Messages in this thread | | | From | "Peter T. Breuer" <> | Subject | Re: useful exercise | Date | Fri, 8 Jan 1999 01:00:11 +0100 (MET) |
| |
"A month of sundays ago Michael Elizabeth Chastain wrote:" > > Hi Peter, > > > if (C_FOO) { > > } > > This has a fatal implementation problem. include/linux/autoconf.h does > not define C_FOO to 0 if the feature is not set; it #undef's C_FOO.
Yes I know. That's why I suggested C_FOO instead of CONFIG_FOO. That way it wouldn't disturb the existing arrangement. A second pass at building autoconf.h (or whatever) would be required. I can build it from .config if req'd. I don't know how autoconf.h is made.
> Or it simply has no C_FOO lines at all because C_FOO might be conditional > on some other variable C_BAR.
.config should have enough data.
> Other than that, an interesting idea. I am interested in anything that > moves error reporting earlier in the development process. Especially if > it moves the error report to a place *before* a patch hits the > Penguin Nuclear Mailbox.
Uh, yes. Some option might be selected to make a trial run with. I mean, if we have something that just affects a file or two, we could try sticking a #ifdef CONFIG_FOO .. #define C_FOO 1 .. #else #define C_FOO 0 #endif in it. Then change the #ifdef CONFIG_FOO's to if(C_FOO) and see what's needed to make it compile.
> Michael Elizabeth Chastain
Peter
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |