> >WRT 'rasops', as a 'side effect' rasops_setfont() does all the work
> >for you, provided you've changed the 'struct rasops_info'. [The name
> >of that function and semantics will change a bit when I fix PR 8121].
> 
> So after I modify the chip to a different res/bpp, I would need to
> call rasops_setfont?

Yeah, but rcons needs to be told too. I haven't even thought about
what's needed yet :).
 
> > > Speaking of rcons, it would be nice to be able to tell
> > > it what to use for fg pixel and bg pixel. Also if it
> >
> >If you're talking about ANSI color, that's possible. For actual pel
> >values, there's nothing there to let you do that at the moment.
> 
> If in 8bpp pseudo-color, I want to be able to specify the indexes
> into the colormap to use for fg and bg.  This is because the firmware
> for the sparcbook uses 0 for white and 255 for black and it would be
> nice if rcons could continue to use the same values.

Hmm... rasops_info.ri_devcmap contains the translations from index 
value to what's actually written to the screen. The 8bpp specific code 
ignores these, because it assumes that the index is the value to be 
written to the framebuffer. If this code brought in line with the rest, 
it'll do what you want (and make fixing part of PR 8121 a synch).
 
- ad