[−][src]Struct crossbeam_utils::CachePadded
Pads and aligns a value to the length of a cache line.
In concurrent programming, sometimes it is desirable to make sure commonly accessed pieces of
data are not placed into the same cache line. Updating an atomic value invalides the whole
cache line it belongs to, which makes the next access to the same cache line slower for other
CPU cores. Use CachePadded
to ensure updating one piece of data doesn't invalidate other
cached data.
Cache lines are assumed to be 64 bytes on all architectures.
Size and alignment
The size of CachePadded<T>
is the smallest multiple of 64 bytes large enough to accommodate
a value of type T
.
The alignment of CachePadded<T>
is the maximum of 64 bytes and the alignment of T
.
Examples
Alignment and padding:
use crossbeam_utils::CachePadded; let array = [CachePadded::new(1i32), CachePadded::new(2i32)]; let addr1 = &*array[0] as *const i32 as usize; let addr2 = &*array[1] as *const i32 as usize; assert_eq!(addr2 - addr1, 64); assert_eq!(addr1 % 64, 0); assert_eq!(addr2 % 64, 0);
When building a concurrent queue with a head and a tail index, it is wise to place them in different cache lines so that concurrent threads pushing and popping elements don't invalidate each other's cache lines:
use crossbeam_utils::CachePadded; use std::sync::atomic::AtomicUsize; struct Queue<T> { head: CachePadded<AtomicUsize>, tail: CachePadded<AtomicUsize>, buffer: *mut T, }
Methods
impl<T> CachePadded<T>
[src][−]
pub fn new(t: T) -> CachePadded<T>
[src][+]
pub fn into_inner(self) -> T
[src][+]
Trait Implementations
impl<T: Sync> Sync for CachePadded<T>
[src]
impl<T: PartialEq> PartialEq<CachePadded<T>> for CachePadded<T>
[src][+]
impl<T: Clone> Clone for CachePadded<T>
[src][+]
impl<T> From<T> for CachePadded<T>
[src][+]
impl<T: Eq> Eq for CachePadded<T>
[src]
impl<T: Send> Send for CachePadded<T>
[src]
impl<T: Copy> Copy for CachePadded<T>
[src]
impl<T: Default> Default for CachePadded<T>
[src][+]
impl<T> DerefMut for CachePadded<T>
[src][+]
impl<T: Debug> Debug for CachePadded<T>
[src][+]
impl<T: Hash> Hash for CachePadded<T>
[src][+]
impl<T> Deref for CachePadded<T>
[src][+]
Blanket Implementations
impl<T> From for T
[src][+]
impl<T, U> Into for T where
U: From<T>,
[src][+]
U: From<T>,
impl<T> ToOwned for T where
T: Clone,
[src][+]
T: Clone,
impl<T, U> TryFrom for T where
U: Into<T>,
[src][+]
U: Into<T>,
impl<T> Borrow for T where
T: ?Sized,
[src][+]
T: ?Sized,
impl<T> Any for T where
T: 'static + ?Sized,
[src][+]
T: 'static + ?Sized,
impl<T> BorrowMut for T where
T: ?Sized,
[src][+]
T: ?Sized,
impl<T, U> TryInto for T where
U: TryFrom<T>,
[src][+]
U: TryFrom<T>,