Skip to content

Latest commit

 

History

History
487 lines (368 loc) · 17.5 KB

README.md

File metadata and controls

487 lines (368 loc) · 17.5 KB

AtomicFU

Kotlin Beta JetBrains official project GitHub license Maven Central

Note on Beta status: the plugin is in its active development phase and changes from release to release. We do provide a compatibility of atomicfu-transformed artifacts between releases, but we do not provide strict compatibility guarantees on plugin API and its general stability between Kotlin versions.

Atomicfu is a multiplatform library that provides the idiomatic and efficient way of using atomic operations in Kotlin.

Table of contents

Requirements

Starting from version 0.24.0 of the library your project is required to use:

  • Gradle 7.0 or newer

  • Kotlin 1.9.0 or newer

Note on Kotlin version: Currently, the kotlinx-atomicfu Gradle plugin only relies on the version of Kotlin Gradle Plugin (KGP) present in the user's project. It's important to note this constraint if your project configures the custom Kotlin compiler version or modifies the Kotlin Native compiler version using kotlin.native.version property.

In case, you cannot upgrade the Kotlin version to 1.9.0 or newer in your project, you can downgrade kotlinx-atomicfu plugin version to 0.22.0. Please note, though, that using the latest version of the plugin and upgrading the Kotlin version is the more recommended approach.

Features

  • Complete multiplatform support: JVM, Native, JS and Wasm (since Kotlin 1.9.20).
  • Code it like a boxed value atomic(0), but run it in production efficiently:
    • For JVM: an atomic value is represented as a plain value atomically updated with java.util.concurrent.atomic.AtomicXxxFieldUpdater from the Java standard library.
    • For JS: an atomic value is represented as a plain value.
    • For Native: atomic operations are delegated to Kotlin/Native atomic intrinsics.
    • For Wasm: an atomic value is not transformed, it remains boxed, and kotlinx-atomicfu library is used as a runtime dependency.
  • Use Kotlin-specific extensions (e.g. inline loop, update, updateAndGet functions).
  • Use atomic arrays, user-defined extensions on atomics and locks (see more features).
  • Tracing operations for debugging.

Example

Let us declare a top variable for a lock-free stack implementation:

import kotlinx.atomicfu.* // import top-level functions from kotlinx.atomicfu

private val top = atomic<Node?>(null) 

Use top.value to perform volatile reads and writes:

fun isEmpty() = top.value == null  // volatile read
fun clear() { top.value = null }   // volatile write

Use compareAndSet function directly:

if (top.compareAndSet(expect, update)) ... 

Use higher-level looping primitives (inline extensions), for example:

top.loop { cur ->   // while(true) loop that volatile-reads current value 
   ...
}

Use high-level update, updateAndGet, and getAndUpdate, when possible, for idiomatic lock-free code, for example:

fun push(v: Value) = top.update { cur -> Node(v, cur) }
fun pop(): Value? = top.getAndUpdate { cur -> cur?.next } ?.value

Declare atomic integers and longs using type inference:

val myInt = atomic(0)    // note: integer initial value
val myLong = atomic(0L)  // note: long initial value   

Integer and long atomics provide all the usual getAndIncrement, incrementAndGet, getAndAdd, addAndGet, and etc operations. They can be also atomically modified via += and -= operators.

Quickstart

Apply plugin

Gradle configuration

New plugin id: Please pay attention, that starting from version 0.25.0 the plugin id is org.jetbrains.kotlinx.atomicfu

Add the following to your top-level build file:

Kotlin
plugins {
     id("org.jetbrains.kotlinx.atomicfu") version "0.26.1"
}
Groovy
plugins {
    id 'org.jetbrains.kotlinx.atomicfu' version '0.26.1'
}

Legacy plugin application

Kotlin
buildscript {
  repositories {
    mavenCentral()
  }

  dependencies {
    classpath("org.jetbrains.kotlinx:atomicfu-gradle-plugin:0.26.1")
  }
}

apply(plugin = "org.jetbrains.kotlinx.atomicfu")
Groovy
buildscript {
    repositories {
        mavenCentral()
    }
    dependencies {
        classpath 'org.jetbrains.kotlinx:atomicfu-gradle-plugin:0.26.1'
    }
}
  
apply plugin: 'org.jetbrains.kotlinx.atomicfu'

Maven configuration

Maven configuration is supported for JVM projects.

