Re: amd64/gcc-3.4 archive: multilib/-m32 support available (compilation of 32bit i386-binaries)

[Andreas Jochens <aj@andaco.de>]

On 04-Aug-11 20:57, Goswin von Brederlow wrote:
> Andreas Jochens <aj@andaco.de> writes:
> I'm opposing any patches getting added to sarge (compared to debians
> sarge). Your patch goes beyond what could still be added to debians
> sarge and way beyond what could be added to base+standard packages
> (which glibc is). One reason for this is so that security updates will
> build cleanly without human intervention needed.
> 
> It adds two more reasons not to use gcc-3.4 for sarge amd64.
> 
> And with sarge being 3.3 the main pure64 port will stay 3.3 since
> there is no upgrade path to 3.4 (yet).

I never said that I wanted to use the gcc-3.4 archive for sarge. At the 
moment I consider the gcc-3.4 archive as an experimental archive which 
mainly serves the purpose to check if it is possible to use gcc-3.4 and 
to find and fix as many gcc-3.4 and amd64 related bugs as possible. 

As I see it, this effort has been quite successful. Please look at
http://bugs.debian.org/cgi-bin/pkgreport.cgi?submitter=aj@andaco.de
for an overview of the number and kind of bugs which were detected 
and fixed during this process.

I am not at all sure what I shall think about amd64 and sarge. Since 
amd64 is not even in sid yet, an amd64 sarge release of amd64 seems 
to be very difficult to achieve. And the gcc-3.3 archive still follows 
sid and not sarge at the moment.

However, I don't think that this has anything to do with the '-m32'
32bit compilation feature. This would be an interesting addition even 
for the gcc-3.3 archive and I do not see any problems adding that.
Of course, a patch is needed for that, but I think that being able to
compile 32bit i386 binaries on amd64 is a feature which will be asked 
for by many users of the amd64 port.

Regards
Andreas Jochens