Re: XSH/XRAT conflict over cancellation points

[Wojtek Lerch <yyyyyy@xxxxxxx>]

Alexander Terekhov wrote:
> > you're free to implement setjmp() in a way that will allow 
> > you to use it for exception handling.
> That's upside down. C compiler shall translate setjmp() macro 
> to some magic C++ try/catch thing and longjmp() to a rather 
> simple C+ throw.
Only if the setjmp() and longjmp() calls were generated by a 
disassembler.  ;-)

Seriously, when you're implementing a compiler, you're free to use 
whatever magic you find appropriate that implements the correct 
semantics.  And C++ does not require longjmp() to have a lot of 
knowledge about C++ exceptions and destructors (18.7#4).