rust/src/librustrt/task.rs
2014-09-09 11:32:58 +02:00

669 lines
24 KiB
Rust

// Copyright 2013-2014 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// 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
// except according to those terms.
//! Language-level runtime services that should reasonably expected
//! to be available 'everywhere'. Local heaps, GC, unwinding,
//! local storage, and logging. Even a 'freestanding' Rust would likely want
//! to implement this.
use alloc::arc::Arc;
use alloc::boxed::{BoxAny, Box};
use core::any::Any;
use core::atomic::{AtomicUint, SeqCst};
use core::iter::Take;
use core::kinds::marker;
use core::mem;
use core::prelude::{Clone, Drop, Err, Iterator, None, Ok, Option, Send, Some};
use core::prelude::{drop};
use core::raw;
use local_data;
use Runtime;
use local::Local;
use local_heap::LocalHeap;
use rtio::LocalIo;
use unwind;
use unwind::Unwinder;
use collections::str::SendStr;
/// State associated with Rust tasks.
///
/// Rust tasks are primarily built with two separate components. One is this
/// structure which handles standard services such as TLD, unwinding support,
/// naming of a task, etc. The second component is the runtime of this task, a
/// `Runtime` trait object.
///
/// The `Runtime` object instructs this task how it can perform critical
/// operations such as blocking, rescheduling, I/O constructors, etc. The two
/// halves are separately owned, but one is often found contained in the other.
/// A task's runtime can be reflected upon with the `maybe_take_runtime` method,
/// and otherwise its ownership is managed with `take_runtime` and
/// `put_runtime`.
///
/// In general, this structure should not be used. This is meant to be an
/// unstable internal detail of the runtime itself. From time-to-time, however,
/// it is useful to manage tasks directly. An example of this would be
/// interoperating with the Rust runtime from FFI callbacks or such. For this
/// reason, there are two methods of note with the `Task` structure.
///
/// * `run` - This function will execute a closure inside the context of a task.
/// Failure is caught and handled via the task's on_exit callback. If
/// this fails, the task is still returned, but it can no longer be
/// used, it is poisoned.
///
/// * `destroy` - This is a required function to call to destroy a task. If a
/// task falls out of scope without calling `destroy`, its
/// destructor bomb will go off, aborting the process.
///
/// With these two methods, tasks can be re-used to execute code inside of its
/// context while having a point in the future where destruction is allowed.
/// More information can be found on these specific methods.
///
/// # Example
///
/// ```no_run
/// extern crate native;
/// use std::uint;
/// # fn main() {
///
/// // Create a task using a native runtime
/// let task = native::task::new((0, uint::MAX));
///
/// // Run some code, catching any possible failures
/// let task = task.run(|| {
/// // Run some code inside this task
/// println!("Hello with a native runtime!");
/// });
///
/// // Run some code again, catching the failure
/// let task = task.run(|| {
/// fail!("oh no, what to do!");
/// });
///
/// // Now that the task is failed, it can never be used again
/// assert!(task.is_destroyed());
///
/// // Deallocate the resources associated with this task
/// task.destroy();
/// # }
/// ```
pub struct Task {
pub heap: LocalHeap,
pub gc: GarbageCollector,
pub storage: LocalStorage,
pub unwinder: Unwinder,
pub death: Death,
pub name: Option<SendStr>,
state: TaskState,
imp: Option<Box<Runtime + Send + 'static>>,
}
// Once a task has entered the `Armed` state it must be destroyed via `drop`,
// and no other method. This state is used to track this transition.
#[deriving(PartialEq)]
enum TaskState {
New,
Armed,
Destroyed,
}
pub struct TaskOpts {
/// Invoke this procedure with the result of the task when it finishes.
pub on_exit: Option<proc(Result): Send>,
/// A name for the task-to-be, for identification in failure messages
pub name: Option<SendStr>,
/// The size of the stack for the spawned task
pub stack_size: Option<uint>,
}
/// Indicates the manner in which a task exited.
///
/// A task that completes without failing is considered to exit successfully.
