Re: [PATCHSET] PC-9800 subarch. support for 2.5.60 (12/34) console

[Osamu Tomita]

Thanks for the comments.

James Simmons wrote:
> 
> > > --- linux/drivers/char/console_macros.h     Sat Oct 19 13:01:17 2002
> > > +++ linux98/drivers/char/console_macros.h   Mon Oct 28 16:53:39 2002
> > > @@ -55,6 +55,10 @@
> > >  #define    s_reverse       (vc_cons[currcons].d->vc_s_reverse)
> > >  #define    ulcolor         (vc_cons[currcons].d->vc_ulcolor)
> > >  #define    halfcolor       (vc_cons[currcons].d->vc_halfcolor)
> > > +#define def_attr   (vc_cons[currcons].d->vc_def_attr)
> > > +#define ul_attr            (vc_cons[currcons].d->vc_ul_attr)
> > > +#define half_attr  (vc_cons[currcons].d->vc_half_attr)
> > > +#define bold_attr  (vc_cons[currcons].d->vc_bold_attr)
> >
> > Bah, console_macros.h should just die.
> 
> Agree. Next developement series it will.
How do I know your chnages? Please points me. URL? ML archive?
This means writing directly them into source file?
 
> > Please set CONFIG_KANJI in the Kconfig file and in general
> > the CONFIG_KANJI usere look really messy.  I don't think it's
> > easy to get them cleaned up before 2.6, you might get in contact
> > with James who works on the console layer to properly integrate them.
> 
> I doubt it. The console system is designed around VGA text mode. It will
> be lots of #ifdef to get it to work. Just a bad mess. I plan someday to
> rework the console system to handle all these cases.
For difference of VRAM mapping, I implemented scr_* function family.
But codes for difference of attribute value remains in mainline sources.
And I need many #if for support japanese character.
How do you think about 2bytes character support?
I think, If kernel suports 2bytes character, many people useing japanese 
chinese korean... may be happy.

Regards,
Osamu Tomita