On Tue, Feb 21, 2006 at 05:09:45PM +0100, Quentin Garnier wrote:
> On Tue, Feb 21, 2006 at 04:22:41PM +0100, Peter Postma wrote:
> > static int
> > hpcapm_match(struct device *parent, struct cfdata *cf, void *aux)
> > {
> > 	struct mainbus_attach_args *ma = aux;
> > 
> > 	if (strcmp(ma->ma_name, hpcapm_cd.cd_name) != 0) {
> > [..]
> > 
> > Or do you suggest to work around this?
> 
> This is completely broken.  There are quite some config(9) abuses in hpc
> land...  I'll have a look at hpcapm users and fix that mess.  Th
> hpcapm/apmdev separation looks dubious, too.
> 

Do you have any idea how long this will take (I'd like to get this all done
before 4.0 will be branched).

I've put a hpcapm(4)'ified j720pwr.c here (untested):
ftp://ftp.netbsd.org/pub/NetBSD/misc/peter/j720pwr.c

Yes, this needs "j720pwr* at j720ssp?" to install the config hooks, but
I'm afraid there's no better solution.

> One thing I don't quite get in your patch though is the extra layer for
> each objects.  hpckbd should attach directly to j720ssp;  the split in
> different source files is fine, though.

Is this a showstopper, i.e. does this need to be fixed before committing?

Thanks,
-- 
Peter Postma