committed: merge with libada-branch

[Nathanael Nerode]

On Tue, 2004-02-10 at 19:20, Zack Weinberg wrote:
> So, what this gives us is a wrapper Makefile in libada/ that invokes
> gcc/Makefile to do all the work.  None of the code is moved.  This is
> a step forward, but the multilibs logic is very fragile, and I doubt
> it can be made to work without physically moving all of the libada
> source code into the libada directory (I could be wrong about this).

In response to which Laurent Guerby wrote:
>The GNAT library has always been built by first
>copying ("ln -s" where available) the appropriate
>set of runtime sources in a clean directory (ada/rts).
>The new libada stuff works the same.
>
>Moving the Makefile code makes sense, but I don't
>see clearly the point in moving the runtime 
>sources around in CVS.

Moving the Makefile code is very important.  Actually, I've been
putting off thinking about merging the libada-branch to mainline
precisely because of this issue.  Once the Makefile code is moved (which
looks pretty hairy to me, but might not be as hairy as it looks), I
can work on cleaning it up, getting multilibs working, etc.

Also, moving runtime sources can be done piecemeal later (yeah, I know how
to do this), but only after the Makefile code is moved.

-- 
Nathanael Nerode  <neroden at gcc.gnu.org>
US citizens: if you're considering voting for Bush, look at these first:
http://www.misleader.org/  http://www.cbc.ca/news/background/arar/
http://www.house.gov/reform/min/politicsandscience/