Re: __va_copy == G_VA_COPY

[Tim Janik <timj gtk org>]

On Thu, 17 Jun 1999 torsten@inetw.net wrote:

> So,
> I located G_VA_COPY in glibconfig.h, and I commented it out.
> This allows glib.h to redefine G_VA_COPY (I assume correctly)
> and now gsm compiles, and therefore gnome core, compiles 
> through.
> 
> Does anyone know why G_VA_COPY was defined in glibconfig.h
> to __va_copy, where __va_copy doesn't exist?
> 
> Or does __va_copy exist, and I simply don't know where 
> it is?


__va_copy is normally provided by gcc, i don't know whether other compilers
define this as well.

> And does commenting out G_VA_COPY in glibconfig.h and letting
> it be redefined in glib.h correctly define the macro?

probably not, glib's configure.in goes through some effort to figure how
to copy va_list variables, and so far we haven't encountered any problems with
its tests (you'd certainly notice earlier if __va_copy really isn't provided
on your system as glib and gtk make extensive use of it).

> And finally, any idea who I contact about this redefinition
> problem?

general G_VA_COPY problems belong to gtk-list@redhat.com or
gtk-devel-list@redhat.com. i don't know what the problem is with the gsm
code in this regard. assuming that your system runs fine w/ regards to glib
gtk and gnome in general, it might be something weird that gsm is doing
with G_VA_COPY

> 
> Torsten
> 

---
ciaoTJ