Re: ofwboot problem on OF2 machines

[Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>]

Frank Wille wrote:

> Izumi Tsutsui wrote:
> 
> > Sorry for old topic, but I tried recent ofwboot on my Apus2000
> > and noticed it didn't work properly, probably due to OF 2.0 quirks:
> > [...]
> >
> > I don't remember how I tested your ofdev.c patch at that time,
> > but the following change causes a problem:
> > [...]
> 
> Ok.
> Unfortunately I have no OF 2.0 machines to test that.

No problem, as noted in cvs log I tested it per your request,
but probably I missed something..

> > If it's for lazy defaults, adding the following ofw_version check
> > makes it work even on my Apus2000. What do you think about this?
> 
> Ok for me. It works on my PowerMac G4.
> 
> Maybe we can remove the strlcat() block from above in the current source (so
> everybody can test it until 7.0), but pull up your patch for netbsd-6, to
> be on the safe side?

I'll commit the patch shortly and send a pullup request,
but If there is no bad side effect on OF3 machines
I'll leave that block.

> Unfortunately there are more problem left with ofwboot: since October 2010
> no ofwboot 1.12 version seems to work on my PowerBook G4 or iBook G4. No
> idea what is wrong there...
> 
> I will have to find some time in the next weeks to debug that.

I'll also dig out my G4 Mac later.

> > Should it be done rather in the above prom2boot() function?
> 
> As far as I understand prom2boot() is only called in the beginning to
> convert the bootpath from the OFW environment. It is not used for the path
> entered on the command line, so it wouldn't help you?

Hmm.

Probably prom2boot() is intended to fix path passed from firmware,
and we can note what device node path should be used on command line
prompt in document?

Anyway, I'll leave it as is for now.

--
Izumi Tsutsui