On Thu, May 23, 2002 at 02:40:35AM +0900, Gilles Filippini wrote:
> > No, Ruby raises SystemStackError in such case, or can do
> > nothing, even displaying the [BUG] message.
> 
> Sure ? I've recently had a segfault signal which IMO was caused by a
> stack overflow. I think it didn't raised SystemStackError because the
> stack overflow occured inside an external C++ library.

I think what he meant was that in this case you should just see a
"segmentation fault" message and not a "[BUG]" message.  This seems to
be consistent with what I've seen:

  $ ulimit -s 1000
  $ ruby -e 'def foo; foo; end; foo'
  Segmentation fault (core dumped)

Paul