View source for Java virtual machine

{{Short description|Virtual machine that runs Java programs}}
{{Redirect|JVM}}
{{Infobox CPU architecture
| name       = Java virtual machine
| designer   = [[Sun Microsystems]]
| bits       = [[32-bit computing|32-bit]], [[64-bit computing|64-bit]]{{efn|32-bit x86 support was removed starting with JDK 21 (Windows)<ref>{{cite web |title=JEP 479: Remove the Windows 32-bit x86 Port |url=https://openjdk.org/jeps/479 |website=OpenJDK}}</ref> and JDK 25 (Linux).<ref>{{cite web |title=JEP 503: Remove the 32-bit x86 Port |url=https://openjdk.org/jeps/503 |website=OpenJDK}}</ref>}}
| introduced = 1994
| version    = {{wikidata|property|preferred|references|edit|Q251|P348|P548=Q2804309}}
| design     = 
| type       = [[Stack machine|Stack]] and [[Load–store architecture|register–register]]
| encoding   = Variable
| branching  = Compare and branch
| endianness = [[Endianness#Big-endian|Big]]
| page size  = 
| extensions = 
| open       = Yes
| registers  = 
| gpr        = Per-method operand stack (up to 65535 operands) plus per-method local variables (up to 65535)
| fpr        = 
}}
[[File:JvmSpec7.png|thumb|Overview of a Java virtual machine (JVM) architecture based on The Java Virtual Machine Specification Java SE 7 Edition]]
A '''Java virtual machine''' ('''JVM''') is a [[virtual machine]] that enables a computer to run [[Java (software platform)|Java]] programs as well as programs written in [[List of JVM languages|other languages]] that are also compiled to [[Java bytecode]]. The JVM is detailed by a [[specification (technical standard)|specification]] that formally describes what is required in a JVM implementation. Having a specification ensures interoperability of Java programs across different implementations so that program authors using the [[Java Development Kit]] (JDK) need not worry about idiosyncrasies of the underlying hardware platform.

The JVM [[reference implementation]] is developed by the [[OpenJDK]] project as [[Open-source software|open source]] code and includes a [[JIT compiler]] called [[HotSpot (virtual machine)|HotSpot]]. The commercially supported Java releases available from [[Oracle Corporation|Oracle]] are based on the OpenJDK runtime. Eclipse [[OpenJ9]] is another open source JVM for OpenJDK.

==JVM specification==
The Java virtual machine is an abstract (virtual) computer defined by a specification.  It is a part of the Java runtime environment. The [[Garbage collection (computer science)|garbage collection]] algorithm used and any internal optimization of the Java virtual machine instructions (their translation into [[machine code]]) are not specified. The main reason for this omission is to not unnecessarily constrain implementers. Any Java application can be run only inside some concrete implementation of the abstract specification of the Java virtual machine.<ref>Bill Venners, ''[http://www.artima.com/insidejvm/ed2/index.html Inside the Java Virtual Machine] {{Webarchive|url=https://web.archive.org/web/20210125092727/http://www.artima.com/insidejvm/ed2/index.html |date=2021-01-25 }}'' Chapter 5</ref>

Starting with [[Java Platform, Standard Edition]] (J2SE) 5.0, changes to the JVM specification have been developed under the [[Java Community Process]] as JSR 924.<ref>{{cite web |url=http://www.jcp.org/en/jsr/detail?id=924 |title=The Java Community Process(SM) Program - JSRs: Java Specification Requests - detail JSR# 924 |publisher=Jcp.org |access-date=2015-06-26 |archive-date=2020-12-24 |archive-url=https://web.archive.org/web/20201224125612/https://jcp.org/en/jsr/detail?id=924 |url-status=live }}</ref> {{As of|2006}}, changes to the specification to support changes proposed to the [[class (file format)|class file format]] (JSR 202)<ref>{{cite web |url=http://www.jcp.org/en/jsr/detail?id=202 |title=The Java Community Process(SM) Program - JSRs: Java Specification Requests - detail JSR# 202 |publisher=Jcp.org |access-date=2015-06-26 |archive-date=2012-02-26 |archive-url=https://web.archive.org/web/20120226185155/http://www.jcp.org/en/jsr/detail?id=202 |url-status=live }}</ref> are being done as a maintenance release of JSR 924. The specification for the JVM was published as the ''blue book'',<ref>''[http://java.sun.com/docs/books/vmspec/ The Java Virtual Machine Specification] {{Webarchive|url=https://web.archive.org/web/20080709010412/http://java.sun.com/docs/books/vmspec/. |date=2008-07-09 }}'' (the [http://java.sun.com/docs/books/vmspec/html/VMSpecTOC.doc.html first] {{Webarchive|url=https://web.archive.org/web/20081012060813/http://java.sun.com/docs/books/vmspec/html/VMSpecTOC.doc.html |date=2008-10-12 }} and [http://java.sun.com/docs/books/vmspec/2nd-edition/html/VMSpecTOC.doc.html second] {{Webarchive|url=https://web.archive.org/web/20110925050249/http://java.sun.com/docs/books/vmspec/2nd-edition/html/VMSpecTOC.doc.html |date=2011-09-25 }} editions are also available online).</ref> whose preface states:
{{Blockquote|[[Sun Microsystems|We]] intend that this specification should sufficiently document the Java Virtual Machine to make possible compatible clean-room implementations. Oracle provides tests that verify the proper operation of implementations of the Java Virtual Machine.}}

