On Wed, 10 Jan 2001, Linus Torvalds wrote:
>
>
> On Wed, 10 Jan 2001, Marco Colombo wrote:
> > >
> > > case xxx:
> > > /* fallthrough */ ;
> > > }
> > >
> > > or something (or maybe just a "break" statement), just so that we don't
> > > turn the poor C language into line noise (can anybody say "perl" ;)
> >
> > Of course, you don't mean that the fallthrough comment and the break
> > statement have the same functionality! (well you put the closing
> > bracket and I agree that for the last case it's the same).
>
> Note that the warning case we're discussing was really only about case
> statements at the end of a compound statement.
>
> In the middle of compound statements we're already fine: it's only the
> corner case of a case "statement" without the statement that gcc
> historically used to accept without warning, and that the gcc people only
> recently noticed that they really shouldn't accept at all.
>
> So that's why a comment and a "break" is equivalent. ONLY for the special
> case of the new compile warning, though, obviously (see the subject line,
> but yes, I should have made that more explicit).
I see. The choice between the comment and the 'break' can be left
to the author, maybe he knows what is the best one (the one that will
let us add a new case with less effort).
If you bother setting a styleguide rule, I believe the latter is more
common.
But still I dislike semicolons *after* comments. Why not:
case a:
...
; /* fallthrough */
case b:
...
; /* fallthrough */
}
It looks just a little less ugly to my eyes (please take a few seconds to
get used at it), maybe it's because it's reminiscent of shell
double-semicolon. In general, I find it more readable when comments
are somewhat "outside" the C code.
[ Note: I still put it after the first case, both for documenting the
fallthrough behaviour, which is a good thing anyway, and to allow
deleting the last case with no thinking of it. ]
I'd even put a tab before the comment.
case b:
...
; /* fallthrough */
}
Ok, that's really a matter of taste. You spend more time in front
of the source than me, and you know better. I'd like you to set a kind
of rule, because it's something that many are not used to see and it's
easier to recognise it if it looks the same everywhere.
.TM.
--
____/ ____/ /
/ / / Marco Colombo
___/ ___ / / Technical Manager
/ / / ESI s.r.l.
_____/ _____/ _/ [EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/
