This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Status of Chill-frontend and Pascal
- From: Frank Heckenbach <frank at g-n-u dot de>
- To: gcc at gcc dot gnu dot org, andreas dot meier_ at gmx dot de
- Cc: doko at cs dot tu-berlin dot de, hebisch at math dot uni dot wroc dot pl
- Date: Mon, 4 Aug 2003 02:58:22 +0200
- Subject: Re: Status of Chill-frontend and Pascal
- References: <19192.1059592628@www58.gmx.net>
- Reply-to: frank at g-n-u dot de, gcc at gcc dot gnu dot org, andreas dot meier_ at gmx dot de, doko at cs dot tu-berlin dot de, hebisch at math dot uni dot wroc dot pl
Andreas Jaeger wrote:
> > Andreas Meier <andreas.meier_@gmx.de> writes:
>
> > [...]
> > What will be done with the Pascal frontend in near future? Will it be in GCC
> > 3.4 or 3.5 or never be integrated?
>
> I propose to ask the Pascal folks. If they're willing to integrate
> Pascal, I see no problem to get it into 3.5. For 3.4 it's too late
> already IMHO.
>
> Btw. I would love to get Pascal in - but currently GPC is maintained
> outside of GCC so the GPC developers should really make a proposal and
> then the GCC developers can answer ;-)
We have some structural issues to solve before I think integration
would be useful. E.g., we still have to support gcc-2.x due to some
bugs in different versions of the backend on various targets.
I suppose you'd also want to have those gcc-2.x conditionals removed
before integration.
Another thing is that the build process currently requires a heavily
modified compiler driver. We have plans to resolve this (shifting
most of the problematic parts to an external utility), so it can be
properly integrated with the other compiler drivers.
These are the major ones, there might be some smaller issues as
well.
I can't really say how long this will take. We don't have fixed
release schedules, but make releases when things are ready. Perhaps
sometime next year will be a time to consider integration. I don't
know which GCC version will be current then, but I'd rather not
hurry now, but do things properly and then see which GCC version
we'll reach.
As far as Chill is concerned (which apparently the original question
was about), I'm aware of one potential conflict, namely set types.
AFAIK, there was some agreement with Peter Gerwinski about how to do
it in the backend. I'm not familiar with all the details, but since
many if not most of the backend patches we use in GPC are related to
sets, I assume back then nobody has pursued it further and merged it
to the official sources. (These patches are available in GPC
distributions or I could send them, if anyone is interested.)
Frank
--
Frank Heckenbach, frank@g-n-u.de
http://fjf.gnu.de/
GnuPG and PGP keys: http://fjf.gnu.de/plan (7977168E)