Struct sunrise_kernel::sync::SpinRwLock

pub struct SpinRwLock<T> where
    T: ?Sized,  {
    lock: AtomicUsize,
    data: UnsafeCell<T>,
}

A reader-writer lock

This type of lock allows a number of readers or at most one writer at any point in time. The write portion of this lock typically allows modification of the underlying data (exclusive access) and the read portion of this lock typically allows for read-only access (shared access).

The type parameter T represents the data that this lock protects. It is required that T satisfies Send to be shared across tasks and Sync to allow concurrent access through readers. The RAII guards returned from the locking methods implement Deref (and DerefMut for the write methods) to allow access to the contained of the lock.

An RwLockUpgradeableGuard can be upgraded to a writable guard through the RwLockUpgradeableGuard::upgrade RwLockUpgradeableGuard::try_upgrade functions. Writable or upgradeable guards can be downgraded through their respective downgrade functions.

Based on Facebook's folly/RWSpinLock.h. This implementation is unfair to writers - if the lock always has readers, then no writers will ever get a chance. Using an upgradeable lock guard can somewhat alleviate this issue as no new readers are allowed when an upgradeable guard is held, but upgradeable guards can be taken when there are existing readers. However if the lock is that highly contended and writes are crucial then this implementation may be a poor choice.

Examples

use spin;

let lock = spin::RwLock::new(5);

// many reader locks can be held at once
{
    let r1 = lock.read();
    let r2 = lock.read();
    assert_eq!(*r1, 5);
    assert_eq!(*r2, 5);
} // read locks are dropped at this point

// only one write lock may be held, however
{
    let mut w = lock.write();
    *w += 1;
    assert_eq!(*w, 6);
} // write lock is dropped here

Fields

lock: AtomicUsize

data: UnsafeCell<T>

Implementations

impl<T> RwLock<T>

pub fn new(user_data: T) -> RwLock<T>

Creates a new spinlock wrapping the supplied data.

May be used statically:

use spin;

static RW_LOCK: spin::RwLock<()> = spin::RwLock::new(());

fn demo() {
    let lock = RW_LOCK.read();
    // do something with lock
    drop(lock);
}

pub fn into_inner(self) -> T

Consumes this RwLock, returning the underlying data.

impl<T> RwLock<T> where
    T: ?Sized, 

pub fn read(&self) -> RwLockReadGuard<T>

Locks this rwlock with shared read access, blocking the current thread until it can be acquired.

The calling thread will be blocked until there are no more writers which hold the lock. There may be other readers currently inside the lock when this method returns. This method does not provide any guarantees with respect to the ordering of whether contentious readers or writers will acquire the lock first.

Returns an RAII guard which will release this thread's shared access once it is dropped.

let mylock = spin::RwLock::new(0);
{
    let mut data = mylock.read();
    // The lock is now locked and the data can be read
    println!("{}", *data);
    // The lock is dropped
}

pub fn try_read(&self) -> Option<RwLockReadGuard<T>>

Attempt to acquire this lock with shared read access.

This function will never block and will return immediately if read would otherwise succeed. Returns Some of an RAII guard which will release the shared access of this thread when dropped, or None if the access could not be granted. This method does not provide any guarantees with respect to the ordering of whether contentious readers or writers will acquire the lock first.

let mylock = spin::RwLock::new(0);
{
    match mylock.try_read() {
        Some(data) => {
            // The lock is now locked and the data can be read
            println!("{}", *data);
            // The lock is dropped
        },
        None => (), // no cigar
    };
}

pub unsafe fn force_read_decrement(&self)

Force decrement the reader count.

This is extremely unsafe if there are outstanding RwLockReadGuards live, or if called more times than read has been called, but can be useful in FFI contexts where the caller doesn't know how to deal with RAII. The underlying atomic operation uses Ordering::Release.

pub unsafe fn force_write_unlock(&self)

Force unlock exclusive write access.

This is extremely unsafe if there are outstanding RwLockWriteGuards live, or if called when there are current readers, but can be useful in FFI contexts where the caller doesn't know how to deal with RAII. The underlying atomic operation uses Ordering::Release.

pub fn write(&self) -> RwLockWriteGuard<T>

Lock this rwlock with exclusive write access, blocking the current thread until it can be acquired.

This function will not return while other writers or other readers currently have access to the lock.

Returns an RAII guard which will drop the write access of this rwlock when dropped.

let mylock = spin::RwLock::new(0);
{
    let mut data = mylock.write();
    // The lock is now locked and the data can be written
    *data += 1;
    // The lock is dropped
}

pub fn try_write(&self) -> Option<RwLockWriteGuard<T>>

Attempt to lock this rwlock with exclusive write access.

This function does not ever block, and it will return None if a call to write would otherwise block. If successful, an RAII guard is returned.

let mylock = spin::RwLock::new(0);
{
    match mylock.try_write() {
        Some(mut data) => {
            // The lock is now locked and the data can be written
            *data += 1;
            // The lock is implicitly dropped
        },
        None => (), // no cigar
    };
}

pub fn upgradeable_read(&self) -> RwLockUpgradeableGuard<T>

Obtain a readable lock guard that can later be upgraded to a writable lock guard. Upgrades can be done through the RwLockUpgradeableGuard::upgrade method.

pub fn try_upgradeable_read(&self) -> Option<RwLockUpgradeableGuard<T>>

Tries to obtain an upgradeable lock guard.

Trait Implementations

impl<T> Debug for RwLock<T> where
    T: Debug + ?Sized, 

fn fmt(&self, f: &mut Formatter) -> Result<(), Error>

Formats the value using the given formatter.

impl<T> Default for RwLock<T> where
    T: Default + ?Sized, 

fn default() -> RwLock<T>

Returns the "default value" for a type.

impl<'a, T> Lock<'a, RwLockReadGuard<'a, T>> for SpinRwLock<T>

fn lock(&self) -> SpinRwLockReadGuard<T>

Locks the lock until the returned guard is dropped. The actual semantics is up to the underlying lock. This trait does not make any guarantee about anything like memory ordering or whatnot. Please read the documentation of the underlying lock type.

impl<'a, T> Lock<'a, RwLockWriteGuard<'a, T>> for SpinRwLock<T>

fn lock(&self) -> SpinRwLockWriteGuard<T>

Locks the lock until the returned guard is dropped. The actual semantics is up to the underlying lock. This trait does not make any guarantee about anything like memory ordering or whatnot. Please read the documentation of the underlying lock type.

impl<T> Send for RwLock<T> where
    T: Send + ?Sized, 

impl<T> Sync for RwLock<T> where
    T: Send + Sync + ?Sized,