eeh.h compile warnings / adbhid.c build failure

[Benjamin Herrenschmidt]

On Tue, 2005-03-08 at 15:10 -0800, Brad Boyer wrote:
> On Wed, Mar 09, 2005 at 09:26:55AM +1100, Benjamin Herrenschmidt wrote:
> > On Tue, 2005-03-08 at 13:56 +0100, Olaf Hering wrote:
> > >  On Thu, Mar 03, Benjamin Herrenschmidt wrote:
> > > 
> > > > There is no ADB bus on a G5, so the driver isn't useful anyway.
> > > > Currently, ppc64 allows you to enable pmac drivers that won't build, but
> > > > they also are useless on G5s. I'll fix that over time though.
> > > 
> > > They are of course not useless. Send this patch to Linus to allow the
> > > mouse button emulation until either someone split it off the ADB driver,
> > > or until someone fixes the stupid userinterfaces in Linux.
> > 
> > Oh well, don't people buy real mice to plug on G5s ? :)
> 
> I bought a very fancy USB pointing device for my G5, but it would be
> nice to support the older stuff. Eventually I'm planning to add
> support for the Griffin iMate, which would give us ADB on anything
> that supports USB. It's just not at the top of my list.

If we go that way, then we should finally bite the bullet and get ADB in
the device model (define an adb bus_type, with proper drivers etc...).
That would also allow a clean mecanism (sysfs properties) for things
like trackpad settings, etc...

> > Anyway, mouse button emulation should be split off adb stuff.
> 
> I'm pretty sure it already is. The last time I looked at it, the
> only tie it had left was presenting itself as an ADB device to
> the input layer (bustype of BUS_ADB). I have to admit I haven't
> tried it, but it ought to work without any of the actual ADB code
> even compiled in.
> 
> I should caveat that by saying that you can't compile the in-kernel
> mouse emulation in at all unless it's on MAC or PPC_PMAC due to the
> fact that it's in the drivers/macintosh directory.
> 
> 	Brad Boyer
> 	flar at allandria.com
-- 
Benjamin Herrenschmidt <benh at kernel.crashing.org>