This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Add -iquote support
- From: Zack Weinberg <zack at codesourcery dot com>
- To: Mike Stump <mrs at apple dot com>
- Cc: Mark Mitchell <mark at codesourcery dot com>, gcc-patches at gcc dot gnu dot org
- Date: Mon, 03 May 2004 14:43:26 -0700
- Subject: Re: Add -iquote support
- References: <81477612-9D46-11D8-B137-003065A77310@apple.com>
Mike Stump <mrs@apple.com> writes:
>> Before we add this additional knob, is there any implementation
>> precedent out there?
>
> That is the _only_ thing driving this _feature_. If we didn't care
> about existing uses (users), we could dodge the issue and not provide
> it.
>
> Another solution would be to change -I- to not so disable current dir
> searching, that would allow us to use -I-, but, that has the cost of
> potentially breaking older uses of -I-. I don't think I like this
> solution, but others might.
The thing is, I'm pretty much convinced that there are no older uses
of -I-, because it just plain doesn't work. System headers rely on
directory-of-the-current-file behavior. GCC's fixincludes mechanism
relies on it. Software tends to rely on it internally, although
that's a weaker argument since software that uses -I- wouldn't.
Tell you what. I am agreed in principle with a patch that
simultaneously adds -iquote and removes -I- entirely (make it issue an
error). For mainline. Provided you make a loud public announcement
of this so we can see if any users come crawling out of the crevices.
This is not a preapproval; I will want to see the patch and check it
for technical issues.
zw