This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Fix preprocessor/5153
On Thu, Feb 21, 2002 at 11:52:49PM +0000, Neil Booth wrote:
> > however, we should add an option
> > which triggers the old behavior, and any time a header file is not
> > found and a #line directive *could* have changed the directory of the
> > current file, we emit a diagnostic suggesting the user try adding that
> > option to CFLAGS.
> >
> > How's that sound?
>
> A bit complicated. I tend to think we should choose one way or the
> other; and provide a switch only if really necessary (we already have
> extreme switch creep in GCC). Seeing as the current behaviour is the
> least confusing (the best argument against the old way IMO), and getting
> the old behaviour is just a matter of adding an -I switch if wanted, is
> there a good reason not to leave things as they are now?
I have no objection. Maybe we should add a couple paragraphs to the
documentation explaining why this was changed in 3.0; I can write
those up later today, if you like.
zw