///
/// If you wish for this result's delivery to block until all
/// children tasks complete, recommend using a result future.
pub type Result = ::core::result::Result<(), Box<Any + Send>>;
pub struct GarbageCollector;
pub struct LocalStorage(pub Option<local_data::Map>);
/// A handle to a blocked task. Usually this means having the Box<Task>
/// pointer by ownership, but if the task is killable, a killer can steal it
/// at any time.
pub enum BlockedTask {
Owned(Box<Task>),
Shared(Arc<AtomicUint>),
}
/// Per-task state related to task death, killing, failure, etc.
pub struct Death {
pub on_exit: Option<proc(Result):Send>,
marker: marker::NoCopy,
}
pub struct BlockedTasks {
inner: Arc<AtomicUint>,
}
impl Task {
/// Creates a new uninitialized task.
///
/// This method cannot be used to immediately invoke `run` because the task
/// itself will likely require a runtime to be inserted via `put_runtime`.
///
/// Note that you likely don't want to call this function, but rather the
/// task creation functions through libnative or libgreen.
pub fn new() -> Task {
Task {
heap: LocalHeap::new(),
gc: GarbageCollector,
storage: LocalStorage(None),
unwinder: Unwinder::new(),
death: Death::new(),
state: New,
name: None,
imp: None,
}
}
/// Consumes ownership of a task, runs some code, and returns the task back.
///
/// This function can be used as an emulated "try/catch" to interoperate
/// with the rust runtime at the outermost boundary. It is not possible to
/// use this function in a nested fashion (a try/catch inside of another
/// try/catch). Invoking this function is quite cheap.
///
/// If the closure `f` succeeds, then the returned task can be used again
/// for another invocation of `run`. If the closure `f` fails then `self`
/// will be internally destroyed along with all of the other associated
/// resources of this task. The `on_exit` callback is invoked with the
/// cause of failure (not returned here). This can be discovered by querying
/// `is_destroyed()`.
///
/// Note that it is possible to view partial execution of the closure `f`
/// because it is not guaranteed to run to completion, but this function is
/// guaranteed to return if it fails. Care should be taken to ensure that
/// stack references made by `f` are handled appropriately.
///
/// It is invalid to call this function with a task that has been previously
/// destroyed via a failed call to `run`.
///
/// # Example
///
/// ```no_run
/// extern crate native;
/// use std::uint;
/// # fn main() {
///
/// // Create a new native task
/// let task = native::task::new((0, uint::MAX));
///
/// // Run some code once and then destroy this task
/// task.run(|| {
/// println!("Hello with a native runtime!");
/// }).destroy();
/// # }
/// ```
pub fn run(mut self: Box<Task>, f: ||) -> Box<Task> {
assert!(!self.is_destroyed(), "cannot re-use a destroyed task");
// First, make sure that no one else is in TLS. This does not allow
// recursive invocations of run(). If there's no one else, then
// relinquish ownership of ourselves back into TLS.
if Local::exists(None::<Task>) {
fail!("cannot run a task recursively inside another");
}
self.state = Armed;
Local::put(self);
// There are two primary reasons that general try/catch is unsafe. The
// first is that we do not support nested try/catch. The above check for
// an existing task in TLS is sufficient for this invariant to be
// upheld. The second is that unwinding while unwinding is not defined.
// We take care of that by having an 'unwinding' flag in the task
// itself. For these reasons, this unsafety should be ok.
let result = unsafe { unwind::try(f) };
// After running the closure given return the task back out if it ran
// successfully, or clean up the task if it failed.
let task: Box<Task> = Local::take();
match result {
Ok(()) => task,
Err(cause) => { task.cleanup(Err(cause)) }
}
}
/// Destroy all associated resources of this task.
///
/// This function will perform any necessary clean up to prepare the task
/// for destruction. It is required that this is called before a `Task`
/// falls out of scope.
///
/// The returned task cannot be used for running any more code, but it may
/// be used to extract the runtime as necessary.
pub fn destroy(self: Box<Task>) -> Box<Task> {
if self.is_destroyed() {
self
} else {
self.cleanup(Ok(()))
}
}
/// Cleans up a task, processing the result of the task as appropriate.
