Re: cursing my yboot'd Lombard

[Michael Schmitz <schmitz@zirkon.biophys.uni-duesseldorf.de>]

> > What should we initialize the palette registers to, in order to make the
> > cursor visible again? 
> 
> Ah, that's another possible approach: initialize the palette such that
> inverting the pixel value inverts the color.
> 
> But I see one problem with that: how to know you're in 16-color text console
> emulation mode or not? You don't want to mess with other palette entries when
> e.g. running X.

Well, judging from the fact that running X or not didn't change anything,
I conclude that X doesn't use that palette entry at all (at least never
wrote it). Or switching consoles saves/restores all palette entries -
shouldn't it do that anyway? 
 
> > What would the proper mask to use be, and would it be the same for all
> > cards and bpp's? 
> 
> It would depend on the card and bpp. For atyfb (yes, I'm to lazy to calculate
> the hex values):
> 
>   - depth 15: (15 << 10) | (15 << 5) | 15
> 
>   - depth 16 (not yet supported): (15 << 11) | (15 << 5) | 15
> 
>   - depth 24: (15 << 16) | (15 << 8) | 15
> 
>   - depth 32: (15 << 24) | (15 << 16) | (15 << 8) | 15

I see (it's actually easier to understand this way, the hex values would 
have been confusing). I'll try to play with it as soon as I find some time
for that. 

	Michael