Declare atomicfu version
<properties>
     <atomicfu.version>0.26.1</atomicfu.version>
</properties> 
Declare provided dependency on the AtomicFU library
<dependencies>
    <dependency>
        <groupId>org.jetbrains.kotlinx</groupId>
        <artifactId>atomicfu</artifactId>
        <version>${atomicfu.version}</version>
        <scope>provided</scope>
    </dependency>
</dependencies>

Configure build steps so that Kotlin compiler puts classes into a different classes-pre-atomicfu directory, which is then transformed to a regular classes directory to be used later by tests and delivery.

Build steps
<build>
  <plugins>
    <!-- compile Kotlin files to staging directory -->
    <plugin>
      <groupId>org.jetbrains.kotlin</groupId>
      <artifactId>kotlin-maven-plugin</artifactId>
      <version>${kotlin.version}</version>
      <executions>
        <execution>
          <id>compile</id>
          <phase>compile</phase>
          <goals>
            <goal>compile</goal>
          </goals>
          <configuration>
            <output>${project.build.directory}/classes-pre-atomicfu</output>
          </configuration>
        </execution>
      </executions>
    </plugin>
    <!-- transform classes with AtomicFU plugin -->
    <plugin>
      <groupId>org.jetbrains.kotlinx</groupId>
      <artifactId>atomicfu-maven-plugin</artifactId>
      <version>${atomicfu.version}</version>
      <executions>
        <execution>
          <goals>
            <goal>transform</goal>
          </goals>
          <configuration>
            <input>${project.build.directory}/classes-pre-atomicfu</input>
            <!-- "VH" to use Java 9 VarHandle, "BOTH" to produce multi-version code -->
            <variant>FU</variant>
          </configuration>
        </execution>
      </executions>
    </plugin>
  </plugins>
</build>

Usage constraints

  • Declare atomic variables as private val or internal val. You can use just (public) val, but make sure they are not directly accessed outside of your Kotlin module (outside of the source set). Access to the atomic variable itself shall be encapsulated.
  • To expose the value of an atomic property to the public, use a delegated property declared in the same scope (see atomic delegates section for details):
private val _foo = atomic<T>(initial) // private atomic, convention is to name it with leading underscore
public var foo: T by _foo            // public delegated property (val/var)
  • Only simple operations on atomic variables directly are supported.
    • Do not read references on atomic variables into local variables, e.g. top.compareAndSet(...) is ok, while val tmp = top; tmp... is not.
    • Do not leak references on atomic variables in other way (return, pass as params, etc).
  • Do not introduce complex data flow in parameters to atomic variable operations, i.e. top.value = complex_expression and top.compareAndSet(cur, complex_expression) are not supported (more specifically, complex_expression should not have branches in its compiled representation). Extract complex_expression into a variable when needed.

Atomicfu compiler plugin

To provide a user-friendly atomic API on the frontend and efficient usage of atomic values on the backend kotlinx-atomicfu library uses the compiler plugin to transform IR for all the target backends:

  • JVM: atomics are replaced with java.util.concurrent.atomic.AtomicXxxFieldUpdater.
  • Native: atomics are implemented via atomic intrinsics on Kotlin/Native.
  • JS: atomics are unboxed and represented as plain values.

To turn on IR transformations set the following properties in your gradle.properties file:

Please note, that starting from version 0.24.0 of the library your project is required to use Kotlin version >= 1.9.0. See the requirements section.

kotlinx.atomicfu.enableJvmIrTransformation=true // for JVM IR transformation
kotlinx.atomicfu.enableNativeIrTransformation=true // for Native IR transformation
kotlinx.atomicfu.enableJsIrTransformation=true // for JS IR transformation
Here are the configuration properties in case you use older versions of the library lower than 0.24.0.

For Kotlin >= 1.7.20

kotlinx.atomicfu.enableJvmIrTransformation=true // for JVM IR transformation
kotlinx.atomicfu.enableNativeIrTransformation=true // for Native IR transformation
kotlinx.atomicfu.enableJsIrTransformation=true // for JS IR transformation

For Kotlin >= 1.6.20 and Kotlin < 1.7.20

kotlinx.atomicfu.enableIrTransformation=true // only JS IR transformation is supported

Also for JS backend make sure that ir or both compiler mode is set:

kotlin.js.compiler=ir // or both

Options for post-compilation transformation

Some configuration options are available for post-compilation transform tasks on JVM and JS.

To set configuration options you should create atomicfu section in a build.gradle file, like this:

atomicfu {
  dependenciesVersion = '0.26.1'
}

JVM options

To turn off transformation for Kotlin/JVM set option transformJvm to false.