///
/// This function consumes ownership of the task, deallocating it once it's
/// done being processed. It is assumed that TLD and the local heap have
/// already been destroyed and/or annihilated.
fn cleanup(self: Box<Task>, result: Result) -> Box<Task> {
// The first thing to do when cleaning up is to deallocate our local
// resources, such as TLD and GC data.
//
// FIXME: there are a number of problems with this code
//
// 1. If any TLD object fails destruction, then all of TLD will leak.
// This appears to be a consequence of #14875.
//
// 2. Failing during GC annihilation aborts the runtime #14876.
//
// 3. Setting a TLD key while destroying TLD or while destroying GC will
// abort the runtime #14807.
//
// 4. Invoking GC in GC destructors will abort the runtime #6996.
//
// 5. The order of destruction of TLD and GC matters, but either way is
// susceptible to leaks (see 3/4) #8302.
//
// That being said, there are a few upshots to this code
//
// 1. If TLD destruction fails, heap destruction will be attempted.
// There is a test for this at fail-during-tld-destroy.rs. Sadly the
// other way can't be tested due to point 2 above. Note that we must
// immortalize the heap first because if any deallocations are
// attempted while TLD is being dropped it will attempt to free the
// allocation from the wrong heap (because the current one has been
// replaced).
//
// 2. One failure in destruction is tolerable, so long as the task
// didn't originally fail while it was running.
//
// And with all that in mind, we attempt to clean things up!
let mut task = self.run(|| {
let mut task = Local::borrow(None::<Task>);
let tld = {
let &LocalStorage(ref mut optmap) = &mut task.storage;
optmap.take()
};
let mut heap = mem::replace(&mut task.heap, LocalHeap::new());
unsafe { heap.immortalize() }
drop(task);
// First, destroy task-local storage. This may run user dtors.
drop(tld);
// Destroy remaining boxes. Also may run user dtors.
drop(heap);
});
// If the above `run` block failed, then it must be the case that the
// task had previously succeeded. This also means that the code below
// was recursively run via the `run` method invoking this method. In
// this case, we just make sure the world is as we thought, and return.
if task.is_destroyed() {
rtassert!(result.is_ok())
return task
}
// After taking care of the data above, we need to transmit the result
// of this task.
let what_to_do = task.death.on_exit.take();
Local::put(task);
// FIXME: this is running in a seriously constrained context. If this
// allocates GC or allocates TLD then it will likely abort the
// runtime. Similarly, if this fails, this will also likely abort
// the runtime.
//
// This closure is currently limited to a channel send via the
// standard library's task interface, but this needs
// reconsideration to whether it's a reasonable thing to let a
// task to do or not.
match what_to_do {
Some(f) => { f(result) }
None => { drop(result) }
}
// Now that we're done, we remove the task from TLS and flag it for
// destruction.
let mut task: Box<Task> = Local::take();
task.state = Destroyed;
return task;
}
/// Queries whether this can be destroyed or not.
pub fn is_destroyed(&self) -> bool { self.state == Destroyed }
/// Inserts a runtime object into this task, transferring ownership to the
/// task. It is illegal to replace a previous runtime object in this task
/// with this argument.
pub fn put_runtime(&mut self, ops: Box<Runtime + Send + 'static>) {
assert!(self.imp.is_none());
self.imp = Some(ops);
}
/// Removes the runtime from this task, transferring ownership to the
/// caller.
pub fn take_runtime(&mut self) -> Box<Runtime + Send + 'static> {
assert!(self.imp.is_some());
self.imp.take().unwrap()
}
/// Attempts to extract the runtime as a specific type. If the runtime does
/// not have the provided type, then the runtime is not removed. If the
/// runtime does have the specified type, then it is removed and returned
/// (transfer of ownership).
///
/// It is recommended to only use this method when *absolutely necessary*.
/// This function may not be available in the future.
pub fn maybe_take_runtime<T: 'static>(&mut self) -> Option<Box<T>> {
// This is a terrible, terrible function. The general idea here is to
// take the runtime, cast it to Box<Any>, check if it has the right
// type, and then re-cast it back if necessary. The method of doing
// this is pretty sketchy and involves shuffling vtables of trait
// objects around, but it gets the job done.
