11:03:34 -07:00
obj: Box<RtioTimer + Send>,
}
2013-07-19 16:03:02 -07:00
2014-06-03 20:09:39 -07:00
struct TimerCallback { tx: Sender<()> }
/// Sleep the current task for `msecs` milliseconds.
pub fn sleep(msecs: u64) {
let timer = Timer::new();
let mut timer = timer.ok().expect("timer::sleep: could not create a Timer");
timer.sleep(msecs)
}
2013-07-19 16:03:02 -07:00
impl Timer {
/// Creates a new timer which can be used to put the current task to sleep
/// for a number of milliseconds, or to possibly create channels which will
/// get notified after an amount of time has passed.
pub fn new() -> IoResult<Timer> {
2014-06-03 20:09:39 -07:00
LocalIo::maybe_raise(|io| {
io.timer_init().map(|t| Timer { obj: t })
}).map_err(IoError::from_rtio_error)
2013-07-19 16:03:02 -07:00
}
/// Blocks the current task for `msecs` milliseconds.
///
/// Note that this function will cause any other receivers for this timer to
/// be invalidated (the other end will be closed).
2013-08-20 09:53:02 -07:00
pub fn sleep(&mut self, msecs: u64) {
self.obj.sleep(msecs);
2013-07-19 16:03:02 -07:00
}
/// Creates a oneshot receiver which will have a notification sent when
/// `msecs` milliseconds has elapsed.
///
/// This does *not* block the current task, but instead returns immediately.
///
/// Note that this invalidates any previous receiver which has been created
/// by this timer, and that the returned receiver will be invalidated once
/// the timer is destroyed (when it falls out of scope). In particular, if
/// this is called in method-chaining style, the receiver will be
/// invalidated at the end of that statement, and all `recv` calls will
/// fail.
///
/// # Example
///
/// ```rust
/// use std::io::Timer;
///
/// let mut timer = Timer::new().unwrap();
/// let ten_milliseconds = timer.oneshot(10);
///
/// for _ in range(0u, 100) { /* do work */ }
///
/// // blocks until 10 ms after the `oneshot` call
/// ten_milliseconds.recv();
/// ```
///
/// ```rust
/// use std::io::Timer;
///
/// // Incorrect, method chaining-style:
/// let mut five_ms = Timer::new().unwrap().oneshot(5);
/// // The timer object was destroyed, so this will always fail:
/// // five_ms.recv()
/// ```
pub fn oneshot(&mut self, msecs: u64) -> Receiver<()> {
2014-06-03 20:09:39 -07:00
let (tx, rx) = channel();
self.obj.oneshot(msecs, box TimerCallback { tx: tx });
return rx
}
/// Creates a receiver which will have a continuous stream of notifications
/// being sent every `msecs` milliseconds.
///
/// This does *not* block the current task, but instead returns
/// immediately. The first notification will not be received immediately,
/// but rather after `msec` milliseconds have passed.
///
/// Note that this invalidates any previous receiver which has been created
/// by this timer, and that the returned receiver will be invalidated once
/// the timer is destroyed (when it falls out of scope). In particular, if
/// this is called in method-chaining style, the receiver will be
/// invalidated at the end of that statement, and all `recv` calls will
/// fail.
///
/// # Example
///
/// ```rust
/// use std::io::Timer;
///
/// let mut timer = Timer::new().unwrap();
/// let ten_milliseconds = timer.periodic(10);
///
/// for _ in range(0u, 100) { /* do work */ }
///
/// // blocks until 10 ms after the `periodic` call
/// ten_milliseconds.recv();
///
/// for _ in range(0u, 100) { /* do work */ }
///
/// // blocks until 20 ms after the `periodic` call (*not* 10ms after the
/// // previous `recv`)
/// ten_milliseconds.recv();
/// ```
///
/// ```rust
/// use std::io::Timer;
///
/// // Incorrect, method chaining-style.
