Page History

...

...

The

...

Jasm format

...

The JASM format is well suited to represent semantics of byte-code instructions within the method of a class.  The JASM format uses mnemonics descriptions for encoding instructions - the same mnemonic language described in Chapter 6 of the  Java VM specification.  The JASM format also uses a Java-like syntax for class/method/field descriptions, access_flag mask values (that are decoded into "public", "protected", etc modifiers), and indexes into the constant pool (that can optionally be replaced with the values).

...

To illustrate the JASM format, consider a small, very familiar Java class, HelloWorld  (HelloWorld.java):

Code Blocknote
theme DJango language java title HelloWorld.java
icon false
HelloWorld.java public class HelloWorld { public static void main(String args[]) { System.out.println("Hello, world!"); } }

...

Now let's consider the JASM encoding of this.  First, you would have to compile this class into it's object form to generate the .class file:

...

 

theme

Tip
Emacs icon language bash title Produce object from Java false
# Compile HelloWorld.java (to produce HelloWorld.class) $javac HelloWorld.java # Disassemble HelloWorld.class (to produce HelloWorld.jasm) $java -jar asmtools.jar jdis HelloWorld.class > HelloWorld.jasm

...

When you open the JASM formatted disassembly (HelloWorld.jasm), you see the following:

Code Blockinfo
themeicon FadeToGrey language cpp title HelloWorld.jasm false
super public class HelloWorld version 50:0 { public Method "<init>":"()V" stack 1 locals 1 { aload_0; invokespecial Method java/lang/Object."<init>":"()V"; return; } public static Method main:"([Ljava/lang/String;)V" stack 2 locals 1 { getstatic Field java/lang/System.out:"Ljava/io/PrintStream;"; ldc String "Hello, world!"; invokevirtual Method java/io/PrintStream.println:"(Ljava/lang/String;)V"; return; } } // end Class HelloWorld

 

In this listing, you can see how the syntax is a mixture of Java declarations (for describing member signatures), Strings (encoded in the format that constants are represented in the constant pool), and Java instruction mnemonics.  This type of format lends itself well to create class files to test incorrect sequences of instructions.

The

...

Jcod format

...

The JCOD format is well suited to represent the structure of a class file (more so than instruction sequences).

...

Let's reuse our HelloWorld example to generate a JCOD disassembly:

Code Blocknote
theme DJango language java title HelloWorld.java
icon false
HelloWorld.java public class HelloWorld { public static void main(String args[]) { System.out.println("Hello, world!"); } }

 

 

Again, you need to compile this class into it's object form to generate the .class file and disassemble the object:

Code Blocktip
themeicon Emacs language bash title Produce object from Java false
# Compile HelloWorld.java (to produce HelloWorld.class) $javac HelloWorld.java # Disassemble HelloWorld.class (to produce HelloWorld.jcod) $java -jar asmtools.jar jdec HelloWorld.class > HelloWorld.jcod

 

 

 

When you open the JCOD formatted disassembly (HelloWorld.jcod), you see the following:

Code Blockinfo
icon false theme FadeToGrey language cpp title HelloWorld.jasm
class HelloWorld { 0xCAFEBABE; 0; // minor version 50; // version [] { // Constant Pool ; // first element is empty Method #6 #15; // #1 Field #16 #17; // #2 String #18; // #3 Method #19 #20; // #4 class #21; // #5 class #22; // #6 Utf8 "<init>"; // #7 Utf8 "()V"; // #8 Utf8 "Code"; // #9 Utf8 "LineNumberTable"; // #10 Utf8 "main"; // #11 Utf8 "([Ljava/lang/String;)V"; // #12 Utf8 "SourceFile"; // #13 Utf8 "HelloWorld.java"; // #14 NameAndType #7 #8; // #15 class #23; // #16 NameAndType #24 #25; // #17 Utf8 "Hello, world!"; // #18 class #26; // #19 NameAndType #27 #28; // #20 Utf8 "HelloWorld"; // #21 Utf8 "java/lang/Object"; // #22 Utf8 "java/lang/System"; // #23 Utf8 "out"; // #24 Utf8 "Ljava/io/PrintStream;"; // #25 Utf8 "java/io/PrintStream"; // #26 Utf8 "println"; // #27 Utf8 "(Ljava/lang/String;)V"; // #28 } // Constant Pool 0x0021; // access #5;// this_cpx #6;// super_cpx [] { // Interfaces } // Interfaces [] { // fields } // fields [] { // methods { // Member 0x0001; // access #7; // name_cpx #8; // sig_cpx [] { // Attributes Attr(#9) { // Code 1; // max_stack 1; // max_locals Bytes[]{ 0x2AB70001B1; }; [] { // Traps } // end Traps [] { // Attributes Attr(#10) { // LineNumberTable [] { // LineNumberTable 0 1; } } // end LineNumberTable } // Attributes } // end Code } // Attributes } // Member ; { // Member 0x0009; // access #11; // name_cpx #12; // sig_cpx [] { // Attributes Attr(#9) { // Code 2; // max_stack 1; // max_locals Bytes[]{ 0xB200021203B60004; 0xB1; }; [] { // Traps } // end Traps [] { // Attributes Attr(#10) { // LineNumberTable [] { // LineNumberTable 0 3; 8 4; } } // end LineNumberTable } // Attributes } // end Code } // Attributes } // Member } // methods [] { // Attributes Attr(#13) { // SourceFile #14; } // end SourceFile } // Attributes } // end class HelloWorld

 

In this listing, you can see how the syntax more directly correlates to the structure of a class file, in the same order and manner which a class file is described in Chapter 4 of the VM specification.   Instructions themselves are not decoded (as they are in the JASM format), and appear as raw byte strings.  Also, notice how references to the constant pool are treated as indexes (instead of being encoded as strings).  As well, the constant pool is layed-out in the beginning of the listing.

...