I have the following source code in Java:

class Outer {
    Nested nested;
    Nested getNested() {
        return nested;
class Nested {
    Inner inner;
    Inner getInner() {
        return inner;
class Inner {
    String foo;
    String getFoo() {
        return foo;

public class NullableTest {
	public static Outer getInitializedOuter(){
		Outer outer = new Outer();
		outer.nested = new Nested();
		outer.nested.inner = new Inner();
		outer.nested.inner.foo = "Jerry";
		return outer;
	/* null pointer exception
private static void way0(){
		Outer outer = new Outer();
	public static void way1(){
		Outer outer = getInitializedOuter();
		if (outer != null && outer.nested != null && outer.nested.inner != null) {

	public static void main(String[] args) {
Get de-assembled byte code via javap:
Navigate to the part for method way1():

According to instruction list explanation in wiki:

0: invokestatic #42 // Method getInitializedOuter:()Ljava8/Outer;
Call static method getInitializedOuter, whose return type is Outer
3: astore_0
Store returned Outer reference to local variable with id 0
4: aload_0
Since in Java source code, the reference outer will be compared against null via outer != null, so here aload_0 loads the reference stored in local variable #0 to stack.
5: ifnull 41

If the current value is null, execution will go to code #41, which is directly return.

or else continue to execute from #8: aload_0 to fetch outer into stack again.
Why repeated call a_load_0 for loading outer reference is needed here?
Again check the instruction list table:

For instruction ifnull, the value in the stack before executed is value, value is gone after ifnull is executed.

From the list above, we can also see another instruction ifnonnull. So here the interesting fact is, according to the analysis so far, the source code below:

if (outer != null && outer.nested != null && outer.nested.inner != null) {

is actually compiled by Java as the execution approach below:

if (outer == null )
if( outer.nested == null )
if( outer.nested.inner == null) 

The usage of LineNumber Table:

LineNumberTable is one of the optional attributes that holds metadata for debugging purposes. In this case, it specifies which offsets in the bytecode correspond to each line in the original source code. This is useful for printing more informative stack traces and for providing features like single step in the debugger. The example above illustrates the mapping relationship from byte code and original source code.

When change way1 from public to private, in javap output you cannot find byte code for way1() itself any more.
To report this post you need to login first.

Be the first to leave a comment

You must be Logged on to comment or reply to a post.

Leave a Reply