Re: Kconfig troubles when using menuconfig - Was: [patch]Re: [linux-usb-devel] linux-2.6.22-rc5-gf1518a0 build #300 failed in zc0301_core.c

["Satyam Sharma" <satyam.sharma@gmail.com>]

Hi,

On 6/24/07, Trent Piepho <xyzzy@speakeasy.org> wrote:
> On Sat, 23 Jun 2007, Satyam Sharma wrote:
> > On 6/23/07, Trent Piepho <xyzzy@speakeasy.org> wrote:
> > > [...]
> > > Now all the usb drivers will gain USB as a dependency directly and can't be
> > > set to something higher than USB.
> >
> > Ok, so we add this as solution 2.(c) to the reply I just sent to Jan :-)
> >
> > But I still prefer 2.(b) -- making the config scripts intelligent so that if a
> > given "menuconfig FOO depends on BAR", then all the "config BAZ"s
> > inside this menuconfig also automatically "depend on" BAR too.
>
> Of course, there currently is no "inside" a menuconfig.  You would have to do
> something like make everything inside an "if FOO / endif" gain not just a
> dependency on FOO, but also gain a dependency on all of FOO's dependencies.

Yes, making the config scripts do what you describe there
automagically is exactly what I meant there.

> > This is simpler in the long run because it requires least amount
> > (actually none) of redundant typing and would continue to work in
> > the future if/when the:
> > [...]
> > configmenu FOO
> > ...
> > endconfigmenu # FOO
> >
> > kind of idiom ...
>
> Like that I suggested here?
> http://article.gmane.org/gmane.linux.kernel/524823
>
> Basically, make menuconfig work like menu does, except the menu itself can be
> turned on and off.  Instead of having menuconfig work like a config, but with
> some kind of "menu" hint.  It seems like the former is more in line with what
> menuconfig is actually used for.

Again, we're in complete agreement here (also like what Jan
mentioned elsewhere on this thread). But then there's often
too much "suggestions" on Kconfig/Kbuild matters on lkml,
but too little code (wonder if that was what got Roman
irritated yesterday). This may be because stuff in scripts/ doesn't
quite get the kind of eyeballs kernel/ or fs/ or even drivers/ gets,
but I guess the only way to get Roman around to our view would
be start submitting patches for the solutions we're all fond of
"suggesting" :-)

Satyam