The most commonly used Java virtual machine is Oracle's [[HotSpot (virtual machine)|HotSpot]].

Oracle owns the Java trademark and may allow its use to certify implementation suites as fully compatible with Oracle's specification.

===Garbage collectors===
{{See also|Garbage collection (computer science)#Java}}
Garbage collectors available in [[Java (software platform)|Java]] [[OpenJDK]]s virtual machine (JVM) include:
* Serial
* Parallel
* [[Concurrent mark sweep collector|CMS]] (Concurrent Mark Sweep)
* [[Garbage-first collector|G1]] (Garbage-First)
* ZGC (Z Garbage Collector)
* Epsilon
* Shenandoah
* GenZGC (Generational ZGC)
* GenShen (Generational Shenandoah)
* IBM Metronome (only in [[IBM]] OpenJDK)
* SAP (only in [[SAP]] OpenJDK)
* Azul C4 (Continuously Concurrent Compacting Collector)<ref name="Tene-Iyengar-Wolf_2011">{{cite book |doi=10.1145/1993478 |chapter-url=https://www.azul.com/files/c4_paper_acm1.pdf |archive-url=https://web.archive.org/web/20170809024656/http://www.azul.com/files/c4_paper_acm1.pdf |archive-date=2017-08-09 |url-status=live |chapter=C4: the continuously concurrent compacting collector |title=ISMM '11: Proceedings of the international symposium on Memory management |date=2011 |author-last1=Tene |author-first1=Gil |author-last2=Iyengar |author-first2=Balaji |author-last3=Wolf |author-first3=Michael |isbn=978-1-45030263-0}}</ref> (only in [[Azul Systems]] OpenJDK)

{| class="wikitable"
|+ Java versions and their Garbage Collectors
|-
! Version !! Default GC !! Available GCs
|-
| 6u14 || rowspan="2" | Serial /<br />Parallel ([[Multiprocessor system architecture|MP]]) || Serial, Parallel, [[Concurrent mark sweep collector|CMS]], ''[[Garbage-first collector|G1]] (E)''
|-
| 7u4 - 8 || rowspan="2" | Serial, Parallel, CMS, G1
|-
| 9 - 10  || rowspan="9" | G1
|-
| 11 || Serial, Parallel, CMS, G1, ''Epsilon (E)'', ''ZGC (E)''
|-
| 12 - 13 || Serial, Parallel, CMS, G1, ''Epsilon (E)'', ''ZGC (E)'', ''Shenandoah (E)''
|-
| 14 || Serial, Parallel, G1, ''Epsilon (E)'', ''ZGC (E)'', ''Shenandoah (E)''
|-
| 15 - 20 || Serial, Parallel, G1, ''Epsilon (E)'', ZGC, Shenandoah
|-
| 21 - 22 || Serial, Parallel, G1, ''Epsilon (E)'', ZGC, Shenandoah, ''GenZGC (E)''
|-
| 23  || Serial, Parallel, G1, ''Epsilon (E)'', ZGC, Shenandoah, GenZGC (default ZGC)
|-
| 24  || Serial, Parallel, G1, ''Epsilon (E)'', Shenandoah, GenZGC, ''GenShen (E)''
|-
| 25  || Serial, Parallel, G1, ''Epsilon (E)'', Shenandoah, GenZGC, GenShen
|-
| colspan="3" | <small> ''(E)'' = ''experimental''</small>
|}

