With Java 9, which is planned to be released in the summer of 2017, (3) will be possible as well. The short answer is, normally, “java” does both (1) and (2). So now our question is how “java” works, as an (1) Interpreter, (2) JIT compiler and/or (3) AOT compiler and when. So, (1) is the job of an Interpreter, (2) is the result of JIT compilation and (3) is the result of AOT compilation.įor the sake of completeness, there is actually one more approach, to interpret the source code directly, but Java does not work like that, but for example Python does. So the compilation is before/Ahead Of execution Time. Before everything, before running the program, compiling the intermediate code to the native code and letting the CPU run it from the beginning.So the compilation is Just In Time of the execution.

Just before executing the intermediate code, compiling it into the native code, and letting the CPU run the very freshly generated native code.

As you know, JVM is executed by running the “java” program. JVM is (or more properly has) a Java language interpreter. In other and better words, “interpreting” the intermediate code.

Executing the intermediate code directly.

In order for Java bytecodes to do some actual work, there are 3 options: It is called intermediate because it is not understandable and executable by the real computing device, the CPU, so it is an intermediate form between the source code and the CPU-executable “native” machine code. Java bytecodes are the intermediate language. It is the purpose of this post to show how all of these are used in the Java Virtual Machine, JVM.Īs you probably know, when you code in the Java programming language, you run the Java compiler (the one triggered by the “javac” program), which compiles the Java source code (.java files), into Java bytecodes (.class files). We also hear “interpreted” languages a lot.

I guess anyone coding Java has heard about JIT (Just In Time) and maybe also AOT (Ahead Of Time) compilation.

Compiling java to ijvm example problems