//
// FIXME: This function is a serious code smell and should be avoided at
// all costs. I have yet to think of a method to avoid this
// function, and I would be saddened if more usage of the function
// crops up.
unsafe {
let imp = self.imp.take().unwrap();
let vtable = mem::transmute::<_, &raw::TraitObject>(&imp).vtable;
match imp.wrap().downcast::<T>() {
Ok(t) => Some(t),
Err(t) => {
let data = mem::transmute::<_, raw::TraitObject>(t).data;
let obj: Box<Runtime + Send + 'static> =
mem::transmute(raw::TraitObject {
vtable: vtable,
data: data,
});
self.put_runtime(obj);
None
}
}
}
}
/// Spawns a sibling to this task. The newly spawned task is configured with
/// the `opts` structure and will run `f` as the body of its code.
pub fn spawn_sibling(mut self: Box<Task>,
opts: TaskOpts,
f: proc(): Send) {
let ops = self.imp.take().unwrap();
ops.spawn_sibling(self, opts, f)
}
/// Deschedules the current task, invoking `f` `amt` times. It is not
/// recommended to use this function directly, but rather communication
/// primitives in `std::comm` should be used.
pub fn deschedule(mut self: Box<Task>,
amt: uint,
f: |BlockedTask| -> ::core::result::Result<(), BlockedTask>) {
let ops = self.imp.take().unwrap();
ops.deschedule(amt, self, f)
}
/// Wakes up a previously blocked task, optionally specifying whether the
/// current task can accept a change in scheduling. This function can only
/// be called on tasks that were previously blocked in `deschedule`.
pub fn reawaken(mut self: Box<Task>) {
let ops = self.imp.take().unwrap();
ops.reawaken(self);
}
/// Yields control of this task to another task. This function will
/// eventually return, but possibly not immediately. This is used as an
/// opportunity to allow other tasks a chance to run.
pub fn yield_now(mut self: Box<Task>) {
let ops = self.imp.take().unwrap();
ops.yield_now(self);
}
/// Similar to `yield_now`, except that this function may immediately return
/// without yielding (depending on what the runtime decides to do).
pub fn maybe_yield(mut self: Box<Task>) {
let ops = self.imp.take().unwrap();
ops.maybe_yield(self);
}
/// Acquires a handle to the I/O factory that this task contains, normally
/// stored in the task's runtime. This factory may not always be available,
/// which is why the return type is `Option`
pub fn local_io<'a>(&'a mut self) -> Option<LocalIo<'a>> {
self.imp.as_mut().unwrap().local_io()
}
/// Returns the stack bounds for this task in (lo, hi) format. The stack
/// bounds may not be known for all tasks, so the return value may be
/// `None`.
pub fn stack_bounds(&self) -> (uint, uint) {
self.imp.as_ref().unwrap().stack_bounds()
}
/// Returns whether it is legal for this task to block the OS thread that it
/// is running on.
pub fn can_block(&self) -> bool {
self.imp.as_ref().unwrap().can_block()
}
/// Consume this task, flagging it as a candidate for destruction.
///
/// This function is required to be invoked to destroy a task. A task
/// destroyed through a normal drop will abort.
pub fn drop(mut self) {
self.state = Destroyed;
}
}
impl Drop for Task {
fn drop(&mut self) {
rtdebug!("called drop for a task: {}", self as *mut Task as uint);
rtassert!(self.state != Armed);
}
}
impl TaskOpts {
pub fn new() -> TaskOpts {
TaskOpts { on_exit: None, name: None, stack_size: None }
}
}
impl Iterator<BlockedTask> for BlockedTasks {
fn next(&mut self) -> Option<BlockedTask> {
Some(Shared(self.inner.clone()))
}
}
impl BlockedTask {
/// Returns Some if the task was successfully woken; None if already killed.
pub fn wake(self) -> Option<Box<Task>> {
match self {
Owned(task) => Some(task),
Shared(arc) => {
match arc.swap(0, SeqCst) {
0 => None,
n => Some(unsafe { mem::transmute(n) }),
}
}
}
}
/// Reawakens this task if ownership is acquired. If finer-grained control
/// is desired, use `wake` instead.
pub fn reawaken(self) {
self.wake().map(|t| t.reawaken());
}
// This assertion has two flavours because the wake involves an atomic op.
