Re: Open Projects and GCC 3.0

[Jeffrey A Law <law at redhat dot com>]

  In message <Pine.LNX.4.30.0012222241020.9846-100000@kern.srcf.societies.cam.a
c.uk>you write:
  > On Fri, 22 Dec 2000, Toon Moene wrote:
  > 
  > > The following large middle-end projects are known to me and have had no
  > > updates for months:
  > 
  > I'll also ask again for the status of live range splitting, promised
  > "soon" for over two years
The LRS pass written by Cygnus in 1997 would need some *major* cleanup to
be brought up to current standards.  In fact, that's the only reason it 
hasn't been contributed -- the Red Hat GCC engineers haven't had the time
to clean the bloody things up.

If someone wants to try and pick up our stuff, I'll send them diffs and cc
gcc-patches for archival purposes.

However, I do not think it is wise to make gcc-3.0 dependent on inclusion
of LRS optimizations.

Probably the most important task is to convert LRS to use (and maintain)
the CFG structures.  The second most important task is to clean up how
we pass information about the split ranges to the debugger.


  > (with, in alternative, a new register allocator
  > by David Miller, also in progress over two years ago).  See URLs in
  > <URL:http://gcc.gnu.org/ml/gcc-patches/2000-11/msg00183.html> for relevant
  > past messages.
Yup.  I'd like to see this code made public too.  David?  You listening?

jeff