Re: [tree-ssa][rfc] plans for nested functions

[dewar at gnat dot com (Robert Dewar)]

> Why do you want to continue to support nested functions in the first
> place? They are an abomination, a gcc-specific extension, and nobody
> sane uses them (yeah, yeah, libobj, but I said _sane_).

There are of course strong arguments in favor of nested functions, and
some would say that any decent language includes them for good reasons.
I hardly think there is any need to repeat the arguments here, since 
they should be familiar to anyone with a reasonable background in 
language design. The arguments nbecome very strong of course in a 
multi-threaded environment.

> Are there really any users that do not deserve to either be fixed
> or just shot in the head?

Certainly the implementation of Ada depends on support for nested
subprograms, since Ada is a language in which the use of nested
subprograms is fundamental.

As for the suggestion of removing them from C, that seems a pretty
drastic step. I suspect that would cause major headaches for a
lot of users, and yes, of course there are many legitimate uses of
nested function.

> Nested functions was one such idea. The
> implementation used to do bad things (icache flushes at run-time,
> for christ sake!).

A level of emotional concern leading to casual profanity seems a bit
extreme in a technical discussion! The trampoline approach is indeed
problematical, but does avoid what would otherwise be distributed
overhead for programmers not using this extension. Now that is 
something that I would expect Linus to get upset over :-)

In fact we identify the relatively inefficient implementation
of nested functions as problematical in the Ada case., but there are
far more worthwhile targets of opportunity (e.g. efficient
arithmetic overflow checking).