On Mon, Jul 11, 2016 at 04:43:57PM -0400, David Malcolm wrote: > On Mon, 2016-07-11 at 22:34 +0200, Jakub Jelinek wrote: > > On Mon, Jul 11, 2016 at 10:23:30PM +0200, Marek Polacek wrote: > > > On Mon, Jul 11, 2016 at 01:18:02PM -0700, Andi Kleen wrote: > > > > > I explained why supporting the classic lint style comment > > > > > wouldn't fly. > > > > > > > > Not convincing, it worked fine for 30+ years of lints. > > > > > > So how can the compiler handle > > > /* Never ever fall through here */ > > > ? > > > > It can't. But it perhaps could handle a couple of documented most > > common > > comment styles, perhaps only if followed by break token, and turn say > > /* FALLTHROUGH */ > > break > > Presumably you mean "case" here? A "break" token is surely the thing > we're *not* expecting, right?
Sure, I meant case token. > ...or maybe we could detect such comments and store, say, a bitmap of > all source_location/location_t values at which we've seen a > "fallthrough" style comment, and then have the warning code query this > bitmap for a comment in the pertinent range. It can be flag on the case token too. The thing is if we care about -save-temps, ccache, distcc and non-integrated preprocessing (I'm currently trying to convince ccache to start using -fdirectives-only, but they don't want to, see https://github.com/ccache/ccache/issues/93 for details). Jakub