[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: console plans

From: Marcus Brinkmann
Subject: Re: console plans
Date: Thu, 14 Feb 2002 19:50:26 +0100
User-agent: Mutt/1.3.27i

On Wed, Feb 13, 2002 at 10:56:45AM +0100, Niels Möller wrote:
> Marcus Brinkmann <address@hidden> writes:
> > The output half is similar.  The only driver initially will be a standard
> > VGA driver (Super VGA might be an option for later).  I have found some
> > documentation on VGA cards (FreeVGA Project), and we will do all the VGA
> > card programming in user space.
> Do you have a plan for how to coordinate access to the hardware with
> other programs, like framebuffers and X?

I forgot to mention that.  Adding to Rolands reply, there is no particular
watch guard in what I considered yet, but I am aware that there is an
interface for virtual console switching in X that is system dependent (and
we don't implement it yet, of course).  Rolands suggestion to mimick the
existing behaviour is the right one of course.
> To me, it would make some sense to use a separate device-like
> translator for representing the vga hardware.

Eventually, you might want something like this, if it isn't the console
server itself already.  If you have the console server and X as the only
players, the mentioned protocol ensures that X can safely grab control over
the graphic card and the keyboard (the console server might in fact act as an
repeater for the keyboard input on /dev/kbd) and hands it back to the console
server, although the console server might have some hotkeys to reset state
in emergency cases (only works if used in repeater mode of course).

> At the very least, it
> would need to keep enough knowledge of what people are doing with it
> so that it, at any time, can restore the hardware to a sane state. It
> might also be desirable to put the hardware banging in a separate
> program, just for limiting the amount of code that needs to run with
> privileges.

My main concerns are the text console (I am quite different from Roland in
that regard :) and that we can run X, because we will have to break support
for it temporarily (unless I implement the repeater very early).

But I am very glad you raised the issue, because it is indeed an important
one, although I will only make sure what exists will be safe, but not as
flexible as you suggest.

It seems to me that getting X to run good enough is sufficient for now. 
Note to myself:  Check what is required to make two or more X displays on
different virtual consoles working.


`Rhubarb is no Egyptian god.' Debian http://www.debian.org address@hidden
Marcus Brinkmann              GNU    http://www.gnu.org    address@hidden

reply via email to

[Prev in Thread] Current Thread [Next in Thread]