RE: Small example of livelock regression in garbage collector for GCJ 3.3 under Win32

[Andrew Haley <aph at redhat dot com>]

Øyvind Harboe writes:
 > > Right, so it's related to the way that null pointer 
 > > exceptions are caught on Win32.  That's weird, because I 
 > > thought that platform generated pointer checks.
 > 
 > 
 > Division by zero can also cause the application to terminate
 > (as well as livelock).
 > 
 > If I add "throw new NullPointerException();" I do not see the problem.

Your libgcj has been built incorrectly.  It is essential that
-fcheck-references and -fuse-divide-subroutine are used at all times
and that libgcj is built with these compiler options.
libgcj/configure.host should set these options where necessary.  As
far as I can see, it does.

Andrew.