Nobuyoshi Nakada wrote:
> This patch makes:
> * if --with-setjmp-type configure option is given, use its value,
> * if _setjmp is available, use it,
> * if sigsetjmp is N/A, use setjmp,
> * use setjmp on Solaris and Cygwin, or sigsetjmp on others.
> 
> Any more reasonable solutions?

This seems reasonable to me.

Also, I wonder if there is any benefit to __builtin_setjmp instead of 
_setjmp?

Also, just thinking, I wonder how hard it would be to make use of dwarf2 
exceptions on gcc via unwind.h?

(I think the API is described here: 
http://www.ucw.cz/~hubicka/papers/abi/node25.html)

Paul