Re: [RFA] breakpoint.c, create_breakpoint, document that the case statement falls through.

[Pedro Alves <pedro@codesourcery.com>]

On Thursday 03 March 2011 04:26:52, Joel Brobecker wrote:
> > 2011-03-02  Michael Snyder  <msnyder@vmware.com>
> > 
> > 	* breakpoint.c (create_breakpoint): Document that case falls through.
> > 
> > Index: breakpoint.c
> > ===================================================================
> > RCS file: /cvs/src/src/gdb/breakpoint.c,v
> > retrieving revision 1.546
> > diff -u -p -u -p -r1.546 breakpoint.c
> > --- breakpoint.c	1 Mar 2011 02:16:56 -0000	1.546
> > +++ breakpoint.c	3 Mar 2011 00:38:01 -0000
> > @@ -7843,6 +7843,7 @@ create_breakpoint (struct gdbarch *gdbar
> >  	default:
> >  	  throw_exception (e);
> >  	}
> > +      /* FALLTHROUGH */
> >      default:
> >        if (!sals.nelts)
> >  	return 0;
> 
> Just some thoughts:
> 
> It's actually never going to fall through, is it? Can we use "break;"
> instead, even if we know it's never going to be reached? I think
> it would make it clearer by not suggesting something that isn't
> supposed to happen (the fall through).

The root problem seems to be that lint hasn't been tought that
throw_exception doesn't return, so it thinks we do fall-through.
There are mechanisms to teach lint about the no-return property
of functions.

-- 
Pedro Alves