/// let mut five_ms = Timer::new().unwrap().periodic(5);
/// // The timer object was destroyed, so this will always fail:
/// // five_ms.recv()
/// ```
pub fn periodic(&mut self, msecs: u64) -> Receiver<()> {
2014-06-03 20:09:39 -07:00
let (tx, rx) = channel();
self.obj.period(msecs, box TimerCallback { tx: tx });
return rx
}
}
impl Callback for TimerCallback {
fn call(&mut self) {
let _ = self.tx.send_opt(());
}
2013-07-19 16:03:02 -07:00
}
#[cfg(test)]
mod test {
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn test_io_timer_sleep_simple() {
let mut timer = Timer::new().unwrap();
timer.sleep(1);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn test_io_timer_sleep_oneshot() {
let mut timer = Timer::new().unwrap();
timer.oneshot(1).recv();
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn test_io_timer_sleep_oneshot_forget() {
let mut timer = Timer::new().unwrap();
timer.oneshot(100000000000);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn oneshot_twice() {
let mut timer = Timer::new().unwrap();
let rx1 = timer.oneshot(10000);
let rx = timer.oneshot(1);
rx.recv();
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert_eq!(rx1.recv_opt(), Err(()));
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn test_io_timer_oneshot_then_sleep() {
let mut timer = Timer::new().unwrap();
let rx = timer.oneshot(100000000000);
2014-06-08 13:22:49 -04:00
timer.sleep(1); // this should invalidate rx
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert_eq!(rx.recv_opt(), Err(()));
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
iotest!(fn test_io_timer_sleep_periodic() {
let mut timer = Timer::new().unwrap();
let rx = timer.periodic(1);
rx.recv();
rx.recv();
rx.recv();
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn test_io_timer_sleep_periodic_forget() {
let mut timer = Timer::new().unwrap();
timer.periodic(100000000000);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
iotest!(fn test_io_timer_sleep_standalone() {
sleep(1)
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
iotest!(fn oneshot() {
let mut timer = Timer::new().unwrap();
let rx = timer.oneshot(1);
rx.recv();
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert!(rx.recv_opt().is_err());
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
let rx = timer.oneshot(1);
rx.recv();
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert!(rx.recv_opt().is_err());
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
iotest!(fn override() {
let mut timer = Timer::new().unwrap();
let orx = timer.oneshot(100);
let prx = timer.periodic(100);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
timer.sleep(1);
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert_eq!(orx.recv_opt(), Err(()));
assert_eq!(prx.recv_opt(), Err(()));
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
timer.oneshot(1).recv();
})
iotest!(fn period() {
let mut timer = Timer::new().unwrap();
let rx = timer.periodic(1);
rx.recv();
rx.recv();
let rx2 = timer.periodic(1);
rx2.recv();
rx2.recv();
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
iotest!(fn sleep() {
let mut timer = Timer::new().unwrap();
timer.sleep(1);
timer.sleep(1);
})
iotest!(fn oneshot_fail() {
let mut timer = Timer::new().unwrap();
let _rx = timer.oneshot(1);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
fail!();
} #[should_fail])
iotest!(fn period_fail() {
let mut timer = Timer::new().unwrap();
let _rx = timer.periodic(1);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
fail!();
} #[should_fail])
iotest!(fn normal_fail() {
let _timer = Timer::new().unwrap();
fail!();
} #[should_fail])
iotest!(fn closing_channel_during_drop_doesnt_kill_everything() {
// see issue #10375
let mut timer = Timer::new().unwrap();
let timer_rx = timer.periodic(1000);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
spawn(proc() {
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
let _ = timer_rx.recv_opt();
});
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
// when we drop the TimerWatcher we're going to destroy the channel,
// which must wake up the task on the other end
})
iotest!(fn reset_doesnt_switch_tasks() {
// similar test to the one above.
let mut timer = Timer::new().unwrap();
let timer_rx = timer.periodic(1000);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
spawn(proc() {
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
let _ = timer_rx.recv_opt();
});
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
timer.oneshot(1);
})
iotest!(fn reset_doesnt_switch_tasks2() {
// similar test to the one above.
let mut timer = Timer::new().unwrap();
let timer_rx = timer.periodic(1000);
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
spawn(proc() {
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
let _ = timer_rx.recv_opt();
});
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
timer.sleep(1);
})
iotest!(fn sender_goes_away_oneshot() {
let rx = {
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
let mut timer = Timer::new().unwrap();
timer.oneshot(1000)
};
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert_eq!(rx.recv_opt(), Err(()));
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
iotest!(fn sender_goes_away_period() {
let rx = {
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
let mut timer = Timer::new().unwrap();
timer.periodic(1000)
};