Java's design philosophy revolves around the assumption of a garbage collector. Unlike languages such as [[C++]] and [[Rust (programming language)|Rust]], deterministic memory management through a <code>delete</code> keyword (as in C++) is not possible. Even introducing such a feature is not possible, due to a lack of ownership, aside from using [[Security of the Java software platform#The sun.misc.Unsafe class|{{mono|sun.misc.Unsafe}}]] or through [[Java Native Interface#Foreign Function and Memory API|{{mono|java.lang.foreign}}]] to allocate/deallocate memory outside the Java heap.<ref name="UnsafeBlog">{{Cite web|title=The Unsafe Class: Unsafe at Any Speed|url=https://blogs.oracle.com/javamagazine/post/the-unsafe-class-unsafe-at-any-speed|website=blogs.oracle.com|author=Ben Evans|date=4 May 2020|publisher=Oracle Corporation}}</ref><ref name="javaFFI">{{cite web|url= https://openjdk.org/jeps/454|title= JEP 454: Foreign Function & Memory API|date=7 April 2025|publisher=OpenJDK}}</ref> Due to Java primarily using heap-based allocation, objects are stored as references, and deletion would result in [[dangling pointer]]s.
<syntaxhighlight lang="cpp">
Foo a = new Foo();
Foo b = a;

delete a;
System.out.println(b);
</syntaxhighlight>

===Class loader===
{{Main|Java class file}}

One of the organizational units of JVM byte code is a [[Class (programming)|class]]. A class loader implementation must be able to recognize and load anything that conforms to the Java class [[file format]]. Any implementation is free to recognize other binary forms besides ''class'' files, but it must recognize ''class'' files.

The class loader performs three basic activities in this strict order:

#Loading: finds and imports the binary data for a type
#Linking: performs verification, preparation, and (optionally) resolution
#*Verification: ensures the correctness of the imported type
#*Preparation: allocates memory for class variables and initializing the memory to default values
#*Resolution: transforms symbolic references from the type into direct references. 
#Initialization: invokes Java code that initializes class variables to their proper starting values.

In general, there are three types of class loader: bootstrap class loader, extension class loader and System / Application class loader.

Every Java virtual machine implementation must have a bootstrap class loader that is capable of loading trusted classes, as well as an extension class loader or application class loader. The Java virtual machine specification does not specify how a class loader should locate classes.

===Virtual machine architecture===
The JVM operates on specific types of data as specified in Java Virtual Machine specifications. The data types can be divided<ref>{{Cite web|url=https://docs.oracle.com/javase/specs/jvms/se8/html/jvms-2.html#jvms-2.2|title=Chapter 2. The Structure of the Java Virtual Machine|access-date=2021-09-15|archive-date=2021-09-15|archive-url=https://web.archive.org/web/20210915050448/https://docs.oracle.com/javase/specs/jvms/se8/html/jvms-2.html#jvms-2.2|url-status=live}}</ref> into primitive types ([[integer]] and [[floating-point]] values) and [[Reference (computer science)|reference]] types. <code>long</code> and <code>double</code> types, which are [[64-bit computing|64-bits]], are supported natively, but consume two units of storage in a frame's local variables or operand stack, since each unit is 32 bits. <code>boolean</code>, <code>byte</code>, <code>short</code>, and <code>char</code> types are all [[sign-extended]] (except <code>char</code> which is [[Sign extension#zero-extended|zero-extended]]) and operated on as 32-bit integers, the same as <code>int</code> types. The smaller types only have a few type-specific instructions for loading, storing, and type conversion. <code>boolean</code> is operated on as 8-bit <code>byte</code> values, with 0 representing <code>false</code> and 1 representing <code>true</code>. (Although <code>boolean</code> has been treated as a type since ''The Java Virtual Machine Specification, Second Edition'' clarified this issue, in compiled and executed code there is little difference between a <code>boolean</code> and a <code>byte</code> except for [[Name mangling#Java|name mangling]] in [[method signature]]s and the type of boolean arrays. <code>boolean</code>s in method signatures are mangled as <code>Z</code> while <code>byte</code>s are mangled as <code>B</code>. Boolean arrays carry the type <code>boolean[]</code> but use 8 bits per element, and the JVM has no built-in capability to pack booleans into a [[bit array]], so except for the type they perform and behave the same as <code>byte</code> arrays. In all other uses, the <code>boolean</code> type is effectively unknown to the JVM as all instructions to operate on booleans are also used to operate on <code>byte</code>s.)

