[LLVMdev] GSoC project questions.

[Alex L]

>
> If you are going to work on improvements in this area, then it would
> help to add them to the proposal (maybe that will move other, not so
> important, features out of scope).
>
That would be a good idea indeed.


> The EQUIVALENCE keyword is deprecated, still I don't know if it is still
> widely used in F77 code. I think one of the main resaons it was deprecated
> is that it voids all type checking

Thanks, I did miss this one.

FWIW, this is still widely used in F77 code (as are common blocks). This
> is, essentially, how you get 'union' in Fortran. I think it is not used
> much in modern Fortran, however. I recommend that we should at least think
> about how to support these various things, but only actually do it up front
> if it is in BLAS.
>
I just did a search through BLAS sources and EQUIVALENCE isn't used there.
I might consider getting rid of it for the proposal. The COMMON blocks
aren't used in BLAS too.

We should double-check with Chris Lattner about license compatibility, but
> regarding basic Fortran I/O, we can probably take some base code from
> libf2c. That should give us the correct Fortran 77 I/O semantics without
> too much pain (http://www.netlib.org/f2c/).
> http://www.netlib.org/f2c/src/notice -- it seems BSD-like. I'm suggesting
> this because my feeling is that most of the work in I/O is really in the
> runtime, not in the semantic-analysis/codegen pieces.
>
That's interesting, thanks for the link. Yes, IO would involve a lot of
runtime work, and I guess that it would probably be a good idea to use
another library for that, at least in the beginning. The license would
allow for that it seems. I think I should definitely mention that the usage
of this library is a possibility in my proposal.




2013/4/24 Hal Finkel <hfinkel at anl.gov>

> ----- Original Message -----
> > From: "Matthieu Brucher" <matthieu.brucher at gmail.com>
> > To: "Alex L" <arphaman at gmail.com>
> > Cc: "LLVM Developers Mailing List" <llvmdev at cs.uiuc.edu>
> > Sent: Wednesday, April 24, 2013 12:55:44 PM
> > Subject: Re: [LLVMdev] GSoC project questions.
> >
> >
> >
> >
> >
> >
> >
> >
> > 2013/4/24 Alex L < arphaman at gmail.com >
> >
> >
> >
> >
> >
> >
> >
> > You may want to check the later versions of the standard also to
> > check if some of your proposal is not deprecated there.
> > I had a quick check with 2008 standard, and I'm pretty sure that all
> > of the stuff that I chose is there, although a lot of it has various
> > additions and changes.
> >
> >
> > The EQUIVALENCE keyword is deprecated, still I don't know if it is
> > still widely used in F77 code. I think one of the main resaons it
> > was deprecated is that it voids all type checking (especially in
> > modules).
>
> FWIW, this is still widely used in F77 code (as are common blocks). This
> is, essentially, how you get 'union' in Fortran. I think it is not used
> much in modern Fortran, however. I recommend that we should at least think
> about how to support these various things, but only actually do it up front
> if it is in BLAS.
>
>  -Hal
>
> >
> >
> >
> > Matthieu
> >
> > --
> > Information System Engineer, Ph.D.
> > Blog: http://matt.eifelle.com
> > LinkedIn: http://www.linkedin.com/in/matthieubrucher
> > Music band: http://liliejay.com/
> >
> > _______________________________________________
> > LLVM Developers mailing list
> > LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20130424/a9b44bc7/attachment.html>