// In the faster version, destructors will fail dramatically instead.
#[cfg(not(test))] pub fn trash(self) { }
#[cfg(test)] pub fn trash(self) { assert!(self.wake().is_none()); }
/// Create a blocked task, unless the task was already killed.
pub fn block(task: Box<Task>) -> BlockedTask {
Owned(task)
}
/// Converts one blocked task handle to a list of many handles to the same.
pub fn make_selectable(self, num_handles: uint) -> Take<BlockedTasks> {
let arc = match self {
Owned(task) => {
let flag = unsafe { AtomicUint::new(mem::transmute(task)) };
Arc::new(flag)
}
Shared(arc) => arc.clone(),
};
BlockedTasks{ inner: arc }.take(num_handles)
}
/// Convert to an unsafe uint value. Useful for storing in a pipe's state
/// flag.
#[inline]
pub unsafe fn cast_to_uint(self) -> uint {
match self {
Owned(task) => {
let blocked_task_ptr: uint = mem::transmute(task);
rtassert!(blocked_task_ptr & 0x1 == 0);
blocked_task_ptr
}
Shared(arc) => {
let blocked_task_ptr: uint = mem::transmute(box arc);
rtassert!(blocked_task_ptr & 0x1 == 0);
blocked_task_ptr | 0x1
}
}
}
/// Convert from an unsafe uint value. Useful for retrieving a pipe's state
/// flag.
#[inline]
pub unsafe fn cast_from_uint(blocked_task_ptr: uint) -> BlockedTask {
if blocked_task_ptr & 0x1 == 0 {
Owned(mem::transmute(blocked_task_ptr))
} else {
let ptr: Box<Arc<AtomicUint>> =
mem::transmute(blocked_task_ptr & !1);
Shared(*ptr)
}
}
}
impl Death {
pub fn new() -> Death {
Death { on_exit: None, marker: marker::NoCopy }
}
}
#[cfg(test)]
mod test {
use super::*;
use std::prelude::*;
use std::task;
use std::gc::{Gc, GC};
#[test]
fn local_heap() {
let a = box(GC) 5i;
let b = a;
assert!(*a == 5);
assert!(*b == 5);
}
#[test]
fn tls() {
local_data_key!(key: Gc<String>)
key.replace(Some(box(GC) "data".to_string()));
assert_eq!(key.get().unwrap().as_slice(), "data");
local_data_key!(key2: Gc<String>)
key2.replace(Some(box(GC) "data".to_string()));
assert_eq!(key2.get().unwrap().as_slice(), "data");
}
#[test]
fn unwind() {
let result = task::try(proc()());
rtdebug!("trying first assert");
assert!(result.is_ok());
let result = task::try::<()>(proc() fail!());
rtdebug!("trying second assert");
assert!(result.is_err());
}
#[test]
fn rng() {
use std::rand::{StdRng, Rng};
let mut r = StdRng::new().ok().unwrap();
let _ = r.next_u32();
}
#[test]
fn comm_stream() {
let (tx, rx) = channel();
tx.send(10i);
assert!(rx.recv() == 10);
}
#[test]
fn comm_shared_chan() {
let (tx, rx) = channel();
tx.send(10i);
assert!(rx.recv() == 10);
}
#[test]
fn heap_cycles() {
use std::cell::RefCell;
struct List {
next: Option<Gc<RefCell<List>>>,
}
let a = box(GC) RefCell::new(List { next: None });
let b = box(GC) RefCell::new(List { next: Some(a) });
{
let mut a = a.borrow_mut();
a.next = Some(b);
}
}
#[test]
#[should_fail]
fn test_begin_unwind() {
use std::rt::unwind::begin_unwind;
begin_unwind("cause", &(file!(), line!()))
}
#[test]
fn drop_new_task_ok() {
drop(Task::new());
}
// Task blocking tests
#[test]
fn block_and_wake() {
let task = box Task::new();
let task = BlockedTask::block(task).wake().unwrap();
task.drop();
}
}