The JVM has a garbage-collected heap for storing objects and arrays. Code, constants, and other class data are stored in the "method area". The method area is logically part of the heap, but implementations may treat the method area separately from the heap, and for example might not garbage collect it. Each JVM thread also has its own [[call stack]] (called a "Java Virtual Machine stack" for clarity), which stores [[Call stack#STACK-FRAME|frames]]. A new frame is created each time a method is called, and the frame is destroyed when that method exits.

Each frame provides an "operand stack" and an array of "local variables". The operand stack is used for operands to run computations and for receiving the return value of a called method, while local variables serve the same purpose as [[Processor register|registers]] and are also used to pass method arguments. Thus, the JVM is both a [[stack machine]] and a [[register machine]]. In practice, HotSpot eliminates every stack besides the native thread/call stack even when running in Interpreted mode, as its Templating Interpreter technically functions as a compiler.

The JVM uses references and stack/array indexes to address data; it does not use [[byte addressing]] like most physical machines do, so it does not neatly fit the usual categorization of [[32-bit]] or [[64-bit]] machines. In one sense, it could be classified as a 32-bit machine, since this is the size of the largest value it natively stores: a 32-bit integer or floating-point value or a 32-bit reference. Because a reference is 32 bits, each program is limited to at most 2{{sup|32}} unique references and therefore at most 2{{sup|32}} objects. However, each object can be more than one byte large, and potentially ''very'' large; the largest possible object is an array of <code>long</code> of length 2{{sup|31}} - 1 which would consume 16 GiB of memory, and there could potentially be 2{{sup|32}} of these if there were enough memory available. This results in upper bounds that are more comparable to a typical 64-bit byte-addressable machine. A JVM implementation can be designed to run on a processor that natively uses any bit width as long as it correctly implements the integer (8-, 16-, 32-, and 64-bit) and floating-point (32- and 64-bit) math that the JVM requires. Depending on the method used to implement references (native pointers, compressed pointers, or an indirection table), this can limit the number of objects to less than the theoretical maximum. An implementation of the JVM on a 64-bit platform has access to a much larger address space than one on a 32-bit platform, which allows for a much larger heap size and an increased maximum number of threads, which is needed for certain kinds of large applications; however, there can be a [[64-bit computing#Pros and cons|performance hit from using a 64-bit implementation]] compared to a 32-bit implementation.

===Bytecode instructions===
{{Main|Java bytecode}}

The JVM has [[instruction (computer science)|instructions]] for the following groups of tasks:

{{flatlist|
* [[Load/store architecture|Load and store]]
* [[Arithmetic]]
* [[Type conversion]]
* [[dynamic memory allocation|Object creation and manipulation]]
* [[stack (abstract data type)|Operand stack management (push / pop)]]
* [[branch (computer science)|Control transfer (branching)]]
* [[subroutine|Method invocation and return]]
* [[exception handling|Throwing exceptions]]
* [[monitor (synchronization)|Monitor-based concurrency]]
}}

The aim is binary compatibility. Each particular host [[operating system]] needs its own implementation of the JVM and runtime. These JVMs interpret the bytecode semantically the same way, but the actual implementation may be different. More complex than just emulating bytecode is compatibly and efficiently implementing the [[Java Class Library|Java core API]] that must be mapped to each host operating system.

These instructions operate on a set of common {{vanchor|abstracted [[data type]]s|DATA_TYPE}} rather the [[native data type]]s of any specific [[instruction set architecture]].

===JVM languages===
{{Main|List of JVM languages}}

A JVM language is any language with functionality that can be expressed in terms of a valid class file which can be hosted by the Java Virtual Machine. A class file contains Java Virtual Machine instructions ([[Java byte code]]) and a symbol table, as well as other ancillary information. The class file format is the hardware- and operating system-independent binary format used to represent compiled classes and interfaces.<ref>{{cite web |url=http://docs.oracle.com/javase/specs/jvms/se7/jvms7.pdf |title=The Java Virtual Machine Specification : Java SE 7 Edition |publisher=Docs.oracle.com |access-date=2015-06-26 |archive-date=2021-02-04 |archive-url=https://web.archive.org/web/20210204093304/https://docs.oracle.com/javase/specs/jvms/se7/jvms7.pdf |url-status=live }}</ref>

