A class may have static member that are initialized. The
code for the initialization is collected into a function
known as a class initializer and given the name "<clinit>".

You can't set a breakpoint in the "<clinit>" routine.
The problem is that code field of the method block for the "<clinit>"
function is set to zero. This field is suppose to contain the
starting pc address for the "<clinit>" method.

The code for the <clinit> method is read into a separate buffer from
the rest of a class's code. This is because the <clinit> code
can only be executed once and is then discarded. This effectively
means we can't set a breakpoint in <clinit> because the code doesn't
stay around.