On Tue, Apr 25, 2000 at 05:49:58PM -0700, Dunlap, Randy wrote:
> Matt,
Sorry, my name isn't "Matt", but "Matti", the ending vovel is
pronounced (but lets not dwell on this side issue here..)
> I am looking over the compiler warnings that you listed.
> BTW, I'll take patches for them if you want to send them.
...
> I don't see these warnings since I'm not using an Alpha.
>
> > keybdev.c: In function `keybdev_event':
> > keybdev.c:136: warning: implicit declaration of function `emulate_raw'
> >
> > A declaration inclusion missing somewhere.
>
> This one looks odd to me. emulate_raw() is arch-dependent and
> is defined 2 times before line 136. One time is for
> #if defined(CONFIG_X86) || defined(CONFIG_IA64) || \
> defined(CONFIG_ALPHA) || defined(CONFIG_MIPS)
>
> and the other one is for
> #elif defined(CONFIG_ADB_KEYBOARD)
>
> So are you on an Alpha but CONFIG_ALPHA is not defined, or what?
> Can you try to explain this?
Proper subtest for this is: defined(__alpha__)
( and defined(__mips__) -- I think )
While arch/i386/config.in file begins with
define_bool CONFIG_X86 y
same isn't true for arch/alpha/config.in, nor for
mips*, sparc*, ppc or m68k.
(IBM i390 isn't yet folded into 2.3 tree, so I can't say
where it stands -- not that it has uses for USB...)
ARM seems to have "CONFIG_ARM"
IA64 seems to have "CONFIG_IA64"
Super-H seems to have "CONFIG_SUPERH"
Mostly this tells of genealogy of the arch/xxxx/config.in
file; they are modelled after ia32, and are newer than
when ever i386 got that definition..
I am not quite sure, should other architectures get
similar generic labels ?
> I'm still looking at the others.
>
> ~Randy
/Matti Aarnio
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
