Subject: Re: Trouble compiling old Ivanenko colorhack
To: Dinsdale Piranha <dinsdale@vegas.infi.net>
From: Stephen C. Brown <sbrown@shellx.best.com>
List: port-mac68k
Date: 03/30/1997 12:12:53
There's a newer version of the colorkit kernel available at:

<ftp://ftp.best.com/pub.s/sbrown/NetBSD/kernels/netbsd.dormouse#2.hqx>

This is compiled from -current source around about January 5th.
There are also source diffs and a tar of the relevant files available
at:

<ftp://ftp.best.com/pub.s/sbrown/NetBSD/colorkit/>

These will likely require changes to work with the -current.  But,
they are closer to functioning than the source from last October.
Taras Ivanenko has also made the colorkit support available as an lkm
(loadable kernel module).  I forget the address to get that right off.
Perhaps Taras can fill you in there....

Steve Brown,
sbrown@best.com

p.s.  There have been several reports of the "dormouse" kernel allowing
        color nubus functionality on the quadra systems.  There have also
        been a couple of reports of quadras not getting this functionality.
        It's likely all dependent on which nubus video card you use.  It
        could also be because of quirks in the nubus support on certain
        quadras as well.  Just don't expect this functionality to be a
        "sure thing".

>I've recently attempted to compile a -current GENERIC kernel with Taras
>Ivanenko's old (10/96) NuBus color video hack patched in.  Everything has
>gone fine until the compiler reaches arch/mac68k/dev/nubus.c; there, it gets
>a bunch of errors and chokes.  Now, I figure that at some point between last
>October and now, *someone* else must have tried to compile this thing; has
>anyone else had any problems?  I've tried to come up with an explanation for
>this and frankly I'm stumped.  Here are the errors, if that'll help...
>
>cc1: warnings being treated as errors
>../../../../arch/mac68k/dev/nubus.c:67: warning: initialization from
>incompatibl
>e pointer type
>../../../../arch/mac68k/dev/nubus.c:72: warning: initialization makes
>pointer fr
>om integer without a cast
>../../../../arch/mac68k/dev/nubus.c: In function `nubusmatch':
>../../../../arch/mac68k/dev/nubus.c:85: dereferencing pointer to incomplete
>type
>../../../../arch/mac68k/dev/nubus.c:85: `BUS_NUBUS' undeclared (first use
>this f
>unction)
>../.../../../arch/mac68k/dev/nubus.c:85: (Each undeclared identifier is
>reported
>only once
>../../../../arch/mac68k/dev/nubus.c:85: for each function it appears in.)
>../../../../arch/mac68k/dev/nubus.c: In function `probe_slot':
>../../../../arch/mac68k/dev/nubus.c:206: warning: implicit declaration of
>functi
>on `bus_peek'
>../../../../arch/mac68k/dev/nubus.c:206: `BUS_NUBUS' undeclared (first use
>this
>function)
>../../../../arch/mac68k/dev/nubus.c:243: warning: implicit declaration of
>functi
>on `bus_mapin'
>*** Error code 1
>
>Stop.
>
>-----
>
>P.S.  Thanks to all those who responded to my question about config!  :)