This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Fix PR preprocessor/6489 (take 2)
- From: Zack Weinberg <zack at codesourcery dot com>
- To: Jakub Jelinek <jakub at redhat dot com>
- Cc: mark at codesourcery dot com, neil at daikokuya dot demon dot co dot uk,gcc-patches at gcc dot gnu dot org
- Date: Thu, 2 May 2002 13:33:40 -0700
- Subject: Re: [PATCH] Fix PR preprocessor/6489 (take 2)
- References: <20020430162510.R32482@sunsite.ms.mff.cuni.cz> <20020430175853.GA26266@codesourcery.com> <20020501000808.A32482@sunsite.ms.mff.cuni.cz> <20020501155107.GF26266@codesourcery.com> <20020502093941.O23056@devserv.devel.redhat.com>
On Thu, May 02, 2002 at 09:39:41AM -0400, Jakub Jelinek wrote:
>
> It handles stuff like #include <foo.h>^M but not \^M. \^M is handled in
> > 50 cases in that spaghetti code, I've started working on a patch but
> it would be huge (what I wrote was about 14K patch and was far from
> finishing it).
> Is cpp -traditional actually performance critical?
> Wonder if a simple pass over a newly read buffer which would
> canonicalize line endings wouldn't be much cleaner and shorter
> (and obviously correct)...
No, it's not performance critical -- and Neil wants to fold it back
into cpplib, which would make the whole issue go away. A prescan pass
would do in the meantime, and as you say, be much easier to validate.
zw