[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: KDE and Gcc 3.0
From:       Stefan Westerfeld <stefan () space ! twc ! de>
Date:       2001-06-18 22:28:34
[Download RAW message or body]

   Hi!

On Mon, Jun 18, 2001 at 10:26:18PM +0200, Nils Holland wrote:
> On Monday 18 June 2001 20:37, Jesper Pedersen wrote:
> > On Monday 18 June 2001 20:24, Nils Holland wrote:
> > > On Monday 18 June 2001 19:56, Mathieu Chouinard wrote:
> > > > Hi,
> > > > Does anybody have compiled KDE 2.2alpha2 with GCC 3.0?
> > > > Mathieu
> > >
> > > No, but as far as I am informed, GCC 3 is not able to properly compile
> > > "real world code" like KDE at the moment. I wonder if the GCC 3 version
> > > that will eventually be released will be able to do this. Originally, the
> > > GCC folks wanted to get 3.0 out of the door in the first half of this
> > > year, but from what I have heard from various people (and from what I
> > > have read on their website in their release-criteria), I think they're
> > > not quite there yet...
> > >
> > > Greetings
> > > Nils
> >
> > GCC-3.0 was released today.
> 
> Ok. I don't see an announcement on http://gcc.gnu.org, but I believe you ;-). 
> Seems that it's time to put it on my test-machine, then. I think that it 
> should not yet be installed on my normal workstations. When one reads the 
> release-criteria on the GCC website, it seems as if GCC 3 should be able to 
> do exactly the same as 2.9.x, only better. Unluckily, since it does not 
> compile all of KDE correctly (and also has problems with other things, as I 
> have heard) this is not the case. So - who's responsible for making it 
> eventually work? Will we have to wait for a maintenance release from the GCC 
> 3 branch, or will the KDE developers make their code work with GCC 3?

Well, my view of this is that the way gcc3 breaks aRts (by miscompiling code
using virtual inheritance) makes it extremely unlikely that working around
the issue by changing the aRts code is a good idea (it might even be
completely impossible). It's also a regression against previous versions of
GCC (i.e. 2.95.2 compiles virtual inheritance just fine). The bug is reported
to the gcc developers, as

  c++/3145: virtual inheritance still creates wrong code

and I hope that it will get fixed there soon.

   Cu... Stefan
-- 
  -* Stefan Westerfeld, stefan@space.twc.de (PGP!), Hamburg/Germany
     KDE Developer, project infos at http://space.twc.de/~stefan/kde *-         
 
>> Visit http://master.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

[prev in list] [next in list] [prev in thread] [next in thread]