There are several JVM languages, both old languages ported to JVM and completely new languages. [[JRuby]] and [[Jython]] are perhaps the most well-known ports of existing languages, i.e. [[Ruby (programming language)|Ruby]] and [[Python (programming language)|Python]] respectively. Of the new languages that have been created from scratch to compile to Java bytecode, [[Clojure]], [[Apache Groovy|Groovy]], [[Scala (programming language)|Scala]] and [[Kotlin (programming language)|Kotlin]] may be the most popular ones. A notable feature with the JVM languages is that they are [[Language interoperability|compatible with each other]], so that, for example, Scala libraries can be used with Java programs and vice versa.<ref>{{cite web |url=http://www.scala-lang.org/old/faq/4 |title=Frequently Asked Questions - Java Interoperability |author=<!--Staff writer(s); no by-line.--> |website=scala-lang.org |access-date=2015-11-18 |archive-date=2020-08-09 |archive-url=https://web.archive.org/web/20200809214018/https://www.scala-lang.org/old/faq/4 |url-status=live }}</ref>

Java 7 JVM implements ''JSR 292: Supporting Dynamically Typed Languages''<ref>{{cite web |url=https://jcp.org/en/jsr/detail?id=292 |title=The Java Community Process(SM) Program - JSRs: Java Specification Requests - detail JSR# 292 |publisher=Jcp.org |access-date=2015-06-26 |archive-date=2020-12-20 |archive-url=https://web.archive.org/web/20201220200733/https://jcp.org/en/jsr/detail?id=292 |url-status=live }}</ref> on the Java Platform, a new feature which supports dynamically typed languages in the JVM. This feature is developed within the [[Da Vinci Machine]] project whose mission is to extend the JVM so that it supports languages other than Java.<ref>{{cite web |url=http://openjdk.java.net/projects/mlvm/ |title=Da Vinci Machine project |publisher=Openjdk.java.net |access-date=2015-06-26 |archive-date=2020-11-11 |archive-url=https://web.archive.org/web/20201111162302/https://openjdk.java.net/projects/mlvm/ |url-status=live }}</ref><ref>{{cite web |url=http://www.oracle.com/technetwork/articles/javase/dyntypelang-142348.html |title=New JDK 7 Feature: Support for Dynamically Typed Languages in the Java Virtual Machine |publisher=Oracle.com |access-date=2015-06-26 |archive-date=2018-09-13 |archive-url=https://web.archive.org/web/20180913101203/http://www.oracle.com/technetwork/articles/javase/dyntypelang-142348.html |url-status=live }}</ref>

===Bytecode verifier===
A basic philosophy of Java is that it is inherently safe from the standpoint that no user program can crash the host machine or otherwise interfere inappropriately with other operations on the host machine, and that it is possible to protect certain methods and data structures belonging to trusted code from access or corruption by untrusted code executing within the same JVM. Furthermore, common programmer errors that often led to data corruption or unpredictable behavior such as accessing off the end of an array or using an uninitialized pointer are not allowed to occur. Several features of Java combine to provide this safety, including the class model, the garbage-collected [[#Heap|heap]], and the verifier.

The JVM verifies all bytecode before it is executed. This verification consists primarily of three types of checks:

* Branches are always to valid locations
* Data is always initialized and references are always type-safe
* Access to private or package private data and methods is rigidly controlled

The first two of these checks take place primarily during the verification step that occurs when a class is loaded and made eligible for use. The third is primarily performed dynamically, when data items or methods of a class are first accessed by another class.

