[Bug 75] Real life brokenness broken?

http://sablevm.org/bugs/show_bug.cgi?id=75





------- Additional Comments From [email protected]  2005-01-15 09:54 -------
In http://sablevm.org/lists/sablevm-user/2004-December/000092.html,
asutosh.gopinath at lycos.com wrote:

I had compiled using real life brokeness. I want to paint Helloworld, works 
fine with 1.1.6 but gives following error with 1.1.7:

import java.awt.*;
import java.awt.event.WindowAdapter;
import java.awt.event.WindowEvent;
import java.io.PrintStream;

public class FirstAWT extends Frame
{
    private class WindowDestroyer extends WindowAdapter
    {

        public void windowClosing(WindowEvent windowevent)
        {
            System.exit(0);
        }

        private WindowDestroyer()
        {
        }

    }


    public static void main(String args[])
    {
        FirstAWT firstawt = new FirstAWT();
        firstawt.setVisible(true);
    }

    public FirstAWT()
    {
        setSize(400, 200);
        setBackground(Color.red);
        setTitle("My first AWT application");
        WindowDestroyer windowdestroyer = new WindowDestroyer();
        addWindowListener(windowdestroyer);
    }

    public void paint(Graphics g)
    {
        g.setColor(Color.black);
        g.drawString("Hello world!", 150, 50);
        System.out.println("paint called");
    }
}



--- stack trace dump (begin) ---
Thread ID: 3, Posix ID: 32771
(GtkComponentPeer.java:-1)  gnu/java/awt/peer/gtk/GtkComponentPeer.gtkWidgetSetC
ursor  n
(GtkComponentPeer.java:461)  gnu/java/awt/peer/gtk/GtkComponentPeer.setCursor

(GtkComponentPeer.java:456)  gnu/java/awt/peer/gtk/GtkComponentPeer.setCursor

(internal frame)
(GtkMainThread.java:-1)  gnu/java/awt/peer/gtk/GtkMainThread.gtkMain  n
(GtkMainThread.java:106)  gnu/java/awt/peer/gtk/GtkMainThread.run
(Thread.java:704)  java/lang/Thread.run
(VMThread.java:120)  java/lang/VMThread.callRun
(Thread.java:390)  java/lang/Thread.callRun
(VirtualMachine.java:121)  java/lang/VirtualMachine.runThread
(internal frame)
--- stack trace dump (end) ---
[jni: fatal error (Local reference capacity exceeded.
** This is a bug in your application or in the GNU Classpath Library.
** A virtual machine is only required to provide 16 native local
** references per JNI function call.  To get more, an application must call
** EnsureLocalCapacity().
** Temporarily, you might want to compile SableVM with the configure option:
** --enable-real-life-brokenness
** See: http://java.sun.com/docs/books/jni/html/pitfalls.html#11229
)]
Aborted




------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.