Date: Mon, 17 Jan 2005 17:16:56 -0500 From: Garance A Drosihn <drosih@rpi.edu> To: obrien@freebsd.org Cc: Robert Watson <rwatson@freebsd.org> Subject: Re: gratuitous gcc warnings: unused function arguments? Message-ID: <p06200726be11e141a97d@[128.113.24.47]> In-Reply-To: <20050117053655.GA96378@dragon.nuxi.com> References: <Pine.NEB.3.96L.1050116210328.50371F-100000@fledge.watson.org> <p06200723be10ba368225@[128.113.24.47]> <20050117053655.GA96378@dragon.nuxi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
At 9:36 PM -0800 1/16/05, David O'Brien wrote: >On Sun, Jan 16, 2005, Garance A Drosihn wrote: > > At 9:05 PM +0000 1/16/05, Robert Watson wrote: >> >On Sun, 16 Jan 2005, David O'Brien wrote: >> > > We're not going to hack GCC to deal with this. That is going way >> > > too far. This is coming up because people are using high WARNS > > > > values in Makefiles. Either back them down to a lower WARNS > > > > value; or we should add -Wno-unused-parameter to WARNS level 3. > > > > > > I'd be fine with simply pushing threshold for unused parameters > > > up a few notches on the warning scale. I'd like to have access > > > to the other interesting warnings are WARNS=3 and WARNS=4 relating > > > to qualifiers, strings, etc. > > >> I think it would be useful to keep that warning "in general", but >> have an option to turn it off. The following seems to work for me, >> assuming we can decide on the best name for a new NO_WUNUSED_ARGS >> option: > >Do you have a piece of code that exhibits this warning but otherwise >could pass at higher warning level? Who? Robert? Or me? I might, but I can't think of any at the moment. However, I have had cases where this warning has pointed out bugs to me -- such as a routine using a global variable because it had misspelled a reference to one of its own parameters. So, "in general" I do think this is a useful warning. And I (personally) prefer using __unused or the '(void)some_arg;' trick to silence the warning. >I'm resistant to adding a lot of NO_FOO_WARNING knobs. Otherwise we >might as well as remove the whole WARNS thing and set each warning a >Makefile wants. If developers would prefer to always turn it off for WARNS=3 and WARNS=4, that is certainly fine with me too. I agree that we do not want a lot of NO_WARN_FOO options, but I can see where this warning might be very annoying for a few (otherwise clean) source files. I wouldn't want those source files to be compiled with WARNS=2 (or to avoid -Werror) for no other reason than to avoid hassles with this one check. -- Garance Alistair Drosehn = gad@gilead.netel.rpi.edu Senior Systems Programmer or gad@freebsd.org Rensselaer Polytechnic Institute or drosih@rpi.edu
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?p06200726be11e141a97d>