> From: Bruce Korb <[EMAIL PROTECTED]>
>
> Kaveh wrote:
> > Another approach would be to do this:
> >
> > if $host == $target && $build == $target
> > Use autoconf to set GETGROUPS_T
> > else
> > GETGROUPS_T = gid_t
> > fi
>
> Good call. I like this solution. It will be
> necessary to use __GETGROUPS_T, however, because
> GETGROUPS_T is in the user name space. "configure"
> can get away with it because it is acting on behalf of
> the user, whereas we are inserting stuff into system
> headers.
That would only be necessary if we make gcc define GETGROUPS_T,
similarly to how we handle __SIZE_TYPE__. I'm not saying we need to
go that far. I suggest we define GETGROUPS_T during the gcc build
instead.
> The reason I am dragging the autoconf folks
> into the discussion is because it might be convenient
> for them to:
>
> #ifdef __GETGROUPS_T
> #define GETGROUPS_T __GETGROUPS_T
> #else
> ....
>
> (i.e. only go through the machinations of the getgroups
> brokenness test when they have to.)
I would argue against this. When we have one autoconf/configure
mechanism for gcc and one for cc, then the cc one tends to bit rot
over time. Its just better software engineering to use the same code
for both cases unless the benefits far outweigh the negatives.
IMHO, I don't see a large benefit here worth teaching gcc about
__GETGROUPS_T like we do for __SIZE_TYPE__.
--Kaveh
--
Kaveh R. Ghazi Engagement Manager / Project Services
[EMAIL PROTECTED] Qwest Internet Solutions
