A small toolkit of classes that support compare-and-swap semantics for mutation of variables.
On top of the JVM, this means dealing with lock-free thread-safe programming.
On top of Javascript / Scala.js using Atomic references is still good because:
1. boxing values in a smart reference with nice helpers for transformations is always a good idea.
2. on the JVM there are times when synchronization, and when used for synchronization, atomic
references can now cross compile to Scala.js
3. compareAndSet is actually a good idea to have even in an asynchronous, non-multi-threaded
environment, such as Javascript, because it takes time into account and time related problems
can happen even without multi-threading
The backbone of Atomic references is this method:
def compareAndSet(expect: T, update: T): Boolean
This method atomically sets a variable to the update value if it currently holds
the expect value, reporting true on success or false on failure. The classes in this package
also contain methods to get and unconditionally set values. In comparison with the JVM version,
these Atomic references do not have methods for weakly setting values (i.e. weakCompareAndSet, lazySet),
since those really make no sense in Javascript.
Building a reference is easy with the provided constructor, which will automatically return the
most specific type needed:
val atomicNumber = Atomic(12L)
atomicNumber.incrementAndGet()
In comparison with java.util.concurrent.AtomicReference, these references implement common interfaces
that you can use generically (i.e. Atomic[T], AtomicNumber[T]). And also provide useful helpers for
atomically mutating of values (i.e. transform, transformAndGet, getAndTransform, etc...).
Other differences with the JVM-variant - in Scala.js you do not have access to the methods meant to
block (spin-lock) the current thread (e.g. waitForCompareAndSet, waitForCondition, etc...), as
the semantics of those operations aren't possible on top of Scala.js
A small toolkit of classes that support compare-and-swap semantics for mutation of variables.
On top of the JVM, this means dealing with lock-free thread-safe programming. On top of Javascript / Scala.js using
Atomic
references is still good because:1. boxing values in a smart reference with nice helpers for transformations is always a good idea. 2. on the JVM there are times when synchronization, and when used for synchronization, atomic references can now cross compile to Scala.js 3.
compareAndSet
is actually a good idea to have even in an asynchronous, non-multi-threaded environment, such as Javascript, because it takes time into account and time related problems can happen even without multi-threadingThe backbone of Atomic references is this method:
This method atomically sets a variable to the
update
value if it currently holds theexpect
value, reportingtrue
on success orfalse
on failure. The classes in this package also contain methods to get and unconditionally set values. In comparison with the JVM version, theseAtomic
references do not have methods for weakly setting values (i.e.weakCompareAndSet
,lazySet
), since those really make no sense in Javascript.Building a reference is easy with the provided constructor, which will automatically return the most specific type needed:
In comparison with
java.util.concurrent.AtomicReference
, these references implement common interfaces that you can use generically (i.e.Atomic[T]
,AtomicNumber[T]
). And also provide useful helpers for atomically mutating of values (i.e.transform
,transformAndGet
,getAndTransform
, etc...).Other differences with the JVM-variant - in Scala.js you do not have access to the methods meant to block (spin-lock) the current thread (e.g.
waitForCompareAndSet
,waitForCondition
, etc...), as the semantics of those operations aren't possible on top of Scala.js