The verifier permits only some bytecode sequences in valid programs, e.g. a [[branch (computer science)|jump (branch) instruction]] can only target an instruction within the same [[method (computer programming)|method]]. Furthermore, the verifier ensures that any given instruction operates on a fixed stack location,<ref>{{cite web |title=The Verification process |url=http://java.sun.com/docs/books/jvms/second_edition/html/ClassFile.doc.html#9766 |work=The Java Virtual Machine Specification |publisher=Sun Microsystems |year=1999 |access-date=2009-05-31 |archive-date=2011-03-21 |archive-url=https://web.archive.org/web/20110321165204/http://java.sun.com/docs/books/jvms/second_edition/html/ClassFile.doc.html#9766 |url-status=live }}</ref> allowing the JIT compiler to transform stack accesses into fixed register accesses. Because of this, that the JVM is a stack architecture does not imply a speed penalty for emulation on [[register machine|register-based architectures]] when using a JIT compiler. In the face of the code-verified JVM architecture, it makes no difference to a JIT compiler whether it gets named imaginary registers or imaginary stack positions that must be allocated to the target architecture's registers. In fact, code verification makes the JVM different from a classic stack architecture, of which efficient emulation with a JIT compiler is more complicated and typically carried out by a slower interpreter. Additionally, the Interpreter used by the default JVM is a special type known as a Template Interpreter, which translates bytecode directly to native, register based machine language rather than emulate a stack like a typical interpreter.<ref>{{Cite web |url=https://openjdk.java.net/groups/hotspot/docs/RuntimeOverview.html#Interpreter |title=HotSpot Runtime Overview - Interpreter|website=OpenJDK|access-date=2021-05-24 |archive-date=2022-05-21 |archive-url=https://web.archive.org/web/20220521024017/https://openjdk.java.net/groups/hotspot/docs/RuntimeOverview.html#Interpreter |url-status=live }}</ref> In many aspects the HotSpot Interpreter can be considered a JIT compiler rather than a true interpreter, meaning the stack architecture that the bytecode targets is not actually used in the implementation, but merely a specification for the intermediate representation that can well be implemented in a register based architecture. Another instance of a stack architecture being merely a specification and implemented in a register based virtual machine is the [[Common Language Runtime]].<ref>{{Cite web|url=https://github.com/dotnet/runtime/issues/4775|title=Why not make CLR register-based? · Issue #4775 · dotnet/runtime|website=GitHub|access-date=2021-05-24|archive-date=2023-04-20|archive-url=https://web.archive.org/web/20230420122729/https://github.com/dotnet/runtime/issues/4775|url-status=live}}</ref>

The original specification for the bytecode verifier used natural language that was incomplete or incorrect in some respects. A number of attempts have been made to specify the JVM as a formal system. By doing this, the security of current JVM implementations can more thoroughly be analyzed, and potential security exploits prevented. It will also be possible to optimize the JVM by skipping unnecessary safety checks, if the application being run is proven to be safe.<ref>{{Cite book |doi=10.1145/320384.320397|chapter=A formal framework for the Java bytecode language and verifier|title=Proceedings of the 14th ACM SIGPLAN conference on Object-oriented programming, systems, languages, and applications - OOPSLA '99|pages=147–166|year=1999|last1=Freund|first1=Stephen N.|last2=Mitchell|first2=John C.|isbn=978-1581132380|citeseerx=10.1.1.2.4663|s2cid=14302964}}</ref>

====Secure execution of remote code====
A virtual machine architecture allows very fine-grained control over the actions that code within the machine is permitted to take. It assumes the code is "semantically" correct, that is, it successfully passed the (formal) bytecode verifier process, materialized by a tool, possibly off-board the virtual machine. This is designed to allow safe execution of untrusted code from remote sources, a model used by [[Java applet]]s, and other secure code downloads. Once bytecode-verified, the downloaded code runs in a restricted "[[sandbox (computer security)|sandbox]]", which is designed to protect the user from misbehaving or malicious code. As an addition to the bytecode verification process, publishers can purchase a certificate with which to [[digital signature|digitally sign]] applets as safe, giving them permission to ask the user to break out of the sandbox and access the local file system, [[clipboard (software)|clipboard]], execute external pieces of software, or network.

Formal proof of bytecode verifiers have been done by the Javacard industry (Formal Development of an Embedded Verifier for Java Card Byte Code<ref>{{cite web|last1=Casset|first1=Ludovic|last2=Burdy|first2=Lilian|last3=Requet|first3=Antoine|date=10 April 2002|title=Formal Development of an Embedded Verifier for Java Card Byte Code|website=Inria - National Institute for Research in Digital Science and Technology at [[Côte d'Azur University#National research organizations|Côte d'Azur University]]|url-status=live|archive-date=3 October 2022|url=http://www-sop.inria.fr/everest/Lilian.Burdy/CBR02dsn.pdf|archive-url=https://web.archive.org/web/20221003184410/http://www-sop.inria.fr/everest/Lilian.Burdy/CBR02dsn.pdf}}</ref>)

===Bytecode interpreter and just-in-time compiler===
For each [[hardware architecture]] a different Java bytecode [[Interpreter (computing)|interpreter]] is needed. When a computer has a Java bytecode interpreter, it can run any Java bytecode program, and the same program can be run on any computer that has such an interpreter.

When Java bytecode is executed by an interpreter, the execution will always be slower than the execution of the same program compiled into native machine language. This problem is mitigated by [[Just-in-time compilation|just-in-time (JIT) compilers]] for executing Java bytecode. A JIT compiler may translate Java bytecode into native machine language while executing the program. The translated parts of the program can then be executed much more quickly than they could be interpreted. This technique gets applied to those parts of a program frequently executed. This way a JIT compiler can significantly speed up the overall execution time.