Configuration option jvmVariant defines the Java class that replaces atomics during bytecode transformation. Here are the valid options:

JS options

Starting from version 0.26.0 transformJs flag does not take any effect and is disabled by default. Please ensure that this flag is not used in the atomicfu configuration of your project, you can safely remove it.

Here are all available configuration options (with their defaults):

atomicfu {
  dependenciesVersion = '0.26.1' // set to null to turn-off auto dependencies
  transformJvm = true // set to false to turn off JVM transformation
  jvmVariant = "FU" // JVM transformation variant: FU,VH, or BOTH
}

More features

AtomicFU provides some additional features that you can use.

Arrays of atomic values

You can declare arrays of all supported atomic value types. By default arrays are transformed into the corresponding java.util.concurrent.atomic.Atomic*Array instances.

If you configure variant = "VH" an array will be transformed to plain array using VarHandle to support atomic operations.

val a = atomicArrayOfNulls<T>(size) // similar to Array constructor

val x = a[i].value // read value
a[i].value = x // set value
a[i].compareAndSet(expect, update) // do atomic operations

Atomic delegates

You can expose the value of an atomic property to the public, using a delegated property declared in the same scope:

private val _foo = atomic<T>(initial) // private atomic, convention is to name it with leading underscore
public var foo: T by _foo            // public delegated property (val/var)

You can also delegate a property to the atomic factory invocation, that is equal to declaring a volatile property:

public var foo: T by atomic(0)

This feature is only supported for the IR transformation mode, see the atomicfu compiler plugin section for details.

User-defined extensions on atomics

You can define you own extension functions on AtomicXxx types but they must be inline and they cannot be public and be used outside of the module they are defined in. For example:

@Suppress("NOTHING_TO_INLINE")
private inline fun AtomicBoolean.tryAcquire(): Boolean = compareAndSet(false, true)

Locks

This project includes kotlinx.atomicfu.locks package providing multiplatform locking primitives that require no additional runtime dependencies on Kotlin/JVM and Kotlin/JS with a library implementation for Kotlin/Native.

  • SynchronizedObject is designed for inheritance. You write class MyClass : SynchronizedObject() and then use synchronized(instance) { ... } extension function similarly to the synchronized function from the standard library that is available for JVM. The SynchronizedObject superclass gets erased (transformed to Any) on JVM and JS, with synchronized leaving no trace in the code on JS and getting replaced with built-in monitors for locking on JVM.

  • ReentrantLock is designed for delegation. You write val lock = reentrantLock() to construct its instance and use lock/tryLock/unlock functions or lock.withLock { ... } extension function similarly to the way jucl.ReentrantLock is used on JVM. On JVM it is a typealias to the later class, erased on JS.

Note that package kotlinx.atomicfu.locks is experimental explicitly even while atomicfu is experimental itself, meaning that no ABI guarantees are provided whatsoever. API from this package is not recommended to use in libraries that other projects depend on.

Tracing operations

You can debug your tests tracing atomic operations with a special trace object:

private val trace = Trace()
private val current = atomic(0, trace)

fun update(x: Int): Int {           
    // custom trace message
    trace { "calling update($x)" }
    // automatic tracing of modification operations 
    return current.getAndAdd(x)
}

All trace messages are stored in a cyclic array inside trace.

You can optionally set the size of trace's message array and format function. For example, you can add a current thread name to the traced messages:

private val trace = Trace(size = 64) {   
    index, // index of a trace message 
    text   // text passed when invoking trace { text }
    -> "$index: [${Thread.currentThread().name}] $text" 
} 

trace is only seen before transformation and completely erased after on Kotlin/JVM and Kotlin/JS.

Kotlin Native support

Since Kotlin/Native does not generally provide binary compatibility between versions, you should use the same version of Kotlin compiler as was used to build AtomicFU. See gradle.properties in AtomicFU project for its kotlin_version.

Available Kotlin/Native targets are based on non-deprecated official targets Tier list with the corresponding compatibility guarantees.

Gradle Build Scans

Gradle Build Scans can provide insights into an Atomicfu Build. JetBrains runs a Gradle Develocity server. that can be used to automatically upload reports.

To automatically opt in add the following to $GRADLE_USER_HOME/gradle.properties.

org.jetbrains.atomicfu.build.scan.enabled=true
# optionally provide a username that will be attached to each report
org.jetbrains.atomicfu.build.scan.username=John Wick

A Build Scan may contain identifiable information. See the Terms of Use https://gradle.com/legal/terms-of-use/.