On Tue, Aug 01, 2006 at 11:29:26PM +0200, Matthias Drochner wrote:

[Just a brief response about the origins; I'll have to get back to
this...] 

> There was a design decision appearently to allow a zero
> pointer argument to setcontext(2), and interpret this
> as _exit(0).

Yes; while this was beyond the requirements of the SUS, the
idea was to keep NetBSD's implementation of setcontext(2) as
close as possible to the SVR4 one.


Best,
- Klaus