unwind question

[Richard Henderson]

On Wed, Nov 20, 2002 at 06:10:02PM +0100, Dietmar Maurer wrote:
> Thats wrong. Any asynchrounous signal can interrupt any code, so this is
> a real problem for us.

You're fooling yourself (or generating really bulky slow code) 
if you are attempting to handle truely asynchrounous exceptions.

GCC doesn't support this notion for exactly this reason.
We _do_ support synchrounous exceptions from SIGSEGV, SIGFPE
and the like, but SIGALRM is right out.


r~