There is no necessary connection between the Java programming language and Java bytecode. A program written in Java can be compiled directly into the machine language of a real computer and programs written in other languages than Java can be compiled into Java bytecode.

Java bytecode is intended to be platform-independent and secure.<ref>David J. Eck, ''[http://math.hws.edu/javanotes/c1/s3.html Introduction to Programming Using Java] {{Webarchive|url=https://web.archive.org/web/20141011192544/http://math.hws.edu/javanotes/c1/s3.html |date=2014-10-11 }}'', Seventh Edition, Version 7.0, August 2014 at Section 1.3 "The Java Virtual Machine"</ref> Some JVM implementations do not include an interpreter, but consist only of a just-in-time compiler.<ref>''[http://docs.oracle.com/cd/E15289_01/doc.40/e15058/underst_jit.htm Oracle JRockit Introduction] {{Webarchive|url=https://web.archive.org/web/20150906145705/http://docs.oracle.com/cd/E15289_01/doc.40/e15058/underst_jit.htm |date=2015-09-06 }}'' Release R28 at 2. "Understanding Just-In-Time Compilation and Optimization"</ref>

==JVM in the web browser==
At the start of the Java platform's lifetime, the JVM was marketed as a web technology for creating [[Rich web application|Rich Web Applications]]. {{As of|2018}}, most web browsers and operating systems bundling web browsers do not ship with a Java [[plug-in (computing)|plug-in]], nor do they permit side-loading any non-[[Adobe Flash|Flash]] plug-in. The Java browser plugin was deprecated in [[JDK]] 9.<ref name="ars-nopluginjdk9">{{cite web |url=https://arstechnica.com/information-technology/2016/01/oracle-deprecates-the-java-browser-plugin-prepares-for-its-demise/ |title=Oracle deprecates the Java browser plugin, prepares for its demise |website=Ars Technica |date=28 January 2016 |access-date=15 April 2016 |archive-date=8 April 2016 |archive-url=https://web.archive.org/web/20160408145301/http://arstechnica.com/information-technology/2016/01/oracle-deprecates-the-java-browser-plugin-prepares-for-its-demise/ |url-status=live }}</ref>

The [[NPAPI]] Java browser plug-in was designed to allow the JVM to execute so-called [[Java applets]] embedded into HTML pages. For browsers with the plug-in installed, the applet is allowed to draw into a rectangular region on the page assigned to it. Because the plug-in includes a JVM, Java applets are not restricted to the Java programming language; any language targeting the JVM may run in the plug-in. A restricted set of APIs allow applets access to the user's microphone or 3D acceleration, although applets are not able to modify the page outside its rectangular region. [[Adobe Flash Player]], the main competing technology, works in the same way in this respect.

{{As of|June 2015}} according to W3Techs, Java applet and [[Microsoft Silverlight|Silverlight]] use had fallen to 0.1% each for all web sites, while Flash had fallen to 10.8%.<ref>{{cite web|url=http://w3techs.com/technologies/history_overview/client_side_language/all/y |title=Historical yearly trends in the usage of client-side programming languages, June 2015 |publisher=W3techs.com |access-date=2015-06-26}}</ref>

===JavaScript JVMs and interpreters===
Since May 2016, JavaPoly allows users to import unmodified Java libraries, and invoke them directly from JavaScript. JavaPoly allows websites to use unmodified Java libraries, even if the user does not have Java installed on their computer.<ref>{{cite magazine|last= Krill|first= Paul|title= JavaPoly.js imports existing Java code and invokes it directly from JavaScript|url= http://www.infoworld.com/article/3069995/java/new-javascript-library-brings-java-to-browsers-without-applets.html|magazine= InfoWorld|access-date= 18 July 2016|date= 13 May 2016|archive-date= 25 July 2016|archive-url= https://web.archive.org/web/20160725024906/http://www.infoworld.com/article/3069995/java/new-javascript-library-brings-java-to-browsers-without-applets.html|url-status= live}}</ref>

===Transpilation to JavaScript===
With the continuing improvements in JavaScript execution speed, combined with the increased use of mobile devices whose web browsers do not implement support for plugins, there are efforts to target those users through [[Transpile|transpilation]] to JavaScript. It is possible to either transpile the source code or JVM bytecode to JavaScript.

