cvs commit: src/sys/dev/an if_an.c if_an_isa.c if_an_pccard.c if_an_pci.c

[Bruce Evans]

On Thu, 10 Apr 2003, M. Warner Losh wrote:

> In message: <Pine.BSF.4.21.0304101114320.31875-100000 at root.org>
>             Nate Lawson <nate at root.org> writes:
> : > @@ -817,9 +806,11 @@
> : >  	 */
> : >  	ether_ifattach(ifp, sc->arpcom.ac_enaddr);
> : >  	callout_handle_init(&sc->an_stat_ch);
> : > -	AN_UNLOCK(sc);
> : >
> : >  	return(0);
> : > +fail:;
> : > +	mtx_destroy(&sc->an_mtx);
> : > +	return(error);
> : >  }
> : >
> : >  static void
> :
> : Extra ; after "fail:"
>
> I feel that it isn't a mistake.  A label labels a statement,

It is not a mistake in the same way that an extra semicolon on a case
statement is not a mistake.  It is just a style bug.

> and this
> label labels a null statement on purpose in case the statements after
> it disappear.

A C programmer would not remove the statements without also removing
the label (since it is part of the first statement), and the statements
can't disappear here since at least a return statement is needed to
return a value.

Bruce