[Gomp-discuss] Welcome!

[Biagio Lucini]

Hi All - Thanks for accepting my invitation to join. Before starting the
discussion, I have to tell you a few things. 

I've asked for some space at savannah, the name of the project is
"gomp" (just because I had to choose one). The nice guys at savannah also
sponsor this mailing list and possible others, give us a cvs server etc.
etc. I don't go into much details, since you probably know better than me.
What I wanted to say is that I have made those arrangements just to start
with something, they can be changed at any time, should you have any
objection. Also, formally I am the administrator of the project, but my
only ambition is to see things done, so if anybody else likes to take the
leadership he is very welcome to do that.

For the moment, the only active thing is this list, whose subscribers are
currently four people (including myself) plus I have two other invitations
pending. I thought that we could start chatting for a while, then a
volunteer (if nobody else I'll do it) can write a sort of preliminary
draft and finally we can start to write a bit of code following the agreed
specification. I'm not a professional software engineer nor a very
experienced programmer, so I'm open to suggestions also about that point.

After this "warm up", I summarize a couple of e-mail that I have received
from you, making specific observations on the direction of the project.
Essentially, Diego suggests to build things on top of the tree SSA branch.
Also Sebastian knows a lot about the project, I guess, since he is listed
among the contributors. Another suggestion by Sebastian was to look at the
OpenMP implementation given in Open64, on which he has been working for a
while. Finally Lars is pleased to partecipate to a project that could help
in building HPC around GNU tools. 

As for me, I am an OpenMP users and a gcc user. So one of my dream is to
have a chance to say that I am an user of a gcc version of OpenMP. 

OK, know for the discussion. Just for a start, I suggest the following
topic. OpenMP has Fortran and C/C++ specifications. A naive approach for a
compiler would be to parse the OpenMP directives either in Fortran or in
C++ and to translate them onto a threaded program in the original
language. If I interpret correctly his point, Diego suggests (and
apparently AFAIK also Open64 uses the same approach) to translate the
source into an intermediate language and then to process that. What are
the advantages of this second approach? 

And, by the way, do you know a suitable candidate threading library that
we could use for threading the application?

Cheers
Biagio