Re: covariant*.C FAILs on GCC 3.4 branch

[Jan Hubicka <jh at suse dot cz>]

> >>>>> Jan Hubicka writes:
> 
> Jan> OK, the covariant.C problem is caused by fact that I forgot to apply the
> Jan> semantics.c patch before the memory reduction patch.
> Jan> I think given the current situation, it is best to hold both patches
> Jan> until AIX is fixed.  Thus I am reverting the PR 12850 fix.
> Jan> Can you please tell me once the problem is fixed so I can ask Mark for
> Jan> re-approval of both patches?
> 
> 	I would recommend the opposite.  If this patch is going to be in
> the GCC 3.4 release branch, we should have it in the branch early so that
> it gets more testing. I can work around the problem on AIX by disabling
> the semantics change and living with the covariant testsuite failures.  Of
> course it's up to Mark what he wants done on the 3.4 branch.

OK, Mark, is it OK to reapply the memory savings patch in combination
with applying the thunk emitting patch for 3.4 branch then?

Thanks for clarification.
Honza
> 
> David