std: Make std::comm return types consistent There are currently a number of return values from the std::comm methods, not all of which are necessarily completely expressive: Sender::try_send(t: T) -> bool This method currently doesn't transmit back the data `t` if the send fails due to the other end having disconnected. Additionally, this shares the name of the synchronous try_send method, but it differs in semantics in that it only has one failure case, not two (the buffer can never be full). SyncSender::try_send(t: T) -> TrySendResult<T> This method accurately conveys all possible information, but it uses a custom type to the std::comm module with no convenience methods on it. Additionally, if you want to inspect the result you're forced to import something from `std::comm`. SyncSender::send_opt(t: T) -> Option<T> This method uses Some(T) as an "error value" and None as a "success value", but almost all other uses of Option<T> have Some/None the other way Receiver::try_recv(t: T) -> TryRecvResult<T> Similarly to the synchronous try_send, this custom return type is lacking in terms of usability (no convenience methods). With this number of drawbacks in mind, I believed it was time to re-work the return types of these methods. The new API for the comm module is: Sender::send(t: T) -> () Sender::send_opt(t: T) -> Result<(), T> SyncSender::send(t: T) -> () SyncSender::send_opt(t: T) -> Result<(), T> SyncSender::try_send(t: T) -> Result<(), TrySendError<T>> Receiver::recv() -> T Receiver::recv_opt() -> Result<T, ()> Receiver::try_recv() -> Result<T, TryRecvError> The notable changes made are: * Sender::try_send => Sender::send_opt. This renaming brings the semantics in line with the SyncSender::send_opt method. An asychronous send only has one failure case, unlike the synchronous try_send method which has two failure cases (full/disconnected). * Sender::send_opt returns the data back to the caller if the send is guaranteed to fail. This method previously returned `bool`, but then it was unable to retrieve the data if the data was guaranteed to fail to send. There is still a race such that when `Ok(())` is returned the data could still fail to be received, but that's inherent to an asynchronous channel. * Result is now the basis of all return values. This not only adds lots of convenience methods to all return values for free, but it also means that you can inspect the return values with no extra imports (Ok/Err are in the prelude). Additionally, it's now self documenting when something failed or not because the return value has "Err" in the name. Things I'm a little uneasy about: * The methods send_opt and recv_opt are not returning options, but rather results. I felt more strongly that Option was the wrong return type than the _opt prefix was wrong, and I coudn't think of a much better name for these methods. One possible way to think about them is to read the _opt suffix as "optionally". * Result<T, ()> is often better expressed as Option<T>. This is only applicable to the recv_opt() method, but I thought it would be more consistent for everything to return Result rather than one method returning an Option. Despite my two reasons to feel uneasy, I feel much better about the consistency in return values at this point, and I think the only real open question is if there's a better suffix for {send,recv}_opt. Closes #11527
2014-04-10 10:53:49 -07:00
assert_eq!(rx.recv_opt(), Err(()));
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
})
iotest!(fn receiver_goes_away_oneshot() {
let mut timer1 = Timer::new().unwrap();
timer1.oneshot(1);
let mut timer2 = Timer::new().unwrap();
2014-06-08 13:22:49 -04:00
// while sleeping, the previous timer should fire and not have its
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
// callback do something terrible.
timer2.sleep(2);
})
iotest!(fn receiver_goes_away_period() {
let mut timer1 = Timer::new().unwrap();
timer1.periodic(1);
let mut timer2 = Timer::new().unwrap();
2014-06-08 13:22:49 -04:00
// while sleeping, the previous timer should fire and not have its
Implement native timers Native timers are a much hairier thing to deal with than green timers due to the interface that we would like to expose (both a blocking sleep() and a channel-based interface). I ended up implementing timers in three different ways for the various platforms that we supports. In all three of the implementations, there is a worker thread which does send()s on channels for timers. This worker thread is initialized once and then communicated to in a platform-specific manner, but there's always a shared channel available for sending messages to the worker thread. * Windows - I decided to use windows kernel timer objects via CreateWaitableTimer and SetWaitableTimer in order to provide sleeping capabilities. The worker thread blocks via WaitForMultipleObjects where one of the objects is an event that is used to wake up the helper thread (which then drains the incoming message channel for requests). * Linux/(Android?) - These have the ideal interface for implementing timers, timerfd_create. Each timer corresponds to a timerfd, and the helper thread uses epoll to wait for all active timers and then send() for the next one that wakes up. The tricky part in this implementation is updating a timerfd, but see the implementation for the fun details * OSX/FreeBSD - These obviously don't have the windows APIs, and sadly don't have the timerfd api available to them, so I have thrown together a solution which uses select() plus a timeout in order to ad-hoc-ly implement a timer solution for threads. The implementation is backed by a sorted array of timers which need to fire. As I said, this is an ad-hoc solution which is certainly not accurate timing-wise. I have done this implementation due to the lack of other primitives to provide an implementation, and I've done it the best that I could, but I'm sure that there's room for improvement. I'm pretty happy with how these implementations turned out. In theory we could drop the timerfd implementation and have linux use the select() + timeout implementation, but it's so inaccurate that I would much rather continue to use timerfd rather than my ad-hoc select() implementation. The only change that I would make to the API in general is to have a generic sleep() method on an IoFactory which doesn't require allocating a Timer object. For everything but windows it's super-cheap to request a blocking sleep for a set amount of time, and it's probably worth it to provide a sleep() which doesn't do something like allocate a file descriptor on linux.
2013-12-28 23:33:56 -08:00
// callback do something terrible.
timer2.sleep(2);
})
}