[distcc] distcc & gcc3.3

[Martin Pool]

On 17 Apr 2006, Dan Kegel <dank at kegel.com> wrote:
> On 4/12/06, Martin Pool <mbp at sourcefrog.net> wrote:
> > http://kegel.com/distcc/2006-03-28/01-distcc-gdb-20051210.patch
> > > and I think it's worth including in mainline distcc.
> >
> > Does that give you byte-for-byte identical output?
> 
> In what sense?
> 
> md5sum of the .o's?  It can't be that,
> since otherwise the patch would be a no-op.
> 
> md5sum of the generated opcodes?  Probably;
> I haven't checked.  But I do compile metric boatloads of
> code with this patch every day, and we haven't seen
> any regression tests fail as a result.
> 
> If you define your criterion a bit more closely, I'll try to
> see if the patch satisfies it.

I meant identical .o files between local and remote compilation, which
is what the original poster seemed to be asking.  Obviously I don't mean
between distcc before and after the patch.

-- 
Martin