This is the mail archive of the
mailing list for the GCC project.
Re: GSoc-2015: Modular GCC
- From: Trevor Saunders <tbsaunde at tbsaunde dot org>
- To: gcc at gcc dot gnu dot org
- Date: Thu, 5 Mar 2015 20:12:26 -0500
- Subject: Re: GSoc-2015: Modular GCC
- Authentication-results: sourceware.org; auth=none
- References: <CAA_ASpVFs7LXS4Pw1Mxo5VCxtEncvYXjYgnULGoRAWnbM3PkYw at mail dot gmail dot com> <20150305102955 dot GA31832 at tsaunders-iceball dot corp dot tor1 dot mozilla dot com> <CAA_ASpWOyWr2Mgy52i6mY5_swyuCP6yuiuBD8s11o2_ipaV3Ag at mail dot gmail dot com>
On Fri, Mar 06, 2015 at 12:21:58AM +0530, Sidharth Chaturvedi wrote:
> I like the idea of making the intermediate representations more
> streamable. But I think this task should also involve creating
> separate front-end and middle-end modules, as then there can be a
> clear distinction of what IR is an input to a module and what IR is
> output from a module(including a more specific structure to each IR).
making the front / middle ends more modular is certainly also a worth
while activity, but I'd advise against merging the issues. They are
each fairly big projects on there own, and its much better to check in
patches that solve one small part of the problem than to have a half way
working patch to fix everything.
> This would give a pipeline structure to GCC. I don't know how much of
> this can be achieved via GSoc, but I would still like to give it a
> try. Any interested mentors? Thanks.
I can help some, but I don't think I am qualified to mentor, or can
commit to spending the time saddly.
> On Thu, Mar 5, 2015 at 3:59 PM, Trevor Saunders <firstname.lastname@example.org> wrote:
> > On Thu, Mar 05, 2015 at 12:43:49AM +0530, Sidharth Chaturvedi wrote:
> >> Hi,
> >> I came across this page while browsing through ideas for GSoc-2015:
> >> https://gcc.gnu.org/wiki/ModularGCC
> >> I like the overall task, but it seems somewhat overambitious. I wanted
> >> to verify if this is still a desirable architectural goal (as the last
> >> update on this page was around 3 years ago). Also, how feasible is
> >> this for GSoc?
> > I'd say it is desirable, various people are working on various bits of
> > it, and hopefully some of them will add there $0.02. I would agree
> > doing everything on that page is completely unrealistic. I'd encourage
> > you to pick a sub area you'd be interested in and see if someone wants
> > to mentor that. Personally I'd really like to see someone take up
> > making intermediate representations more machine streamable. There's
> > already machinary to print most data, but its not in a machine readable
> > form.
> > Trev
> >> --
> >> Regards
> >> Sidharth Chaturvedi
> Sidharth Chaturvedi