Re: [PATCH] x86/efi: Add EFI framebuffer earlyprintk support

[Matt Fleming <matt-HNK1S37rvNbeXh+fF434Mdi2O/JbrIOy@public.gmane.org>]

On Mon, 14 Oct, at 09:57:26AM, Ingo Molnar wrote:
> 
> * H. Peter Anvin <hpa-YMNOUZJC4hwAvxtiuMwx3w@public.gmane.org> wrote:
> 
> > On 10/12/2013 10:49 AM, Ingo Molnar wrote:
> > > 
> > >>>> +static void early_efi_write_char(void *dst, char c, int h)
> > >>>> +{
> > >>>> +	const u8 *src;
> > >>>> +	u32 fgcolor = 0xaaaaaa;
> > >>>
> > >>> That's RGB grey, right? Why not 0xffffff for a very visible white?
> > >>  
> > >> The VGA earlyprintk code uses the equivalent grey, AFAIK, which is why I 
> > >> picked this value.
> > > 
> > > The VGA code should be changed to white too I suspect ;-)
> > > 
> > 
> > For compatibility with the classical text console we use light grey as 
> > the default color and a double-stroked font (which was necessary on the 
> > CGA display since it didn't have enough bandwidth to handle 
> > single-stroked fonts well).  The problem with changing that to white is 
> > that you end up with a mismatch between the earlyprintk console and the 
> > "real" console, *or* we change the behavior of everyone's consoles...
> 
> Btw., I have no problem at all with making the early console look separate 
> and making it clear when we switch to the 'real' console.
> 
> earlyprintk is a debug method, and more information can only be good. 

Have we reached a consensus here? I've no strong opinion either way on
the colour of the text.

-- 
Matt Fleming, Intel Open Source Technology Center