Regressions on mainline\

[Jan Hubicka]

> Andrew Pinski wrote:
> 
> >This was almost likely caused by:
> > 
> >
> Thanks for analyzing.
> 
> >2004-04-28  Jan Hubicka  <jh@suse.cz>
> >
> >       PR c/15004
> >       * function.c (do_warn_unused_parameter): Break out form ...
> >       (expand_function_end): ... here; warn only when not using 
> >       cgraphunit.
> >       * function.h (do_warn_unused_parameter): Declare.
> >       * cgraphunit.c: Include function.h.
> >       (cgraph_finalize_function): Do unused parameter warning.
> >       * Makefile.in (cgraphunit.o): Depend on function.h
> >But I think it is a C++ front-end bug which deals with cloning as it is 
> >not warning about
> >an unused parameter in a constructor.
> >
> Whether or not that is true, it is Jan's responsibility to deal with the 
> problem.
> 
> It is not acceptable to completely ignore this kind of breakage 
> indefinitely.  Two months is far too long.

I was completely unaware of any testsuite regressions caused by this
patch (it definitly passed full testing on i686 according to the logs).
I will try to look into it soon (I need to prioritize things now, so I
will look on it after the 3.4.1 problems)

Honza
> 
> -- 
> Mark Mitchell
> CodeSourcery, LLC
> (916) 791-8304
> mark@codesourcery.com
>