Compiling the JVM bytecode, which is universal across JVM languages, allows building upon the language's existing compiler to bytecode. The main JVM bytecode to JavaScript transpilers are TeaVM,<ref>{{cite web |url = http://teavm.org/ |title = TeaVM project home page |publisher = Teavm.org |access-date = 2015-06-26 |archive-date = 2015-06-27 |archive-url = https://web.archive.org/web/20150627045343/http://teavm.org/ |url-status = live }}</ref> the compiler contained in Dragome Web SDK,<ref>{{cite web |url = http://www.dragome.com/ |title = Dragome Web SDK |publisher = Dragome.com |access-date = 2015-06-26 |archive-date = 2015-08-01 |archive-url = https://web.archive.org/web/20150801074231/http://dragome.com/ |url-status = live }}</ref> Bck2Brwsr,<ref>{{cite web |url = http://wiki.apidesign.org/wiki/Bck2Brwsr |title = Bck2Brwsr - APIDesign |publisher = Wiki.apidesign.org |access-date = 2015-06-26 |archive-date = 2015-06-27 |archive-url = https://web.archive.org/web/20150627062724/http://wiki.apidesign.org/wiki/Bck2Brwsr |url-status = live }}</ref> and j2js-compiler.<ref>Wolfgang Kuehn (decatur). [https://github.com/decatur/j2js-compiler j2js-compiler] {{Webarchive|url=https://web.archive.org/web/20130929131436/https://github.com/decatur/j2js-compiler |date=2013-09-29 }} GitHub</ref>

Leading transpilers from JVM languages to JavaScript include the Java-to-JavaScript transpiler contained in [[Google Web Toolkit]], J2CL,<ref>{{Cite web |title=J2cl: Transpile Java to JavaScript for Closure Compiler {{!}} Hacker News |url=https://news.ycombinator.com/item?id=19768844 |archive-url=https://web.archive.org/web/20210102081604/https://news.ycombinator.com/item?id=19768844 |archive-date=2021-01-02 |access-date=2025-10-31 |website=news.ycombinator.com |language=en |url-status=live }}</ref> Clojurescript (Clojure), GrooScript (Apache Groovy), Scala.js (Scala) and others.<ref>{{cite web |url = https://github.com/jashkenas/coffeescript/wiki/list-of-languages-that-compile-to-js |title = List of languages that compile to JS · jashkenas/coffeescript Wiki · GitHub |publisher = Github.com |date = 2015-06-19 |access-date = 2015-06-26 |archive-date = 2020-01-31 |archive-url = https://web.archive.org/web/20200131233044/https://github.com/jashkenas/coffeescript/wiki/List-of-languages-that-compile-to-JS |url-status = live }}</ref>

==See also==
{{Portal|Computer programming}}
* [[Common Language Runtime]]
* [[List of Java virtual machines]]
* [[List of JVM languages]]
* [[Comparison of Java virtual machines]]
* [[Comparison of application virtualization software]]
* [[Automated exception handling]]
* [[Java performance]]
* [[Java processor]]
* [[K virtual machine]] (KVM)

==Notes==
{{notelist}}

==References==
{{Reflist|30em}}

{{Refbegin}}
* ''[http://java.sun.com/docs/books/vmspec/2nd-edition/jvms-clarify.html Clarifications and Amendments to the Java Virtual Machine Specification, Second Edition] {{Webarchive|url=https://web.archive.org/web/20060110040426/http://java.sun.com/docs/books/vmspec/2nd-edition/jvms-clarify.html |date=2006-01-10 }}'' includes list of changes to be made to support J2SE 5.0 and JSR 45
* [http://www.jcp.org/en/jsr/detail?id=45 JSR 45] {{Webarchive|url=https://web.archive.org/web/20060205094707/http://www.jcp.org/en/jsr/detail?id=45 |date=2006-02-05 }}, specifies changes to the class file format to support source-level debugging of languages such as [[JavaServer Pages]] (JSP) and [[SQLJ]] that are translated to Java
{{Refend}}

== External Links ==
* [https://web.archive.org/web/20250907204443/https://stuefe.de/ Thomas Stüfe's blog - JVM Engineer at Red Hat]
* [https://github.com/vbauer/java-blogs?tab=readme-ov-file Java Blogs List]

{{Java Virtual Machine}}
{{Java (software platform)}}
{{Sun Microsystems}}

[[Category:Java virtual machine| ]]
[[Category:Java (programming language)|Virtual machine]]
[[Category:Java platform software]]
[[Category:Java specification requests|Virtual machine]]
[[Category:Stack-based virtual machines]]