On 25 Feb 00 at 14:33, Mike A. Harris wrote:
> >up to 255xMUCH screen (VT supports at most 255 columns, so no problem),
> >but standard VGA pixclocks are 25 or 28MHz, so without chipset knowledge
> >you can have at most 90 characters line. And with 16 lines character
> >cell height you can have at most 30 lines (60 lines with 8x8 chars).
> Exactly. Which is why my proposal is to have a userland app
> inform the kernel of the necessary modelines, fonts, clocks to
> use. Thus the kernel does not start out knowing about anything
> more than it does now. Your startup scripts run an app which is
> something like SVGAtextmode which informs the kernel how to
> change to the modes that will be used, and then tells the kernel
> what mode each VC is supposed to be in.
> Do you see anything I'm missing in the above? I think it is
> possible, but I'm not 100% sure. I'd like other people's
> feedback.
How do you tell to kernel how to switch pixclock? It is done by non-vga
registers...
> >> 1) Puts MINIMAL code in the kernel to do what needs to be done.
> >Dig vgafb somewhere, copy vgaHWinit & vgaHWrestore from matroxfb
> >and glue it together.
> Ok, can anyone tell me where to get the vgafb code? If it does
> what I'm looking for, I'd be glad to have something to start
> with.
Look into linux-2.1.125 package... I believe that it was in...
> >> 2) Does not bloat the kernel with details of every video card
> >> known to man or mode tables.
> >
> >If you will not, you cannot switch pixclock to reasonable values,
> >you can have 12.6, 14.2, 25.3 and 28.5 MHz dotclock and nothing else.
> Is this really true? Could I not have a userland interface which
> a special app could tell the kernel these things at runtime? If
> it is impossible, I'd like to know why, or is it just really
> complex?
Through non-VGA registers only. And you must change it when you change
pixclock. And if every console has different size, you must change
pixclock...
> >fbset passes complete videomode information to kernel. But kernel
> >has to pass it to hardware - and for this it must know about
> >hardware. Standard VGA is only standard VGA. Face to it.
> Hmm. So a userland app would have to know what video card you
> have. This could be done with autodetection, or by the user
> picking a card. I don't see this a problem because you need to
> do the same for X, or for any other OS out there anyways. Once
> the card is known by a userland app, it can inform the kernel of
> this somehow no? I mean not with what we have now, but we could
> extend the kernel in this manner no?
Type of card is not interesting. How to program such beast is interesting
thing...
> >(borrow from vga16fb or matroxfb) and which programs palette (vga16fb).
> >Only thing you have to add is pixclock programming code for unusual
> >dotclocks.
> Ok, then it does sound feasible. The dotclock code can be stolen
> from SVGATextMode, as long as it is GPL I guess. At least it can
> provide me with an idea of the details necessary.
Yes. But it is not up to date with latest hardware.
> Just point me to the vgafb code, and I'll have a look-see. That
> is from the 2.1.x era correct?
I have to go to to train station and left Prague just now, unfortunately.
If you will not find it before Monday, ask me again and I dig through
my archives.
Petr
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Feb 29 2000 - 21:00:13 EST