Peter Samuelson wrote:
>
> [Greg Banks]
> > [...CONFIG_SERIAL and CONFIG_PCMCIA warnings...]
> >
> Hmmm, either I missed those in your earlier messages, or you didn't
> post them.
Probably I didn't post them. What I posted was a small subset of the full log.
> > > + dep_tristate ' I2C bit-banging interfaces' CONFIG_I2C_ALGOBIT $CONFIG_I2C
> >
> > Are you sure want this one there?
>
> I didn't like it either, but it's needed in a couple odd places. What
> would you suggest - moving the whole i2c menu up?
Not all the way up to the new menu, but before the bits that depend on them,
which are in drivers/video/ drivers/media/video and drivers/ieee1394 IIRC.
> Thanks, your "oracle" feedback is much appreciated.
I'm hoping to have an RPM out this weekend so you can do the augury yourself.
> > @@ -210,2 +210,5 @@
> > 2 CONFIG_FB
> > + 2 CONFIG_KMOD
> > + 2 CONFIG_MODULES
> > + 2 CONFIG_MODVERSIONS
> > 2 CONFIG_RTC
>
> What does that mean? All I did there was to combine two toplevel
> menus into one. Did this do something bad?
Apparently. In the stock kernel:
warning:arch/mips/config.in:224:"CONFIG_KMOD" has overlapping definitions
warning:init/Config.in:19:location of previous definition
warning:arch/parisc/config.in:53:"CONFIG_KMOD" has overlapping definitions
warning:init/Config.in:19:location of previous definition
Did I mention Gordian knot?
> > -36 overlapping-definitions
> > +38 overlapping-definitions
> > 11 CONFIG_SOUND_CMPCI_FMIO
> > @@ -261,2 +263,3 @@
> > 2 CONFIG_PARPORT
> > + 2 CONFIG_USB
>
> OK, I see CONFIG_USB in arch/cris/drivers/Config.in - is there another
> instance I'm missing?
There's two in the same file, lines 185 and 189.
> > -8 different-compound-type
> > -3290 total
> > +10 different-compound-type
> > +3055 total
>
> different-compound-type?
Please ignore that one. It's an artifact of the way I check for symbols
not declared anywhere at all, related to config.in's using the same banner
for a menu and a comment.
Greg.
--
the price of civilisation today is a courageous willingness to prevail,
with force, if necessary, against whatever vicious and uncomprehending
enemies try to strike it down. - Roger Sandall, The Age, 28Sep2001.
[Greg Banks]
> > > > + dep_tristate ' I2C bit-banging interfaces' CONFIG_I2C_ALGOBIT $CONFIG_I2C
> > >
> > > Are you sure want this one there?
> >
> > I didn't like it either, but it's needed in a couple odd places. What
> > would you suggest - moving the whole i2c menu up?
>
> Not all the way up to the new menu, but before the bits that depend on them,
> which are in drivers/video/ drivers/media/video and drivers/ieee1394 IIRC.
It should be possible to take I2C back out of init/Config.in, in that
case. I'll do that in my tree.
> > > + 2 CONFIG_KMOD
> > > + 2 CONFIG_MODULES
> > > + 2 CONFIG_MODVERSIONS
> > > 2 CONFIG_RTC
> >
> > What does that mean? All I did there was to combine two toplevel
> > menus into one. Did this do something bad?
>
> Apparently. In the stock kernel:
>
> warning:arch/mips/config.in:224:"CONFIG_KMOD" has overlapping definitions
> warning:init/Config.in:19:location of previous definition
> warning:arch/parisc/config.in:53:"CONFIG_KMOD" has overlapping definitions
> warning:init/Config.in:19:location of previous definition
Weird. These should have triggered all along - any idea why they
didn't?
Well, they're fixed in my tree. It looks [trivial], but this one
makes me uneasy. I mean, it's such an obvious bug - the toplevel
"Loadable module support" menu appears twice - that I almost think it
was somehow intentional.
> Did I mention Gordian knot?
Yes, I believe you did. Does that make ESR Alexander the Great? (:
> > OK, I see CONFIG_USB in arch/cris/drivers/Config.in - is there another
> > instance I'm missing?
>
> There's two in the same file, lines 185 and 189.
Right - they're mutually exclusive, so I thought it might only count
as one. Anyway, fixed in my tree.
> related to config.in's using the same banner for a menu and a
> comment.
mainmenu_option next_comment ... now *there's* a bit of syntax that
needs to change. Even config-language.txt agrees.
Peter