[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gomp-discuss] gomp and where to start.
From: |
Steven Bosscher |
Subject: |
Re: [Gomp-discuss] gomp and where to start. |
Date: |
05 Feb 2003 11:43:38 +0100 |
Op wo 05-02-2003, om 10:42 schreef Lars Segerlund:
>
> I think I was unclear ( again :-) ), I thought that we could have
> private copies of tree.{def,c,h} and so on , and use the path to overide
> the default places the build system searches for it, in this way we
> could live in our own subdir and would have considerably less troble
> keeping the tree's in sync ( cvs trees ).
>
The changes we're looking at are not that big, you know.
I have the following changes right now for the C front end:
Makefile.in | 3
flags.h | 3
toplev.c | 5 +
tree.h | 12 ++
c-openmp.c | 174 ++++++++++++++++++++++++++++++++++++++++++++++++++++
5 files changed, 197 insertions(+)
c-openmp.c is the #pragma handler. It'll need some minor changes to
some more c-* files (c-pragma, c-semantics, etc.) but in all I don't
think we'll have to touch more than 12 files or so in the C front end.
For the C++ front end it's probably even less. Most changes will
probably go into new files anyway, and those never cause merge trouble
:-)
Diego said he'll do the merge, *and* he's willing to work on the
middle-end; he shouldn't have much trouble merging his own work.
So the change sets are small enough that I think merging one CVS branch
with the other will be quite easy.
The advantages of using CVS need no explaining I suppose.
Greetz
Steven